Governing Mobility Through Exemptions: Cross-National Dependencies,Immigration Policy,and Migrant Labour in South African Historical Perspective

Abstract

Over the last century, the South African state has periodically engaged in the practice of ‘exempting’ various migrants from their otherwise irregular immigration statuses. Always backed by official legislation, exemptions represent one way by which dominant capitalist interests have relied on the legitimacy of the state to meet their labour needs by sometimes employing undocumented migrants from the Southern African region. Through insights from sub-imperialism and bordering, this paper discusses historical case examples from policy articulations, parliamentary debates, secondary literature and archival materials. By exploring cross-national relationships of exploitation and differentiation, the paper argues that exemptions should be understood as attempts by which the contradictions of ubiquitous informal cross-border mobility and employment in a regime of unfree regional movement might be resolved. Exemptions also attest to the challenge of governing human mobility in a region invested with a historically vast infrastructure of producing, attracting as well as exploiting cheap migrant labour.

Introduction

In the year 2022, the South African government took steps to end a 13-year special arrangement in which about 180,000 Zimbabwean migrants have had an exemption permit to live and work in the country. ¹ The exemption permit sought to regularise undocumented Zimbabwean migrants, offer amnesty to those in possession of fraudulently acquired South African documents, as well as address perceived abuse of the asylum system by economic migrants (Nzima and Moyo, 2023). The pending termination of the Zimbabwean Exemption Permit (ZEP) is in accordance with section 31(2)(b) of the Immigration Act (Act 13 of 2002), a provision on ‘exemptions’ that allows the minister of Home Affairs to pronounce if, when, and for how long such exemptions should apply.

Those migrants who seek to continue to reside and work in South Africa initially had until the end of 2022 to apply for other permits as spelt out in an initial South African Department of Home Affairs Immigration Directive (Department of Home Affairs, 2021). This grace period has since been extended several times, the latest of which has pushed the termination date to 29 November 2025 (Chirume, 2023; Department of Home Affairs, 2022; Republic of South Africa, 2023). Of note, however, is that immigration exemptions are not a recent novelty in South Africa. In the last 46 years from 1977, for instance, there have been no less than eight similar exemptions extended to different nationals whose legal status in the country had become problematic in one way or another (Coetzer, 1979; Nzima and Moyo, 2023). The practice of exempting sections of the migrant population from their otherwise irregular immigration statuses draws attention to the discretion, flexibility and ambivalence that sometimes characterises South Africa’s migration governance regime, a point already alluded to in academic debate (Moyo, 2019; Thebe, 2017, personal communication). Their historical ubiquity is however mobilised in this paper to address broader issues of migration and migrant labour governance in the country.

The current run of exemptions, just like most others implemented previously, presents two issues of interest for this discussion. The first issue is the migrant labour question which, for more than a century and a half, has underpinned the regional competitiveness of South Africa’s economy through cheap wage labour and institutionalised discrimination most epitomised by the ‘colour bar’ (Yudelman and Jeeves, 1986). This paper argues that discretionary exemptions, together with a combination of prolific deportations and open anti-immigrant sentiments, all attest to an enduring challenge of governing mobility within wider and entangled patterns of regional as well as global neoliberal dependency. Corollary to this, the second issue has to do with the post-apartheid dilemma of enhancing internal cohesion, national development and global economic integration through a combination of exclusionary citizenship, an austere regime of unfree regional mobility and state led neoliberal economic growth (Lesufi, 2006; Moyo, 2019; Peberdy, 2001). This is an important question, as it arises in a context of largely unmet political promises that are evidenced by deepening inequality, poverty and unemployment, as well as the informalisation of the country’s economy and cross-border mobility (Schierup, 2016; Tshabalala, 2017). This second point therefore also demands a nuanced examination of the notion that exclusionary citizenship and a restrictive migration regime were merely an inevitable outcome of post-apartheid South Africa (see Neocosmos, 2006).

To understand how the governance of migrant labour has over the years been dealt with through exemptions, this discussion adopts an approach that embraces both structural dependency and relationships of exploitation that go beyond national explanations. ² Through a selective reference to South African immigration policy articulations, parliamentary debates, secondary literature and archival materials beginning in the early 1900s to the recent present, this paper explores the forms of conflict and accommodation the exemptions have sought to address. Using analytical tools availed by the notion of bordering, the paper locates the particular functions of exemptions in a system of global as well as regional dependency where forms of legal protection for cross-border mobility and for migrant labour appear both flexible and uncertain. With this point, the paper also challenges the proposition by some scholars that exemptions are proof that immigration policy in South Africa is merely ad hoc and reactionary.

The methodological import of bordering helps fashion understandings of the management of political, economic and social relations as well as migrant subjectivities whose outcomes reinforce complex and ongoing global, racial as well as gender hierarchies. Exemptions work well not just in reproducing local and global relations of differentiation (Tervonen et al., 2018), but also in the circulation ³ of migrants (Nail, 2016) within a socio-spatial dialectic that highlights the fluidity of capitalistic accumulation and institutionalised difference (McIntyre and Nast, 2011). The management of accumulation and exploitation, visible in the circulation of migrant labour through the policy of exemptions, has important implication for thinking the ways in which socio-political and economic processes reproduce each other in the governance of South African migration. Beyond this, approaching exemptions as a tactic of bordering (De Genova, 2017) serves other purposes as well. This paper makes a case for how historical and current relationships between undocumented movement and migration policy uncover regulatory norms that some might see as ill adapted to trends in regional cross-border movement. After a brief theoretical discussion of South Africa’s sub-imperial position in the region, the rest of this discussion will however demonstrate, starting from the early colonial era through apartheid, the political transition to black rule, and all the way to the post-1994 era, that exemptions and associated quasi-formal strategies of mobility governance are geared towards the management of a vulnerable migrant labour force, and are thus positioned as the link between broader accumulation processes and a socially exposed and cheap regional labour frontier.

Cross-National Dependencies and the Making of a Southern African Labour Periphery

General discussions of the political economy of labour, to paraphrase Smith (2016), often posit a global division of labour that has shaped the working class particularly through the coercive suppression of its international mobility. This division is partly underpinned by a dialectic of dependency between national economies and a world capitalist system, and partly informs classic and current discussions in both development and political economic theory (Valencia, 2017). Although it was Ruy Mauro Marini who first presented the generally accepted axiom that the history of underdevelopment is the history of the development of the world capitalist system, this dialectic is not static. Its evolution from imperial and colonial to commercial export-based, to financial-industrial, to technological and industrial, through to current neoliberal dependency is well developed in Latin American scholarship, and may be applicable to other part of the developing world, not least in the Southern African case. Dependency, it is argued, also extends imperialism in a new form, with new features, – a neo-imperialism – spanning socio-economic, political, technological, cultural as well as military formations (Valencia, 2017: 45). Nevertheless, among other things, this dialectic is characterised by the constant creation and extension of new peripheries by the global division of labour as well as its export in cheap commodity form from peripheral to core accumulation regions.

Valencia (2017) writes that a fusion of classical as well as contemporary imperialism and dependency best define neo-imperialism precisely through the neoliberal logics of labour market deregulation and associated informalisation (Slavnic, 2010), thus driving new forms of dependency and underdevelopment in a global capitalist and neo-imperialist context. Within this schema, subordinate and dependent economies may define new enclaves for accumulation and the reproduction of ‘dependent neoliberalism’ within regional peripheries partly through a constant refashioning of the logic of migrant labour disposability (after Yates, 2011). If its role may serve to advance further the agenda of accumulation, such disposable migrant labour may be inferred to be an important substructure of sub-imperialism.

Regarding Marini’s theorisation, it is nevertheless important to guard against the kind of reductionism that sees sub-imperialism as uni-directional, as the conduit or alternatively a proxy for, say, ‘western’ imperialism. It is equally important to note that sub-imperialism should not be assumed to exhibit similar characteristics everywhere, for to suggest that would be to ignore the specific historical processes that define global dependencies that are characteristic of different regions and economies of the world. In the Brazilian case, for instance, Marini (in Valencia, 2017) was careful to present the entanglement of American foreign policy with American multinationals in the country, together with class struggle, the nature of capitalist accumulation, as well as the rootedness of the Brazilian state in partly autonomous capitalist social relations, that at the time was definitive of the sub-imperialism that obtained there.

With due regard for the danger of transposing Marini’s framework to the Southern African context, it may be helpful to briefly trace which features may present South Africa as a sub-imperial power. Following both Samson’s (2009) and Bond’s (2004) analyses, it could be noted, for instance, that South Africa’s industrialisation drive of the inter- and post-war years was partly underpinned by the support of the American state for apartheid (Samson, 2009). Due to that support, it was thus possible for American and other multinationals to not just establish a base in South Africa, but also to launch their products and operations in the Southern African region and the African continent.

The apartheid state’s determination to preserve white minority rule, as noted by David Simon (1991), was itself also important for both American foreign policy and multinational companies’ involvement in the region. This and other factors help define South Africa’s sub-imperialism beyond both Marini’s economism and early South African colonialism. American consultants, international financial institutions, as well as the strength of neoliberal ideology have later all had a hand in shaping the African National Congress’s (ANC) adoption of such macro-economic policies as Growth, Employment and Redistribution (GEAR) after the political watershed of 1994. In the intervening period to the global financial crisis of 2008, class struggle at regional level, as seen in the collective actions of employees of South African multinationals in countries like Zambia and Mozambique (Miller, 2004, 2005) themselves embedded in cross-border flows of capital and investment strategies, has had a role in shaping South Africa’s sub-imperialism.

Other contested social relations, such as race and gender and their relation to class, have also been key in shaping South Africa’s self-preservation and domination of the region across the colonial, apartheid, as well as the post-1994 eras. Apart from that, some scholars (for instance, Miller, 2004) have suggested that South African ‘capital’ now partly seeks to legitimate itself as African by appealing to the idea of the African Renaissance, even as South African multinationals maintain rigid racial hierarchies and stereotypes in their managerial structures, and thus affecting regional labour relations (Miller, 2005). This is further complicated by a selective Pan-Africanism in South Africa (Khan, 2021), which sometimes oscillates between xenophobic nativism at home, and arrogant, classist and neo-colonialist socio-political relations with the rest of the continent.

Regarding labour, it is equally important to highlight some shifts in the South African political economy from colonial times to the present. Scholars tend to agree that attempts to create jobs through redirecting a labour scarce, capital intensive economy of the colonial and apartheid times to a labour intensive one, for instance through state-led efforts such as the Reconstruction and Development Programme (RDP), did not bear the intended results. The replacement of the RDP with GEAR was partly the outcome of this, with the result that through the late 1990s to the early 2000s, the wish to stimulate investment in skills development and higher wages took planning on a high productivity, technology-driven growth trajectory (Moyo, 2019). The weight of the Mineral-Energy-Complex (MEC), as well as international competition, would however conspire to stifle growth and underpin the kind of fragility that eventually led the economy towards financialisation, which shrunk further skills development and kept wages low, entrenching neoliberal deregulation, casualisation and informalisation (Marais, 2014; Reddy, 2014; Schierup, 2016; Theron, 2010). Today, as the South African economy embraces the digital era, it is not yet clear how the polarising effects of technology on the global labour market obtain in the country. Nevertheless, concerns already exist that wages and conditions of employment in the digital platform sector fall below both domestic legislation and minimum International Labor Organization (ILO) standards (Naidoo, 2020). These are important issues of consideration in relation to labour organisation and (de)regulation going forward.

Given the forgoing discussion, it would not be difficult to see that labour, and particularly migrant labour, has always remained structurally redundant to accumulation in the South African economy, as it might have elsewhere. This is an important strand of the ‘human-as-waste’ argument (Yates, 2011), where the global dependencies that drive capitalist accumulation work along the ‘logic of human disposability’, and where migrants fare the worst (Schierup, 2016). It is along this terrain that immigration policy exemptions and the migrant labour governance regime in South Africa and region should be followed and explored.

Most regimes of mobility governance, both in advanced and developing economies, have tended to tilt towards the doctrine of the free movement of persons (Nita, 2013) as this is considers a relatively less complicated way to deal with international human mobility. In Southern Africa, however, the regime of mobility governance has remained contradictory. Thus, while regional economic policy frameworks have inherited the neoliberal ‘benefits of migration’ view, the free movement of labour in the region remains largely restricted (Tshabalala, 2023). The role of private capital is seen as fundamental in this regard, where the demand for labour places stronger regional economies at a comparative advantage in detecting the terms of human mobility to their benefit. This paper argues that it is primarily these contradictions and disparities that have been informing South Africa’s policy of exemptions over an extended period of time.

Colonial Capital, the State and Migrant Labour Recruitment

The truism that capitalism everywhere thrives on cheap labour began to play itself out on an industrial scale with the discovery of commercial quantities of diamonds and gold in the South African interior in the mid- to late-19th century. Of note was the nature and expansion of settler capitalism that took root in the country and the region, which related directly to the mining economy in the interior itself as well as to forms of socio-economic organisation in the capitalist and the non-capitalist economies of Southern Africa (Crush et al., 1991; Harries, 1994; James, 1992; Jeeves, 1986; Moodie, 1994). This expansion initially rested on the resolution of the ‘native labour question’, ⁴ even if it neither resolved the emerging labour shortages, nor discouraged expansionary capital investments in mining.

Concerning early prospecting and extraction, deeper gold-ore mining required sophisticated and expensive technology, which spooked the global gold share market. The ensuring profitability crisis was initially resolved by reducing the wages of African workers. When technological solutions were later introduced, the persistence of deep-level mining necessitated heavy investment, which phased out small claimholders in favour of conglomerates. The need for more labour fostered by heavy investment, coupled with the reluctance of Indigenous populations in participating in waged work on the gold belt (the Rand), improved neither wages nor the working conditions.

The cost structure of the mines, the low grade of the gold ore and the fixed international gold price meant that only the labour cost could be adjusted (Johnstone, 1976). Cutting wages, expanding the geographical area from which the labour was recruited and establishing a monopsony in labour recruitment through the establishment of the Chamber of Mines in 1889, the Witwatersrand Native Labour Association (WNLA) in 1901, as well as the Native Recruitment Corporation (NRC) in 1912, were some of the strategies employed to engage such labour at low cost. ⁵ After initial resistance from the colonial government, the mining conglomerates prevailed in establishing a regional reserve that saw the formal recruitment of mine labour from parts of present-day Mozambique, Lesotho, Swaziland, Zimbabwe, Zambia, Malawi, Botswana, and for brief periods, India and China.

To be sure, such recruitment agreements have always been bilateral in nature, involving agreements between the South African employers (with approval from the state) and respective colonial governments in the region. The agreements represented the first appearances of the policy of exemptions as it relates to migrant labour. While they ensured the competitiveness of South African mining (and later agriculture) on the global market, the exemptions themselves made possible the presence of black workers in the South African industrial core where they were in fact robustly policed and otherwise territorially excluded. They also worked for a slave wage, a significant part of which they could only access after they returned to their reserves at the end of their temporary work contracts (Khan, 2021). In this sense, exemptions dramatise the contradictions that fashion the kinds of dependency, inequality, and exploitation that have helped establish a South African industrial core and a regional labour periphery with all the markings of enclave or sub-imperial power in a Southern African global periphery.

Centralised labour recruiting in colonial South Africa did not proceed without its problems. From its earliest days, the labour recruitment scheme as championed by the Chamber of Mines and its allies resulted in conflict between the mine owners (the Randlords) and the colonial government, then led by Paul Kruger (Harries, 1994). Disagreements revolved around the state’s labour policies, railway rates, the high cost of mine explosives, as well as his taxation policy on one hand, and the miners’ recruitment of foreign labour, on the other hand. Further conflicts and dissatisfaction partly led to the South African war (1899–1902) and Kruger’s overthrow soon after (Johnstone, 1976; Marks and Trapido, 1979; Phimister, 1993; Porter, 1990). Here can be found an indication of the power of miners over the state in establishing their hold over political and economic relations between South Africa and the region, and between the needs of global capital and regional labour at a Southern African economic enclave.

Thus, the passing of the Native Labour Regulation Act in 1911 at the behest of the Chamber of Mines established the form of labour that characterised the migrant labour system for decades to come. ‘There would never have been deep-level gold mining industry in South Africa’, argues Crush et al (1991), ‘if the large numbers of low-wage, unskilled migrant miners had not been recruited from throughout the subcontinent’ (p. 1). Thus, recruiting cheap labour from the Southern African hinterland became a South African economic norm for years to date. And while the Act legislated the recruitment and employment of cheap migrant labour primarily in the mines, other industries followed suit by replicating these policies all across the country’s larger economy.

The migrant labour benefits that accrued from this Act for the wider colonial economy persisted even in the face of other disputes that emerged in the political arena, for instance regarding poor working conditions at the mines. One such dispute pertained the ban in 1913 on the recruitment of migrants from areas north of latitude 22° south of the equator. The ban, whose aim was to arrest high pulmonary tuberculosis induced mortality rates at the mines, affected all Africans, or ‘tropical natives’, from the extreme north of South Africa as well as from the rest of the continent. Playing out at legislative levels, the dispute revolved around the widespread and continued employment of these so-called prohibited natives in the rest of the economy, instead of them being detained or repatriated to their territories of origin for being in South Africa without permitting work documents. Francis Musoni (2018) has suggested that by failing to produce a clear strategy for the implementation of the ban, the colonial government was only interested in playing politics while leaving employers with enough legal leeway to offer employment to such natives as a matter of de facto exemption. At one level, the centralised recruitment of migrants from the region had partly succeeded in generating a greater pull for clandestine migration into South Africa (Bolt, 2015). At another level, the ban on regional migrants redirected them for employment in other sectors of the economy such as agriculture. At yet another level, their employment became much more circumscribed, a situation that could only worsen the conditions of their employment.

The apartheid government later inherited and entrenched this system from the 1950s through the 1980s. Labour disputes that emerged were met with violent clampdowns as well as more aggressive recruiting beyond the northern borders (Crush et al., 1991; Maloka, 1997). After the revitalisation of labour recruitment, and mainly through the efforts of WNLA in the north, 80% of total African mine labour was made of foreign migrants by 1973 (Maloka, 1995: 218). This proceeded while the apartheid government institutionalised the control it held on the internal movement of both Black South African and foreign labour. In this regard, therefore, recruiting functioned as an exemption on African labour specifically geared towards fulfilling shortages while at the same time retaining forms of segregation and exploitation of such labour, a situation that has remained the same over the years.

Such exploitation materialised under a ‘workplace racism’ founded on a policy that promoted a ‘two gates system’ (Fine, 2014: 332). Only White migrants, mainly from Western Europe, could enter urban South Africa legally, that is, through the front gate. Blacks were confined to the back gate, which barred those deemed undesirable while allowing low-wage migrant workers on temporary contracts (Segatti, 2011). Caught up in this informal migrant labour system, resulting ‘illegal’ migrants were at times ignored, although they were sometimes marshalled into forced work in mines, on farms and in manufacturing, often later to be arrested and deported (Crush and McDonald, 2001). This racialised management of migrant labour reinforced the dual nature of industrial relations in apartheid South Africa (Clercq, 1979). So, while exemptions for White immigrants earned them and their unions recognition in the White government’s mechanism of control, in the process guaranteeing them higher wages and monopoly control over access to skills, exemptions for Blacks achieved the opposite, dispersing them to White-owned farms, mines and factories, where they often found representation through shop-floor, management-dominated committees, a practice still prevalent today.

Separate Development, Co-Dependency and Unfree Movement in South(ern) Africa

Soon after it came to power in 1951, the Daniel Malan government passed the Regional Territorial and Tribal Authorities Act (or Black Authorities Act 68 of 1951) as a way of manufacturing foreign, ethnically based political identities (Mamdani, 1996) through its flagship policy of separate development. By confining Black South Africans to poor and overcrowded territorial enclaves known as ‘Homelands’ or native reserves, the apartheid state could thus denationalise them and entrench their exploitation as cheap labour. Through what Mamdani (1996) has called decentralised despotism, co-opted traditional leaders often mobilised the recruitment of mine labour within their local jurisdictions, appointed their overseers at the labour compounds, and collected ‘tributes’ from returning workers (Bond, 2018). The place of territorial segregation in the successes of a labour system of that magnitude therefore depended on a multiplex of other alliances that took a regional scope, linking together the South African and regional colonial governments, traditional leaders, the mines, as well as formal and informal labour recruiters.

Within this complex, Jeeves (1986) identifies two elements that are important to flag with regard to the mobilisation and employment of migrant labour. The first is that labour recruiters, labour contractors, foreign governments and traditional leaders increasingly had a say in labour issues obtaining on the Rand. The second was that in spite of self-proclaimed achievements as evidence of the triumph of free market capitalism, leaders of industry constantly courted state assistance to legislate for and administer the organisation of Black labour supply. The relationship between informal, traditional and regional institutions and the state regulative mechanisms in shaping the regional migrant labour system is such that while the miners colluded with the state to put in place coercive labour and mobility legislation, the alliances they established with informal as well as neighbouring jurisdictions in pursuit of a ‘steady’ supply of labour resulted in an unstable, expensive and conflict-ridden labour relations. The seeming tolerance of a coercive and conflictual migrant labour environment by the state, concludes Jeeves (1986: 13), left the South African government in a position where it supported the miners’ efforts to recruit migrant labour, while attempting to avoid complete identification with them. This is the specific import of the policy of exemptions as they historically emerge in the Southern African political economy of migrant labour. They do not emerge as an anomaly or in contention with an otherwise unambivalent labour migration regime. They represent the enlightened disinterest of the state in the historic exploitation of such labour.

In the book Insiders and Outsiders, Nyamnjoh (2006) presents a case of similarly ambiguous relations between undocumented migrants, informal labour recruiters, and state officials in the domestic care industry in wealthier parts of northern Johannesburg. His analysis represents a critique of the apartheid state’s ability to circumscribe the movement, domicile and work of Black labour in urban South Africa. He demonstrates how White residential areas that apartheid sought to protect from Black menace through legislated segregation actively employed the same Blacks, most of them undocumented migrants. The inner workings of this relationship between documentation, employers and coercive and yet often inefficient attempts of the state to control their presence there finds elucidation in a lengthy passage taken from Coetzer (1979: 103). Coetzer muses,

Rhodesian (Zimbabwean) blacks were expected to apply for Workers’ Travel Documents through their local Bantu Affairs Commissioners, who would communicate direct with the appropriate district commissioner in Rhodesia who, in turn, would insert the applicant’s particulars in a Worker’s Travel Document and post it to the Commissioner in South Africa for endorsement and delivery to the applicant. (However,) there were communication difficulties between applicants and South African officials, resulting in long and often fruitless correspondence between the later and Rhodesian district commissioners, besides which it had dawned on our blacks that provided they could keep a jump ahead of officialdom and the Police, their illegal status was in many ways preferable to the circumscribed and regulated existence of blacks in registered employment. Today still (late 1970s), the communications network and alarm system perfected by the Rhodesians in Johannesburg’s northern suburbs in the late sixties is a legend of efficiency; when the West Rand Administration Board’s labour inspectors pounce, their prey, miraculously, is nowhere to be found. Obviously there is a measure of collusion with their clandestine employers.

Tailor-made and yet loosely enforced, immigration policies such as Worker’s Travel Documents, the ban on tropical natives, or the ZEP, thus provide an avenue of ambivalence through which a wider and precarious mobility and employment of undocumented migrants could be conceived. If at the beginning the capitalists fought to suppress the Black wage, their success laid the ground for a broader system by which business could engage undocumented migrant labour whose generally lacklustre policing could be construed as a veneer of tacit state support.

Mirroring the colonial government before, apartheid’s coercive but slack policing of mobility thus oversaw growing competition for migrant labour by other primary and secondary industries outside of mining, including agriculture. Such industries also adopted the long-standing model where agreements were primarily between employers and workers (Fine, 2014). Colonial and apartheid states established the parameters of the engagement of foreign labour, but often left the terms of such engagements to employers so that

[o]ver time a parallel, more informal and unregulated system emerged as the state signed localized agreements to allow commercial farmers to recruit temporary and cross-border migrants and in later periods sometimes offered illegal labour migrants the option of working on commercial farms rather than being deported. (Crush, 2008, in Fine, 2014: 331)

This particularly came to be the case in agricultural capitalism in the extreme north, an arid agricultural belt between latitude 22° and South Africa’s northern border with Zimbabwe. In keeping with the precedence in the early mining industry, large scale-agrarian capitalism in this area grew out of a need by capitalists to assert control over both territory and labour (Bolt, 2012). While apartheid institutionalised coercive control on movement, in reality, it has been near impossible to stem historic cross-border labour migration coming into South Africa from Zimbabwe, Mozambique and beyond. By establishing farms as capitalist enterprises, apartheid farmers succeeded in establishing their enterprises as local economic hubs (Bolt, 2012). Although the state has a long history of policing this border, including in the farm compounds, their efforts underscore the limits of effective control on cross-border movement. Exemptions that have come to characterise the employment of farm labour in this area and elsewhere, including the current ZEP, attest to this challenge. Meanwhile, the agrarian economy in the north has proven to be of enough significance for the state to establish a special economic zone in the area, introducing a new dimension to the policy of exemptions.

Agrarian Capitalism, ‘Special Employment’ and the Informalisation of Migrant Labour

The Special Employment Zone that was established along the northern border in the 1980s, under the special exemption provisions of the Aliens Control Act (of 1913), demonstrates an important aspect of the reshaping of the regulative framework that governs migrant labour. The employment zone exempted farmers from the usual bureaucratic channels of recruiting workers (Rutherford, 2008). This meant that Zimbabwean migrants could seek employment in this area if they were in possession of a BI-17 special permit that linked them directly to a specific farmer (Bolt, 2012). How this connection between employee and farmer was established came largely to depend on an individual farmers’ foremen, often Black migrants, who invariably doubled up as informal recruiters in their own right. This further personalised the regulation of labour, enabling the development of networks between South African farms and Zimbabwean villages, from which the labour originated, through these and other middlemen. Webs of kinship networks emerged across the border and between farm populations, elevating foreman into important proxies for capital in the labour reserves. These relations, concurs Bolt (2012), are largely still at play to this day.

Such ambivalent attitudes of the state towards cross-border migration in an austere regime of unfree movement underscore the informalisation of the structures and relationships that bring together the needs of the state in governing territory and those of the capitalists in accessing labour along banks of the Limpopo River (Bolt, 2012; Bradford, 1993). A vestige of the colonial past, devolving the governance of migrant labour and its mobility to capital interest often reorients state interest from mere lack of capacity towards political self-preservation. It may be read as both abrogation of official responsibility by the state and maximisation of profit by capital. In the last instance, it has historically worked well to resolve contradictions between the two interests by exploiting a vulnerable and desperate labour force.

On evidence of archival records (Tshabalala, 2017), one sees that (a) the agrarian enterprise that developed there depended heavily on and was largely built by undocumented immigrant labour; (b) that the presence of informal recruiters of that labour from other parts of the country was rampant but formally, if only rhetorically, opposed by both farmers and the state; and (c) that alongside these developments, and through what came to be known as the Farm Labour Scheme, the state actually developed a network of depots from as early as the late 1940s from which to collect informally recruited labour to be processed by agents of the state at the main depot located in Louis Trichardt (Bolt, 2012). Farm employment today continues to illustrate the manner in which the state attempts to assert territorial control, ‘deal’ with the informalisation of mobility, as well as promote economic development through quasi-formal means.

The special employment zone has also seen, at one point, the opening of an informal border post known as Gate 17 (Bolt, 2012). The gate catered for the farms directly by processing labour that was headed there from directly across the border. In this way, migrants seeking work at the adjacent farms on the other side of the Limpopo river were during this time able to bypass labour recruitment requirements that otherwise apply in immigration law. A times Gate 17 was avoided, so that migrants would be issued with permits by farmers themselves, who use their special status as border employers. Through the same special exemption, farmers only have to formally issue their employees with employment agreements, which workers would take to the immigration officials for post hoc documentation (Crush, 2000; Lincoln and Maririke, 2000). Even if Gate 17 has closed and the special employment zone discontinued, farmers have succeeded in establishing themselves as centres of enough economic importance and autonomy to often ignore, resist or dictate the terms of farm labour engagement to border and other relevant state authorities (Rutherford, 2008; Rutherford and Addisson, 2007). At a local everyday level, Bolt (2012: 126) observes that ad hoc agreements between farmers, the police and the army are not only locally effective in bringing a veneer of stability in terms of the protection of quasi-documented farm labour, they ensure that farmers shape border policing, and to some degree, migration governance.

Migration governance during the political transition and after 1994 has faced more or less similar challenges. The renegotiation of formal labour recruitment agreements between South Africa and its neighbouring countries (Jeeves, 1986) that instituted the strategy of ‘internalisation’, saw the Chamber of Mines shift its efforts towards recruiting from the local South African labour pool. The workers’ unions, especially the National Union of Mineworkers (NUM), had in the 1980s also won crucial battles in improving wages, working conditions, and the renovation and ‘modernisation’ of mine compounds (Crush et al., 1991). Combined with drops in real revenue, the mines had lost more than 200,000 workers by 1987 (Maloka, 1997). About the same time, not only was migrant labour being captured by other secondary industries, it was also increasingly becoming mixed.

Unable to prevent undocumented movement into the country, the ANC government has had to also deal with hardening attitudes towards migration policy reform from within South African business. Its attempts to fine businesses that employ undocumented foreigners have been ignored (Segatti, 2011), and its efforts to force the business community to cooperate with its migration reform agenda significantly resisted. The year 2002 marked a moment of note in this regard. The current Immigration Act (of 2002) had to go through a prolonged vetting process before it passed through Parliament. When it eventually did, some of its key features, such as levies on businesses that hired foreigners, the establishment of immigration courts and an immigration branch that would be separate from the Department of Home Affairs, were opposed and eventually omitted from the Immigration Act (Crush, 2008). If the aim was to ‘get tough’ on undocumented movement, as has been the state position from since the apartheid days, then that position is largely rhetorical. A stringent migration governance regime that is however characterised by a variety of recurring exemptions, opens regular cross-border mobility as well as the formal employment of migrants to potential abuse.

Such a governance regime also promotes creative strategies for circumventing regularised border crossings, and in turn feeds into informal cross-border mobility and employment in South Africa. Although estimates vary widely, the numbers of undocumented migrants currently residing and working in the country’s formal and informal sectors have long been thought to run into the millions (Minnaar et al., 1995). In this historical context, immigration exemptions have less to do with regularising mobility as they function as a reminder that the state has not completely disappeared in the governance of labour migration. Its apparently curtailed capabilities fulfil the same accumulation functions as they have done in the last century and a half.

Conclusion

The expansive recruitment infrastructure that was set up by the erstwhile Randlords, developed to recruit labour from as far afield as Malawi and northern Mozambique (Jeeves, 1986), together with the enduring gaps in migration management, have laid the ground for a recurring pattern of largely circular undocumented cross-border mobility of economic opportunities in South Africa. The shifts towards broader informalisation that now define parts of the South African economy (Schierup, 2016; Theron, 2010), which are seen to be part of a broader neoliberalisation of the region’s economies (Bond, 2018), make the general employment of undocumented labour to appear as if it is a distinctly post-apartheid phenomenon. While it presents visible evidence of the ubiquity of undocumented migration into an economy that is now heavily services and informal oriented (Jinnah, 2018), this discussion has argued that ambivalent migration governance has historically played a significant role in sustaining South Africa’s competitive economic edge in the region.

This sub-imperial position of South Africa maps onto the dominant role it has played (together with Botswana and Namibia) in rejecting the idea of free movement within the regional economic community (Nshimbi and Fioramonti, 2014). This author has argued elsewhere (Tshabalala, 2023) that resistance to a protocol of the free regional movement of persons mostly succeeds in maintaining a contradictory, asymmetrical and increasingly securitised approach to migration governance. Although both Botswana and Namibia are starting to soften their stance, a regime of unfree movement has fashioned the exploitation of migrant labour as a matter of historical routine.

To reiterate, the exemptions that are routinely extended to migrants, although sporadic, do not represent ‘short-term’, ‘temporary solutions’ that are meant to marginally relieve migrant illegality, as scholars such as Thebe (2017) have argued. Exemptions partly demonstrate how the neoliberal Southern African state, historically and presently, has been tending to extend itself into the informal sphere through the ambivalent role exemptions play in migration governance as one of the ways it can spruce up its economic competitiveness while maintaining a veneer of regulative efficiency.

From the earliest days of its industrialisation, the South African economy has relied significantly on a mode of migration governance that only partially resolves irregular migrant statuses as a core component of how it functions. Rather than arguing that exemptions represent a way by which an overwhelmed state periodically eases the burden of undocumented migrant labour on its systems, this paper contends that the oversupply of migrant labour is a structural effect of an economy that has invested massively in an infrastructure designed to tap a cheap migrant labour periphery. Such an infrastructure generates a twin phenomenon of undocumented cross-border mobility that is accompanied by an enlightened reluctance to institute an efficient framework for governing its labour – or offer a clear pathway towards full citizenship – resorting instead to piecemeal exemptions to keep up an appearance of legitimacy. As the above examples have demonstrated, a combination of the ready availability of cheap labour recruited from the region, a profit motive that often elides minimum labour standards, as well as organic and complex cross-national dependencies combine to subsidise official migration governance efforts. So far, they have all worked to further circulate and precarize migrants in the manner that both Nail (2016) and Mezzadra and Neilson (2013) see as fundamental to the functioning of institutions and social relations of exploitation and differentiation that service capitalist and neoliberal accumulation in a Southern African global periphery.