INTRODUCTION

Since South Africa's democratic transition in April 1994, the new ANC-led Government has confronted challenges brought on by its new role in the international system. After decades of raising barriers to the outside world and being shunned in various official and informal ways, the country is now (re)joining multilateral organisations and (re)establishing diplomatic ties. Another consequence of South Africa's burnished reputation is the 'influx' of foreigners. The media and certain vocal politicians are up-in-arms about 'illegal immigrants', who are blamed for a range of scourges, from crime to unemployment. While international relations theorists trumpet 'new security issues', South Africa experiences them on a daily basis. Migration is but one of these concerns.

The issue of (both legal and illegal) migration intersects various international 'regimes', that is, clusters of international norms and institutions that regulate state policies in particular issue areas. For example, some immigrants qualify under international norms as refugees, a crucial source of transborder population movement in Southern Africa. In 1993, South Africa finally entered into its first agreement with the United Nations High Commission for Refugees (UNHCR). International human rights norms also play a critical role in the politics of migration. For example, a new culture of constitutionalism has created a flurry of popular demands for legal protections, also raising the issue of the extent to which foreigners qualify for human rights safeguards. In the 1994 elections, no less, certain temporary residents acquired the right to vote. Expectations for additional rights have escalated.

There are numerous reasons to expect the new democratic Government to be more responsive to international normative pressures than the apartheid regime. The ANC, for example, spent three decades in exile, working within international organisations and building diplomatic ties.1 Having seen the power of international norms (through global sanctions) in aiding their own rise to power,2 ANC leaders should be expected to be more sensitive to the state's international reputation. Yet, democracy also means that domestic protest is stronger and broader, and hence more influential than before. Not all members of the new Government spent those years in exile, most notably unionists. Many divisions within the liberation movement are now played out on different grounds. Immigration policy has emerged as one of these contested terrains.

In the midst of pragmatic policy concerns, the government also confronts a meta-political quandary. Does South Africa 'owe' anything to other countries in the region, which suffered from overt military aggression, destabilisation, economic coercion, and more subtle political costs for supporting the anti-apartheid movement? And if so, should migrants - including refugees – be allowed freely into the country? Who should get citizenship rights? Who deserves entitlement claims, including education and health care, in the face of budgetary constraints? Contemporary debates over these questions highlight overarching issues of political community and identity.

Looking at the ways in which the South African state seeks to regulate population flows enables the exploration of the extent to which international norms and organisations influence both policy-making and processes of identity formation. Comparing current policy-in-formation with the 1920s, furthermore, offers perspective on whether these 'new issues' actually represent unprecedented global normative commitments. Evidence from the history of migration puts doubt on the recent stress on globalisation as something unique, or more forceful, in the post-Cold War era. International pressures have always been, and will continue to be crucial factors in South Africa's migration policy.

NORMS, IDENTITY AND MIGRATION

In general, the analysis of international regimes offers insights into two types of political processes: policy-making and identity formation.3 The former stresses the effects of international norms and institutions on the behaviour of states while the latter explores the social construction of agents. Rather than giving a broad survey of the literature on norms, however, the focus here will fall on the constitutional dimensions of the migration issue, particularly current debates on globalisation, sovereignty and post-national polities.

'Globalisation' has become the trend of the 1990s as policy-makers and theorists in hegemonic states finally acknowledge the extent of international constraints on state action. But for post-colonial states, globalisation may be nothing new: they have never enjoyed the sovereignty of international relations theory but rather, struggled with state-building.4 South Africa, as a former British Dominion, falls somewhere between the two extremes, having enjoyed early self-government (albeit not representative of the majority of the population), but without autonomy in foreign policy.

A renewed debate over sovereignty accompanies concerns over globalisation. Many international relations theorists increasingly stress the role of meaning in constituting states. Yet, each places a different emphasis on the process by which social construction operates. Some focus on the social construction of threat.5 Others challenge the foundational role of the domestic analogy.6 Those rooted in international law traditions consider rights and obligations.7 Despite their differences, these critiques of conventional realist notions of sovereignty all question distinctions between 'us' and 'them' and reject the reification of states.

In sum, it is too easy to take state boundaries for granted. Sovereignty, both conceptually and physically, is more porous than most conceptions imply. Yet, to the extent that these scholars pursue empirical research, most focus on traditional issues: weapons, intervention, security communities, international trade, and environmental hazards. In many cases, the social foundations of sovereignty and state-building get lost because of a persistent state-centrism. Noteworthy exceptions are those increasingly acknowledging an important role for transnational social movements.8

These debates lead to three sets of questions about social construction and migration. The first concerns the state, the second, international norms, and the third, agency.

Questioning the nature and origins of threats leads to an examination of the notion of 'we-ness', that is, the boundaries and characteristics of political communities. Defining friends and enemies has symbolic and pragmatic implications. For migrants, the process of 'naming' demarcates legal status.9 The categories involved, such as national origin versus cause of displacement, fuel political debate and identity formation (for example, in some circumstances, xenophobia).

Nor are these categories simply discursive – resources are at stake. Rather than disputing friends and enemies in terms of weapons, migration turns debates over 'we-ness' inward, encompassing crucial welfare dimensions (including health care, education, employment opportunities, as well as political and civil rights). Focusing on migration enables us to explore questions of sovereignty and the state by encompassing both the external and internal dynamics of social construction. Thus we can escape both the state-centrism of an emphasis on strategic interaction, as well as the domestic focus implicit in most studies of nationalism.

International normative frameworks are one component of the categorisation or naming process. For example, migrants are evaluated as individuals or national groups in part due to prevailing international regimes. While state policy generally follows broader global trends, variations in 'incorporation regimes' across countries are significant. Regional variations are also bound to be significant.10 For example, the Organisation of African Unity (OAU) convention on refugees contains a broad-brush definition, which acknowledges flight from political instability, economic hardship, and environmental disruptions, in addition to political persecution as defined more narrowly by the UN convention.

The historical evolution of norms regulating population flows provides insights into the diffusion of global social understandings and their implications for states' policies and evolving identities. But we should not expect norm diffusion to be either linear or inherently benevolent; it is not necessarily 'progressive' in either sense. Rising xenophobia and violence against (legal and illegal) migrants around the world are sobering reminders that globalisation is not necessarily a 'good' process.

Normative diffusion raises a third issue for international politics and theory: the nature of agents. Migration is clearly about non-state actors, and these population flows have crucial implications for state policy. But what are the political differences between formal non-government organisations (NOGs) versus non-organised agents of change? In other words, should we take into account migrants themselves as social actors, or simply the various interest groups, NGOs, bureaucracies and international organisations which mobilise around (or against) international norms?

To explore these broad issues raised by looking at migration from the perspective of international relations theory, the following sections briefly survey immigration policy in South Africa during the twentieth century. Three themes are explored: the way in which immigrants are defined as threats, the role of international norms, and the range of non-state actors involved in public debate and the policy-making process.

MIGRATION IN THE INTERWAR PERIOD

The origins of the contemporary international migration regime, to the extent that one exists, can be traced to the interwar period. Two aspects of regime formation are evident: the emergence of global regulatory norms (in the wake of refugee flows from World War I) and the growth of international organisations to implement these new understandings. Prior to the late 1910s, most states adopted open migration policies, often encouraging emigration (from Europe) or immigration (in host countries and colonies).

South Africa was one colonial destination for migration. Following its establishment as the Union in 1910, South Africa retained Commonwealth ties to Britain (particularly in foreign affairs). Migration policy highlights the difficulties of state-building during this period. Serious regional tensions, as well as racial prejudices, are evident in policy-makers' attempts to create and regulate a new South African identity. The compromises and statutes of the interwar period set the foundations of migration policy for the rest of the century.

One theme that persists in immigration policy has been high barriers to Africans. Initially, immigration policy did not explicitly address this question since Africans were not allowed official entry as permanent residents.11 Africans were, brought in regularly, however, as migrant workers, especially from neighbouring Mozambique and Malawi (formerly Nyasaland).12 But since migrant labour falls under a different international economic and legal framework (primarily through bilateral agreements between South Africa and Portugal), it will not be considered here.

Because of the consensus on African exclusion, the prime focus of early immigration debates was the status of Indians, who were brought to South Africa as indentured labourers. Themes raised in these debates foreshadowed apartheid restrictions implemented following the National Party election victory in 1948.

The original Union policies were comparable, in a sense, to international agreements as they sought to co-ordinate four provincial perspectives. Cape, Natal, Orange Free State and Transvaal all had different cultural and economic concerns. Yet, they shared a common view of the importance of maintaining the predominance of white society. While the growing presence of Indians presented a threat (albeit geographically concentrated in Natal), important differences emerged between the provinces about how to regulate white population flows as well – with a marked prejudice against newcomers from eastern and southern (as opposed to western and northern) Europe.

One key legal provision that provides insight into cultural preferences is the language criterion for immigration, implemented as one of the very first restrictions in 1902. Early legislation, such as the Cape Immigration Act (No 30) of 1906, required that an immigrant be able to write out an application form in any European language. At the time, Yiddish counted. Newcomers also had to be in possession of visible means of support (at the time, defined as 20 pounds). The statute contained a further provision for religious or political persecution. The laws in the three other colonies generally corresponded with these Cape statutes, with some variations reflecting local concerns. Natal remained particularly concerned about Indian and Chinese immigration. The Orange Free State, desiring to increase its white population, had lighter restrictions. Transvaal primarily faced African migration.

The Union formulated policy within the Commonwealth context. Not only did this entail direct British influence on colonial policy, but also attempts to co-ordinate immigration flows throughout the Empire, with direct involvement of the various voluntary societies (precursors to contemporary NGOs) that encouraged and supervised immigration.13

Asian, particularly Indian, migration rose as the primary issue of contention between South Africa and Britain. (Periods of anti-Jewish sentiment also flared, but less persistently.) Within the context of world trends, immigration to South Africa was relatively marginal: its Indian population, however, was the largest in the Empire. As India gained an increasingly significant role in the Commonwealth, Britain remained caught between its commitment to the rights of its subjects to travel freely within the Empire and its recognition of the rights of self-government for the Dominions. For example, Australia granted full citizenship to Indians in 1918. At the 1921 Imperial Conference, Britain urged citizenship rights for Indians throughout the Commonwealth. In contrast, South African restrictions provoked concern which flared repeatedly at subsequent Commonwealth gatherings.14 The Orange Free State, for example, prohibited Asians from farming, trading, or owning property. (Following the 1919 Peace Conference, pressure also increased to accept Jewish war refugees, with mixed results. Some of those with valid passports – often difficult for Jews to obtain – headed to Johannesburg.)

Although sympathetic to racial concerns, Imperial policy-makers would not allow explicitly discriminatory provisions. Language remained a useful technique for circumventing such sensitivities. By 1906, for example, Indian languages were rejected for the educational requirement. Measures to encourage Indians to depart voluntarily also skirted Imperial norms. Transvaal and Natal, for example, applied a special tax to indentured workers who did not return to India. Later, the Department of the Interior implemented elaborate (and costly) repatriation schemes to entice Indians to return to India (with varying success). Within this context, Gandhi emerged as a powerful leader who extracted some compromises from the provincial and later Union governments.15

Opponents of the Indian presence emphasised two themes, one economic, the other cultural. The 1910s and 1920s marked a particularly difficult economic period in white South African history. The 'poor white problem' resulted in many Afrikaners flooding the mines and urban areas seeking work as their farms went under. Union organisers thrived, ultimately leading to a bloody confrontation with mine owners and the government. (Many blamed 'Bolsheviks' for the uprisings, thus fuelling anti-Jewish prejudice and restrictions on Russian Jewish immigration as well.) During these lean years, Indians represented an economic threat, especially in trading. In addition, opponents of immigration emphasised the lack of cultural assimilation, evident, for example, in debates over the legality of non-Christian marriages.16 'White civilisation', they argued, was threatened.

The 1920s, in general, marked a period of growing xenophobia, exacerbated by increasingly restricted immigration policies in the United States, which forced would-be migrants to seek other destinations. Following international trends, South Africa adopted a national quota system in 1930, inaugurating a new era of treating immigrants by country of origin, rather than personal circumstances (that is, by group rather than as individuals).17 South Africa followed in the nativist tide of the times.

The numbers, however, do not substantiate white fears of 'swamping'. But as repatriation efforts ebbed, segregationists grew increasingly concerned with the South African-born Indian population, which had even fewer incentives to return to their ancestral homeland – and they could not legally be deported.18 Restrictive domestic legislation increased, foreshadowing both apartheid legislation and future conflict within the United Nations over the treatment of Indians in South Africa.

Immigration policy contained two fundamentally domestic components perhaps unique to the South African setting. Firstly, due to the federation of four distinct colonies at the time of Union, immigration policy also regulated the flow of people between provinces, not simply from 'outside'. The precursors of apartheid influx control, in other words, are already evident by the turn of the century. For example, Indians were not allowed free inter-provincial movement until 1975.19 Secondly, the question of indentured Indians quickly shifted to the question of Indians born within South Africa as demographic trends began to favour the latter. Restrictions on citizenship rights, even more draconian during the apartheid era, thus also emerged as early as the turn of the century.20 Perhaps it is no mere coincidence that the founders of apartheid, notably future Prime Ministers Malan and Verwoerd, served as immigration ministers earlier in their careers.

POST-APARTHEID MIGRATION POLICY

As this historical overview has shown, the regulation of population flows is not new. The apartheid era even included restrictions on internal movement (such as population registration, pass laws, 'independent' bantustans) that were harsher than many other states' policies regulating cross-border migration. The democratic Government has already eliminated the most egregious of these internal policies, but thus far has retained much of the legislative legacy for dealing with external migration. One noteworthy change was the 1990 reduction of the voltage on South Africa's electrical fence on the border of Mozambique from a lethal to stun level.

The extent to which policy continuity will prevail, or alternatively, whether critics advocating a more open policy will make political inroads, depends in part on the strength of international norms (across a number of relevant regimes), the new Government's responsiveness to international expectations, and the extent to which domestic and transnational political actors utilise this international normative framework. While it is too early to give a definitive answer (the government is only now preparing a green paper on migration policy, the precursor to a more formal white paper proposing draft legislation), debates over the past few years highlight trends and warning signs.

The current furore over 'illegal' migrants entering South Africa conflates a number of fairly distinct issues, and the growing xenophobia accompanying this debate further blurs the understanding of the stakes in the debate. Examining this situation through a focus on international norms, however, helps to unravel various themes. As in earlier times, the debate over migration concerns questions of citizenship and entitlements within a discourse of rights (either to be granted or denied).

At the heart of the rhetoric is the term 'illegal'. Yet, while South Africa is currently overturning apartheid legislation right and left, there is a curious silence about the derivation of categories of migrants. The criteria of (il)legality is another apartheid legacy.

Recent immigration legislation, notably the Aliens Control Amendment Act of 1995 and the South African Citizenship Act of 1995, modifies the Government's powers to deport illegals and alter the rules of residency. Despite government claims, these new laws strengthen the state's powers and further limit the possibility of migrants gaining legal status.21 Furthermore, the administrative framework for implementing immigration policy – a system of regional committees appointed by the minister – remains fundamentally consistent with initial Union institutions. Thus, South Africa's current policy is a legacy, not simply of apartheid, but of the interwar period.

That immigration policy needs to adapt to the contemporary international and domestic context is obvious, but it remains unclear how this should happen. The international normative context – regimes – demystifies the situation somewhat. Two themes have persisted throughout the 20th century in migration policy: refugees and human rights. Both are evident in contemporary South African debates.

The status of refugees is one area of policy that has shifted substantially since 1990. South Africa reached an agreement with the UNHCR in 1991 to supervise the repatriation of (anti-apartheid) exiles. Then, after years of denying any obligations in particular to Mozambicans streaming across the border (caused by South Africa's own destabilisation policies), South Africa reached a second accord with the UNHCR in 1993 granting Mozambicans refugee status. Converting them from illegal immigrants to refugees enabled displaced Mozambicans to claim international aid for the first time.

However, the extent of this new turn in policy is limited. This became clear when South Africa announced in July 1996 that it would no longer extend refugee status to Mozambicans after the first of January 1997. In its defence, the Government claimed that the civil war had ended and migrants no longer could justifiably claim flight from persecution. Numerous difficulties with this position have been pointed out in the public debate. These include the persistent problem of landmines that make a return to farming dangerous, if not lethal; the uncertainty of the current political settlement between RENAMO and FRELIMO; and the lease of land to South African corporations which previously belonged to these refugees.

In response to the announced withdrawal of the right to international aid, refugees have already been marching in protest, demanding to remain in South Africa. Thus, international norms are transforming the debate, teaching 'illegal' migrants about the existence of international rights and resources, placing unprecedented pressures on the South African Government to alter its policies. How this political situation will resolve itself is uncertain, but the terrain of debate has changed.

A second dimension of international pressure arises regionally, as members of the Southern African Development Community (SADC) call for a regional migration policy. So far the South African Government has resisted a liberal policy of a free-flow of populations in the region, but also has stopped short of resisting this as a longer term goal. Once again the lines are blurred between refugees and economic migrants, keeping the domestic debate in South Africa hostage to rising xenophobia in a time of budgetary limitations and high unemployment. Yet, important domestic groups – notably the Council of South African Trade Unions (COSATU) – strongly support a co-ordinated regional plan to address questions of employment, development and migration.22 Thus migration policy is clearly linked to broader issues of regional integration and South Africa's economic hegemony.

The recent marches by refugees also highlight the other dimension of the new debate over migrants: human rights. For obvious reasons, the extent to which migrants themselves can actively organise is seriously constrained. But commentators have championed their cause by challenging the persistent denial of basic human rights – another legacy of apartheid. With the adoption of a new Constitution and the creation of a human rights commission, South Africans finally have institutional recourse. One of the primary concerns is the treatment of illegals while in detention, and the general lack of transparency and appeal in processing immigrants.

In this debate over human rights, migrants are just one dimension of the new politics of constitutionalism, itself a product of new international norms reinforcing democracy. Some evidence has emerged to indicate a loosening notion of sovereignty. For example, many migrant labourers resident in South Africa in 1994 were allowed to vote in the first universal suffrage elections.

This is not an attempt to make strong claims about the overthrow of sovereignty and the traditional boundaries of the state in South Africa.23 Yet, the extent to which international norms are penetrating the domestic political process must be emphasised, both framing a new discourse of rights and offering new goals for mobilisation. South Africans are not yet embracing a post-national identity, but there are signs that the international processes moving in that direction in other states are having some effect in what has been an adamantly insulated and isolationist state for almost fifty years.

IMPLICATIONS

This overview of South Africa's immigration policies in the 1920s and 1990s raises a number of bigger theoretical questions. It recognises population pressures as part of the processes of 'social construction' at the heart of a more sociological and less economistic approach to world politics. An analysis of migration sheds light on the agency side of these processes, in part by highlighting the role of actors which are not formally organised (that is, migrants themselves). As 'agents', migrants respond to opportunities and constraints created by international social structures, create pressures on states, and subvert state authority. These are just a sampling of the types of activities by which non-state actors construct and reconstruct state identities and interests.

This institutional approach also adds to the study of migration, particularly highlighting the need to understand international normative structures. Like many areas in international relations, migration analyses tend to be dominated by implicit, if not explicit realist assumptions. Alternatively, other disciplinary approaches tend to emphasise economic dimensions, such as changes in global production patterns and competitiveness (especially for understanding labour flows). To the extent that any of these approaches take a more social or cultural perspective, they tend to be micro-level studies of migrants' decisiom-making processes or local community relations. International relations theory has much to add to these analyses, including explicit attention to state theory.

The evolution of immigration policy has substantial implications for the Southern African region. South Africa is no longer insulated and isolated. It needs to develop new policies which are sensitive to international normative constraints, as well as to the regional context. To the extent that we can expand our conceptual tools and empirical understanding, we can hope to develop more effective policies which address and perhaps reconcile multiple economic, social and ethical demands.

ENDNOTES

For a comprehensive study of the ANC's foreign policy, see S Thomas, The Diplomacy of Liberation: The Foreign Relations of the ANC since 1960, I B Taurus, London, 1996. On the rivalry between the ANC and the South African state for diplomatic recognition, see A Klotz, Diplomatic Sanctions against South Africa, in N Crawford and A Klotz (eds.), How Sanctions Work: South Africa (forthcoming).



A Klotz, Norms and Sanctions: Lessons from the Socialization of South Africa, Review of International Studies, 22, April 1996, pp. 173-190.



F Kratochwil and J G Ruggie, International Organization: A State of the Art on the Art of the State, International Organization, 40 (4), Autumn 1986, pp. 753-775; A Wendt, Collective Identity Formation and the International State, American Political Science Review, 88(2), June 1994, pp. 384-396; A Klotz, Norms in International Relations: The Struggle against Apartheid, Cornell University Press, Ithaca, NY, 1995; P Katzenstein (ed.), Culture and Security, Columbia University Press, New York, 1996.



For a provocative argument that stresses the role of international norms and institutions in constructing states, see R H Jackson, Quasi-States: Sovereignty, International Relations and the Third World, Cambridge University Press, New York, 1990.



See, for example, A Wendt, Anarchy is What States Make of It, International Organization, 46(2), 1992, pp. 391-425; D Campbell, Writing Security, University of Minnesota Press, Minneapolis, 1992; contributions in Katzenstein, op. cit.



See especially R B J Walker, Inside/Outside: International Relations as Political Theory, Cambridge University Press, New York, 1993.



For example, F Kratochwil, Norms, Rules and Decisions, Cambridge University Press, New York, 1989, and Citizenship: On the Border of Order, in Y Lapid and F Kratochwil (eds.), The Return of Culture and Identity in IR Theory, Lynne Rienner, Boulder, 1996, pp. 181-197.



See especially C Lynch, Beyond Appeasement: Interwar Social Movements and Change in World Politics, manuscript, Northwestern University, Chicago, Illinois, June 1996.



M R Sinclair, Unwilling Aliens, manuscript, University of Western Cape, Bellville, June 1996.



On domestic incorporation regimes, see Y Soysal, Limits of Citizenship, University of Chicago, Chicago, 1994.



The exception was those from neighboring British protectorates. See E Bradlow, Immigration into the Union 1910-1948: Policies and Attitudes, unpblished Ph.D. dissertation, University of Cape Town, 1978, p. xvi.



For a concise overview of the history of migrant labor, see A H Jeeves and J Crush, The Failure of Destabilisation Experiments on South African Gold Mines, in Crossing Boundaries: Mine Migrancy in a Democratic South Africa, IDASA, Cape Town, 1995, pp. 2-13.



For a detailed study of the work of some of these immigration societies, see C Swaisland, Servants and Gentlewomen to the Golden Land: The Emigration of Single Women from Britain to Southern Africa, 1920-1939, University of Natal, Pietermaritzburg, 1993.



Bradlow, op. cit., pp. 111-112, 127.



Ibid., pp. 10-16, 113-116.



In other words, the question of the rights of more than one wife (and children) to immigrate remained contentious and highlighted cultural differences; ibid., pp. 83-93.



Ibid., pp. 221-226.



Ibid., pp. 116-120, 124.



Ibid., p. 184.



Ibid., pp. 14, 20-21.



Government Gazette, Aliens Control Amendment Act No 76 of 1995, 364(16741), pp. 1-41; South African Institute of Race Relations, Annual Survey 1995/96, SAIIR, Johannesburg, 1996, pp. 5-7.



See H Kotze and L Hill, Illegal Immigration in a Democratic Southern Africa: Polemic and Perception in the Search for Policy, paper presented at the biennial Conference of the South African Political Studies Association, 27-29 September 1995, pp. 11-12.



For example, South Africa's policies are no where near as dramatically post-national as those analysed by Soysal, op. cit.