Introduction

Movements of people from Africa to Europe have dominated mainstream media headlines and political discourses in recent decades. Migration flows between African societies have received considerably less attention, despite being much larger in scale and at least as consequential for politics on the continent. Likewise, debates about restrictive migration policies, eroding asylum rights, and fortifications along the borderlands of Europe, Australia, and North America (Jones 2017) have concealed the fact that migration within Africa has often been met with similar and ever-expanding institutional measures of regulation, control, and containment from the colonial period until today. East Africa in particular has been at the center of ebbs and flows of numerous migration movements that, during the 1990s, reached a peak through the forced displacement of hundreds of thousands from Somalia, the Democratic Republic of Congo (DRC), Ethiopia, Sudan, Rwanda, and Burundi into neighboring countries, in particular Kenya and Tanzania. By 2000, Tanzania was hosting around 702,000 refugees and asylum seekers, while Kenya was home to over 219,000 (UNHCR 2004). Since then, Kenya's refugee population has soared to 490,000, while 337,000 remain in Tanzania today.¹

Colonially defined borders had long-lasting effects on these contemporary movements. Despite being often closely related to the citizens of their countries of asylum—through shared culture, identity, and language—displaced people were automatically classified as noncitizens, “aliens,” and foreign “others” (Daley 2013). Encarnación Gutiérrez Rodríguez argues that this “dichotomy between citizens and migrants is embedded in a racializing logic produced within social relations shaped by the enduring effects of colonial epistemic power” (2018: 25). Although the aftereffects of empire may indeed be most noticeable in the encounters of black and brown migrants with white immigration bureaucracies that regulate and monitor their circulation in the global North (Browne 2015; Mayblin 2017; Wardle and Obermuller 2019), many state institutions and migration policies in Africa are anything but free from such discrimination, racialization, and systemic exclusion. Tayyab Mahmud contends that living with the specter of the migrant “is to live with desires and anxieties of the state and the nation. It is also to live with the heritage and genealogies of empire and imperialism” (1997: 633). Instead of transforming the colonial modes of managing and hierarchizing noncitizens, many postcolonial African states actually perpetuated parochial and racialized categories of mobility, despite at times experimenting with progressive politics. Even though Tanzania and Kenya have gone through temporary periods of “open door policies” toward refugees, exiles, and labor migrants (Chaulia 2003; Verdirame and Harrell-Bond 2005), their current migration regimes suggest a continuing preoccupation with the rigid categories of race, ethnicity, and belonging for managing mobile Africans. In August 2018, Kenya's Department of Immigration Services launched a new hotline for national citizens to report “suspicious foreigners” in an orchestrated attempt to crack down on “illegal immigration” (Muraya 2018). Half a year later, Tanzanian state authorities conducted house-to-house searches for “illegal aliens” in the country's western regions, bordering the DRC, Rwanda, Burundi, and Uganda, and further announced a comprehensive “verification” of citizenship for local residents in the months that followed (Karashani 2019). In recent years, both countries have experienced a notable surge in popular and, not least, government-led anti-refugee rhetoric that has often adopted a racialized and colonial language that positions migrants in opposition to a more “civilized” host state. Refugees, asylum seekers, and other migrants are on the one hand typically held responsible for environmental degradation, spreading diseases, and overstretching local infrastructure (Aminzade 2013: 305), or fostering political instability and terrorism on the other (Mwangi 2019; Brankamp 2020). However, such popular fears and resentments in the two countries are embedded in, and inextricably linked to, longer histories of state institutions differentiating between more or less desirable migrant bodies along similar lines. In this article we therefore offer a critical analysis of the (post)colonial migration regimes in Kenya and Tanzania, and trace enduring processes of differential valuation of human worth, economic benefits, and racialization that have been part and parcel of government policies in the colonial and postcolonial periods. Migration regimes are not simply “formal expression[s] of collectively determined objectives concerning the regulation of the movement of people” (Cornelissen 2009: 355), but also encompass certain legal and discursive productions of migrants as either unproductive, perilous, or profitable subjects. As Cassandra Veney notes, migration regimes in the two neighboring countries are especially noteworthy, because they share a history of British colonial rule, both transitioned into postcolonial one-party states, yet chose diverging developmental paths after independence (2007: 2). Edward Mogire observes that their respective geostrategic location has rendered understanding the migration dynamics in the two countries more urgent (2009: 16). However, unlike many other countries in the region, Kenya and Tanzania have experienced long stretches of relative political stability and have therefore mostly received rather than generated refugees, asylum seekers, or migrants themselves. Most recently, this has given more impetus to institutional efforts for identifying and monitoring Africans on the move by capturing their biometrical data under the guise of “enhancing” migration management and “empowering” migrants themselves (IOM 2018). The Kenyan and Tanzanian governments have been actively soliciting funds and expertise from donors and agencies, such as the International Organization for Migration (IOM) and the United Nations High Commissioner for Refugees (UNHCR), to reiterate aspirations of “orderly” global migration and tailor their enforcement capacities to coping with bodies that are deemed irregular, illegitimate, or simply “out of place.” We argue that this recent proliferation of border technologies in Africa, which rely on the “machine-readable body” (Ploeg and Sprenkels 2011), are but the latest stage in longer histories of differential migration governance on the continent that continue to impact which bodies are expected (and allowed) to be mobile while others are not. Since formal colonial rule, Africa has been coded as a “racially identified space” (Ford 1994: 1845) to which Africans should ideally be confined. However, even within the continent itself, movements of colonized subjects were geographically restricted through governmental delineations of reserves, estates, frontiers, colonial territories, and corridors. This ordering of Africans through racial and ethnic classifications, and the fixing of them to both imperial spheres of influence and ethnic territories, denied their right to have multiple identities and to choose their autonomous mobility and place of dwelling (Weitzberg 2017). No such restrictions were placed on bodies racialized as white, who could explore, conquer, and, above all, move unimpededly. The administrative bifurcation of colonial African states into a civic and a customary domain (Mamdani 1996) was precisely about authorizing or denying this very mobility to the colonized. Colonial biopolitics dictated that nonwhite bodies only move at the behest of capital, colonial authorities, and certainly never on their own volition.

This article builds on previous scholarly assertions of the need to analyze postcolonial histories of migration and asylum regimes in order to make sense of continuities in the present (Marfleet 2007; Mayblin 2017; Tošić and Lems 2019). This is particularly crucial for work on refugees and forced migration, in which “history has always been notable by its absence” (Marfleet 2007: 136). Focusing on Kenya and Tanzania, we contend that a number of contemporary laws and institutions that organize land, ethnic territoriality, migration, and asylum in fact perpetuate colonial legacies of differential mobility, though often in modified form. The first section hence traces the management of labor migrations in colonial and postcolonial Kenya and Tanzania, emphasizing the capitalist political economy that determined the (un)desirability of certain kinds of Africans moving between territories. The second section expands this discussion to refugees and displaced people seeking asylum in neighboring countries, but who are also monitored, excluded, and securitized as dangerous, racialized subjects in need of spatial fixation. The third section draws on Francis Nyamnjoh's (2013, 2017) critique of dominant approaches to studying African migrations and points out the limits of controlling migration on the continent and the emergence of more convivial social worlds that found precedence in Tanzania's enduring support for African liberation fighters between 1963 and 1994.

Contributing to literature on migration and mobility in Africa (Bakewell and Jónsson 2013; Bakewell and Landau 2018; Nyamnjoh 2013), this article offers a concrete analysis of migration regimes in Kenya and Tanzania that combine ethnoterritorial notions of citizenship, national security concerns, and policies of global migration governance, all of which ultimately perpetuate colonial dispositions of race, territory, and graduated mobility. It concludes that studies of contemporary migration in Africa should be attentive to the coloniality of contemporary policies and practices and should seek solutions that draw from Africans’ histories of conviviality.

African Bodies as Labor: Labor Migration in the Colonial and Postcolonial Eras

Our analysis of colonial and postcolonial labor regimes in Tanzania and Kenya demonstrates how colonial conceptions and management of African bodies as labor has crucially informed postcolonial understandings of belonging and mobility. We contend that colonial views on the need to make “out-of-place” Africans productive equally permeates contemporary discourses about refugee economies and the utilization of refugee labor (see Betts et al. 2016; see also Farah, this volume). Colonial authorities utilized racialized and ethnicized ideologies to categorize African people according to their propensity for different types of labor, especially their ability to perform more arduous, harmful, and militaristic tasks in the service of empire. The legacies of colonial regulations and policies, aimed at instituting tightly controlled migrant labor regimes and managing African presence in spaces deemed “non-native,” continue to reverberate with postcolonial policies and political discourses. Colonial labor controls drew on the racialized myth of the “lazy” (unproductive and idle) “native” who had to be forced out of “his” lethargy. Individualized wage labor was thus understood as a powerful tool to break especially African men from what were seen as “primitive” kinship ties and social solidarities (Orde-Browne 1946). During the 1920s, colonial authorities in the Tanganyika Territory² introduced legislation that targeted men who were deemed “unproductive.” The 1923 Destitute Persons Ordinance thereby redefined who was considered destitute in the eyes of the colonial state—“any person without employment and unable to show that he has visible and sufficient means of subsistence”—and compelled them to either find employment or face imprisonment and forced labor on behalf of the state. Mobilizing Africans to work in colonial enterprises thus led to the emergence of further bureaucratic differentiations between so-called resident natives, non-natives, and foreigners.³ The introduction of identity documentations aided this categorization process and enabled employers to manage African workers, but it ultimately had broader societal implications, including the proliferation of Eurocentric naming patterns or the hardening of ethnic territorial boundaries. Despite imposed colonial regulations, Africans often remained agential in the migration process and instrumentally used the categorization of racialized laboring bodies to navigate employment, both within and across colonial borders.

Kenya and Tanzania had slightly different histories of managing circulations of labor and ideas of belonging. Land in colonial Tanganyika was relatively plentiful and the population density low, creating a greater reliance on attracting an African labor force from both residents of the area as well as neighboring colonial territories. Labor used on the plantations in Kenya, however, was largely sourced from Africans who had previously been displaced from their land and forced into overcrowded reserves, thus creating a “pool of cheap labour” for the white settler economy (Overton 1990: 163). This alienation of land from the local African communities, and their reconstruction as mere “squatters” on Kenya's European-owned farms, made the colony's labor regime more punitive than in Tanzania, where many workers were imported into the area without having prior claims to the land they tilled. Kenya's Master and Servants Ordinance of 1906 introduced restrictive labor laws that benefited European employers through limiting their responsibility toward African workers, while leaving the employees with few protections of their own. For the colonial establishment, “labour at a ‘primitive’ stage of development was thought to require ‘primitive’ forms of labour law” (Anderson 2000: 461). The Resident Native (Squatters) Ordinance of 1918 further entrenched tenant duties of African populations working on European-occupied land and penalized failures to provide labor in accordance with existing legislation. The Employment of Natives Law, which included a series of seven ordinances passed between 1910 and 1920, further specified the terms of labor contracts, usually to the detriment of the African employees.

After World War II, a new registration system was introduced that required every adult African male to carry a certificate stamped by the employer, containing information on the person's “name, father, district, tribe, location sublocation, group, clan, circumcision age, date, signature of official issuing with place for empowers’ signature and fingerprints of all ten fingers” (Orde-Browne 1946: 150). The increasingly draconian nature of these laws contributed to no small extent to the rise of the Land and Freedom Army (so-called Mau Mau) during the 1950s (Berman 1990). Migrants from other colonial territories settled in Kenya's main city of Nairobi, but it is not known exactly how many plantation workers were from outside the colony. The shortage of labor during the 1940s, as settlers increased production as part of the war effort, saw tea plantations in Kenya requesting permission to recruit labor from Ruanda-Urundi (present-day Rwanda and Burundi) (Orde-Browne 1946).

Labor migration in the whole of East Africa was extensive during the colonial period. Audrey Richards's (1952) study of migrants in Buganda (present-day Uganda) identified workers who came from modern-day Burundi and Rwanda, Tanzania, the Congo, Sudan, and Kenya to work for African farmers as well as European and Asian enterprises, and draws on a 1948 survey that counted foreign workers from 21 ethnic groups. The colonial demarcation of “ethnic territories” fixed African subjects in place, but since control of labor was intrinsic to the development of commodity production, labor scarcity meant that the colonial authorities sought to mobilize Africans (largely men) for either forced labor or as regular wage laborers, while controlling the flow and direction of these different migrations. Governmental mechanisms, including punitive hut and poll tax cash exactions as well as forced recruitment for the construction of roads and administrative buildings, also led to a growth of voluntary movements and direct enrollment of men to work in mines and on settler plantations. Nonetheless, Africans often avoided colonial recruiters and migrated independently.

Labor departments were established and regulations on recruitment and employment conditions were implemented to provide a steady and stable labor force to mines and settler plantations. In Tanganyika, Joe Lugalla (1989: 147) notes that Pass laws were introduced for every African (over 18 years) in 1913, requiring them “to have a special card which specified his or her engagement locality.” Various Masters and Servants Ordinances from 1923 to 1931 governed the recruitment, employment, and conditions of workers. In 1926, the Tanganyika administration set up a labor department specifically to recruit labor for government work and for the sisal and cotton estates. In some parts of the territory labor recruitment had begun as early as 1907 for work on caravans and in railway construction. Laurent Sago (1983) notes that in the Buha area of western Tanganyika local chiefs were instrumental in persuading villagers to migrate. From the 1930s to the 1960s the Waha were among the most prominent migrant labor groups (Sago 1983). After the depression of the 1930s the colonial authorities saw only active labor recruiting as solving the problem of wage labor scarcity. The formation of SILABU (Labour Bureau of the Tanganyika Sisal Growers Association) in 1944 and later the Labour Supply Corporation signified a deliberate attempt to capture labor, especially what was termed “distance labor.”

Long-distance migration between colonial states was common, as people fled forced labor regimes, but also sought work to pay taxes. In Tanganyika, migrants from Belgian (Rwanda and Burundi) and Portuguese (Mozambique) territories formed the most common group of “distance labor” and were provided with free housing. Ethnic and racial stereotyping and hierarchies meant that workers were defined by their aptitude for hard work. The Hutus from Burundi were stigmatized as “dirty” but also hard workers who could do the most arduous tasks on plantations. Hutus escaping racial stereotyping in Belgian colonial trusteeships found that these representations followed them into British colonial territories.

From the 1920s, “distance labor” from Ruanda-Urundi contributed significantly to the labor market in Tanganyika, Uganda, and the Belgian Congo. In 1935, 50 percent of the 2,361 workers recruited in the Kigoma area of Tanganyika were from Belgian territory (Western Province, Annual Report, 1942). However, the exact magnitude of the labor force migrating to Tanganyika was difficult to determine due to the uncontrolled nature of the flow and confusion with locals from the same ethnic group that straddled the border region. By 1946 the demand for labor led the colonial labor advisor Major G. St. J. Orde-Browne to recommend the channeling of migrants from Ruanda-Urundi into well-defined routes, and their maintenance for two weeks in reception camps with medical facilities. Reception centers were already established at the main crossing points into Uganda. This was achieved with the cooperation of the Belgian colonial authorities (Manpower Committee 1951). By 1955 labor recruiters from British territories were actively engaged in Ruanda-Urundi. By 1950, the Belgians estimated that over 675,000 migrants from Belgian territory were in Uganda and Tanganyika, of which 157,000 were in Tanganyika (UN Trusteeship Council 1950). Between the 1948 -1957 census years the Barundi recorded the highest percentage increase (35.3) of any group in Tanganyika. The sisal estates of Tanga, Coast, and Dar es Salaam districts were the main destinations.

From the 1920s to the 1960s, the sisal industry was the largest employer in Tanganyika, with over 386,000 workers in 1956, of which 52,000 were from other territories (Guillebaud 1958). Men were contracted for specified tasks or periods of employment, often 18 months to three years, and were expected to return to their home once the contract ended. This system, known as Kipande (piecework), was harsh. Working conditions on the estates were generally poor and resulted in high labor turnover and absenteeism. Workers were penalized and subjected to fines, corporal punishment, or imprisonment if they failed to complete their contracts. Under such terms of employment, desertion was the most popular form of resistance (Raikes 1978).

After Tanganyika's independence in 1961, the sisal plantations continued to be the major employer of immigrant labor. However, the collapse in the price of sisal, mechanization, stabilization of the workforce, and immigration restrictions effectively reduced migrant labor. The socialist ideology of the independent Tanzanian state had an adverse impact on the wage labor market, such that by the 1967 census, the Barundi population in Tanzania had declined to 114,605 (Egero 1979). As Philip Raikes (1978: 288) points out, the postcolonial state “officially disapproved” of labor hiring, as being “inconsistent with its policies.” Even in the 1980s, the working conditions of the migrants from Rwanda and Burundi appeared to have improved very little since the 1920s. Raikes (1978) notes that Banyarwanda and Barundi, who were normally denied access to land in Tanzania due to fear of them settling permanently because of land shortages in their countries of origin, were among the worst paid and most badly treated labor force in the country.

Little information is available on the settlement and social integration of labor migrants into host societies in the region except the research carried out by Richards (1952) in the former kingdom of Buganda in Uganda. In Buganda ethnic hierarchization placed migrants from Rwanda and Burundi at the bottom of the social order. In Tanganyika, the colonial state attempted to improve conditions on sisal plantations so that workers could settle, but to no avail. As the workers were predominantly young men, they did sometimes settle in the territory and gained citizenship at independence. These migrants virtually merged into the host population. Recent contestations over citizenship of people considered Tanzanian tend to target the descendants of colonial labor migrants.

In the postcolonial period, gender differences did not play a significant role in Kenya and Tanzania's approach to migrant labor, except with respect to women's marital status. During the colonial period, while recruitment focused on men, wives migrating with husbands were seen as a way of stabilizing the workforce at its destination. Single women (unmarried and widowed) were discouraged from migrating, but many fled patriarchal control in rural areas, moving to cities independently. Despite single women in towns being labeled prostitutes, these women were able to become traders and shop/bar owners—providing services for men (Stitcher 1982). Using accusations of prostitution to control women's mobility was another practice that persisted well beyond the ending of formal colonial rule.

African mobility was encouraged only to colonial spaces of capital accumulation. Colonial authorities also attempted to control African migration to urban centers, where African men were needed for menial roles within the colonial urban economy. In 1944, the colonial state in Tanganyika enacted the Townships (Removal of Undesirable Persons) Ordinance, which empowered colonial district commissioners “to remove (via a ‘Removal Order’) undesirable persons in the town area under [their] jurisdiction” (Lugalla 1989: 134). From that time and well into the independence period, unemployed men were rounded up and returned to rural areas in both Kenya and Tanzania (Lugalla 1989; Stitcher 1982). In addition, Tanzania introduced a series of acts to address “idle” citizens. In 1976, the government launched Operation Kila Mtu Afanye Kazi, and Tanzania's Human Resource Deployment Act of 1983 (popularly known as the Nguvu Kazi Act) empowered authorities to register all employees and to round up unemployed men in Dar es Salaam and repatriate them to rural areas. The imposition of economic liberalization in the 1980s and 1990s and the subsequent growth in unemployment and the size of the urban poor made it difficult for states to distinguish between the economic situation of long-term residents and recent rural-to-urban migrants. Consequently, forced repatriation was not politically feasible amid a decline in states’ control over the labor force.

The formation of the East African Community in 1999 and the adoption of its Common Market Protocol in 2010 indicated a coordinated effort by postcolonial states to enable and further manage labor migration within the region. Articles 7, 9, and 10 of the protocol provide for the movement of persons, travel documents, and the free movement of workers, respectively. In terms of migrant workers Burundians remain the largest group within the region, and Kenya and Uganda are the two major destinations for international migrants (IOM 2017). Admittedly, the out-migration from Burundi reflects the political instability in the state since the 2015 elections and the country's dire economic situation. While some of these migrants could be classified as refugees, the international community's use of the term “mixed migrations” serves to include the variety of reasons for which people migrate.

Colonial Bodies “Out of Place”: Refugees, Exiles, and Empire

Contemporary refugee regimes in Tanzania and Kenya have emerged from overlapping legacies of colonial laws and regulatory frameworks, international and regional legislations introduced after independence, and the more recent national politics of mobility, each of which displayed a noticeable concern for closely managing the belonging, bodies, and (im)mobilities of Africans on the move. While the British colonial authorities sought to control labor migration through a mixture of force, incentivization, and deterrence, refugees and asylum seekers posed an even more serious challenge for governmental regulation and a colonial will-to-order, because of the per se uncontrolled and involuntary nature of their migration.

Europe's colonial expansion in Africa was among the main driving forces behind large-scale displacements, both within and across colonial boundaries (Gatrell 2013: 226). Although populations who would today be legally classified as “refugees” historically entered Kenya and Tanzania, it was not until the postwar period in the 1940s that their position began to be increasingly codified in law. The dispersal of Polish refugees across Tanganyika, Kenya, and Uganda after 1942, for instance, precipitated racialized anxieties for the British authorities, which tried to spatially segregate white arrivals from the African residents around them (Piotrowski 2007). British Tanganyika's Defence Regulations Act of 1946 and the Refugee (Control) and (Expulsion) Ordinance No. 3 of 1949 were therefore aimed primarily at governing refugees of European origin who had found temporary asylum in the region. The first gave the colonial governor sweeping powers to expel refugees from Tanganyika's territory, while the second centered on restricting the movement of foreign exiles domestically. Article 7 of the Defence Regulations Act defined a refugee as a “person who entered any part of East Africa during the war in pursuance of an arrangement made by any government in East Africa for the reception of persons evacuated from war areas, and has been permitted to enter the territory, without observance of the immigration laws.” It was thereby officially stipulated for the first time that refugees had to reside in designated areas, and the administrative post of “camp commandant” was established specifically for their management and spatial control. It was not long until this was again put to the test with the arrival of Tutsi refugees following their political oppression in Rwanda in 1959 (Long 2012).

Colonial Kenya had no specific legislation pertaining to refugees, although the British colonial authorities instead passed a number of laws that could restrict the movements of “resident” and “nonresident” Africans, which included those who would in retrospect qualify as “refugees.” The 1925 Vagrancy Ordinance generally impeded the free movement of African populations in the territory, while the 1948 Immigration (Control) Ordinance was specifically aimed at managing the in-migration from other colonies and codifying the conditions of their formal residency in the colony of Kenya. In the wake of Italy's invasion of Ethiopia in 1935, British Kenya faced an unprecedented influx of Ethiopians fleeing the conflict, who were first discouraged from entering Kenyan territory—and delegitimized as “deserters” and renegade “tribesmen”—before eventually being allowed entry and subsequently housed in camps near the town of Isiolo (Shadle 2019; Wilkin 1980). In addition, Kenya received substantial numbers of displaced populations from Rwanda and Burundi throughout the 1920s and 1950s, as well as from Sudan after the 1955 mutiny and during the ensuing Anyanya war (Otunnu 1994; Verdirame and Harrell-Bond 2005: 2). What transpired through these parallel colonial histories of managing displacement was the notion that not all refugees were the same in the eyes of the colonial state, and there was a keen interest in affording considerable mobility to some bodies, but not others. Banishment and isolation in distant parts of the colonial territory had always been part of the standard colonial practice of dealing with African subjects who resisted or threatened the colonial order. Kenyan Mau Mau fighters, for example, were sent to the remotest part of southwestern Tanzania (now the Katavi region). As Uma Kothari and Rorden Wilkinson (2010: 1400) note—writing about the Seychelles as a place for political exiles—“exiling was used as a technique of colonial rule.” An inversion of this policy toward colonial dissenters was the containment of incoming refugees along the margins of colonial territories. In German and later British Tanganyika, the authorities were opposed to the idea of admitting Africans from neighboring colonies seeking refuge, unless they could simultaneously be exploited as laborers. Permissions (or denials) to move were issued exclusively at the behest of the colonial government, which always weighed the imagined attributes of certain ethnicized identities as industrious or idle, potential economic burdens or gains, humanitarian responsibilities, and the potential harm caused to the prevailing political order. Even after political independence, the Kenyan and Tanzanian states would, to a large extent, retain this institutional aspiration of ordering racialized bodies in space.

Managing “Foreign” Africans: Refugees and Exiles in Postcolonial Kenya and Tanzania

British colonial legislation laid down the basic tenets of the respective future national laws dealing with immigration, dissidents, and refugees (Ayok 1985). Independent Tanzania passed a series of acts between 1965 and 1969—the Extraditions Act (1965), Tanzania Statutes (1965), and the Fugitive Offenders Act (1969)—all of which had a bearing on the rights and obligations of refugees and asylum seekers. The main feature of the 1966 Refugees (Control) Act was to enable the Tanzanian state to impose a blanket classification of entire designated populations as “refugees” and had a provision for enforcing “settlement lock-ups and the custody of persons therein” as a matter of public security. This act remained in force until it was superseded by the 1998 Refugees Act, which, although containing some progressive changes and dispensing of the explicit term “control” in its title, continued to champion equally restrictive and prohibitive clauses.

Kenya and Tanzania's acceptance of the colonial borders after independence—and both states’ advancement of nation-building projects—necessitated the reinforcement of distinctions between “resident” and “foreign” Africans, which were being increasingly expressed through the language of citizenship. The criteria for citizenship were contested, as both countries had hosted migrant laborers, political exiles, and refugees at different times throughout the colonial period and were multiracial societies with numerous ethnic groups straddling the political boundaries. Citizenship determination varied across postcolonial Africa and was either based on residency in the respective territory at the time of colonial partition or independence, descent (belonging to a recognized ethnic group in the colony), or naturalization after a specified period of lawful residency. Political manipulation and emerging clientelism further cemented binary distinctions between supposed “strangers” and “natives,” “immigrants” and “autochthones,” “outsiders” and “authentic” citizens (Dorman et al. 2007; Geschiere 2009).

This was even further complicated by existing colonial racialized categorizations that, for instance, classified Kenya's ethnic Somalis as “non-natives,” despite their demonstrable presence in the northern areas of the country for decades, even centuries (Weitzberg 2017). Somali claims to citizenship in the country remained ambiguous at best, and were further complicated by successive influxes of Somali refugees from outside Kenya's territory during the 1990s (Scharrer 2018). Brennan notes accordingly that Tanzania's colonial legal code was imbued with paternalistic notions of supposedly protecting a “native” population against encroachments from “foreigners.” But the categories of “native” and “non-native” themselves were all but self-evident, and urban Indian, Arab, and Somali populations were alternatingly classified as either (Brennan 2012: 24–25). Colonial preoccupations with uncontrolled mobilities and ambiguous belonging of populations were carried over into the postcolonial states, affecting especially refugees, forced migrants, and precarious citizens. In the 1970s, Tanzania enacted discriminatory land laws that were based on a similar colonial-era divide between “indigenous” and “nonindigenous” citizens (Aminzade 2013: 3). The unfinished “business” of citizenship was therefore deeply entangled with colonial power relations and ethnoracial identities, and remained an “incomplete legal project” (Brennan 2012: 22) whose exclusionary effects were felt most severely by refugees and asylum seekers.

Political uncertainty in the wake of decolonization caused renewed concern among “Western” observers and their African allies, who feared the possibility of dissidents forming alliances with communist groups. This inconvenient reality shaped attitudes toward African refugees in particular and had to be speedily addressed through internationally recognized legal conventions to monitor, control, and direct their movements and thus prevent such alignments. In September 1969, the Organization of African Unity (OAU) therefore adopted its Convention Governing Specific Aspects of Refugee Problems in Africa, which was a composite of the United Nations 1951 Refugee Convention and recommendations of the 1967 protocol, with recognition of Africa-specific causes of flight that included external aggression, occupation, and foreign domination (Ayok 1985; OAU 1969). The UN provided a technical advisor and further support in drafting the OAU Convention.

By channeling the question of refugee settlement through the UN, a multilateral solution could be financed and executed in the overarching interests of maintaining the international political order in the face of potential crisis situations caused by refugee displacement (Metclaffe 1967, cited in Brooks and El-Ayouty 1970: 76). Many postcolonial African governments even favored a situation in which African mobilities would be controlled again through external institutions that often relied on the backing of former colonial powers. In a speech to the 1967 United Nations conference on refugees, Diallo Telli, the first secretary-general of the OAU, described refugees as an international problem, because of “the potential threats it holds to stability, peace and security in Africa and the world.”

The internationalization of Africa's emerging “refugee question” contributed to the continent's displaced no longer being able to move undetected between neighboring states. UNHCR's humanitarian aid thus became a tool to control African refugee mobility. The rather ambiguous activities of the United Nations on the continent often provided “obscure moral support for liberation, while at the same time pursuing programs which encourage a posture of rehabilitation with the promise of a new life, serving free Africa, and not the cause of the liberation for which they seek support” (Metclaffe 1967, cited in Brooks and El-Ayouty 1970: 74). Reliance on aid from Europe and North America meant that independent African states had to largely adopt solutions to displacement that their former colonial powers favored. The OAU was encouraged to recognize the specificity of African refugee situations, where physical disasters like drought, flood, famine, disease, or poverty caused people to cross international borders.

Kenya ratified the UN Refugee Convention Relating to the Status of Refugees in 1966 as well as the OAU Convention in 1969, and included a first form of refugee status recognition as “Class M” entry permits under the Immigration Act of 1967 (Maina 2016). But the veneer of progressiveness in the OAU's refugee definition did not obscure its more restrictive codes on freedom of movement and restricting political rights of refugees. This was articulated through the convention's principal “distinction between a refugee who seeks a peaceful and normal life and a person fleeing his country for the sole purpose of fomenting subversion from outside,” which effectively opened the door for state repression. Kenya and Tanzania's refugee policies tended to reflect the prevailing political ideology of each host state and their alignments in the international arena. A number of bilateral and multilateral agreements on the control of refugees and dissidents between neighboring East African countries have affected the continuing safety of displaced communities in Kenya and Tanzania, who could then be forcibly repatriated on the grounds of political opposition activities. Refugees were vulnerable to instability and political change within their country of asylum, as successive governments often had different attitudes to the same refugee group. Economic crisis reinforced this scapegoating, and refugees began to be labeled as economic saboteurs, rendering their positions increasingly precarious.

Since the 1990s, the management of refugee mobility in Kenya and Tanzania has taken an unparalleled turn toward using notions of “security.” While cumulative arrivals of displaced populations have always invoked an imaginary of excess and threat during and after colonial times—describing refugee movements as “incursions” or “invasions” (Sequeira 1939)—the new security paradigm was concerned with the spectre of “terrorism.” This was a clear continuation of popular perceptions of unsolicited African migrants as “destabilizing” influences during the 1980s. Then, colonial territoriality was upheld as a safeguard of postcolonial state power, while forced migration seemed to empirically question the “identity containers” that Africa's nation builders sought to create. While this exclusion was rooted in colonial differentiations between desirable (authorized and for labor) and undesirable (unauthorized and without public revenue) cross-border mobilities, displacement between independent states reignited fears of “foreign” African bodies that would misuse the safety of exile in order to launch armed struggles against governments in their countries of origin, making them “refugee warriors” (Zolberg et al. 1989: 275)—a term for rebel movements or the military wings of political parties that evokes primitivity and atavism, and which, in so doing, delegitimizes their insurgencies.

The sheer magnitude of cross-border displacements during the 1990s, and the adverse effects of previous structural adjustment on state capacities (Loescher and Milner 2004; Veney 2007), galvanized the Kenyan and Tanzanian governments into pursuing more hard-line policies toward hosting refugees. Kenya introduced counterterror measures that focused disproportionately on Somali refugees, who were singled out as particularly prone to religious extremism, while South Sudanese refugees were perceived as generally troublesome or “unruly” (Hyndman 2000; Mwangi 2019). In a similar vein, Burundians in Tanzania were depicted either as “lazy” populations who would put strains on the national budget, or as a fifth column made up of violent insurgents in the midst of a peace-loving host nation (Malkki 1995).

The latest iteration of differentially governing migrant mobilities across the region are internationally funded projects for biometric borders and refugee monitoring systems. Not least with respect to the ethnoracialized nature of aid and mobility management, Katja Jacobsen notes that the comprehensive rolling out of “humanitarian refugee biometrics has rendered new domains of life intervenable” (2017: 545). Once again, this raises the spectre of differential valuation of African bodies and the decision to immobilize some while enabling the conditional movement of others. Kenya and Tanzania have been at the forefront of these attempts to create a “modern” migration regime that is regulated, verifiable, institutionally legible, and ultimately caters to the political needs of states and intergovernmental agencies, including the IOM and UNHCR. In conjunction with Ethiopia, the two countries are, at the time of writing, spearheading a regional scheme as part of the European Union (EU)/IOM Joint Initiative in the Horn of Africa that is designed to curb “irregular” migration flows along their shared borders and further along the transit route to South Africa. This illustrates the continuing preoccupation of the Kenyan and Tanzanian governments, and their financial backers in international agencies, with managing the circulation and presence of African bodies “out of place.” As the next section will highlight, however, this is not an inevitability, and there are viable alternatives to the politics of containment and control in Africa.

The Limits of Control: Conviviality and Potentiality

Colonial categorizations of Africans as belonging and not belonging have shaped both Kenyan and Tanzanian policies toward (im)migrants, asylum seekers, and refugees. Like their colonial predecessors, many independent African states have continuously sought to control the mobility of citizens and noncitizens within their territories. In this section, however, we explore alternatives to (post)colonial legacies of exclusion that find precedence in the histories of the two countries. In contrast to the binary assumptions that underpin narrow postcolonial approaches to migration and asylum in Kenya and Tanzania, Nyamnjoh reminds us that “being an insider or an outsider is always work in progress, is permanently subject to renegotiation and is best understood as relational and situational” (2013: 654; see also Fiddian-Qasmiyeh, this volume). Rather than seeking to ossify the parochial nature of contemporary migration regimes in the two countries, we wish to understand this renegotiation over time through the emergence of conviviality and solidarity.

With ethnic groups and “kinship” ties frequently straddling colonially imposed borders, many postcolonial African states have had difficulty controlling organic cross-border mobility. Instead, lived experiences on the continent are interwoven with notions of flux at the territorial and symbolic “frontiers” that empirically question the compartmentalization of Africans into contained units of belonging and reimagine them as what Nyamnjoh (2017: 258) terms “frontier beings.” He notes that “frontier Africans … straddle myriad identity margins and constantly seek to bridge various divides in the interest of the imperatives of living interconnections, nuances and complexities made possible or exacerbated by the evidence of mobilities and encounters.” As opposed to reinforcing colonial binaries at the border, this position involves “experimenting with multiple layered and shifting identities” (ibid.: 264). In fact, under colonialism and for most of the postcolonial period, ordinary people living in the region had a more nuanced approach to dealing with “foreigners” than their respective states, not least through the incorporation of strangers without kinship ties into precolonial societies (Kopytoff 1987). The effects of colonialism on attitudes to “strangers” often hark back to how laboring bodies were differentiated or positioned in the colonial ethnic hierarchies and economy. In Tanzania, even though people on plantations stigmatized Burundian workers, in the border regions they were often even indistinguishable from local Waha communities, and their pronunciation of certain Kiswahili words became a common way of detecting who was brought up speaking French (Daley et al. 2018). In Kenya, differences between ethnic groups were likewise increasingly cast in the language of noncitizenship, hence depicting conationals derogatively as “immigrants” (Jenkins 2012). When humanitarian actors sought to distinguish refugees in Kenya and Tanzania from local citizens, with the help of a draconian state machinery, mobility in the respective borderlands was further securitized and previously cosmopolitan communities fractured.

Nyamnjoh (2017) contends that essentializing claims to belonging are at odds with realities on the continent, and criticizes scholarship on Africa that focuses on bounded forms of rootedness in a national heartland rather than on wider solidarities and convivial relationships (see Malkki 1992). Research has inadvertently legitimized state policies that seek to strengthen the relationship between national territory and identity, subliminally fueling calls to “root out” foreigners and xenophobic attacks against Africans from other parts of the continent. We argue that research on migration in Africa needs to reconceptualize migration as a decidedly political act and deploy concepts that recognize the common humanity of Africans.

During Nyerere's presidency (1963–1985), Tanzania lent its support to various southern African liberation movements, including the Front for the Liberation of Mozambique (FRELIMO), Zimbabwe African National Union (ZANU), South West Africa People's Organisation (SWAPO), African National Congress (ANC), and the Pan-African Congress (PAC). In contrast to contemporary expectations of public resentment toward migration, this was actually popular with Tanzanians themselves. African states like Tanzania have always feared that discourses on liberation and the struggles against imperialism, colonialism, and racism may place the focus firmly on the internal contradictions within their own society. Consequently, independent states were more reluctant to criticize injustices in African-dominated countries. Nyerere's Tanzania illustrated the contradictions of a state that sought to promote human dignity and rights within spaces of white minority rule, while repatriating dissidents to African-dominated regimes like Kenya and Burundi. The return of approaches that recognize the solidarities of African peoples would mean breaking away from Eurocentric understandings of belonging and accepting that alternatives reside in the everyday potentialities of African societies.

Nyamnjoh employs the concept of conviviality to suggest alternative ways of being and belonging that are not antagonistic, exclusive, or violent. Conviviality, he explains, “depict[s] diversity, tolerance, trust, equality, inclusiveness, cohabitation, coexistence, mutual accommodation, interaction, interdependence, getting along, generosity, hospitality, congeniality, festivity, civility and privileging peace over conflict, among other forms of sociality” (Nyamnjoh 2017: 264). In this spirit, he calls for convivial scholarship that challenges exclusionary ideas surrounding the mobility of Africans. Considering Tanzania's historical experience of supporting subversive mobilities in southern Africa, there exists a potential for reimagining African social worlds as driven by solidarity and association rather than a coloniality that views ethnic identities as immutable and posits that Africans should be fixed “in their place.” While the prospect of greater freedom of movement, as adopted in the East African Common Market Protocol, may indicate new ways of viewing migration in postcolonial Africa, it again emphasizes the differential valuation of African bodies as labor, some of which are highly skilled and desirable while others are not.

Conclusion

In this article we have argued that considering long-term sociohistorical trajectories is essential to understand contemporary hegemonic approaches to migration in Africa. Such perspectives reveal the colonial dispositions of dominant conceptualizations of migration as well as current policies and practices. Using the cases of labor migrants and refugees in Kenya and Tanzania, we showed how the colonial state perennially constructed African bodies that could legally only move as “labor” and how refugees became conscripted into a social category that excluded them from the body politic of the postcolonial nation-state. Colonial and postcolonial African states adopted policies to govern migration movements largely in accordance with Eurocentric frameworks, but with international and regional security concerns in mind. A system of differential mobility under colonialism was seamlessly carried into the postcolonial era with the advent of ostensibly “new” security challenges that included terrorism, uncontrolled border crossings and solidarities, and the enduring violability of domestic citizenship regimes. Biometric technology and border controls are modern responses to markedly old problems that governments on the continent have faced when trying to seek international allies for reinforcing parochial visions of belonging and mobility. With haunting parallels to colonial administration, under which docile labor migrants were valued to a degree that unauthorized and forced migrants were not, contemporary refugees in Kenya, Tanzania, and elsewhere throughout the global South are portrayed as only redeemable in public discourse through their reimagination as working, and therefore profitable, bodies.

Finally, we contend that at times African states have been able to transcend external priorities by developing and implementing policies that challenged white domination on the continent. Because freedom, decolonization, and independence were often conceptualized as recapturing the state from people racialized as white, Africa's political elites were not able to perceive how white supremacist and Eurocentric thought had permeated and structured their understanding of mobility in their own societies. This led to the development and reproduction of policies that were counter to African everyday realities and that denied liberty and autonomous decision making to the majority of people. Drawing on Nyamnjoh, we make a case for convivial scholarship that recognizes the pluralities and cosmopolitanism in the ways in which African lives are lived.