As global migration increases, two-thirds of Africa’s migrants move neither to Europe nor to North America but rather to other African countries (Gonzalez-Garcia et al. Reference Gonzalez-Garcia2016), where they often encounter hostility and xenophobia. Surveys show high levels of anti-immigrant sentiment in some African countries, particularly in southern Africa, but relatively low levels in others, including Ethiopia, Nigeria, and Uganda, which host large migrant populations (Crush & Pendleton Reference Crush and Pendleton2004; World Values Survey 2017). This variation in attitudes is influenced by many economic, cultural, and political factors (Whitaker & Giersch Reference Whitaker and Giersch2015). In countries such as South Africa and Côte d’Ivoire, tensions between citizens and foreigners have turned violent. The cross-border movement of peoples also has contributed to international conflict, including Rwanda’s military involvement in the Democratic Republic of Congo and Kenya’s intervention in Somalia.
As anti-foreigner attitudes have increased, many governments have imposed restrictive immigration policies. Botswana, for example, constructed an electrified border fence and deported undocumented migrants while at the same time imposing narrow guidelines for asylum seekers. Tanzania has closed its western border several times, forcibly repatriated thousands of refugees, and imposed limits on other migrants still in the country, even as it has granted citizenship to thousands of refugees from Burundi and Somalia. Kenya concentrated refugees in several large border-area camps and periodically threatened to expel Somali refugees. Thus, even in countries that have historically welcomed immigrants, refugee and immigration policies have shifted. This is not to suggest, however, that past governments were motivated by “traditional African hospitality” (Kibreab Reference Kibreab1985). Countries that maintained open-door policies generally did so to reap the economic benefits that came from an abundance of cheap labor, including the opportunity to open new lands for cash crop production. One-party governments in Côte d’Ivoire and Tanzania even allowed foreigners to vote in elections, with the additional votes boosting their tallies. For decades, then, immigration policies in Africa have been determined largely by economic and political interests.
This article offers an appraisal of the factors that shape immigration policy in African countries by comparing government responses to Somali refugees. Despite a long history of Somali migration throughout the region, policies toward this group have varied significantly. I argue that host governments have responded differently to Somali refugees depending on the relative importance of economic concerns, political considerations, security threats, and normative commitments. This study focuses on three host countries: Kenya, the first destination for many Somali refugees and home to the largest population of migrant Somalis; South Africa, the target destination for many Somalis seeking better opportunities within the continent; and Tanzania, which became a permanent destination for a small population of Somali Bantu refugees due to a unique historical connection.Footnote 1
After reviewing existing literature to identify factors that may influence the immigration policies of African governments, I systematically examine these factors in each host country. This comparative case analysis shows that the relative influence of these factors varies by country, with resulting effects on the extent and scope of immigration measures. In Kenya, Somalis have long been treated as a security threat, a perception that has intensified due to recent Al-Shabaab attacks. Thus, the government has imposed restrictions on Somali refugees that do not apply to other immigrant groups. In South Africa, where economic and political competition have shaped broader xenophobic attitudes, Somalis are treated as a small part of a larger foreign national population that is seen as having disproportionate economic influence. Finally, in Tanzania, normative considerations led the government to grant citizenship to Somali Bantus even as other refugees have faced increasing restrictions on their movement and activities.
Comparative Immigration Policy
Existing comparative immigration policy literature is based primarily on Western countries, but points to several factors that could help explain the differing responses to Somali migration within Africa. The first is the host country’s political and economic context. Pro-immigration interest groups tend to be better organized, but anti-immigration groups increase their pressure on policymakers during periods of economic recession (Freeman Reference Freeman1995; Givens & Luedtke Reference Givens and Luedtke2005; Hopkins Reference Hopkins2010). In some countries, political parties legitimize and mobilize xenophobia(Art Reference Art2011; Rydgren Reference Rydgren2008), but where immigrants are seen by party leaders as a potential support base, permissive policies can emerge (Schain Reference Schain2008). Often, the translation of attitudes into government policy depends on which constituencies are necessary to build a winning electoral coalition (Money Reference Money1997, Reference Money1999). These theories help explain recent differences between Côte d’Ivoire and Ghana with respect to the politicization of immigration (Whitaker Reference Whitaker2015) and could similarly be useful in understanding government responses throughout Africa to Somali migration.
Another factor highlighted in past literature is the link between migration and security (Adamson Reference Adamson2006; Weiner Reference Weiner1993; Weiner & Russell Reference Weiner and Russell2001). Security concerns have long played a role in shaping the policies of African governments toward refugees (Jacobsen Reference Jacobsen1996; Whitaker Reference Whitaker1999), especially when influxes are large and relations between home and host countries are contentious (Lischer Reference Lischer2006). In a post-9/11 world, non-refugee immigrants are increasingly seen as posing a similar threat, particularly after high-profile terrorist attacks in places such as Paris, Nairobi, and Kampala.Footnote 2 To the extent that migration policy is used to manage security threats, it is a function of the level and type of threat that countries face (Rudolph Reference Rudolph2003).
Past research also points to the influence of international norms on immigration policy, including international agreements that obligate host governments to behave in certain ways toward migrants and refugees (Jacobsen Reference Jacobsen1996; Joppke Reference Joppke1998; Soysal Reference Soysal1994). Diffusion effects also play a role as governments react to immigration policies adopted by other countries (Timmer & Williams Reference Timmer and Williams1998). Indeed, African leaders often highlight the exclusionary policies of Western countries to deflect criticism of their own (Whitaker Reference Whitaker1999).
In seeking to explain differing responses to Somali migration within Africa, therefore, past literature suggests a need to examine several factors: economic competition, political calculations, security threats, and normative commitments. Host government policies are likely influenced by some combination of these factors, the relative balance among which may change over time. With this as a starting point, then, the next section turns to Somali migration within Africa and the policy responses in three host countries.
Somali Migration within Africa
As noted in the introduction to this forum (Ibrahim, Malik, & Wielenga Reference Ibrahim, Malik and Wielenga2020), the dispersal of the Somali people out of their original homelands occurred in three waves. Of these, the most significant out-migration of refugees from Somalia took place in the late 1980s and early 1990s, when a civil war and the collapse of Siad Barre’s authoritarian regime led to widespread insecurity, with various armed factions competing for power. The ensuing famine in 1991 to 1992 forced many more Somalis to flee. International military interventions from 1992 to 1995 failed to resolve the conflict, but the security situation in Somalia stabilized in the late 1990s under an arrangement described as “governance without government” (Menkhaus Reference Menkhaus2006). Violence escalated again in the 2000s, with the rise to power of the Union of Islamic Courts in 2006 and its ouster by Ethiopian troops six months later. The subsequent emergence of Al-Shabaab’s violent insurgency against the precarious internationally-backed Somali government prompted a renewed exodus. Some refugees started to return when Al-Shabaab lost ground, in part due to the African Union Mission in Somalia (AMISOM) after 2007, but several devastating attacks in recent years show that the security situation in Somalia remains tenuous.
Over the past three decades, Somalia’s diaspora has grown to nearly two million people (United Nations 2017).Footnote 3 Most Somali refugees fled first to neighboring Kenya and Ethiopia, which continue to host nearly half a million Somalis each. Some eventually found their way to other countries in Africa, the Middle East, and beyond.Footnote 4 Indeed, for many Somali families, dispersion to multiple cities and countries has been a survival strategy to protect members and sustain remittances (Ikanda Reference Ikanda2018b; Lindley Reference Lindley2010). Significant attention has been given to the resettlement of Somali refugees in western countries (Balakian Reference Balakian2017; Roble & Rutledge Reference Roble and Rutledge2008; Yusuf Reference Yusuf2012), and the dreams that many Somalis have of being resettled overseas (Horst Reference Horst2006; Ikanda Reference Ikanda2018b, Reference Ikanda2018a), but very few are granted this opportunity.Footnote 5 As a result, nearly 64 percent of Somali’s international migrants reside elsewhere within Africa (United Nations 2017).
With the largest economy in the region, South Africa is a preferred destination for many African migrants, including those from Somalia. While some Somalis have moved there through formal resettlement and immigration procedures, many have put themselves at the mercy of human smugglers and corrupt border officials to work their way along the “southern route” from the Horn of Africa to the perceived “promised land” (McCormick Reference McCormick2016). The Somali population in South Africa increased tenfold from 1995 to 2017, though it still represents less than one percent of the total foreign-born population in the country (United Nations 2017). In between the Horn and South Africa, various other countries have become home to smaller populations of Somali refugees, including some who have abandoned the trek further south.
African host countries vary significantly in the way they have handled the presence of Somali refugees. In Kenya and Ethiopia, Somalis have been for the most part confined to large-scale refugee camps close to the border. Kenyan authorities have periodically rounded up Somalis living in cities and moved them to the camps (Anderson & McKnight Reference Anderson and McKnight2015; Balakian Reference Balakian2016). The Ethiopian government blocks refugees from getting work permits or business licenses, though it has recently pledged to allow some out-of-camp jobs (Smith et al. Reference Smith2019). In contrast, Somali refugees in Uganda are permitted more freedom of movement and employment beyond their settlements, and a significant portion live in Kisenyi, Kampala’s “Little Mogadishu” (McSheffrey Reference McSheffrey2014). A recent law in Djibouti similarly allows refugees to move around for work and school, though many Somalis continue to reside in camps (Smith et al. Reference Smith2019). In South Africa, Somalis live and work throughout the country, especially in urban areas, and the Mayfair suburb of Johannesburg has become home to a sizeable Somali community (Ripero-Muñiz Reference Ripero-Muñiz2017). And in northeastern Tanzania, Somalis were housed first in the Mkuyu refugee camp before being allocated farmland in the larger Chogo settlement and eventually offered citizenship.
What explains this wide variation in African government responses to Somali migration? Clearly, raw numbers are an important part of the story. As the countries bordering southern Somalia, where the worst of the violence has occurred over the years, Kenya and Ethiopia have received by far the largest numbers of Somali refugees. This has undoubtedly shaped their policy responses, as explored in more detail in the Kenya case study below. Even so, Djibouti also borders Somalia and has received a significant number of refugees, especially as measured per capita, but it has moved toward more progressive policies.Footnote 6 And Tanzania has received a relatively small number of Somali refugees but initially kept them encamped, while Uganda and South Africa allowed larger numbers of Somalis to integrate into host communities. Beyond just the numbers, then, the variation in government responses appears to reflect differing approaches toward refugees and migrants broadly and toward Somalis in particular.
As a result of this migration throughout the region, Somali refugees are an ideal population for examining the various immigration policies of African governments and the different forms those policies take, allowing a comparison of the same group across multiple contexts. After effectively controlling for some key characteristics of the migrant population, one can focus on differences among the host countries that shape their divergent approaches toward Somalis and other immigrant populations. Drawing on existing accounts, field visits, and parliamentary debates in Kenya, South Africa, and Tanzania, and building on the comparative literature discussed above, I examine economic, political, security, and normative factors influencing these countries’ respective responses to Somali migration.
Kenya
With a total population of nearly 50 million, Kenya is home to more than one million foreign-born individuals; Somalia, Uganda, South Sudan, Tanzania, and Ethiopia are the top five countries of origin (United Nations 2017). Approximately half of the immigrants in Kenya are refugees, asylum-seekers, and other people of concern to the United Nations High Commissioner for Refugees (UNHCR) (UNHCR 2018). To understand the recent escalation of hostility toward immigrants and refugees in Kenya, and especially the crackdown on Somali refugees, it is necessary to examine the broader context in which these changes have taken place.
Since the early 1990s, Kenya has undergone a prolonged period of economic and political liberalization. After some initial challenges associated with the process of structural adjustment, the economy has grown significantly over the past two decades, and above the African average in recent years, but the gains have not been realized by everyone. There is a deep divide between the burgeoning middle class and millions of poor Kenyans who eke out their livings in urban slums and underserved rural areas. Yet some immigrants, including refugees from Somalia, have taken advantage of free market reforms to monopolize retail trade in the Eastleigh section of Nairobi and other urban areas and to establish extensive trading networks throughout the region (Little Reference Little2003; Weitzberg Reference Weitzberg2017), generating resentment among some locals.
As Kenyan politics also have become more competitive, particularly since a transition to democracy in 2002, politicians have sought to win power by mobilizing voters largely along ethnic lines in order to build winning ethnic coalitions. These dynamics have fueled heated debates about which ethnic groups “belong” in the Kenyan polity, with much attention focused on the significant population of Kenyans of Somali ethnicity (designated in the 2009 national census as “Kenyan Somalis”) and their long and complicated history of inclusion in and exclusion from the Kenyan state (Weitzberg Reference Weitzberg2017). At roughly 6 percent of the population, Kenyan Somalis currently have relatively little political clout, and their support is not necessary for victory, which undermines any political pressure for more inclusive immigration policies. Some observers believe that the Kenyan government is not willing to grant citizenship to Somali refugees because their sheer numbers would alter the calculations for assembling winning ethnic coalitions (Warah Reference Warah2016). Within these complex economic and political dynamics, people have become receptive to periodic efforts by politicians to win support by lashing out at foreigners, both those who are foreign-born and those who are perceived as foreign due to their ethnicity.
Even more than economic concerns, political discourse in Kenya has emphasized the security risks of immigration. Somali refugees in particular have been defined as a security problem for decades. Indeed, official suspicion of Somalis goes back to at least the 1960s, when Kenyan Somalis were perceived as supporting regional efforts to build a Greater Somalia. There have been periodic attempts since then to differentiate between citizen and non-citizen Somalis in Kenya, generally in the name of national security (Lochery Reference Lochery2012; Weitzberg Reference Weitzberg2017). In the 1990s, when war and famine in Somalia generated a large-scale exodus across the border, Kenyan officials emphasized the security threat posed by the sheer numbers of refugees. The Dadaab refugee complex in northeastern Kenya opened with roughly 90,000 Somali refugees in the early 1990s and peaked at more than half a million refugees in 2011. After the 1998 bombing of the United States Embassy in Nairobi that killed 224 people (and the subsequent 9/11 attacks in the U.S.), American-backed counterterrorism operations in Kenya focused on the country’s Muslim minority, particularly people of Somali ethnicity (Whitaker Reference Whitaker2008b).
In recent years, a series of terrorist attacks by people with connections to Al-Shabaab (see Inman Reference Inman, Carpenter and Lawrance2018), an insurgent group based in Somalia, has further exacerbated Kenyan security concerns. In 2011, a spate of kidnappings and a renewed refugee exodus prompted Kenya to send troops into Somalia, later joining forces with the United States-backed African Union mission there. In response, Al-Shabaab stepped up its attacks on Kenya (Anderson & McKnight Reference Anderson and McKnight2015), most notably with assaults on Nairobi’s Westgate Mall in 2013, Garissa University in 2015, and the Dusit D2 Hotel complex in Nairobi in 2019, which together killed 235 people. Kenyan officials insist that Somali refugees helped plan these attacks although they have not produced evidence, and several people involved in the most recent attack were Kenyans of non-Somali origin, causing analysts to warn of a new wave of homegrown terrorism (AFP 2019). Many Kenyans have called on their government to withdraw troops from Somalia, but President Uhuru Kenyatta has so far resisted.Footnote 7 His reluctance to make this move is motivated perhaps not only by security concerns but also by the distraction the intervention has provided from his own domestic political woes.Footnote 8
Periodically, and often before national elections, the Kenyan government has threatened to close refugee camps and expel residents from the country, as explored further in Balakian’s (Reference Balakian2020) article in this forum. With each threat, human rights activists have reminded government officials of their obligations under international law, and foreign donors such as UNHCR and the United States government have pledged additional funding to support the camps. Whether motivated by normative commitments or millions of dollars in foreign aid, the government repeatedly has allowed camps to remain open or moved residents to other camps within the country.Footnote 9 After yet another pledge in May 2016 to close refugee camps later that year, human rights organizations challenged the order in Kenyan courts. In February 2017, the High Court blocked the government’s plan to close Dadaab, with Judge John Mativo arguing that “specifically targeting Somali refugees is an act of group persecution, illegal, discriminatory, and therefore unconstitutional” (Gettleman Reference Gettleman2017). The government vowed to appeal the decision on the premise that its primary responsibility was to provide security to all Kenyans and that the camps had become “a haven for terrorism and other illegal activities” (Government of Kenya 2017). From the perspective of government officials, security concerns supersede any normative commitments to refugee protection.
In Kenya, both Somali immigrants and Kenyan Somalis have been targets of suspicion. Consistent with the view that they are a security threat to the nation, government policy has confined Somali refugees in remote camps and periodically closed the border to prevent a further influx. There have been frequent efforts to round up refugees (and some Kenyan citizens) from Somali-dominated urban areas and send them to Dadaab, especially since the launch of Operation Usalama Watch in April 2014. Such actions have led to charges by human rights groups that the government is scapegoating Somalis (Amnesty International 2014). Shakedowns have become so frequent that the Eastleigh neighborhood of Nairobi is described colloquially as a cash machine for the Kenyan police (Balakian Reference Balakian2016).
Importantly, recent round-ups and similar policies have focused on Somalis, despite the continued presence in the country of thousands of refugees and immigrants from South Sudan, Ethiopia, and elsewhere. This is presumably because of the special security threat that people of Somali ethnicity are seen as posing in light of recent attacks by Al-Shabaab. At times in the past, however, other refugee groups in Kenya have been the targets of security operations based on similar perceived threats. In 1990, for example, refugees from Uganda and Rwanda were rounded up and many were deported, after Rwandan exiles launched an armed incursion into Uganda (Veney Reference Veney2007). Along similar lines, the tone and volume of debates in parliament and among foreign policy officials in Kenya varied significantly from 1963 to 2010 depending on the country of origin of the refugees under discussion (Abdelaaty Reference Abdelaaty2017).
An analysis of recent parliamentary transcripts in Kenya confirms a focus on refugees over other types of immigrants and a particular emphasis on security in discussions about Somalis. In 2016 and 2017, according to Kenya’s Parliamentary Hansard Archive, there were 237 sessions of the National Assembly, the lower house of Parliament, including joint sessions with the Senate.Footnote 10 Most parliamentary business in Kenya is conducted in English, though some members opt to speak in Swahili. A search of parliamentary transcripts yields a total of 1,360 references to migrant(s) (wahamiaji in Swahili), refugee(s) (wakimbizi), or foreign national(s) during that two-year period. Among these, 1,188 (87 percent) were references to refugees, suggesting a heavy focus in policy debates on refugees, who represent just half of all immigrants in the country. Indeed, there were relatively few debates in the Kenyan parliament during this time about immigration beyond the issue of refugees.
Over this period, Somali(s) (Wasomali) or Somalia were mentioned 297 times in Kenyan parliamentary debates, far more than South Sudan (112), for example. More than a third (37 percent) of these references concerned the Kenyan military deployment in Somalia, which has become increasingly controversial since Al-Shabaab killed roughly 200 Kenyan soldiers at an AMISOM base near Al Edde in January 2016. About 24 percent of the references discussed Somali refugees and the security risks posed by Somalis in Kenya, and another 10 percent focused on Kenyan Somalis, including citizenship questions related to their ancestry. Fewer than 6 percent of the references concerned economic connections with Somalis or Somalia. A few quotes highlight the emphasis on security in parliamentary discussions about Somalis:
The refugees have brought more problems to our people. Dadaab has been a kindergarten for terrorists. …We want the Somali Government to stabilise their country, so that these people can go back. …We are not guaranteed of our security with these refugees in Kenya. (Hon. Kamama Asman Abongutum, November 23, 2016)
…there is a big buffer zone between us and Tana River which is a forested area which is seriously infested with people of Somali origin. They are armed and are always attacking our people. (Hon. Marcus Mutua Muluvi, January 25, 2017)
The Al Shabaab are still roaming the streets of this nation. …there is a training camp in Eastleigh of the Al Shabaab that is ongoing. Those are the people who have been attacking our military camps in Somalia and our people at the borders. (Hon. Johana Ngeno Kipyegon, February 23, 2017)
Even as many parliamentarians spoke about Somalis as a security threat, a few challenged this assumption. Notably, one opposition member blamed Kenya’s own policies for the security issues associated with Somalis:
In Kenya, we have allowed refugees to be stuck in refugee camps. That is our encampment policy. You can see the security risk it has bred when three generations of people have lived in one place since 1990s. They are born in the camps and are basically imprisoned. That has brought about some of the unique security problems we have seen, apart from the fact that we share a border with Somalia. (Hon. Kenneth Odhiambo Okoth, June 7, 2017)
Kenya’s approach to Somali migration thus has been driven primarily by security concerns associated with the ongoing threat from Al-Shabaab. To be sure, economic and political competition have created a supportive context for restrictive policies, and normative commitments have constrained government behavior at times, but security considerations have dominated policy debates on this issue. The result has been a policy that differentiates among immigrants based on their country of origin, applying more restrictions to some than others. Somalis have been the primary target in recent years, as emphasized by the High Court ruling, but rhetoric and policies have changed over time in response to shifting security threats.
South Africa
Since the end of apartheid and the transition to a democratically-elected government in 1994, South Africa has become a popular destination for migrants from throughout the continent and the world. Although there is significant debate about the actual number of immigrants in a country where many cross the border without legal documentation or overstay their visas, the United Nations (2017) shows a total migrant stock in South Africa of more than four million, with the largest populations coming from Zimbabwe, Mozambique, Lesotho, Namibia, and the United Kingdom. In contrast to Kenya, where refugees represent nearly half of all immigrants, fewer than 7 percent of immigrants to South Africa are refugees or asylum seekers (UNHCR 2018). Nearly all of the roughly 31,000 Somalis in South Africa were either resettled there as refugees or sought asylum upon arrival; they represent 11 percent of the refugee population in the country (United Nations 2017; UNHCR 2018).
As the number of immigrants has increased, South Africa has experienced a rise in xenophobic sentiment that must be understood in the broader context of economic and political change (Landau Reference Landau2011; Neocosmos Reference Neocosmos2010; Nyamnjoh Reference Nyamnjoh2006; Steinberg Reference Steinberg2018). In a country where most people have only recently been granted full rights of citizenship, there is intense debate over the proper distribution of benefits that come from membership within the polity (Claassen Reference Claassen2016). The promises of the end of apartheid have not been realized, with millions of black South Africans still living in poverty and continued economic and residential segregation. Following earlier economic growth, a recession since 2008 has only exacerbated frustration. Under these conditions of inequality, many South African citizens are especially resentful of the growing number of immigrants from other African countries (Gordon Reference Gordon2015), some of whom have found economic success in South Africa. This includes the small but concentrated population of Somalis, who started arriving in the mid-1990s and gradually established enclaves in cities throughout the country (Steinberg Reference Steinberg2018), including the Mayfair neighborhood of Johannesburg that was formerly home to an established South Asian Muslim community (Ripero-Muñiz Reference Ripero-Muñiz2016, Reference Ripero-Muñiz2017; Thompson Reference Thompson2016). Working first as hawkers, often in poor areas considered by many South Africans to be dangerous, some Somalis eventually became successful shopkeepers and wholesalers (Thompson Reference Thompson2016; Steinberg Reference Steinberg2018), using their connections to build commercial networks throughout the country.
National politics have been dominated by the African National Congress, but many local races have been highly competitive, and the ruling party itself has been fraught with divisions. In this uncertain context, some high-profile politicians have lashed out at immigrants, using them as scapegoats for various social ills and deflecting blame for their own policy failures (Gordon Reference Gordon2018; Landau Reference Landau2010).Footnote 11 Elite rhetoric has legitimized underlying hostility against immigrants, which has periodically risen to the level of violence. In May 2008, anti-foreigner riots broke out near Johannesburg and spread to other parts of the country, resulting in the deaths of more than sixty people, mostly immigrants. These were neither the first nor the last violent attacks on immigrants in South Africa, though they attracted international attention to the problem of xenophobia there. With these complex economic and political dynamics fueling questions about who really “belongs” in the country (Mosselson Reference Mosselson2010; see also Crush Reference Crush2001; Gordon & Maharaj Reference Gordon and Maharaj2015; Landau Reference Landau2010, Reference Landau2011; Neocosmos Reference Neocosmos2010; Nyamnjoh Reference Nyamnjoh2006; Peberdy Reference Peberdy2001), polls show that a large majority of South Africans favor restrictive immigration policies (Crush & Pendleton Reference Crush and Pendleton2004; World Values Survey 2017).
To the extent that security enters the debate about immigration in South Africa, the focus is largely on crime. There is little threat of a military attack from a neighboring country, and an armed insurgency by immigrants seems unlikely, reducing any push for policies that differentiate among immigrant groups. Instead of being perceived as a national security threat, however, immigrants to South Africa are seen as a “criminal threat” (Crush & Peberdy Reference Crush and Peberdy2003). Indeed, many South Africans regard foreigners as a leading cause of high crime rates in the country (Landau Reference Landau2010). This view has been promoted by elite rhetoric (Landau Reference Landau2010; Neocosmos Reference Neocosmos2010) and reinforced by media coverage (Danso & McDonald Reference Danso and McDonald2001; McDonald & Jacobs Reference McDonald and Jacobs2005). In 2002, for example, the Director General of the Department of Home Affairs testified that “approximately 90% of foreign persons who are in the RSA [Republic of South Africa] with fraudulent documents …are involved in other crimes as well” (Masethla Reference Masethla2002).
In the face of such claims, immigration rights advocates have noted that many foreigners in South Africa are themselves victims of crime (Crush & Peberdy Reference Crush and Peberdy2003). Immigrant-owned businesses and shops in poor townships throughout South Africa have been attacked, looted, and burned (Steinberg Reference Steinberg2018; Thompson Reference Thompson2016). Even so, one study in a poor area of Cape Town shows that levels of violent crime against foreign shopkeepers (80 percent of whom in that area are Somali) are not significantly higher than against South African shopkeepers (Charman & Piper Reference Charman and Piper2012). Moreover, perpetrators of such crimes in poor areas are rarely brought to justice, regardless of whether the target is a foreigner or not (Steinberg Reference Steinberg2018). These dynamics suggest that discussions about immigrants need to be examined in the wider context of persistent crime and violence in the country (Charman & Piper Reference Charman and Piper2012). Regardless of the actual statistics, there is little doubt that concerns about crime have contributed to xenophobia in this context.
South Africa is party to many international human rights agreements and has one of the most progressive constitutions in the world (Sunstein Reference Sunstein2001), but its normative commitments have done little to restrain hostility toward foreigners. In 1999, the South African Human Rights Commission launched a “Roll Back Xenophobia” campaign in partnership with the National Consortium for Refugee Affairs and UNHCR. Three years later, after some training sessions and research studies, the campaign ended due to lack of funding (Southern African Migration Project 2008). As xenophobia escalated in the 2000s, researchers and activists produced countless analyses of the problem and its causes, though little was done to address underlying issues. In the run-up to the 2010 World Cup hosted by South Africa, senior officials sought to rein in anti-immigrant hostility to improve the country’s image, but these top-down efforts had minimal effect (Gordon Reference Gordon2017). Although South Africa’s normative commitments have not prevented xenophobia, they may have nonetheless kept the government from taking more drastic actions against immigrants.
Although refugees represent a small fraction of the total immigrant population in South Africa, textual analysis of parliamentary transcripts suggests that they have been the focus of disproportionate attention in policy debates. The South African Hansard Archive includes 113 transcripts of the National Assembly, the lower house of the South African Parliament (including joint sessions with the National Council of Provinces), from 2016 and 2017.Footnote 12 Parliamentary debate is conducted in English, though members occasionally use one of ten other official languages; when they do, the transcript typically includes an English translation. Over these two years, there were 452 references to migrants, refugees, foreign nationals, or foreigners, 155 of which (34 percent) were specific references to refugees. This is largely because of an extensive debate, especially in March 2017, about proposed amendments to the 1998 Refugees Act. Interestingly, much of this debate centered around the widespread concern that economic migrants were taking advantage of South Africa’s asylum policies and falsely claiming refugee status:
[Immigration] is placing a huge burden on South Africa, as asylum seekers and refugees have freedom of movement, can work and study in the country and are, in addition, able to access a range of social services. Large numbers of economic migrants arrive under the guise of refugees and lodge asylum applications, clogging an already over-burdened system. Irregular migration is largely impacting on the country’s economically depressed communities, which are affected by escalating rates of unemployment. Irregular migrants settle in these communities creating real and perceived competition for scarce resources, resulting in tension between host communities and non-nationals. (Hon. Cheryllyn Dudley, May 25, 2016, emphasis added)
Empirical evidence indicates that the majority of asylum seekers do not qualify. The presence of undocumented foreign nationals poses both economic and security threats to the country, which the government has to deal with. There are also challenges relating to legislation, and regulation of access to citizenship by foreign nationals. Therefore, there is a need to balance the inward flow of low-skilled labour to curtail the negative impact it has on domestic employment. Thus, the Bill seeks to limit the large numbers of undeserving asylum seekers who are, in fact, economic migrants, if not refugees, from justice administration in their own countries. (Hon. Donald M. Gumede, March 15, 2017, emphasis added)
Even in parliamentary debates that were nominally about refugee policy, the focus was still on the negative economic effects of migrants. Most Somalis are legally recognized as refugees, or are in the process of seeking asylum, but this does not stop some South Africans from (incorrectly) accusing them of violating their legal status by conducting business (Charman & Piper Reference Charman and Piper2012). In the halls of Parliament, though, Somalis have received very little specific attention, perhaps due to their small numbers. In 2016 and 2017, Somalis or Somalia were mentioned only 13 times in parliamentary transcripts. All three references in 2016 were to the need for Somali interpreters trained in legal jargon to assist in court cases. In 2017, most of the references came when the National Assembly extended condolences to the people of Somalia after several terrorist attacks in that country. In stark contrast to Kenya, then, Somali refugees in South Africa are not the focus of special attention among members of parliament.
Overall, South Africa’s immigration policy is shaped primarily by economic competition and political calculations, which have fueled demands for more restrictive measures. The government has focused on the economic dimension of immigration by restricting access to work permits, requiring substantial domestic ownership, discouraging labor migration, and defining most foreigners as migrants instead of refugees, making them eligible for deportation. As in other countries, restrictive policies have focused on the economic sectors in which immigrants are predominant (Honig Reference Honig2016). Within this context, Somalis are seen primarily as an economic threat. Many South Africans resent Somalis for their success in building business networks that have thrived even as unemployment rates remain persistently high (Thompson Reference Thompson2016). This dynamic is especially apparent in Mayfair, where Somali-owned businesses dominate the wholesale and merchandizing sectors. But Somalis are not alone; Nigerians, Mozambicans, Zimbabweans, Angolans, and others have been the target of resentment and sporadic violence. The South African case thus provides evidence of economic concerns driving a general policy in which restrictive measures apply to immigrants broadly regardless of their country of origin.
Tanzania
For decades, Tanzania was recognized for its open-door policy (Whitaker Reference Whitaker1999). Refugees received tracts of land to farm food and cash crops; over time, many were granted citizenship. This changed in the 1990s, when the refugee population grew to more than one million with the escalation of conflicts in Rwanda, Burundi, and Congo. The government at that time confined refugees to designated camps, banned them from agricultural production, and closed its western border to prevent further influxes. In an unprecedented move in 1996, the Tanzanian army forcibly repatriated more than 500,000 Rwandans (Whitaker Reference Whitaker, Steiner, Gibney and Loescher2003). Although the number of migrants has dropped significantly, Tanzania continues to host more than 400,000 refugees and asylum seekers (UNHCR 2018). These represent the vast majority (84 percent) of an estimated total migrant stock in Tanzania of nearly half a million people; the most important countries of origin are Burundi, Congo-Kinshasa, Kenya, Congo-Brazzaville, and Mozambique, with Somalia tenth on the list (United Nations 2017).
As in other countries, the shift in Tanzania’s immigration policy occurred at a time of economic and political liberalization that generated both competition and uncertainty. After the implementation of structural adjustment reforms, and especially with the dismantling of marketing boards, the government no longer benefited as much from the sale of cash crops produced by refugees (Daley Reference Daley1992; Veney Reference Veney2007; Whitaker Reference Whitaker1999). A drop in foreign aid in the early 1990s also contributed to widespread economic frustration. With the legalization of opposition political parties in 1992, some politicians sought to garner support by blaming refugees for the country’s economic challenges (Rutinwa Reference Rutinwa1996; Veney Reference Veney2007). In the context of these various changes, there was growing support among many Tanzanians for more restrictive immigration policies.
Just as important as these economic and political dynamics, though, were security concerns associated with the presence of more than one million refugees from neighboring countries. The temporary closure of the western border in 1995 was motivated in part by rising tensions with Burundi and a renewed influx of migrants from that country, though such concerns were amplified in the context of a heated political campaign leading up to Tanzania’s first multiparty presidential election that year. In December 1996, the decision to forcibly repatriate half a million Rwandans was clearly a security move; Rwandan officials reportedly threatened to attack the refugee camps in Tanzania as they had done to the camps in Zaire/Congo just a few weeks earlier (Whitaker Reference Whitaker, Steiner, Gibney and Loescher2003).Footnote 13 To avoid being drawn into a broader regional war (one that ultimately engulfed Congo for more than a decade), the Tanzanian government opted to send home the Rwandan refugees, but not refugees from other countries. As seen in the case of Kenya, then, security concerns led to a policy that differentiated among refugee groups.
The forced repatriation of Rwandans and restrictive measures toward other refugees still living in Tanzania prompted criticism from human rights activists. Concerned about the country’s longstanding reputation for honoring international legal obligations with respect to refugees, officials argued that Rwanda was safe for repatriation and compared their actions to similarly restrictive immigration policies in the United States and Europe (Whitaker Reference Whitaker1999). They also suggested that international donors were partly responsible for the repatriation because of significant drops in funding levels for the refugee camps (Whitaker Reference Whitaker2008a). The fact that Tanzanian officials felt the need to justify the shift in refugee policy implies that normative commitments still had some influence, but other priorities clearly took precedence.
In this context, the influx of more than 4,000 refugees from Somalia in the mid-1990s provided an opportunity for the Tanzanian government to demonstrate that it was still committed to international norms of refugee protection. Even as it was closing the border and forcibly repatriating refugees out west, the government admitted Somali refugees and hosted them in the Mkuyu camp in northeastern Tanzania. Some arrived on their own by boat, having moved further south after passing through Mombasa, while others were formally resettled to Tanzania from Dadaab (Bannon & Wolfcarius Reference Bannon and Wolfcarius2009; Lehman & Eno Reference Lehman and Eno2003). In 2003, there were more than 3,000 Somali refugees remaining in Tanzania.Footnote 14 These were transferred to a more spacious settlement known as Chogo, complete with school, health clinic, and market, where each family received several acres of land to farm (Bannon & Wolfcarius Reference Bannon and Wolfcarius2009). Soon thereafter, the Tanzanian government started accepting applications for citizenship from these Somali refugees. The process was put on hold in 2010 over allegations of corruption, but was eventually completed in 2014, with nearly 3,000 Somalis having received certificates of Tanzanian naturalization (Kabendera Reference Kabendera2014).
Tanzania’s extension of citizenship seems to stand in marked contrast to the ways in which other governments in Africa, including Kenya and South Africa, have handled Somali migration. The important twist in this case, however, is the historical connection that these particular Somalis had to Tanzania. In the eighteenth century, when much of the East African coast was controlled by the Sultanate of Zanzibar, slaves were captured in present-day Tanzania and Mozambique and taken to what is now Somalia (Lehman & Eno Reference Lehman and Eno2003). Although slavery has long since ended, the descendants of these slaves, known as Somali Bantu, continued to face marginalization and persecution due to their cultural and physical differences; many spoke the languages of their ancestors and did not integrate into Somali society (Lehman & Eno Reference Lehman and Eno2003). When violence broke out and people fled Somalia in the 1990s, a small fraction of refugees traced their ancestry to ethnic groups in Tanzania (Besteman Reference Besteman2012; Menkhaus Reference Menkhaus, Hoehne and Luling2010). Indeed, some Somali refugees spoke Kizigua, a language of northeastern Tanzania, when they arrived (Bannon & Wolfcarius Reference Bannon and Wolfcarius2009). This historical connection generated public sympathy and support for the idea of extending citizenship to this group (UNHCR Tanzania 2010). Their limited political engagement in Somalia also mitigated against any potential security concerns. The Tanzanian government thus was able to naturalize this group of Somali refugees with little fear of political backlash. Importantly, though, the Somali Bantu population was not the only group of refugees naturalized by Tanzania in recent years. Around the same time, nearly 170,000 Burundian refugees who had been living in Tanzania since 1972 also were granted citizenship (Kabendera Reference Kabendera2014).
The Parliament of Tanzania has given ample attention to refugees and immigration issues more broadly. The country’s Hansard Archive includes 166 transcripts of proceedings in the unicameral parliament from 2016 and 2017.Footnote 15 Parliamentary business is generally conducted in Swahili, though members occasionally use English. During this two-year period, a total of 1,230 references to refugee(s) (mkimbizi/wakimbizi), migrant(s) (mhamiaji/wahamiaji), or foreign nationals (wageni) appear in the parliamentary record. Of these, approximately 42 percent referenced refugees, a far smaller portion than in Kenya, despite the much higher proportion of refugees as a percentage of all immigrants. More frequently (48 percent of instances), Tanzanian parliamentarians used the term wageni to refer to foreigners.
To some extent, this difference reflects the fact that refugees were not the main focus of parliamentary discussions about immigration in Tanzania during this period; significant attention was given to issues regarding work permits for foreign nationals and economic competition between citizens and foreigners, among other topics. On another level, though, the frequent use of the term wageni is interesting because of the multiple English translations of this Swahili word. While wageni is used for “foreign nationals” or “foreigners,” it also has a broader connotation as “guests.” Indeed, each session of parliament devotes several minutes to welcoming various groups of wageni (guests) in the gallery that day. Each use of the word wageni in the transcripts was thus checked individually and all references to parliamentary guests were removed from the above tally. Even with these instances dropped, wageni were mentioned more often than wakimbizi (refugees), and the term was frequently used in parliamentary discussions about immigration.Footnote 16 In a country where most immigrants are refugees, and where many refugees have been granted citizenship over the years, the frequent use of a word in Swahili that translates as both “foreigner” and “guest” seems fitting.
Given the relative numbers of refugees from each country, it is perhaps unsurprising that Tanzanian MPs were much more likely to mention Burundi (528 times) than Somalia (65 times) during this period. To the extent that there was discussion about Somalia, much of it focused on continued insecurity in that country and the risk of piracy (maharamia) off the coast (see Samatar, Lindberg, & Mahayni Reference Samatar, Lindberg and Mahayni2010; Weldemichael Reference Weldemichael2019). Although this was several years after the mass naturalization of Somali Bantu refugees, the Minister of Home Affairs announced on May 9, 2017, that two additional Somali nationals had been granted citizenship. During this period, Tanzanian officials also expressed concern about the number of migrants passing through the country on their way to destinations further south:
In that area, Tanzania is bordered by the countries of Kenya and Uganda. The security situation along this 1,221-kilometer border was calm despite the problem of an influx of illegal immigrants from Ethiopia, Eritrea, and Somalia who infiltrate the country on their way to areas in the southern part of Africa.Footnote 17 (Hon. Hussein Mwinyi, Minister of Defence and National Service, May 16, 2017)
Generally speaking, though, the issue of Somali migration did not attract a significant amount of attention in policy discussions in the Tanzanian parliament.
Overall, Tanzania’s immigration policy has been shaped by a combination of factors, including security concerns related to large-scale influxes from Rwanda and Burundi and broader conflict dynamics in the region. On the issue of Somali migration in particular, however, Tanzanian policy has been driven mainly by a normative commitment to the integration of a vulnerable population with a unique historical connection to the country. Economic and political competition in recent years have created a fertile environment for the circulation of anti-immigrant rhetoric, but there was widespread sympathy among Tanzanians for the Somali Bantu population and support for their naturalization. The extension of citizenship around the same time to a much larger population of Burundi refugees also was generally accepted given the length of time that the group had lived in Tanzania, even as policymakers remained concerned about more recent arrivals from Burundi and other countries. These patterns suggest that both security considerations and normative commitments influence immigration policy in the African context, particularly when a large portion of immigrants originate as refugees.
Conclusion
This article has shown that host governments in Kenya, South Africa, and Tanzania responded differently to Somali refugees, depending on the relative balance of economic considerations, political calculations, security threats, and normative commitments. Where security concerns are dominant, as in Kenya, policies have differentiated between Somalis and other groups of immigrants based on their perceived level of risk.Footnote 18 Where economic competition and political calculations are primary drivers, as in South Africa, restrictive measures have applied broadly across Somalis and other immigrant groups, targeting economic sectors in which immigrants enjoy comparative strength. And in situations where there have been normative pressures to provide a more permanent solution for refugees, as in Tanzania, citizenship was extended to Somali Bantu refugees even as restrictions were applied to other refugees in the country. Thus, in many ways, Somalis have been defined differently depending on the host country context: as a security threat in Kenya, as an economic threat in South Africa, and, for a small and very specific group of Somali Bantu refugees, as long-lost compatriots in Tanzania.
Although these factors have shaped host government policies toward Somalis in recent years, their relative importance can change substantially over time. This is clearly illustrated in the case of Kenya, where restrictions on Somali refugees have increased considerably since Kenya’s deployment of troops to Somalia in 2011 and the resulting effort by Al-Shabaab to extend the war into Kenya itself (Anderson & McKnight Reference Anderson and McKnight2015). In Tanzania, security concerns were heightened in the 1990s due to the massive influx of refugees from Rwanda and Burundi and the emergence of a broader regional war. That sense of threat has declined more recently, though, as reflected in part by the limited focus on refugees in parliamentary debates. Lingering economic problems and a growing sense of inequality in South Africa have amplified tensions in that context and generated additional pressure for restrictive immigration policies. These cases suggest that economic factors may fuel hostility toward immigrants under conditions of recession, while security concerns may be primed when a host country is at war.
This research has focused on responses to Somali refugees in three host countries in Africa, but it has broader implications for scholarship on comparative immigration policy. Many studies have portrayed immigration policies on a basic continuum from restrictive to permissive (Givens & Luedtke Reference Givens and Luedtke2005; Timmer & Williams Reference Timmer and Williams1998). As shown in this article, however, immigration policy can be permissive in some areas and restrictive in others, and/or can apply differently across immigrant groups. In Kenya, as emphasized by the High Court ruling, recent crackdowns have targeted Somali refugees due to security concerns related to Al-Shabaab. Similar considerations prompted the Tanzanian government to close its border and forcibly repatriate refugees, even as it extended citizenship to many refugees from Somalia and Burundi. Immigration policy thus consists of many different measures that may at times contradict one another (Timmer & Williams Reference Timmer and Williams1998). In highlighting its multiple dimensions, therefore, this article moves beyond open-versus-closed dichotomies and explores a broader range of policy configurations available to host country governments (Honig Reference Honig2016).
Finally, this research suggests that scholars need to pay more attention to the different dimensions of a country’s immigration policy, particularly the extent to which it differentiates among immigrant groups based on their country of origin and/or applies to some parts of the immigration process (such as admission) more than others (such as integration). These are not simply theoretical distinctions; they have practical implications. In situations where security concerns prompt harsh limitations on refugees, for example, international actors should seek to reduce security threats for the host country while strengthening refugee protection. Similarly, in contexts where xenophobia and restrictive policies are driven by a sense of economic threat, emphasis should be placed on addressing underlying issues of poverty and inequality. By developing a more nuanced understanding of the factors that shape immigration policy in Africa and the different forms such policies take, the international community as a whole can work toward better addressing the concerns of both immigrants and hosts.
Acknowledgments
The author would like to thank Ahmed Ibrahim, Aditi Malik, and Cori Wielenga for organizing a 2016 workshop in Pretoria on Somali migration and for moving the papers presented there toward this special issue. Their persistence, support, and friendship are greatly appreciated. The author is also grateful to the ASR editors and anonymous reviewers for their detailed comments and suggestions.
