Congolese refugees informing this study interpreted the notion of “protection space” largely through their various everyday encounters and mundane experiences of urban space. This article emphasizes an inherently spatial and scalar reading of refugees’ discourses of their protection and insecurity in their city of exile Kampala, Uganda. Conceptually, my examination focuses on physical, imagined, lived, and relational elements of space. Refugees’ conceptualizations of urban space are, for analytical purposes, discussed at the micro-, meso-, and macro-scales. It is concluded that the commonly held understanding of “protection space” as a largely institutional space between refugees and protection institutions only provides us with a one-sided understanding of the concept. Thus, when “protection space” is interrogated, refugees’ understandings of urban space in its multiple forms and scales have to be incorporated into the analysis.
The 2009 urban refugee policy of the United Nations High Commissioner for Refugees (UNHCR)1 is centred on the notion of “protection space”. Even though calls for studies investigating the spatiality of refugee protection have repeatedly been made,2 there is a clear need to improve our understanding of the spatiality of urban refugee protection. Arising from this gap in the literature, the aim of this article is to provide an alternative account of “protection space” in the context of Kampala compared to the ones presented, for instance, by UNHCR3 and Sandvik4 who perceive “protection space” largely as institutional space between refugees and the protection institutions. In this article, the limitations of this institutional perspective are shown and a complementary understanding of the multi-scalar everyday spaces of protection in the city is presented.
There is a lack of an agreed or legal definition of “protection space”, but in recent years this concept has gained increasing popularity in the “humanitarian vocabulary”. Indeed, the 2009 UNHCR urban refugee policy is centred on the concept of “protection space”.5 The refugee agency defines it as “the extent to which there is a conducive environment for the internationally recognized rights of refugees to be respected and upheld”.6 Furthermore, it is perceived to consist of two elements: physical space and action space.7 Physical space is limited by geographical or territorial borders. Action space refers to the area in which a refugee moves and makes decisions about his/her life. The idea of surrogacy in action space has been stated as essential, as it refers to the possibility given to an organization or other entity to act on behalf of the refugee (i.e. providing protection). Therefore, investigating refugees’ relationships with the various institutions that provide protection for them in the city is needed.8 Yet, by treating “protection space” solely as something between refugees and UNHCR, we miss the importance of the everyday mundane encounters and spatial struggles expressed by refugees. Subsequently, this largely institutional, dualistic definition of “protection space” argued for by UNHCR is challenged in this article, as it is shown not to provide a fully purposeful representation of refugees’ commonly accepted understanding of the link between protection and space in the urban context.9
UNHCR has also argued that protection space is not fixed but rather time-sensitive and dynamic because it expands according to changes in the political, economic, social, and security environments.10 Therefore, it is understood that macro-level changes in regional and national spheres can influence the sustainability of refugee protection at the city level. Yet the analysis of micro-level production and contestation of “protective spaces”, and the related struggles over agency and power, are often missing from the studies and evaluations of the extent of “protection space”. This article aims to shed light, in particular, on these micro-, meso-, and macro-scale processes producing protective spaces in the urban environment. These scalar distinctions are, however, used as an analytical tool to unravel refugees’ discourses about urban space and thus do not systematically feature their consideration of space as such.
In line with UNHCR’s conceptualization of “protection space”, in an article from 2012, Sandvik discusses the concept of “protection space” with a focus on refugees’ precarious relationship with UNHCR as a potential provider of formal protection. Sandvik analyses “protection space” largely as “refugees’ experiences of rejection and mistreatment as physical mappings of Kampala, in which the creation and closure of urban space give meaning to the idea of protection space”.11 To my reading, her analysis needs to be complemented with an additional interpretation of “protection space” for a number of reasons. First, as already mentioned, Sandvik’s examination of urban protection focuses essentially on the relationship between refugees and UNHCR. Yet, what I aim to demonstrate here is that rather than the institutional management relationship, refugees’ sense of urban protection is affected by their everyday encounters with Ugandan inhabitants of the city at different scales. Second, Sandvik does not sufficiently theorize “space” in her article and thus her analysis of it remains somewhat unclear. This may be due to the fact that her work is socio-legal in nature. Moreover, as she perceives “protection space” as the “contestation over the creation and closure of physical spaces” and as the “bonds between the visible and invisible cityscapes of Kampala”,12 it seems to me that her investigation is focused merely on the physical and the imagined elements of “space”. My attempt here, however, is to provide a theoretically rigorous reading of “space” by drawing from mostly geographical literature on it, thus extending theoretical discussions on “protection space” in the context of refugee studies. Lastly, Sandvik’s empirical analysis is based on the situation in Kampala in 2004–2005, thus with a focus on the history of urban refugee management. Yet, in Uganda, there have been significant changes since then in terms of both refugee policy and legal frameworks. Given this, my aim is to provide a more recent analysis of the refugee context in Kampala by presenting rich empirical analysis based on my data from 2010–2011. In summary, I suggest that Sandvik’s institutional analysis of “protection space”, while relevant, needs to be complemented with another take on it as argued for in this article.
Furthermore, the review of existing − and growing − research on refugees in Kampala justifies the inherently spatial approach to my analysis of “protection space”. With regard to Kampala, the case study for this research, recent studies on refugees have covered the overall situation of urban refugees,13 youth,14 young men,15 non-formal education,16 refugee livelihoods,17 the church and the bars as forms of communities,18 and refugees’ struggles for institutional protection provided by UNHCR.19 Clearly missing are studies with an explicit focus on urban space in its multitude of forms, such as physical, imagined, lived, and relational and in its multiple scales.
In this article, I present an alternative view of “protection space” with a particular focus on urban space and refugees’ everyday life in different forms of shelter, neighbourhoods, and areas of Kampala. I pose and answer two research questions in this article. First, how do Congolese refugees’ understand the notion of “protection space”20 in the context of their everyday life in the city of Kampala, and second, how do the physical, imagined, lived, and relational elements of “space” feature their narratives of urban protection? Essentially I argue that rather than solely examining the relationship between refugees and institutions, such as UNHCR, there is a need to analyse the everyday encounters that take place at multiple spatial scales in refugees’ city of exile. Only then can we begin to grasp refugees’ sense of protection in a city.
Finally, the term “refugee protection”21 is used in this article in a broad sense, as defined by the refugee informants themselves and as discussed in current scholarship. First and foremost, it includes the traditional refugee assistance, physical and legal protection, and “durable solutions”22 (i.e. the upholding of both general human rights and refugee-specific rights).23 However, since refugeehood is largely characterized by insecurity, I consider the sense of ontological security – which can be explained as “the notion of safety, routine and trust in a stable environment”24 – as being central to the understanding of refugees’ experiences of protection. Often, this sense of security, built on trust, is established in community settings.25 In summary, this article shows how refugees conceptualize “protection space” largely in terms of their self-protection and community-based protection arrangements.
2. “SPACE” IN REFUGEE STUDIES
Within the social sciences and humanities, a significant “spatial turn” has taken place. This has been manifested in a greater recognition of the significance of the spatial in understanding social phenomena and also in the increasing use of spatial concepts. A review of the literature on refugees and space suggests that there are at least two on-going debates. Understanding these wider debates is helpful in my attempt to rethink the connection between “protection” and “space” in an urban context. First, there is an extensive scholarly debate on the nature of refugeehood in relation to territorially defined notions of nation state and identity, and space is often referred to in this debate. Second, there is growing body of literature on the spatiality of humanitarian intervention, or refugee protection in particular, often interrogating the notion of “humanitarian space”. Whereas the latter debate treats “space” mostly as a metaphor and thus remains under-theorized, the former debate, in particular, provides elements to consider when the notion of “protection space” is deconstructed. To sum up, I suggest that there are a number of useful issues to be borrowed from these two debates in terms of investigating the link between protection and urban space as is attempted in this article.
Regarding the first debate on the nature of refugeehood in relation to territorially defined notions of nation state and identity, according to previous studies, space, and its relation with people, should not be considered solely as a territorial or physical entity; it is also deeply metaphysical26 and imagined.27 Space is, moreover, created from the multiplicity of social relations.28 When discussing the nature of space, the “sense of place” is yet another critical component to be incorporated into the analysis. The sense of place is important not only in relation to an individual refugee’s sense of identity but also affects refugees’ collective identity29 in the camp or a city of exile. Second, given this diverse and fluid character of space, the analysis should be focused on the activities that take place in these various spaces.30 Particular activities taking place in space are place-making projects.31 It is understood that “displacement is not just about the loss of place, but also about the struggle to make a place”.32 Since various place-making projects take place in a given displacement situation simultaneously, it is, again, important not to focus solely on the physical understanding of space but also on the social nature of space.33 Third, the “the ways in which power works through the organization and conceptualization of space and movement”34 needs to be considered, because space, like protection, is always political and characterized by struggles over power and representation. Therefore, it is essential to not only focus on those who are assumed to be powerful in producing spaces of protection (i.e. the protection institutions, including humanitarian agencies and authorities); the analysis should also address host communities’ and refugees’ agency and power35 in the creation and contestation of “protection spaces”.
The second debate on refugees and space is centred on the issue of the spatiality of refugee protection, or more broadly humanitarian intervention. As already mentioned, it can be argued that humanitarian language has experienced a “spatial turn” in terms of the rapid growth in the use of spatial concepts exemplified by the extensive use of concepts such as “humanitarian space” and “protection space”, which have increasingly replaced the more legal language of protection.36 The list of spatial concepts in humanitarian action is extensive, but what is characteristic of these notions is the understanding of “space” mostly as metaphorical or as physical. Therefore, it is important to examine these “spatial metaphors, which render space innocent and ignore the relations of power which always run through material (what might be considered as ‘real’) spaces”.37 Other researchers have demonstrated how conceptualizations of “humanitarian space” are frequently focused on the “operating environment” or “agency space”, thus highlighting the role of global humanitarian actors.38 Therefore, the role of various local actors, including refugees, is often missing from most discussions of humanitarian space.39 Yet it is important to put people and local actors, such as faith-based communities,40 at the centre of humanitarian action, in particular in urban settings.41
The analysis of “humanitarian space” could also benefit from rescaling because “the humanitarian space […] is also influenced by the geographical locations”.42 This is in line with the argument that there should be more research directed towards “understanding how the micro-level geography of protection space (camps, urban squatter slums etc.) interacts with the macro-level spatial geography of protection”.43 The idea of investigating refugees’ understandings of urban space from various spatial scales has been put forward, for example, by Nah44 who emphasizes the significance of the micro-scale analysis. Nah, in her spatial analysis of urban refugees in Malaysia, identified various spaces that refugees use and share in the city and she recognized that refugees talk about similar problems in similar locations. Thus, there are particular micro-scale urban spaces crucial to refugees’ everyday lives that should be examined in relation to policy-making. Recently, a meso-scale investigation focused on safe and unsafe areas of a city in regard to refugee protection was also advocated by organizations such as UNHCR45 and Women’s Refugee Commission.46 Nonetheless, there is still a lack of studies that integrate these micro- and meso-scale analyses into the examination of the entire city and its macro-scale location in the national and regional context. The aim of this article, therefore, is to show how these three geographical scales are vital in order to analyse refugees’ conceptualizations of “protection space”.
3. THEORIZING “SPACE” AND RECONSIDERING THE CONCEPT OF “PROTECTION SPACE”
In this article, I conceptualize space as physical, imagined, lived, and relational according to the current theories of space widely referred to in the field of human geography. The French social theorist and critical urban thinker Lefebvre47 has provided in his humanist-inspired Marxist analysis another useful way to understand space in a particularly dialectical manner through his thesis on space as a social product. Lefebvre has worked extensively on analysing urban everyday life and the role of space in the survival of capitalism and his most extensive writings on space can be found in his book The Production of Space from 1974 (translated into English in 1991). He argues that it is useful to comprehend space as physical or perceived (“spatial practice”; actions and everyday routines that produce space) and as imagined or conceived (“representations of space”; the abstract space of professionals and technocrats, etc.). However, according to him there is yet another distinct way in which we encounter space in our everyday lives: “representational space”. This refers to the social space or lived space that is produced in everyday social interactions, i.e. the “lived space of sensations, the imaginations, emotions, and meanings incorporated into our everyday lives and practices”.48 With this term, Lefebvre, as explained by McDowell and Sharp, refers to “those everyday representations of spaces linked to the clandestine or underground side of social life as articulated in cultural products […] often produced as resistance to representations of abstract space”.49 Representational spaces are the spaces of “human action and conflict”50 and the “sites of resistance and counter-discourses that have either escaped the purview of bureaucratic power or manifest a refusal to acknowledge its authority”.51 In summary, lived space “represents a person’s actual experience of space in everyday life”.52
Closely related to the social understandings of space, a relational conceptualization of space has been proposed by a number of scholars.53 For Harvey, the relational view of space:
Lefebvre’s55 theorization of space also holds a relational conception of space as a starting point for “the mental, physical and social dimensions of space are understood as internally related within an open totality”.56 Relational space has been the basis of much recent work in critical human geography, and post-structuralist scholars, such as Massey57 and Rose,58 have, for instance, argued that new theorizations of space emphasize the heterogeneity of local contexts and the place-relatedness of our knowledge about the world. This relational understanding of space perceives it as the product of interrelations, heterogeneous and plural. Relational space is “the space of social relations, metaphorical spaces, spaces of […] emotion, and dream and aspiration”.59 Massey has furthermore suggested:
holds there is no such thing as space outside the processes that define it. Processes do not occur in space but define their own spatial frame. […] The very formulation implies that it is impossible to disentangle space from time. We must therefore focus on the relationality of space/time rather than of space in isolation.54
The physical, imagined, lived, and relational elements of space are understood as intertwined. Thus, the relational aspect of space not only refers to relations between people but also to people’s relations to the physical, built, urban environment (e.g. the cityscape) and the imaginations and everyday lived experiences of it. A few refugee scholars have adopted Massey’s theorization on space in their various analyses.61 A particular reading of relational space is presented by Ramadan in his theorisation of spatializing the refugee camp.62 Some aspects of this theorization can, I suggest, also be applied when refugee protection in urban space is investigated. According to Ramadan “spatialising the camp, understanding how it is constituted and functions spatially, is a way of grounding the geopolitics of the everyday”.63 Ramadan proposes that:
First, that we recognise space as the product of interrelations; as constituted through interactions, from the immensity of the global to the intimately tiny. […] Second, that we understand space as the sphere of the possibility of the existence of multiplicity in the sense of contemporaneous plurality […] Third, that we recognise space as always under construction. […] [I]t is always in the process of being made.60
Accordingly, the camp – like urban space – is diverse, dynamic, and always in a process of changing. Refugees’ relations with the wider regional geopolitics and relations with the protection institutions also shape the relational space of the camp or city.
the camp is […] who and what is in the camp, how they interrelate and interact. Camp space is produced out of the relations between and the practices of people (as individuals, families, institutions and organisations), and those subjectivities […] in turn are produced by these interrelations and the space that they are simultaneously constructing.64
To conclude, different conceptualizations of space are taken into account when the production and contestation of “protection spaces” in Kampala are explored in this article. This is because if one dimension of space is given analytical priority, this will, according to Lefebvre, lead toward an intellectual dead-end.65 Thus, as Harvey proposes, I argue that thinking space through different conceptualizations, created through the practices of urban refugee protection, is important as we should keep the “concepts in dialectical tension with each other and to think the interplay among them”.66
In the following section, I briefly discuss the data collection and analysis methods. After this, I describe the context of Congolese refugees living in the capital city of Kampala. My main analysis focuses on refugees’ conceptualization of “protection space” with regard to three geographical scales: the micro-scale (i.e. home and shelter), the meso-scale (i.e. neighbourhoods and relations with Ugandan neighbours), and the macro-scale (i.e. the entire city and its location within the region). The different elements of space (as physical, imagined, social, and relational) are examined within this analytical scalar framework. Lastly, I present some concluding thoughts on the revisited concept of “protection space”.
4. DATA AND METHODS
The data used in this article was collected during 8 months of fieldwork in 2010–2011 from individual Congolese refugees, refugee communities, and officers from the protection institutions. The data-collection methods include a combination of semi-structured interviews (N = 74 individual refugees, 18 refugee community leaders, 16 Ugandan authorities, and 22 officers of protection institutions), observation in various refugee communities and protection institutions, and focus group discussions (FGDs) mostly with refugee community members (N = 13).
I applied an information-oriented selection of samples, which is a form of purposeful sampling. This sampling method fits with the case-study approach as it aims to “maximize the utility of information from small samples and single cases”.67 The 74 Congolese refugees individually interviewed for this study had left the Democratic Republic of Congo (DRC) between 2000 and 2011. The majority of the informants (95 per cent) were from the eastern parts of the country, namely the Kivu provinces (Figure 1). Most of them had an urban background and were relatively well educated. The majority of the informants were married (45 per cent) or single (43 per cent). Some were also widowed or unaccompanied minors (5.5 per cent each). Sixty-eight per cent of the individuals interviewed were men and 32 per cent women.
Overall, the refugees I interviewed for this study were between 15 and 64 years old, the average age being 32 years. Sixty-two per cent of the informants had refugee status, 28 per cent were asylum-seekers, and ten per cent were not registered with the authorities. Seventy-seven per cent of the refugees had not been in the rural refugee settlements over the course of their stay in Uganda. Most of the refugees who identified themselves as a member of a particular ethnic group were Bashi (23 per cent), Banyamulenge (17 per cent), Barega (10 per cent), Nande (10 per cent), or Bembe (10 per cent). The overwhelming majority of my participants identified themselves as Christians (97 per cent) but they represented different denominations − mostly Pentecostal or denominational, but also Catholic.
Most of the individual and collective interviews with Congolese refugees were conducted in Kiswahili (and some in French) and translated into English by an interpreter. Many were also conducted in English without an interpreter. More than 90 per cent of the interviews were audio recorded with the informant’s permission and subsequently transcribed. The interviews lasted between 40 and 90 minutes. I interviewed some of the participants several times. After the initial coding of the data in NVivo 9 programme, I relied on aspects of discourse analysis to analyse my data. Discourse analysis is seen as a particularly appropriate data analysis method in studies on marginalized or excluded people.68
5. CONGOLESE REFUGEES LIVING IN THE CAPITAL CITY OF KAMPALA
Kampala means a “hill of Impalas” derived from a Luganda word impala, a type of antelope. Kampala is both the administrative and commercial capital city of Uganda. The city was initially set on seven planned hills, but later on it expanded to 24 hills that have developed largely without planning.69 The urban population in Uganda has grown rapidly from less than 0.8 million in 1980 to five million in 2012. This means that the urban population has increased more than six times during the given time period. Whereas the level of urbanization was still low, the rate of urban growth was unusually high.70 In the 2002 census “urban area” was defined as gazetted cities, municipalities, and town councils as per the Local Government Act 2000. There were by definition 75 urban areas but only one city in Uganda – Kampala. Kampala accounted for 72 per cent of the urban population of the region and 40 per cent of Uganda’s urban population in 2002.
As a consequence of the decentralization policy implemented widely in Uganda, Kampala City is administrated through “a five-tier hieracrchy of elected leadership”.71 The City of Kampala, headed by the Lord Mayor (LC5), is administratively divided into five divisions: Central, Makindye, Kawempe, Nakawa, and Rubaga (Figure 2). These divisions are further divided into 99 parishes, sub-divided into 802 zones. The local councillors (LCs), who are elected by all registered residents of the particular areas, are in charge of their zones (LC1), parishes (LC2), and divisions (LC3). The LC1s, for instance, keep records on who lives in their zones and they write recommendation letters for the inhabitants. Also, when civil disagreements take place, the LC1s try to settle them. Criminal cases are, however, dealt with by the police. The LCs also supervise the local defence forces responsible for the overall security of their area. In addition the LC1s nominate eight people to make up his/her executive and form the Local Council. Among the council members are representatives of youth, women, the elderly, and the disabled.72
The population of Uganda has changed over time due to both international migration and internal population flows. Ugandans inhabiting Kampala were largely Baganda, the people of the Buganda Kingdom. There were, however, a number of other ethnic groups, such as Banyankole, Besoga, Bakiga, Iteso, Langi, and so forth.73 Some of the ethnic groups, such as Acholis and Karamojong, have largely moved to Kampala as internally displaced people due to the conflicts in the north and east of the country.
Thus, the city of Kampala has historically been diverse. The census from 1959 shows that Kampala had a varied population with significant numbers of people from neighbouring countries, Asia, and Europe.74 Subsequently, international migrants coming to the city over the decades have changed the overall ethnic composition of the city, although they have not had a significant impact on the size of the population.75 In 2002, there were 333,000 immigrants in Uganda and of these 100,000 were recent immigrants.76 In total, all immigrants constituted 1.4 per cent of the total Ugandan population of 24.4 million in 2002. More than 90 per cent of the recent immigrants were from neighbouring countries,77 and one-third of them were from the DRC.78
It can be stated that “migration in Africa has always been a rather complex phenomenon, with various categories of migrants engaged in different forms of mobility”.79 Economic reasons have caused most of the recent internal and international migration into Uganda. Employment, marriage, education, retirement, and natural and man-made disasters, including war and conflicts, have also caused migration flows. Refugees have, at times, formed a significant proportion of immigrants coming to Uganda. Thus, the causes of migration to Uganda have been various and interrelated. This pattern of mixed migration is typical of the Great Lakes region, as was shown in a recent study by Bakewell and Bonfiglio.80
Urban areas have, in particular, been centres for people migrating for better employment and education. The immigrants were also more urbanized in Uganda than the locals: in 2002, the proportion of the immigrant population staying in urban areas was 17 per cent, whereas only 12 per cent of nationals lived in urban areas.81 Out of the recent immigrants, 26 per cent were located in urban areas. However, in terms of research, “in Africa the integration of international migrants in metropolitan areas has hardly been explored beyond South Africa”, although “this immigration and settlement is changing the face of some [African] cities”.82
Refugees coming to Uganda have largely followed these general migration and population trends and, consequently, they are increasingly found in urban areas. In Africa, this urbanization trend of refugee movement has had a long history. In the mid-1980s, it was estimated that one-third of all of the refugees in Africa were found in cities and towns.83 At the end of 2013, according to UNHCR statistics, 62 per cent of refugees in sub-Saharan Africa lived in rural refugee camps or settlements.84 Thus, there are less urban refugees in Africa than, for instance, in Asia and the Middle East. The steady increase in the number of urban refugees in sub-Saharan Africa over the past decades was partly due to the changing nature of conflicts in the area, as explained by Lammers:
It is not, however, only refugees originating from urban areas who choose to stay in the cities and towns of their country of exile. Refugees fleeing rural areas increasingly contribute to the urbanization of refugee flows. Thus, it is clear that refugee movements in Africa, and beyond, are closely linked to the wider migration trends that have an impact on urbanization.
In the course of the 1970s, confrontations between African states as well as internal resistance movements came to have more impact on urban areas. Rebel movements started to draw much of their support from an urbanised and largely youthful population. More and more people from urban areas were forced to flee in search of international protection elsewhere, and it is hardly surprising that many of them chose to settle in towns and cities in their countries of asylum.85
The number of refugees in Kampala expanded rapidly towards the end of the 2000s. In July 2008, the number of recognized refugees and asylum-seekers in Kampala was around 20,000; by 2011 the number had nearly doubled. In addition, an estimated 50,000 unrecognized refugees were living in the city.86 In July 2011, of the total 150,000 recognized refugees and asylum-seekers in Uganda, more than 26 per cent – that is 39,921 – lived in Kampala.87 The majority of refugees in Uganda were from the DRC, South Sudan, Somalia, Burundi, Rwanda, Ethiopia, and Eritrea.88
In July 2011, a total of 80,221 registered Congolese refugees and asylum-seekers lived in Uganda, with 18,075 (23 per cent) living in Kampala. Of the nearly 40,000 registered refugees and asylum-seekers in Kampala, 45 per cent were Congolese, with 48 per cent women and 52 per cent men.89 The number constantly increased due to ongoing violence in the DRC, particularly in the eastern parts (Figure 1). The DRC has experienced two decades of armed conflict and consequent forced migration both inside and outside its borders. The intertwined regional, national, and local armed conflicts since the early 1990s have caused the deaths of nearly 4 million people and displaced around the same number.90 There were different “vintages”91 of Congolese living in Ugandan refugee settlements and cities. Some of them have lived in Uganda since the 1960s,92 whereas a number of them fled the DRC during the First (1996–1997) and Second (1998–2003) Congo Wars.93 Eastern Congolese had also been fleeing to Uganda due to military operations since 2008 and in the post-election violence after 2011. On 29 October 2010, the governments of Uganda and the DRC signed the tripartite agreement with UNHCR to govern Congolese refugees’ voluntary repatriation.94 However, since security in eastern Congo is still highly unstable, repatriation there remains highly unlikely.
Uganda is a party to the 1951 Convention Relating to the Status of Refugees and its 1967 Protocol, and is also a signatory to the 1969 Organization of African Unity (OAU) Convention Governing the Specific Aspects of Refugee Problems in Africa. Uganda replaced its heavily criticized 1960 Control of Alien Refugees Act (CARA) with the 2006 Refugee Act, which is more in line with international laws governing refugee protection. CARA was “perhaps the first piece of legislation in Africa specifically relating to refugees”.95 In 1998, the Government of Uganda launched its Self-Reliance Strategy, which emphasized placing refugees in rural settlements, not fenced camps, where they could become self-sufficient through agriculture. This strategy has, however, largely been proven unsustainable.96
The 2006 Refugee Act has been praised for its progressive nature by the Government of Uganda and UNHCR, but there are, however, “some deficits […] which potentially erode the progressive and protection orientation of the Act and threaten to lower its compliance with international protection standards considerably”.97 These deficits mostly relate to freedom of movement and residence, freedom of association and expression, and the right to work.98 There were also delays in the actual implementation of the 2006 Refugee Act because even though it entered into force in 2008, the regulations necessary for it to become operative were only finalized in 2010.
Despite the overall positive changes in refugee law in Uganda, and the adoption of the amended 2009 UNHCR urban refugee policy99 by the UNHCR country office, both of which recognized cities as legitimate places for refugees to reside, “what exists today – and for the foreseeable future – is a policy that focuses assistance and protection on refugees living in settlements, and not those refugees who chose, for various reasons, to live outside such restrictive spaces”.100 Moreover, even though the 2006 Refugee Act was “lauded on the African continent as a progressive piece of legislation”, in practice it “retains the settlement policy and places undue restrictions on the right of refugees to free movement”.101 An explicit permission is required to leave the settlement in order to live in a city, and this “provision in effect curtails a refugee’s right to choose his/her place of abode and his/her freedom of movement”.102 Subsequently, refugees in Kampala were expected to be self-sufficient since there existed only minimal protection. Thus, it could be argued that at the time of fieldwork Uganda was still heavily reliant on its rural settlement policy, and the situation of urban refugees had not, according to many NGOs working on the issue, improved significantly despite the newly adopted legal frameworks and policies. Given this “institutional absence” in Kampala, refugees commonly suggested that their everyday life and sense of protection was more impacted by the city and their relations with other inhabitants. In the following sections the micro-, meso-, and macro-scale discourses of “protection space”, and the different elements of space (as physical, imagined, social, and relational), are investigated in detail for a more nuanced understanding of refugees’ sense of protection and insecurity in their city of exile.
6. MICRO-SCALE “SPACES OF PROTECTION” – HOME AND SHELTER
6.1. Rented houses
Refugees found physical shelter in various places in the city. In this section I examine two different places representing a “home” or a shelter in Kampala: rented houses and churches. In regard to each micro-space, my aim is to analyse refugees’ spatial experiences of living in these forms of shelter with a focus not only on the physicality of these spaces, but also the imagined, social, and relational aspects of these shelters.
Most of the refugees lived in relatively poor physical conditions. Many Ugandans, however, faced similar housing problems to the refugees because these two groups shared the same residential areas. The deprived living conditions were due to the overall rapid and uncontrolled urbanization that has taken place in Kampala over recent decades.103 Given these conditions, “African host-governments see the situation in their urban centres being exacerbated by the presence of refugees who are said to compete with nationals for scarce employment opportunities and social services”.104 This was also to some extent the situation in Kampala. Refugees, however, argued that on top of the shared urban challenges they also experienced specific problems relating to their particular status. In the next section the focus is on the spatial elements that refugees’ suggested particularly characterized their housing issues and consequent insecurity.
The most common form of physical shelter for the Congolese refugees was a rented house. A house is a vital form of shelter in urban areas and providing refugees access to secure housing during their stay in exile is important because at its best a proper shelter can enable the recovery of sustainable livelihood, provide protection, offer privacy and dignity, and support household and community coping strategies.105 Thus, the physical structures of a house can have an impact on refugees’ understanding of shelter also as an imagined, social, and relational construct.
The social and relational aspects of a house as a micro-scale “protection space” can be examined by looking at the social relations between the people who share this domestic space. People with whom refugees shared their house ranged from members of their nuclear family to extended family, friends, and acquaintances. Single men and women often shared their house with friends from their church or, in the case of the Banyamulenge Tutsis, people from the same ethnic community. The living conditions were often crowded. For instance, a young female asylum-seeker lived in Nsambya in a small, one-room house with her sister, her sister’s husband, and their nine children. The space was, according to her, “shameful”106 because they only had one room for all 12 of them. Thus, the relational space refugees identified was not only limited to the social relations among people but was extended to the relations between people and physical structures (i.e. walls and other material structures of the house).
Refugees’ social, domestic spaces were extended to their Ugandan landlords as demonstrated by the example of paying rent. At the end of the month, finding the money for the rent was difficult. Rent for a small one- or two-room house varied between 50,000 and 200,000 UGX per month (13–50 GBP) depending on the neighbourhood and facilities. Refugees categorically suggested that Ugandan landlords charged them more than Ugandans. This was done, refugees argued, because Ugandans perceived the Congolese as rich because of the natural resources in the DRC. Congolese refugees have experienced a similar kind of discrimination based on their assumed wealth in other African cities, such as Nairobi.107
In terms of the physical infrastructure, housing in Kampala was challenging and many of the refugees could only afford to rent a single-room house in the low-income neighbourhoods of the city. Most of the population (54 per cent) of the city, however, lived in one-roomed houses (muzigos).108 Many informants lived in crowded houses without electricity, sanitation, or running water. Sanitation facilities were commonly shared with other people living in the same neighbourhood. This meant that the washroom facilities were outside the house. This was seen as a security risk, especially by refugee women and children who were concerned about their physical security and health.109 Thus, the shared physical and social spaces both impacted the refugees’ overall sense of protection in the context of their immediate micro-scale neighbourhood.
Refugees’ living arrangements were often not straightforward and identifying their sense of home in the city of Kampala was challenging. Subsequently, their imaginations of “home” had to be examined in detail and it could not be assumed that their physical shelter and sense of belonging correlated. For example, a young Munyamulenge asylum-seeker boy lived with his family in Kasubi. Despite having his family in Kampala, he was unable to sleep at his parents’ house. The house they rented was just one room and in total the family had nine members. Thus, his father had arranged for the male siblings to sleep at the house of a member of their congregation. The boy reflected on his living arrangement accordingly (emphasis added):
This practice of not being able to sleep in a place perceived as a “home” was shared by others. Some of the refugee men, for instance, explained that they could not stay in their rented house with the rest of their family due to security reasons. By occasionally sleeping elsewhere, they aimed to protect their families from a possible attack. This spatial strategy indicated an innovative approach to securing the imagined safe space of a shelter by replacing the physical space where people slept. In another example, a 23-year-old male refugee who was in Kampala alone, did not sleep where he perceived his “home” to be. His living arrangements included sleeping in a church, but spending a significant amount of time at his pastor’s place in another parish. Consequently, according to him, the pastor’s house represented a home to him (emphasis added):
We sleep there and in the morning we go back home. […] The house is close by. It is like a ten-minute walk from my home. We normally go there like 9PM. We try to go there when it is not too dark. I always go with my other brothers. I leave that place at 6.30AM.110
Moreover, his explanation of the pastor’s house was characterized by a sense of belonging to a family that he did not have elsewhere in Kampala. Therefore, having access to this physical space of protection had affected his sense of home and family – both of which he lacked in the church space where he slept.
In Kireka, this house is for the pastor. This is the pastor’s house because every day I go there […] My home would be the pastor’s house; in the church I only sleep […] My pastor told me that instead of spending all day in the church, “you can visit me and my house. And you can stay inside the fence. After that you can go where you are sleeping”. […] Those people are my pastor’s children. This family loves me so much.111
A number of refugees had experienced evictions leading to a sense of insecurity about losing their shelter. In humanitarian situations, Boano has observed that “loss of property and housing is not only severe material damage, but it also carries a powerful symbolic erosion of security, social wellbeing and place attachment. […] This loss seriously erodes the very meaning of life and its continuity in space and time”.112
As seen from this reflection, refugees’ loss of shelter can significantly affect their sense of protection bound in time and space. This was also the case in Kampala. The reasons why people were evicted or otherwise had to move within the city were various, and some were related to their status as refugees. Analytically, many of these reasons reflected refugees’ understandings of space as physical, imagined, social, and relational. Some of the refugees had been forced to move to a new place because the people that they lived with were resettled. Others moved because of exceptionally poor living conditions and insecurity in particular neighbourhoods. This was often caused by difficult relations between refugees and their Ugandan neighbours. The inability to pay rent was also a continuous security risk. Moreover, refugees suggested that Ugandan landlords discriminated against them due to their status and due to their large families. Sometimes the refugees who were unlawfully evicted tried to turn to the protection institutions or LC1s for help. However, more often they were provided with a temporary shelter by a church. In the following section I discuss refugees’ experiences of staying in church buildings.
6.2. Church buildings
Churches provided an important sense of community and belonging for the majority of the Congolese refugees informing this study. In this section, I focus on the idea of the church as a physical shelter, but also aim to demonstrate the imagined, social, and relational elements of the church as a shelter. In previous studies it has been recognized how local faith communities, including churches, can have “key resources that equip them to provide basic services [in humanitarian situations, and] religious buildings may be used for storage, information hubs, shelter and protection”.113 The church also has a long tradition of using its buildings as a “sanctuary”.114 In recent years, this sanctuary movement has experienced resurgence, in particular in the Global North.
In general, shelter was provided for refugees by Congolese and Ugandan denominational and Pentecostal churches.115 In 2011, there were at least 50 Congolese churches in the city.116 The majority of these churches were physically located in Katwe and Nsambya parishes where most of the Kivutian Congolese lived. An estimated 80 per cent of all of the refugees interviewed had stayed in a church at some point during their stay in Kampala, mostly upon their arrival. This was due to the fact that only a few of my refugee informants had a specific plan to come to Uganda or more specifically to Kampala, because they did know anyone in the city prior to their arrival. Thus, the lack of social space in the form of social networks with compatriots, for instance, forced refugees to rely on the physical spaces of protection, such as the local church buildings. Yet, given the importance of the church, upon their arrival in the city refugees who were looking for their lost relatives focused their search on these refugee churches. Subsequently, the physical church building often also acted as a space for recreating social relationships.
In this article, I use one of the churches located in Rubaga parish as an example to discuss the church as a space of protection. The particular church was at the time of the fieldwork attended and led by both Ugandans and Congolese. Initially in 1994 it was set up by a Congolese bishop who at the time lived in Uganda. Throughout the history of the church refugees have attended and lived in the church.
The church provided the refugees with a combination of support because the Ugandan pastor had realized “that the church cannot continue by only giving spiritual guidance without meeting their [the refugees’] physical needs”.117 Therefore, the church provided a physical place where refugees could sleep: it functioned both as a secular space of shelter and as a sacred space of spiritual services and worship. The sacred space of the church was also typically conceived as a relational space between people and God by the refugees and the pastors. Drawing boundaries between secular and sacred spaces is often not straightforward,118 and reconciling these two functions was argued both by the refugees and by the pastors to have potentially jeopardized the more sanctified element of the church building. The pastors were, however, willing to do this because they firmly believed that providing shelter for refugees was what the Bible taught them.
The church rented a small wooden building along the busy Rubaga road. In order to pay the rent they relied on their member’s donations and ties, and “even the refugees give when they can, just like the Bible says: ‘give what you have’”.119 Overall, the Ugandan pastor believed that the physical shelter, and the associated sense of belonging they were able to provide, reinforced refugees’ sense of protection in Kampala:
This example also shows how the Ugandan pastor conceptualized “refugee protection” in spiritual terms, like many of the refugees taking part in this study. He also saw that his church offered refugees a sanctuary based on what the scriptures taught regarding refugee protection.
When the refugees come to the church they feel more protected. When they are in the church, the church gives them remedy of the word of love. But in the refugee camp they do not really mind about that. They just give you protection, and you are under the arm of the Government. But here they are freer. I think that if the church could give them physical protection, I think that the refugees would feel more protected than in other areas.120
During the summer of 2011, seven Congolese refugees stayed in this small church. I will discuss the unique and shared experiences of living in the church through the eyes of Robert and his family.121 One day I met Robert122 by accident on the street near the Office of the Prime Minister. He initiated our conversation by asking me: “Hey sister, are you born again?” After speaking with him, he invited me to visit his church where he lived and worshipped. Robert was in his early 40s and he was a Bangubangu from South Kivu. He had a wife and a 2-year-old daughter. Robert had fled the DRC with his wife because of a family conflict that was caused by his father’s polygamy. They had stayed in the church for 2 years, since their arrival in Kampala in 2009. Initially someone at the Old Kampala police station had advised him to find this particular church, which was known in the city as one of the oldest “refugee churches”. Robert and his family’s everyday rhythm was shaped by the fact that they occupied a church building: “We sleep in the church and early in the morning we wake up because we pray. We sleep on the mattress after the people have left”.
Robert was not only sleeping in the church but also considered himself as an evangelist of the church. He had worked as a preacher in the DRC and wanted to continue spreading the gospel in Uganda. His wife also sang in the worship choir. Often refugees who stayed at churches contributed to the activities with their music, teaching, and other gifts. This reinforced their perception of the church as a social and relational space where they interacted not only with each other but also with God.
Even though the church provided Robert and his family with a place to sleep, he did not feel safe in the church building or in the city of Kampala at large. This was due to his fear of persecution spilling over from the DRC to Uganda. For him this persecution could have taken the form of a spiritual battle:
Subsequently, as it is important to pay attention to refugees’ conceptualization of “protection” in terms of the spiritual realm, it is necessary to reflect on refugees’ understanding of insecurity in supernatural and spiritual matters. Schnoebelenhas, for instance, argued that “workers at international organisations and non-governmental organisations must be aware of the tenacity of witchcraft beliefs, the very real threat they can create for individuals, and be willing to provide protection”.123 Thus, the physical space of the church may have been for some refugees, such as Robert, also a space of imagination and fear.
I do not feel peace even here in Kampala, because there are many problems here in Kampala because of insecurity. […] In Africa we have the witchcraft spirit. They can do something bad to you to [make you] disappear.
The case study of this particular church shows how refugees remained compelled to sleep in churches for various reasons. Even though refugees commonly found a sense of physical protection based on the provision of shelter and belonging to a community of believers, they also raised concerns about the lack of protection in regard to the spiritual battles and the associated sense of metaphysical insecurity. Thus, it was important to recognize the church buildings as “sacred spaces” and as “geographical locations that instil and perpetuate a variety of feelings that affect human experiences and activities [such as protection and trust]”.124 As seen in the case of Kampala, given the limited space, lack of privacy, and, above all, the use of a sacred space as a physical shelter, refugees often perceived that they lived in a church as a last resort before either becoming homeless or shifting to the rural refugee settlements. To sum up, as seen from this analysis, the physical, imagined, social, and relational aspects of a church building as a shelter were intertwined and consequently needed to be analysed jointly.
7. MESO-SCALE “SPACES OF PROTECTION” − NEIGHBOURHOODS AND RELATIONS WITH UGANDANS
7.1. Selecting a neighbourhood
Most of the refugee informants (N = 74) lived in the Makindye and Rubaga divisions of the city (Figure 2). Thirty per cent of them lived in the Katwe parishes and 28 per cent in the Nsambya parishes of the south-central part of Kampala (Figure 3). These two areas were dominated by refugees from eastern Congo. Eastern Congolese also lived in Rubaga parish (8 per cent) and Makindye parish (8 per cent). Banyamulenge Tutsis were typically clustered in the Nakulabye and Kasubi parishes. People from western Congo had taken up residence predominantly in Najjanankumb parish on the outskirts of the city. In addition, a few refugees lived in Old Kampala, Kireka, Mengo, Namugona, Kawempe, Wandegeya, Kibuli, and Ndeeba (Figure 2). As the presence of other Congolese was perceived either as a positive or a negative aspect of a neighbourhood depending on refugees’ individual protection concerns, it can be analytically suggested that refugees’ perception of the social or relational space of a particular neighbourhood had an impact on the selection of their physical living space at the meso-scale.
In other studies it has also been concluded that in urban displacement settings, neighbourhoods are chosen on the basis of a range of criteria, including security, proximity to relatives and friends, and affordability, among other reasons.125 Finding a place to live in Kampala often happened through churches or other social connections. Most informants, however, felt that they had not freely chosen their neighbourhood but rather just happened to find themselves there. This was articulated, for instance, by a 31 year old male refugee:
For many refugees their home and the immediate neighbourhood were important meso-scale spaces of protection. For some these immediate areas provided a feeling of safety whereas others perceived the neighbourhoods as insecure spaces. If possible, refugees wished to move to areas that they perceived as generally secure. Yet the experience of changing neighbourhoods because of insecurity was common. For instance, a 51 year old Mushi refugee man127 who had lived in Kampala since 2010 had to move on several occasions due to neighbourhood-related insecurity. His family of 10 people had adopted a strategy of shifting places for security reasons because he was concerned about intimidation by the people who had made them flee Bukavu in the DRC. In less than 2 years in Kampala, the family had moved from Namugono to Kasubi and then to Katwe where they lived at the time of the interview. At times the refugee man slept in his church to protect his family’s safety by being separated from them. Later, the church rented a small house for the family in Kasubi. Their home was, however, attacked and consequently the man started to sleep in the church again. In Kasubi one of his daughters was poisoned and died as a result. After this, the family decided to move from Kasubi to Katwe where they had found relative safety, at least temporarily. This spatial protection strategy was characterized by the aim of finding stable social and imagined space of protection through altering the physical space of the neighbourhood.
Life just takes you and you find yourself. When, for example, I was staying in Old Kampala I found someone who said: “Oh, I need someone to work for me, I have a certain compound to arrange, so I would need someone to be there for two months.” Then I go there by that condition.126
If possible, some of the refugees wished to live close to the protection institutions and the police. Being close to the potential places of employment was also a criterion for those who could afford to choose where to stay. Living in the outskirts of the city was cheaper and more spacious but having to use public transport became a financial burden. Refugees who lived in the outskirts of the city, such as in Najjanankumb, were keen to emphasize the rural characteristic of the neighbourhood: the natural environment, animals, and the peacefulness of living away from the busy city centre.
Others who lived in more central areas of the city appreciated the services and shops found in their neighbourhoods, but criticized the noisiness and overcrowding of their living areas. Refugees who lived in more sparsely populated higher-income areas, such as Wandegeya, Rubaga, or Mengo had a completely different experience of living in Kampala. This was due not only to the physical side of their neighbourhood, but to the social relations in different neighbourhoods of the city that also had an impact on refugees’ sense of protection and insecurity. In the following section, I will discuss the relations that refugees had with their immediate Ugandan neighbours.
7.2. Relations with Ugandan neighbours
Neighbourhoods “comprise the smallest unit of urban social territory and political organisation”,128 and analysing refugees’ shared social territory with the Ugandan residents in their neighbourhoods is essential. Thus, the analysis of neighbourhoods as social and relational space includes examining the particularized social trust that lies between people who know and interact with each other.129 Yet, the forms of social trust reflected in refugees’ discourses often feature physical and imagined elements of space.
Refugees typically perceived that there were two main reasons for feeling insecure in Kampala. On the one hand, some of them feared the mismanagement and ignorance of the protection institutions, and on the other hand, they felt unsafe because of “neighbours and our neighbourhood”.130 In general, refugees had a rather neutral relationship with their immediate Ugandan neighbours. At times, they described how their neighbours helped them in various ways. More commonly, however, they described their relations in a negative manner. Refugees understood that they had “an obligation to be OK with the Ugandan”.131 Yet some of them argued that they only interacted with their Congolese, not Ugandan, neighbours. Consequently, some of them suggested that they were “not really living with those [Ugandan] neighbours”.132
Refugees frequently suggested that their Ugandan neighbours blamed them and their children for any difficulties that took place in the neighbourhood. Consequently, some were afraid to leave their children outside the house unattended because their neighbours might harass or abuse them. Often refugees, especially single mothers, locked their small children inside the house when they were gone. Consequently, the physical space of the home was used by refugees to manage the social or relational spaces between refugees and Ugandans. Some informants had, moreover, experienced theft by neighbours and others had negative experiences of working for their Ugandan neighbours, as expressed by a refugee woman during one of the FGDs:
Most commonly refugees suggested that their Ugandan neighbours harassed them vocally. This verbal abuse commonly reflected the Ugandans' unclear understanding of why the Congolese refugees were in Uganda. According to a young refugee man, who had lived in Kampala since 2010, his neighbours would ask: “They are Congolese, so why are they here? What is the reason for that?”134 According to him, and many others, their neighbours vocally encouraged them to go back to the DRC. These forms of verbal abuse and discrimination, which took place in the space of the social and relational, largely affected refugees’ imaginations of the neighbourhoods they lived in.
We are not loved by our neighbours. Whenever you work for them, at the end of the day they will not even pay you for the work that you have done. They just keep telling me that: “You are Congolese. You are here as our workers. We cannot afford to pay you anything.” So, we are confused. They keep us like slaves.133
At times, verbal harassment turned into physical violence. An example of such a situation occurred in a particular compound in Nsambya parish where one Ugandan man terrorized several Congolese refugees. The three refugee families living in immediate physical proximity to this man had reported the harassment and abuse to the LC1 but nothing had changed. The Congolese had turned to each other for their communal protection, as explained by a 39-year-old female asylum-seeker who had lived in the city since 2003:
Another young single mother, who was new to Kampala, had also experienced the threatening behaviour of this Ugandan neighbour:
We have one neighbour here who is disturbing us a lot and he is saying that he is going to kill us, all of us Congolese in this place. […] He sometimes spits in our faces while talking to the children and you can see him moving around with a big stick. He is just talking, but if you try to reply, then he can knock you with that. Someone who says he will kill you, you have to fear. For me, I can say maybe the Congolese can assist me but not Ugandans; they would be on the side of the other fellow.135
This young woman had considered moving to the refugee settlement with her baby because of her fear. She assumed that moving within the city would not provide a solution because according to her “many Ugandans in Kampala are not happy to have refugees living among them”. This example illustrates how refugees living in the same compound may have created protective responses to forms of shared insecurity in their immediate neighbourhoods. It also demonstrates the fact that even if Ugandans and Congolese lived in close physical proximity to each other, the social space between them may have been extremely wide and produced by negative interactions.
I am not very safe in Kampala ever since there are neighbours who disturb us. The man will show us who he is. He is our Ugandan neighbour. He is a drinker and a smoker. He has said that we have to fear. He can decide to do whatever. Personally I have not yet been attacked. He sometimes tries but he has not touched me yet.136
8. MACRO-SCALE “SPACES OF PROTECTION” − SITUATING KAMPALA WITHIN NATIONAL AND REGIONAL POLITICS
The national and regional economic and political situation had an effect on refugees’ general sense of “protection space” in Kampala. This is because in African urban politics, it may be “difficult to disaggregate local processes from national processes, and there can be complex interconnections between the two”.137 Thus, it is important not only to focus on the micro-scale analysis of the experiences of protection and insecurity articulated by the refugees, but also to investigate how the city-level sense of insecurity was related to the wider political and economic contexts.
As already mentioned, the national and regional, and even global, economic decline had its effect on refugees’ sense of insecurity in Kampala. It also affected the social spaces between the hosts and the refugees. Whereas the refugees felt that they were unable to sustain their basic needs, the Ugandans often argued against having more refugees living in the city precisely because of the economic disadvantages that they were believed to bring. A refugee woman, for instance, expressed her concerns during one of the FGDs about how Ugandans had told her that refugees were to blame for the high inflation:
Ugandans reacted strongly to the increasing price of living in Kampala. During the time of the research, there were regular demonstrations that often led to riots. Most of the demonstrations were initiated by the Ugandan opposition and they started as “walk-to-work” protests led by the opposition leader Kizza Besigye. Soon these peaceful demonstrations against the increased cost of urban living, however, became violent in parts of the city. Before the demonstrations began, there was an important political event, the general elections, which also had an impact on refugees’ general sense of protection in the city.
For the last two days with this inflation, the Ugandan currency weakened against the [US] dollar. They have accused us as being responsible for the inflation and wherever you go, you can be slapped: “refugees are the one who have created the inflation.” Whenever you are using public transport, everyone else will pay less but from you they ask for more.138
Uganda had presidential and parliamentary elections in February 2011 and local elections in March 2011. Refugees, in general, feared election time because they received threats from some of the Ugandans inhabiting Kampala. Refugees were more concerned about the presidential elections than the local elections because the overall national politics in Uganda affected them more than local politics.139 The threats that the refugees received during the election period were mostly verbal and were centred on the idea that if Museveni lost the elections, the Ugandan opposition would either evict the refugees from the country or kill them. Refugees were generally seen by the Ugandans as supporting President Museveni and his National Resistance Movement party, which had been in power since 1986. Appreciation for the Government of Uganda and President Museveni was expressed by some of the refugees:
The elections caused widespread concern among the Congolese refugees and had a particularly negative impact on social spaces constructed between refugees and their Ugandan hosts. Refugees were, in particular, accused by Ugandans of coming to the country with the purpose of voting. By law, refugees were prohibited from voting or taking part in Ugandan politics in any way. The misperception of refugees’ political involvement, according to one of the refugee community leaders, was related to the cessation clause on Rwandan refugees signed by Uganda, Rwanda, and UNHCR in 2010. Given the precarious situation of the Rwandan refugees, some of them had begun to pretend to be Congolese in order to gain refugee status. The political alliance between Rwanda and Uganda was also recognized by the Banyamulenge Tutsis who were particularly cautious about their position in regard to the elections. According to them the fact that Ugandans saw them as Banyarwanda, who had the right to vote because they had Ugandan citizenship, made them feel unsafe, as explained by a Munyamulenge man during a FGD:
I say, “thank you”. Here the Government of Uganda is not bad because the government is trying to put us in a good condition, but the problem we have here is some of the people we have here. They are not good because all the time there is a problem which happens; they say “ha, we will show you after the elections”. So, this has put us in a big fear. But we are here in Kampala because we do not have somewhere else to go. But this is not a place where we want to remain.140
Before the presidential elections in February 2011, refugees took some precautions in case violence broke out in the city. The various physical forms of precaution were used by refugees to sustain the imagined and social “protection spaces” of the city. However, since the Ugandans were also uncertain about their safety during this time, how refugees physically prepared for the elections was not that different from their hosts. Some refugees spent extended time, even a week,142 indoors during the election period. They also faced challenges in relation to storing emergency food and water as explained by a female refugee: “I fear because people say that we must buy the stock of food because something bad will happen, and I have nothing at home because of money. God himself knows”.143 Others avoided crowded places in the city to evade possible riots, and some intentionally reduced their movement in the city. Thus, the imagined space of fear had a significant impact on refugees’ use of the physicality of the city space.
The main challenge is the confusion when we are called Banyarwandan. In a political way Uganda can have a compatibility with Rwanda, but we are the ones who undergo the problems. The citizens of Uganda call us the Banyarwandan. Yet we are not them. Uganda sees that we are not refugees; that we are here only to support the one who is in power in Uganda. There is a political misunderstanding between Museveni and the opposition, and we are the ones who suffer the problems.141
Some refugees wanted to leave Kampala during the elections, but according to them there was “no way to go, because we are struggling for daily bread, so where to get money for transport. We are left in God’s hand to whatever could happen”.144 To reinforce their protection, refugees were advised by the protection institutions not to move around on voting days because then they might be accused of voting. One of the refugee leaders also advised his community members on how to behave during the election period:
The Refugee Law Project (RLP)146 also instructed the refugees on how to act during the elections. During one of the many community policing events147 that the RLP organized together with the Ugandan police forces in the different neighbourhoods of Kampala, it was made very clear that refugees were in no privileged position in regard to being protected during the elections. This was because “security is for everyone, and no special security will be given to refugees. Police protects every citizen of this country just including refugees”.148
We inform Congolese regarding this election that people should not be involved in it; even putting on this [campaign] T-shirt or doing public campaigning. People are not supposed to participate. If you can buy something, have at least some stock in your house. That period you can be safe with your family.145
Besides the elections, refugees were concerned about the riots and demonstrations that took place in Kampala. In 2011, especially after Museveni’s victory, the Ugandan opposition led by Kizza Besigye encouraged the so-called “walk-to-work” demonstrations and “ride-drive-and-hoot” campaigns, which opposed the increasing prices of food and fuel and more generally government politics. In the spring and summer of 2011 these demonstrations occasionally turned into physically violent riots that the Government of Uganda banned and aimed to prevent with more violence.149 Goodfellow has suggested that these riots were inspired not only by the broader political events of the Arab Spring but also the increasing trend of “mobilisation of urban social groups into protests and riots [which] has institutionalized what might be termed ‘noise’ as the most meaningful form of political participation”150 in Kampala.151
In refugees’ accounts these “walk-to-work” riots were associated with some of the previous unrest, such as the Kabaka riots in 2009,152 the terrorist attack that took place in July 2010 by Al-Shabaab,153 or the Kasubi tombs unrest in 2010.154 Having experienced these previous disorders, some of the refugees reflected on these past events:
During the 2011 demonstrations associated with the aftermath of the elections and increasing economic turmoil, some of the refugees experienced not only verbal but physical violence. A Congolese refugee man, for instance, reported that his son had been injured during the riots. Prior to this incident he had already lost one of his children in the violence in the DRC and another one had been abducted in Kampala. The child who barely survived the riots was his only remaining living child:
So many concerns. Even during the riot of King Kabaka [in 2009] a big number of Congolese were threatened with death, and we even lost two people who were killed. So, now that we are going into the period of election, we have no peace of mind. We are thinking about what will come out.155
The demonstrations also affected some of the refugees’ income as they were unable to sell their products on the streets of Kampala or they were robbed during the riots and lost what they were selling. Generally, refugees felt that they were accused of being part of the riots even though they had neither initiated nor taken part in them.157
One morning our pastor [who they stayed with] sent the children to go and buy iron sheets. When they reached that place, it was the day of the demonstration. Those demonstrations took tyres and they burned them on the streets. They forced the children to pick those tyres and burn them. And put up a road block. The children say “no, me, I am just a refugee, I cannot involve in this.” When he said that he is a Congolese refugee, the demonstrators pushed him to the fire and started to beat him. So, he was burned on one side of his body.156
Additionally, the national and regional political situation, beyond the elections and the demonstrations, had an impact on refugees’ sense of physical, imagined, social, and relational spaces of protection. At the time of my fieldwork, many of the Congolese refugees had become increasingly afraid that the governments of the DRC and Uganda would force all of the refugees back to the DRC. This imagined space of fear was associated with the cessation clause that was to be implemented in 2013 in regard to Rwandan refugees, and to the tripartite agreement signed in 2010 between the DRC, Uganda, and UNHCR regarding the procedures for voluntary repatriation of Congolese refugees. Uganda’s long history of involvement in the conflicts in the DRC also made some of the refugees question how welcome they were in Kampala. This was articulated, for example, by a female refugee during a FGD:
Given these political perceptions and the prevalent institutional mistrust, the Congolese refugees at large did not trust the Government of Uganda or UNHCR to uphold the obligations of the voluntary repatriation agreement. Subsequently, refugees were concerned about being physically forced to repatriate before it was truly safe for them to return to the DRC, in particular to its eastern territories.
We are still living in insecurity because this is one of the countries that attacked our country, Congo, and the Ugandan military during the war raped our mothers, sisters and wives and killed our parents, friends and relatives. We came here in exile, and reaching Uganda we find that it is the country which provoked those kinds of insecurities in our lives and in our country. So, we can be here but without security. We came here seeking asylum as refugees while knowing that we are in a country which made us refugees today.158
As seen from the above examples, refugees’ sense of urban security was not only influenced by micro- and meso-scale factors but also by the wider national and regional political and economic spaces. In particular, the economic decline, elections, riots, and the signing of the voluntary repatriation agreement affected their physical, imagined, and social sense of security in Kampala. Thus, understanding refugees’ macro-scale insecurity discourses was important for understanding what the notion of “protection space” meant in their lives.
In order to deconstruct and rethink the concept of “protection space” in the context of Congolese refugees in Kampala, I asked how the physical, imagined, lived, and relational elements of space condition refugees’ experiences of living in their city of exile. The analysis of refugees’ understandings of the spatiality of their protection in the city of Kampala clearly showed that all of these different elements of space were intertwined.159 My interrogation of these elements of space was scalarly framed around the micro-, meso-, and macro-scales.
The physical spaces of protection manifested themselves at the micro-scale in terms of shelter and different buildings and also at the larger geographical level as the physical structures of the city and Kampala’s distinct position within the East Africa region. There was, however, overlap and interplay between these different scales, and overall, they were used in this study as a device of analysis rather than refugee-driven categories. The physicality of the city provided refugees with protection but also made them feel more insecure than safe at times. Imagined space was a key element in discussing refugees’ sense of protection. Refugees’ urban imaginations were characterized by a sense of safety and insecurity. This was caused by their fear of persecution spilling over from the DRC, their widespread mistrust of the other inhabitants of the city, and spiritual understanding of protection.
The social, relational spaces of protection were characterized by refugees’ distinct representational spaces and everyday encounters with the other city inhabitants at various spatial scales from the home to the entire city of exile. Refugees’ heterogeneous experiences of the lived and everyday urban spaces were crucial for their understandings of urban protection and insecurity. In this article, the analysis focused on the relational spaces between refugees and other inhabitants of the city. This analysis was important given the fact that most of the studies in Kampala, such as Sandvik’s,160 have focused on refugees’ relations with the protection institutions, namely UNHCR. Yet, I argue that the everyday mundane encounters and relations with the city's inhabitants are those that perhaps matter the most.
In sum, in order to be able to examine the concept of “protection space” as understood by refugees, their discourses of different elements of space had to be interrogated in detail. Thus, I conclude that adopting the multiple conceptualizations of space was necessary for the understanding of protection in urban space. Overall, these four elements of space featured refugees’ discourses of protection in a generally equal manner, and therefore one or a few of the elements cannot be said to have dominated. When paying sufficient attention to refugees’ discourses of space, we can begin to grasp the complexities and multiplicities of the widely used concept of “protection space” – a concept that is not solely institutional but very much embedded in the mundane practices and experiences of the everyday life in the city of exile, and the idea of self-protection and community-based protection arrangements.161
All in all, this article calls for a reconceptualization of the notion of “protection space” with rescaling at the centre of analysis. Unlike as indicated by the policies on urban refugees, such as those of UNHCR,162 the notion of “protection space” is both empirically and theoretically problematic. This is because investigating refugees’ experience of urban space provides us with a more nuanced understanding of the concept of protection. Refugees’ experiences of urban space are often bounded and limited, and in view of that, I propose that refugees’ experiences of protection can be investigated by examining “space”. This approach, again, can have an important role in deepening our knowledge of urban refugee protection.
The author would also like to acknowledge the supervision by Emeritus Professor Roger Zetter and Associate Professor Patricia Daley during her doctoral studies at the University of Oxford. This article was originally written as part of an unpublished DPhil thesis examining the notions of protection, space, and trust in regard to urban refugees in Uganda. The work was financially supported by the Oskari Huttunen Foundation, the Finnish Cultural Foundation, the Emil Aaltonen Foundation, and the Alfred Kordelin Foundation.