1Migration has been a significant feature of African history for centuries. [1] The consolidation of empires, inter-communal conflict, shifting cultivation, and the slave trade all resulted in significant migration both from and within Africa. [2] The emergence of the modern refugee phenomenon in Africa may, however, be linked to the anti-colonial struggle and the independence of most African states in the late 1950s and early 1960s, largely for two reasons. First, “the process of decolonisation brought new and powerful political forces into play and released forces which had been checked or suppressed during the colonial period.” [3] Second, this increase in the number of refugees occurred in the context of newly independent states, internationally recognized borders, and changing relations with the international system, dramatically affecting the context within which post-colonial states responded to refugees.

2Through the 1960s and 1970s, many African states responded by integrating refugees into zonal development programs and by encouraging refugees from independent African states to be integrated into the national economy. This approach largely changed in the late 1980s and through the 1990s as African states were subjected to externally-imposed democratization and structural adjustment, combined with a dramatic increase in refugees from post-Cold War conflicts and with declining support from donors for refugee assistance programs. [4] Largely in response, and faced with the containment policies of states in the global North, many African states turned to more restrictive asylum policies, frequently containing refugees in isolated and insecure camps or forcibly returning refugees to their country of origin.

3The past decade, however, has witnessed greater diversity in the policy responses of African states to the mass arrival and prolonged presence of refugees. How can we understand the recent diversity of the asylum policies of African states in this historical context? To answer this question, I draw on the complex histories of asylum policy in Kenya and Tanzania—two of Africa’s largest refugee hosting states. Tanzania has hosted large refugee populations since the late 1950s and has demonstrated policies of inclusion—notably, the naturalization of some 162,000 Burundian refugees [5]—and of exclusion—notably, the expulsion of Rwandan refugees from its territory in 1996. [6] More recently, Tanzania has withdrawn from the UN-backed Comprehensive Refugee Response Framework (CRRF), which would have sought to further include refugees into national development planning. [7] For its part, Kenya became a major refugee hosting state in the 1990s, with the near-simultaneous arrival of significant numbers of refugees from Somalia, Ethiopia and Sudan. Kenya’s approach has been largely to contain refugees in isolated and insecure refugee camps, but this approach contradicts the country’s policy of including refugees in the national economy, a policy it pursued with refugees from Uganda in the 1980s. [8]

4I argue that the variations in asylum policies within and between Kenya and Tanzania are the result of both continuity and change. They represent continuity in that they continue an approach that formulates and implements asylum policies in response to the opportunity structure perceived by the governing regime. They represent change in that the opportunity structure confronting African states has changed significantly over time. As I argue below, African states have shifted from a context of being vulnerable and dependent on Western donors to a position where they are more able to assert their autonomy domestically, regionally and internationally. While this represents an important change in the factors that contribute to asylum policies in Africa, they can only be fully understood by situating the politics of asylum in African within a broader historical context. As argued by this collection, responses to refugees have deep historical roots, which we ignore at our peril.

5How can we categorize periods of asylum policy in Africa? Rutinwa argues that asylum policies in post-independence Africa may be broadly “classified into two periods.” [9] The first period, from the early 1960s to the 1970s, has been described as a time when African states “proved remarkably generous in their response to refugee influxes.” [10] Since the 1980s and through the 1990s, however, Rutinwa notes “a marked shift in refugee politics in Africa”, with states becoming “less committed to asylum.” [11] Before examining more contemporary periods, it will be useful to first understand the foundations of this early transition.

Open asylum policies: 1960s and 1970s

6In contrast with those who argue that the response of African states to refugees in the 1960s and 1970s is best explained in terms of “traditional hospitality”, [12] a closer reading of African history illustrates that asylum policies have been influenced by political concerns since the emergence of an independent African state system in the early 1960s. In fact, asylum policies have been largely motivated by concerns of domestic politics, national security and international relations. This is not to say that a period of open asylum policies in Africa did not exist, but that it existed for specific political and strategic reasons. As such, the shift to more restrictive asylum policies by African states in the late 1980s was arguably not the result of a new approach to refugees, but the result of the changing political context within which refugee movements occurred—both within the African context, and globally as states in the global North adopted more aggressive policies to contain refugees in their region of origin. [13]

7During the 1960s and 1970s, the overwhelming majority of refugees in Africa were to be found south of the Sahara and could be divided broadly into two categories. The first category included refugees from wars of national liberation in minority and colonial-governed states. By the early 1970s, 57 percent of the estimated one million refugees in Africa “originated mainly from the territories still under Portuguese rule, namely Angola, Guinea-Bissau and Mozambique […and] were living in three neighbouring countries sympathetic to the liberation cause, namely Zaire, Senegal and Tanzania.” [14] These refugees, and the liberation movements themselves, were often granted refuge by neighboring states, both as a gesture of African solidarity and as a means of highlighting the deficiencies of the colonial regimes. The second category of refugees was the result of “explosive internal, social and political situations, which existed long before the colonial period but which came to a climax only after independence was reached.” [15] Crises in Zaire, Rwanda, Burundi, Sudan, Ethiopia and Uganda prompted significant refugee movements in the Great Lakes region and in the Horn of Africa, and equally significant challenges for neighboring states. Throughout the 1960s and early 1970s, for example, Sudan believed that refugee populations in neighboring countries provided a base for insurgent groups. In response, the Sudanese military entered border areas in Uganda and Ethiopia, seized refugees and returned them to Sudan, resulting in a significant strain on relations between Sudan and its neighbors. It was these pressures that largely motivated the development of the OAU Convention on Refugees in 1969 as African states sought to supplement the legal regime resulting from the 1951 Convention relating to the Status of Refugees. [16]

8Despite these inter-state concerns, the majority of these refugees did not live in camps, but were integrated into local and national economies through the “zonal development” approach to refugee settlement. [17] As argued by Kironde, “zonal development and planning [was] generally accepted [by African states] as the best method of consolidating the economic and social condition of refugee settlements as well as integrating them into the economic and social systems of the country.” [18] The majority of refugees in Africa were accommodated in local settlements at the end of the 1970s, with more than 90 percent of refugees living in settlements in many of the largest asylum countries, including Zaire and Tanzania. As part of the refugee settlement approach, refugees were allocated land, granted the right to work and encouraged to pursue self-reliance through a range of economic activities. Such approaches were evident in the cases of Kenya and Tanzania.

9It was not until 1987 that the number of refugees in Kenya exceeded 10,000. The majority of these refugees were from Uganda and from more distant conflicts in Rhodesia (now Zimbabwe) and Rwanda. Many refugees brought skills as doctors and teachers that contributed to Kenya’s development and relative prosperity. Recognizing the benefits of hosting refugees, Kenya pursued an open asylum policy. Individual status was granted by a government agency; and refugees enjoyed freedom of movement, access to employment markets and the benefits of many of the social rights detailed in the 1951 Convention. Indeed, “the pre-1990 relief system in Kenya can be described as hospitable.” [19] During this period, the refugee population in Kenya was relatively small, and many refugees possessed technical skills that they were allowed to practice to benefit the Kenyan economy. As outlined by Kagwanja:

10

“Those with employable skills and talents were absorbed into the economy as shopkeepers, artisans, professionals and as high school and university teachers. […] Subsequently, these groups contributed to the economic growth of Kenya and tempered hostility against asylum seekers.” [20]

11Soon after achieving independence in 1962, Tanzania established a reputation as one of the most hospitable countries of asylum in Africa. Through the 1960s and 1970s, it hosted tens of thousands of refugees fleeing wars of national liberation in Southern Africa and, during the same period, post-colonial conflict and repression in neighboring states. Tanzania provided refugees with land; and refugees were encouraged to achieve self-sufficiency, with many entering the country’s workforce. A number of authors have linked Tanzania’s willingness to host refugees from neighboring states during this period to its view that refugees were “a human resource whose settlement in sparsely populated regions could be instrumental in sparking future economic development.” [21] In this way, refugees were received within the context of Tanzania’s ujamaa model of villagization, through which Tanzania’s rural population was organized into self-reliant, communal villages. [22] In contrast, refugees fleeing wars of national liberation were hosted as an extension of solidarity with these struggles, notwithstanding costs to Tanzania. For example, camps for South African refugees in Tanzania managed by South African liberation groups were “regularly attacked by South African armed forces”, [23] while “Portugal made punitive incursions against Tanzania and heavily mined the Mozambique-Tanzania border” in retaliation for Tanzania’s support to FRELIMO (Frente de Libertação de Moçambique). [24]

12Why did Kenya and Tanzania, along with other African states, respond with such open policies, notwithstanding the costs associated with the granting of asylum? It may be argued that there was a dual imperative for providing asylum to refugees in the 1960s and early 1970s. Asylum was granted to refugees fleeing wars of national liberation as part of the ideology of Pan-Africanism and as a means of highlighting the failures of colonial and minority-ruled territories. Likewise, providing asylum to refugees through the zonal development approach was seen as a means of attracting international assistance to underdeveloped areas of newly independent states. Such an approach, however, was premised on manageable numbers of refugees, sustained international assistance and the ability to insulate the governing regime from the security implications associated with hosting politicized refugee populations. Also significant was the fact that this was the era of “monopoly statehood” in Africa, where state leaders were able to pursue policies in the absence of domestic or international opposition. As refugee numbers increased, international assistance waned; and as security concerns increased, coupled with broader changes in the African political and economic landscape, willingness to host refugees diminished.

Emergence of restrictive policies: 1980s and 1990s

13By 1980, there were over four million refugees in Africa. As the continent witnessed additional refugee movements, especially from Chad, Ethiopia and Zaire, and as the number of refugees in Africa continued to climb, local solutions, including settlements and zonal development approaches, became more difficult to sustain. By the end of the 1970s, host countries began to argue that refugees had become an “open-ended burden.” Stein argues that while this change in perspective was the result of increased refugee numbers and the difficulties of finding quick solutions for refugees, it was largely compounded by the fact that many host countries had “less sympathy for, and solidarity with, refugees fleeing from their independent neighbour than for those fleeing from imperialism.” [25]

14These refugee movements also occurred in a changing international context. During this period, many African countries experienced slow economic growth, rapid population growth, balance of payments deficits and a climbing debt burden by the early 1980s. These factors placed additional pressures on host states, resulting in economic and political change that included the imposition of structural adjustment programs. [26] At the same time, new refugee movements, especially in Southern Africa and the Horn of Africa, emerged as part of the globalization of the Cold War. In contrast with the robust Western response to refugees in other regions, however, Loescher argues that “the politicization of refugee problems during the 1980s precluded an easy solution to the refugee dilemma”, especially in Africa. Unlike the Western response to the Indochinese crisis, there was no political will for the large-scale resettlement of refugees from Africa. [27] Unlike the Western response to the refugee situation in Latin America, donor governments no longer appeared willing to support large-scale local integration programs in Africa. Finally, the disengagement of the superpowers from Africa by the end of the 1980s left many of the proxy conflicts unresolved, thus frustrating repatriation efforts and leaving many refugee populations in limbo. Notwithstanding the large-scale repatriation to Namibia, significant refugee populations from Angola, Mozambique, Ethiopia, Sudan and elsewhere remained in exile at the end of the 1980s.

15By 1990, Africa’s refugee population was over 5.8 million. The vast majority of these refugees were no longer accommodated in settlements, but in refugee camps. Some, like Loescher, believe that this approach was sought by host states, who,

16

“concerned about the security risks of hosting refugees from neighbouring countries and about the enormous strains on local economic, political and physical resources, viewed refugee camps as the most convenient way to segregate refugees and to limit their impact on the local community.” [28]

17This shift was an early indication of changes in the asylum policies of African states. [29]

18These trends continued into the 1990s, which witnessed refugee movements in Africa unprecedented in their scale and complexity. Conflict in the Horn of Africa, the Great Lakes region and West Africa almost simultaneously resulted in the influx of hundreds of thousands of refugees into neighboring states. While a resolution of the conflict in Mozambique allowed for the repatriation of some 1.7 million refugees between 1992 and 1996, the proliferation of conflict across the continent saw Africa’s refugee population reach 6.7 million in 1994. Notwithstanding the scale of the problem, it appeared as though Western powers were reluctant to become involved in these crises, especially as their geo-political interests in Africa were diminished by the end of the Cold War. As a result, donor states made modest contributions to funding appeals for refugee situations in Africa, while also implementing domestic and regional policies to contain refugees in their region of origin. This response was particularly stark when compared to Western engagement in other regions, as “the level of international attention and humanitarian assistance to [Africa’s] refugees and displaced people was miniscule compared to other high-profile conflicts”, especially in the Balkans. [30] In response, a number of host states across Africa became “alarmed by the economic, environmental, social and security costs of hosting mass influxes of refugees” and “took steps to exclude asylum-seekers from their territory and to ensure the rapid—and in some cases involuntary—repatriation of refugees.” [31]

19In fact, by the mid-1990s, many believed that African states, generally, had begun “to retreat from the fundamental principles of international refugee law” [32] and that African states had shifted from the “open-door” asylum policies of the 1960s and 1970s to a more generally restrictive approach to asylum. Starting in the early 1990s, there was a dramatic rise in examples of refoulement, rejection at the frontier, expulsion of refugees and proposals to contain refugees in countries of origin. At the same time, there was a marked decline in standards of protection of refugees by a majority of African states, and a retreat from solutions for refugees other than repatriation.

20Rutinwa [33] and Crisp [34] were the first to attempt a systematic explanation for the cause of this continent-wide shift throughout the 1990s. Both agreed on the over-riding significance of the sheer numbers of refugees in Africa, rising from 1 million in the early 1970s to over 6 million in the mid-1990s. While this number had dropped to 3.1 million refugees by 2003, this reduction in numbers did not necessarily lead to a simplification of the refugee problem in Africa. 2.3 million (75 percent) of Africa’s refugees in 2003 were trapped in protracted refugee situations, having been in exile for more than 5 years with no prospect of a solution to their plight. [35] Unlike the 1960s, when most refugees were accommodated in settlements, or the 1990s, when most refugees were in emergency situations, the overwhelming majority of African refugees in 2003 had been contained in Africa’s 170 refugee camps. [36]

21Coupled with the protracted nature of many of these refugee situations was declining international support. Crisp noted that “donor states can be said to have exacerbated the decline in protection standards in Africa by making it increasingly clear that they are no longer prepared to support long-term refugee assistance efforts.” [37] Diminishing international support for long-term refugee programs, coupled with a wider recognition of the range of burdens borne by host states, resulted in the perception by a number of African states that they had come to bear a disproportionate share of the refugee burden. This perception was only exacerbated by the range of policies adopted by states in the global North during this period to contain refugees in their region of origin and to promote “regional responses” to further containment policies. [38] Moreover, UNHCR seemed either unwilling or unable to counter the policies of states in the global North, and arguably facilitated policies of containment. [39]

22Crisp and Rutinwa also emphasize changes in the nature of refugee movements and populations in Africa as another set of causes for the rise in restrictive asylum policies. Both highlight the fact that the majority of refugees in contemporary Africa are no longer fleeing wars of national liberation and decolonization, but rather civil wars taking place in independent countries. As a result, Crisp argued that “the new generation of African exiles have not been able to count on the support and solidarity offered to refugees in earlier years.” [40] These changing circumstances resulted in an increased perception of refugees as a security concern for host states and was increasingly cited by states as a justification for their restrictive asylum policies.

23Crucially, however, both Crisp and Rutinwa emphasized the relationship between external factors, including democratization and economic liberalization, and the rise of a more restrictive asylum policy in Africa. Crisp argued that “there is growing evidence of a linkage between the process of democratization on the one hand and the decline in refugee protection standards on the other.” [41] Both emphasize that prior to the 1990s, authoritarian and one-party states in Africa were able to offer asylum to individuals or groups without having to answer to domestic opposition. In contrast, the opening of the political process across Africa led to growing xenophobia in many African countries, often fueled by the effects of austerity and structural adjustment programs imposed by international financial institutions such as the IMF and the World Bank, which “forced governments to curtail free services to their populations, [making] it difficult for the people to accept the same services being provided free to refugees.” [42] In the context of democratization, Rutinwa argues that “the result has been the adoption of anti-refugee platforms by political parties which result in anti-refugee policies and actions by governments.” [43] Both argue that such factors must consequently be taken into account when considering the asylum policies of African states. Such factors were certainly at play in the changing asylum policies pursued by Kenya and Tanzania.

Kenya’s shift to containment and abdication

24Kenya’s approach to refugees changed dramatically with the arrival of the first Somali refugees in 1989. In late September 1989, 3,000 Somali refugees crossed into Kenya at Liboi, in the North Eastern Province. This was the third large group of Somalis to cross into Kenya in 1989, and they received the same treatment as previous groups. [44] They received no medical or material assistance, and many died as a result of injury or starvation. The refugees were beaten and harassed by Kenyan police and eventually forced back into Somalia. In addition, the government prevented humanitarian agencies from accessing the Somalis, from providing them assistance and from determining their refugee status. [45]

25From 1989 until the fall of the Barre regime, the Kenyan military maintained a large presence at the border and off the Kenyan coast to prevent the arrival of Somali refugees. With the exception of former government officials and military officials associated with the Barre regime, who “fled Mogadishu by air and took up residence in Nairobi hotels”, [46] thousands of Somalis were prevented from entering Kenya, through forcible return at the border and by “push-backs” of boats carrying Somalis. [47] In March 1991, there were an estimated 16,000 refugees in Kenya. That number rose to 39,000 by July, and to 92,200 by December 1991. [48] The rate of arrival continued to climb the following year, as the refugee population in Kenya reached an estimated 246,000 in May 1992 and peaked at 427,278 by the end of 1992. [49] The refugee population in Kenya had grown by 430,000 in just two years.

26In the years that followed, Kenya pursued a new refugee policy centered on two principles: “abdication” of responsibility for refugees to UNHCR [50] and the containment of the refugee population on the periphery of the state. Notwithstanding the principle that the primary responsibility for refugees lies with the host state, and in contrast to its engagement prior to 1991, Kenya withdrew “from active management of refugee affairs, leaving matters practically entirely in the hands of UNHCR.” [51] In response, UNHCR assumed “a number of responsibilities that would more properly fall within the government’s area of responsibility, de facto monopolizing refugee administration in Kenya.” [52] As Helton argues, “a deliberate choice was made by Kenyan government officials in the 1990s to largely cede refugee affairs to UNHCR.” [53] Second, Kenya pursued the policy of containing refugees in camps, located in isolated and insecure border regions. According to the Refugee Consortium of Kenya (RCK), “the government implements a refugee encampment policy” whereby refugees “are obliged to reside in a camp while awaiting a durable solution.” [54] Refugees must reside in camps to qualify for assistance, and those refugees found outside the camps are classified as illegal aliens and subject to deportation. This stands in contrast to its pre-1991 policy of allowing freedom of movement for refugees to facilitate their self-sufficiency and local integration. Through this policy, Kenya sought to situate the refugee “problem” as far as possible from the political core of the state and to insulate the regime from the presence of refugees.

27What drove Kenya’s new approach to refugees? First, the government claimed that the sheer magnitude of the refugee influx in the early 1990s overwhelmed its refugee procedures. Second, Kenya pointed to a lack of support from the international community as a justification for keeping refugees in camps, stating that the presence of refugees placed a strain on the environment and public services. Finally, the government blamed the presence of refugees for a rise in crime and insecurity, both in refugee-populated and in urban areas, in addition to the proliferation of small arms in Kenya.

28While the basis of these claims all deserve closer scrutiny, [55] it is also important to note how a number of broader historical and political factors, unrelated to the presence of refugees, defined the political space within which the Kenyan government formulated its response to the protracted presence of Somali refugees in Kenya through the 1990s. These factors include the colonial inheritance of a conflictual relationship between the government in Nairobi and the North Eastern Province of Kenya, formerly the Northern Frontier District. The region has seen a long history of underdevelopment and repression of its ethnic-Somali inhabitants, driven by fears of irredentism and threats to the territorial integrity of modern Kenya. [56] It is these practices, rooted in colonialism, and the subsequent history, which not only precluded the integration of Somali refugees in Kenya, but also contributed to Somali refugees being viewed as a threat by the Kenyan state.

29Likewise, refugee policy has been affected by the nature of domestic politics in Kenya. Specifically, the process of democratization since 1992 has witnessed the entrenchment of anti-refugee sentiments and stigmatization of the Somali community within Kenyan politics. As argued by the Refugee Consortium of Kenya:

30

“Experience has shown that in the run-up to elections many politicians will not hesitate to manipulate the refugee situation as an electioneering gimmick. Members of Parliament have been known to distort facts and stereotypes and vilify refugees as the sole source of increased crime and insecurity, proliferation of illegal arms and scarcity of resources. They have even been known to point to humanitarian assistance to refugees in the camps as evidence that refugees allegedly enjoy a better lifestyle than the locals. Blaming refugees detracts attention from their own responsibilities towards their constituents.” [57]

31These pressures towards restrictive policies, however, needed to be balanced against Kenya’s relations with the international donor community, upon whom Kenya’s economy was largely dependent. In November 1991, in the midst of the arrival of significant numbers of Somali refugees, the international community suspended aid to Kenya, citing the need for democratization and improved human rights. Kenya’s economy was heavily dependent on aid, [58] and the government had little choice but to comply. In late 1991, the Kenyan ruling party, KANU, revised the Constitution and legalized political parties. As a further concession to donors, the first multi-party elections were scheduled for late 1992. [59] At the same time as these reforms, violence erupted in the Western and Rift Valley Provinces, as Kalenjin youths displaced Kikuyu farmers from their land. While it was later argued that the “land clashes” were state sponsored, and intended to disrupt opposition strong-holds in the run-up to the elections, [60] the result was nonetheless the internal displacement of an estimated 300,000 Kenyans. Responding to this displacement posed a significant challenge to the Government of Kenya. [61]

32In the midst of these events and sustained pressure from the donor community, the challenges posed by the arrival of over 400,000 Somali refugees was almost a blessing in disguise. It could be argued that with the combined pressure of land clashes and democratization, it was good for the regime to have refugees in the country, as having refugees allowed the regime to show the international community that they were doing something for human rights by hosting the refugees. Indeed, the generous quantity of asylum afforded by the governing regime in 1991 and 1992, in stark contrast to its approach in 1989, prior to the suspension of aid, won praise from the donor community and resulted in the release of urgently-needed assistance to Kenya. [62] In this way, Kenya’s approach to Somali refugees was largely conditioned by the historical context of the state’s relations with its own ethnic Somali population and the balance between domestic pressures resulting from democratization and the demands of the international donor community, upon whom Kenya was reliant for its economic survival.

Tanzania’s shift from inclusion to exclusion

33Like Kenya’s experience with refugees from Somalia in the 1990s, Tanzania received hundreds of thousands of refugees as a result of civil war and genocide in Rwanda. The scale of this movement was staggering: between April 28 and 29, 1994, alone, some 250,000 Rwandan refugees crossed the Rusumo bridge into Ngara, Western Tanzania. It was “the largest and fastest movement of refugees in modern history.” [63] Eventually, there were approximately 700,000 Rwandans living in camps in the Kigoma and Karagwe regions of Tanzania. Tanzania’s refugee population climbed from 292,100 at the end of 1992 to 883,300 at the end of 1994. Conditions in both the camps and in the surrounding regions were marked by increasing instability by mid-1994. Within the camps, it was reported that suspected perpetrators of the genocide and members of the Former Government of Rwanda (FGOR) “mingled freely with genuine refugees.” [64] Tanzania concluded that insecurity in the border regions had reached “unacceptable levels.” [65] Environmental degradation also reached alarming levels, and the local population watched “with increasing frustration as large tracts of forest and agricultural land were destroyed.” [66] Shortly thereafter, Tanzania’s long-standing “open-door” refugee policy changed dramatically.

34Tanzania’s first change in policy came on March 31, 1995, when, almost a year after the initial influx, and in the lead-up to the country’s first multi-party elections, the government closed its border with Burundi to prevent additional Rwandan refugees from fleeing renewed conflict in Burundi. Then, “apparently encouraged by the failure of the international community to criticise the repatriation from Zaire”, [67] Tanzania decided to expel the Rwandan refugee population in December 1996. The Tanzanian government and UNHCR issued a joint statement on December 6, stating that all Rwandan refugees were expected to repatriate by December 31, urging that “all refugees make preparations to return before that date.” [68]

35Many refugees requested a delay in the deadline to determine for themselves the viability of return to Rwanda; but as it became clear that the deadline would not be reconsidered, “Rwandans sought other methods to avoid repatriation.” [69] During the night of December 11, an estimated 35,000 refugees left the camps and started moving further into Tanzania. This movement continued the following day. On December 13, the Tanzanian army intervened, “forcing as many as 200,000 fleeing refugees to turn around and retrace their steps.” [70] At the same time, the camps were closed and the Rwandan refugees were “herded down the road toward the border” [71], with the first refugees crossing back into Rwanda on December 14. This mass exodus in reverse caused “widespread dismay in the humanitarian world”, [72] and numerous human rights organizations were quick to criticize the repatriation. [73]

36Many authors have examined Tanzania’s decision to both close its border with Burundi and expel the Rwandan refugees. [74] Most explanations highlight a range of factors, including multi-partyism and the changes in Tanzanian political culture in the 1990s, consequent changes in foreign policy, the magnitude of the Rwandan and Burundian refugee populations, the nature of the populations, external and internal security, and the lack of international support. This section briefly outlines these arguments.

37First, it is important to emphasize that Tanzania’s change in policy came in the context of domestic political changes within Tanzania, not only as a result of externally-imposed economic liberalization and the end of the ujamaa model in the mid-1980s, but also in the run-up to the country’s first multi-party elections. In February 1992, a special congress of Tanzania’s ruling party, CCM, [75] endorsed the principle of a multi-party system of government, partially as a result of pressure from donor institutions. The first multi-party elections were scheduled for October 1995. During the ensuing campaign, “the politicians of all parties […] decided to play to popular sentiments by promising to send all the refugees back.” [76] CCM portrayed itself as “the party of nationalism, peace and stability”, claimed that a vote for the opposition would lead to ethnic divisions in Tanzania, and “lost no time in […] suggesting that to abandon CCM would be to invite disaster.” [77]

38While these political changes in Tanzania were necessary for the change in hosting policy, they alone were not sufficient. Attention must also be paid to specific features of the Rwandan refugee population. First, the scale of the Rwandan influx must be appreciated, with some 700,000 refugees arriving in Tanzania in less than a year. Second was the perception that a solution for the Rwandan and Burundian refugees “remained elusive [… a fact which] began to fuel Tanzanian frustrations.” [78] A third factor was the widely-held suspicion that the camps were harboring FGOR elements who were fleeing not persecution, but justice. These elements maintained a strong influence over the population within the camps, as “bona fide refugees were held hostage, intimidated and even killed when they expressed the wish to return home.” [79] The activities of these elements also extended beyond the camp boundaries, leading to more general concerns of law and order in the surrounding areas. [80] These concerns were compounded by declining support from the international donor community. While it has been argued that the initial response to the emergency was impressive, [81] this support waned over time, leaving the local authorities to supplement food aid, and leaving the local environment and population vulnerable. [82] This led Rutinwa to conclude that the “failures of the international community to give adequate assistance to Tanzania was the main reason for the closure of the border” with Burundi, marking the change in Tanzania’s refugee policy. [83]

39Even before the 1996 expulsion of the Rwandan refugees, government Ministers cited these concerns as adequate justification for a change in asylum policy. Speeches from the Ministries of Foreign Affairs and Home Affairs around the time of the border closure in 1995 suggest significant agreement between government departments on the nature of the problem. Citing the strain on the environment, the local economy, and infrastructure, and concerns relating to internal and external security, Ministers claimed that “the problem has reached saturation point and can no longer be sustained.” [84] Reflecting on Tanzania’s history of hosting large numbers of refugees, it was argued that “though traditionally the local population has always been sympathetic to refugees, all this is now changing under the strain of the current load.” [85] The Ministers of Home and Foreign Affairs both claimed that “experience has shown that measures such as the granting of permanent asylum and citizenship to the refugees are not a panacea for a permanent solution to the refugee crisis” and that “the solution lies in the countries of origin rather than the countries of asylum.” [86] There was agreement on repatriation as the only viable solution for refugees in Tanzania, [87] as well as early calls for the establishment of safe zones in countries of origin as a substitute for asylum. Significantly, both Ministries also accused the international community of “double standards” when condemning Tanzania for its policies, in light of restrictive policies in the West:

40

“There is a tendency by the International Community when dealing with the refugee crisis to impress upon the weaker countries of asylum to live up to their humanitarian obligations at the expense of their national rights and interest. The primary duty of a state is to preserve itself and to protect its citizens and their rights. Countries of asylum, be they weak or strong, poor or rich, have a sovereign right to exercise this duty. The international community must respect this right without bias or double standards.” [88]

41Given the range of challenges stemming from the Rwandan influx and based on the government position articulated in the speeches cited above, it seems clear that Tanzania’s position on the refugee question changed as a result of the influx of refugees in the early 1990s. The emphasis on self-sufficiency and local settlement was replaced by a focus on repatriation, while the integration of refugees into the Tanzanian political community was replaced by a desire to exclude them. Most significantly, refugees are characterized no longer as an asset to Tanzania, but as a burden.

Explaining the rise of restrictive policies: The politics of asylum in Africa

42The histories of Kenya and Tanzania illustrate how political factors unrelated to the presence of refugees tend to play a greater role in the formulation of asylum policies than refugee-specific factors, highlighting the need to understand the politics of asylum in Africa. [89] In the case of Kenya, the history of conflict between the state and Somali irredentism, coupled with pressure from the international donor community, put pressure on the weakened governing regime at the time of the arrival of the Somali refugees. In Tanzania, the introduction of multi-party elections in 1995, coupled with the end of the ujamaa model in 1985, resulted in challenges to the CCM regime and a rearticulation of the idea of the Tanzanian state in the midst of the refugee crisis in the Great Lakes region. In both cases, feelings of vulnerability have contributed significantly to the formulation and implementation of particular asylum policies.

43Through this lens, our understanding of the asylum policies of African states in the 1990s logically begins with an understanding of the characteristics of the host state and governing regime at the time, and of how those factors influenced state behavior. Reflecting on this period, Ayoob argues that in the context of Third World states, “state behaviour […] is largely determined by the insecurity that is aggravated by the overwhelming feeling of vulnerability, if not impotence, among its state elites.” [90] Given this condition, the primary objective of Third World regimes is to “reduce the deep sense of insecurity from which [they] suffer domestically and internationally.” [91] In turn, this sense of vulnerability highlights the calculus that shaped the politics of asylum in Africa through the period leading to the 1990s. Kibreab argues that states are more likely to host refugees if they perceive their presence as economically or politically beneficial. [92] Similarly, Loescher argues that states are more likely to grant asylum when it is likely to increase their power relative to other actors who might otherwise challenge the regime. [93] More generally, he argues that:

44

“The formulation of refugee policy involves a complex interplay of domestic and international factors at the policy-making level and illustrates the conflict between international humanitarian norms and the sometimes narrow self-interest calculations of sovereign nation states.” [94]

45The vulnerability of African states played a significant role in shaping these “self-interest calculations”, and largely defined the political space within which asylum policies were formulated. Given this vulnerability, it may therefore be argued that, at minimum, African states were only willing to implement open asylum policies when it was not to the detriment of the host state or governing regime. Conversely, African host states adopted restrictive asylum policies in cases where the arrival and presence of refugees negatively affected state boundaries, state institutions or governing elites, or weakened “the capacity of states and regimes to act effectively in the realm of both domestic and international politics.” [95]

Recent asylum policies: Continuity or change?

46More recent trends in asylum policies in Africa, however, bring into question the enduring relevance of “vulnerability” as the lens through which the asylum policies of African states can be fully understood. In fact, some states have pursued very open asylum policies, including the promotion of self-reliance for refugees, the inclusion of refugees in national and local economies, and pathways to naturalization and the acquisition of citizenship—all in the midst of continued pressures from the arrival of large refugee populations and fickle responses from the international community. Other states have adopted more restrictive policies, by closing borders, denying refugees key rights and threatening the forced return of refugees to their country of origin. As with earlier periods, the cases of Kenya and Tanzania help illustrate these instances of continuity and change.

47Kenya’s recent response to Somali refugees represents a continuation of its policy of the 1990s. Somali refugees remain largely contained in refugee camps in Northeastern and Northwestern Kenya. While Kenya continued to limit the quality of asylum for refugees by limiting freedom of movement and the right to work, it was generally understood to have maintained the quantity of asylum by allowing refugees to remain on its territory, even in the midst of more large-scale arrivals in 2011. Betts argues that this approach was motivated by a continued desire to maintain its international reputation and ensure access to humanitarian resources from the international community, despite the fact that Somali refugees in the period around 2011 were fleeing both on-going insecurity in Somalia and the consequences of drought and climate change. [96]

48More recently, however, Kenya’s willingness to maintain a high number of Somali refugees on its territory has waned, especially as the securitization of Somali refugees has increased. Somali refugees in Kenya have increasingly been associated with Kenya’s insecurity resulting from a porous border with Somalia, the activities of al-Shabaab in Kenya, and attacks by al-Shabaab on Kenyan territory, especially since the attack on the Westgate Shopping Centre in Nairobi in 2013 and the attack on a university campus in Garissa, just 100 km from the Dadaab refugee camps. In the wake of these attacks, the Government of Kenya announced in May 2016 its intentions to close the Dadaab camps and disband its Department of Refugee Affairs. Significantly, UNHCR and the donor community seemed unable to dissuade Kenya from pursuing repatriation, given the prevailing insecurity in Somalia. In fact, Kenya, Somalia and UNHCR negotiated a Tripartite Agreement to facilitate repatriation in 2015. Ultimately, it was a ruling of Kenya’s High Court in February 2017 that blocked the closure of the Dadaab camps and the repatriation of Somali refugees.

49This represents important continuity and change in the case of Kenya’s asylum policies. The government’s securitization of Somali refugees and limitations on their rights represent clear continuity from the approach across governments over the past quarter century. Where there has been change, however, has been the seemingly limited ability of external actors to influence the actions of the Kenyan government. This may be explained, at least in part, by the declining reliance of Kenya on Western donors, given increased Chinese investment coupled with the declining moral authority of European donors as a result of restrictive asylum policies. More generally, it follows the trend of Kenya’s increased assertiveness and sense of autonomy from the pressures of Western donors, especially following investigations into election-related violence in 2007, indictments of senior government officials by the International Criminal Court (ICC), and suggestions by President Kenyatta in 2016 that Kenya could withdraw from the ICC altogether.

50Tanzania’s recent asylum policies have represented both continuity and change. In apparent contrast to its restrictive policies of the 1990s and its 2003 National Refugee Policy, the Government of Tanzania announced in 2007 its willingness to offer naturalization as part of a comprehensive solution for Burundian refugees who had been in Tanzania since 1972. [97] By June 2010, 162,156 applications for naturalization had been approved. The process then stalled, with the government announcing in late 2011 that it had suspended the process of naturalization. Then, after three years of uncertainty, Tanzania announced in October 2014 that it had decided to go ahead and fully implement the program. Tanzania’s President, Jakaya Kikwete, granted citizenship to all approved applicants, along with their dependents born after the application process, on October 14, 2014.

51External factors played only a very marginal role in Tanzania’s initial decision to offer citizenship to Burundian refugees. Instead, the decision to facilitate naturalization was motivated more by changing relations between Tanzania and Burundi, the political abilities of UNHCR’s Representative to Tanzania, dynamics within Tanzania’s ruling party, and Tanzania’s prior history of naturalizing Rwandan refugees in the 1980s. In contrast, Tanzanian electoral politics, power struggles within the CCM, and the shifting position of President Kikwete within Tanzanian politics best explain the lack of progress in implementing the naturalization program between 2010 and 2014. The fact that President Kikwete was not able to run for reelection in 2015, coupled with his desire to prevent Tanzania from reneging on an international commitment, also partially explain why citizenship was ultimately granted in 2014. Many of these factors are bound-up in broader shifts in the local, national and regional politics of Tanzania since independence.

52Past models of regime vulnerability seem inadequate to fully explain the continuing evolution of the asylum policies of African states such as Kenya and Tanzania. In fact, it would seem problematic to now argue that African regimes can be characterized as vulnerable and unable to act autonomously from the interests of external actors. While Kenya’s recent policies have been widely condemned by Western donors and UNHCR, external actors have seemed unable to exert independent influence to change these policies, in contrast to their influence on asylum policies in the 1990s. Likewise, in Tanzania, the government has pursued an approach to asylum that has taken many Western observers by surprise—with its willingness to offer citizenship to Burundian refugees, its suspension of the process, its ultimate issuance of citizenship certificates and its recent withdrawal from the UN’s Comprehensive Refugee Response Framework. While different in outcome, neither Kenya nor Tanzania have behaved as states that are vulnerable or that lack full sovereignty in the international system.

53In turn, this raises important questions about the need to develop new approaches to understand the politics of asylum in Africa and the extent to which recent policies are expressions of continuity or change in the long historical arc of Africa’s experience with refugees. A closer examination of recent developments, however, suggests that the changing nature of African asylum policies represents both continuity and change. Policies represent continuity as African states continue to formulate and implement their asylum policies largely in response to the perceived interests of state elites and in response to a range of domestic, regional and international challenges and opportunities. This logic is not limited to African states. In his examination of asylum policies in liberal democracies in the global North, Gibney concludes that:

54

“How much any state—or to be more specific, any government—can do for refugees will be determined largely by the possibilities afforded by its domestic political environment, and that environment will be shaped by a changing array of social, institutional and economic forces, both domestic and international in origin.” [98]

55This logic has endured in African states since independence, [99] thus reflecting continuity. What has arguably changed has been the inputs to this logic and the changing constellation of challenges and opportunities at the domestic, regional and international levels perceived by African states, and the increased differentiation in the ways in which these challenges and opportunities are experienced by African states. These changes include changes in the political economy of development in African states, [100] the influence of resource extraction on the behavior of governments [101] and the rise of new external actors seeking engagement with, and offering forms of support to, African states, regimes and elites. [102]

56All this points to the need for sustained scholarly engagement in the evolution of the asylum policies of African states, avoiding collective generalizations and more fully situating our understanding of the politics of asylum in Africa within a nuanced and careful reading of the historical, political and economic context within which asylum policies are developed. Many of these contexts have deep historical roots and represent remarkable continuity in their logics. While the policies of states in the global North have contributed to the containment of refugees in their region of origin, and while UNHCR has made various attempts to influence the asylum policies of African states, a closer reading of the history of asylum in Africa illustrates how deeper political factors and interests have conditioned the policies adopted by African states in response to the mass arrival and prolonged presence of refugees. It can, therefore, be concluded that a more historicized reading of asylum policies in Africa can only help in our quest to better understand the responses of African states and to support the protection of refugees in Africa and a potential solution to their plight.