How far does ‘societal security’ travel? Securitization in South African immigration policies

Abstract

Responding to political developments in Europe during the 1990s, the Copenhagen School drew on speech act theory to argue that state leaders represent certain issues, including immigration, as existential threats to society. Two decades of friendly amendments and vociferous critiques have raised questions about how well the Copenhagen School’s core concept of ‘societal security’ travels outside Europe. To assess the scope of this ‘securitization’ framework more systematically, we examine South Africa, a democracy that recently liberalized its immigration policies despite ethno-nationalist and racist traditions. Specifically, we test four claims: (1) that official discourses will target certain foreigners as an existential threat to collective identity; (2) that bureaucracies will consistently institutionalize these discourses; (3) that identity-oriented groups will be crucial to any societal contestation over these discourses; and (4) that successful securitization produces regionalization. These securitization claims hold up well, even though the nature of threats to societal security shift over time. Keeping in mind that no theory is without weaknesses, we recommend wider integration of the societal security concept into comparative studies of immigration policy, especially in democracies outside Europe.

Keywords

immigration national identity securitization societal security South Africa xenophobia

Introduction

Responding to European political developments in the 1990s, a nascent Copenhagen School proposed that state leaders can use speech acts to characterize an existential threat to a ‘valued referent object’, thereby leading a ‘significant audience’ to accept unprecedented ‘extraordinary measures’ (Buzan et al., 1998: 23–27; Wæver, 2011: 469).¹ A lively debate ensued over the concept of ‘societal security’ and related processes of social construction across a range of issues, including the impact of immigration on collective identity (see e.g. Bigo, 2002; Boswell, 2007; Herd and Löfgren, 2001; Huysmans, 2000; Van Munster, 2009). Yet, after two decades of friendly amendments and vociferous critiques, this ‘securitization’ framework has arguably become divorced from its original underpinnings in speech act theory (Balzacq, 2011a; Ciutâ, 2009). Consequently, some analysts now employ the term to refer simply to a generic sense of threat (Adamson, 2006; Chebel d’Appollonia, 2012), whereas others lament the theory’s devolution into an overly narrow view of the social construction of security (McDonald, 2008: 566).

Furthermore, the relevance of the securitization framework outside Europe remains uncertain. Advocates offer fruitful extensions to North America (e.g. Salter, 2008a; Watson, 2009), but skeptics challenge its utility in non-Western autocracies (e.g. Vuori, 2008; Wilkinson, 2007). Applications to a plethora of non-comparable issues, ranging from airports and refugees in Canada to social unrest in Kyrgyzstan and China, complicate any assessment. Alternatively, sympathetic applications suggest that democratic states offer the most suitable terrain (Tjalve, 2011; Williams, 2011). However, at least for the issue of immigration, former settler societies such as Canada lack crucial similarities with European countries, most of which have only grappled with significant population inflows since the 1950s (Cornelius et al., 2004). These contrasting views of potential geographic or institutional limits on scope underscore the need for more systemic testing across non-European cases.

Accordingly, we examine responses to immigration in South Africa – a settler state that has greater commonalities with European countries than Canada – to assess whether the core concept of societal security successfully ‘travels’ or whether it suffers from too much conceptual ‘stretching’ (Sartori, 1970). Our query rests on the plausibility of securitization: South African immigration policies were based throughout the 20th century on strong traditions of ethno-nationalism, the key Eurocentric characteristic initially invoked by Wæver (1993: 22–23, 40).² Tracing changes over time within one case of a ‘securitized’ state–society relationship also holds constant numerous other factors, such as corporatist links between economic and political spheres, in order to focus on geography. In addition, South Africa’s implementation of non-racial franchise in 1994, followed by selective reforms to its immigration policies, allows us to explore how a deepening of democracy might weaken or ‘desecuritize’ dominant discourses.³

Specifically, we test four claims embedded in the securitization debate:

State leaders will articulate existential threats to societal identity through speech acts – the original Copenhagen formulation (Buzan et al., 1998).

Although such speech acts might invoke exceptional responses, a discourse of threat is more likely to be intensified within everyday practices of bureaucracies – the so-called Paris School position (Bigo, 2002).

Identity-based contestation should play a crucial role in any societal challenges to this state-led discourse – a general constructivist insight (Doty, 1998).

Successful securitization, exemplified by European integration, leads to an expansion of societal identity beyond national boundaries (Huysmans, 2000).⁴

While detailing the rationale for each of these four tests in the next section, we address one additional methodological issue: the need for context-sensitive metrics (Balzacq, 2011b: 35–38). Each empirical assessment is then delineated in the subsequent section on South Africa.

Ultimately, we find that the securitization framework holds up well, despite some weaknesses. Discourses of existential threat employed by South African leaders, alongside corresponding bureaucratic practices, do capture crucial tensions between immigration and national identity. In addition, identity-oriented groups have indeed led societal contestation over these discourses, although other influential societal actors should not be overlooked. Even the lack of regional integration in Southern Africa follows the securitization logic, although this outcome diverges empirically from the high degree of European integration. In the final section, therefore, we call for more sustained engagement between securitization and other immigration theories, especially in Asian and Latin American democracies, to determine how much farther the concept of societal security can travel.

Testing securitization

Theory testing challenges the epistemologies of many securitization theorists (Gad and Petersen, 2011: 320–322; see also Balzacq, 2011a; Guzzini, 2011). Yet, even if they are not termed hypotheses, explicit claims are readily evident in the extensive debate over the key concept of societal security and the processes through which collective identity might be constructed. A wide range of critics have targeted the Copenhagen School tenet that state leaders construct societal security by identifying an issue (e.g. immigration) or group (e.g. immigrants bringing new cultural practices) as an existential threat to national identity. Alternatively, some stress that the presence (or marginalization) of other societal, political, and professional actors may be more significant (Bigo, 2002; Hansen, 2000; Huysmans, 2000; Salter, 2008a). In addition, underlying disagreements over speech acts and practices suggest that gradual intensification, other forms of political communication, or broader social structures might matter as much or more than episodic speech acts (see e.g. Balzacq, 2005; Ciutâ, 2009; Stritzel, 2007; Ciutâ, 2009; Watson, 2009; Williams, 2003).

Demarcating more precise claims is the first step. Table 1 untangles the theoretical stakes that have emerged during the past two decades of debate, with illustrative references to key interventions. Although individual authors might place themselves in another category, or even distance themselves from the securitization label, we treat all these variants as modifications within a common framework that views threats as social constructions and accepts that collective identities matter. From there, substantial disagreements emerge, which we cluster into three categories: actors, actions, and outcomes. These disagreements lead to both process-oriented and outcome-oriented assessments, which we outline in the remainder of this section. In addition, to locate the characteristics of a discourse most relevant to each test, we draw upon the extensive qualitative methods literature, which has already grappled with the relationship between speech and action.

Table I.

Core assumptions of securitization theories.

	Copenhagen School	Modifications
Actions	Speech acts	Silence (e.g. Hansen, 2000)
		Images (e.g. Williams, 2003)
		Physical action (e.g. Wilkinson, 2007)
		Practices (e.g. Balzacq, 2005)
Actors	State as speaker	Bureaucracy (e.g. Bigo, 2002)
Actors	Society as audience	Societal groups (e.g. Huysmans, 2000)
Outcomes	Extraordinary measures	Linguistic structures (e.g. Stritzel, 2007)
		Iterative processes (e.g. Salter, 2008a)
		Risk management (e.g. Van Munster, 2009)
		Social contestation (e.g. Doty, 1998)

Test 1: Articulations of threat

We start by applying the original Copenhagen School formulation, summarized in the middle column of Table 1. Consistent with the notion of a successfully securitized society, South Africa has a long history of excluding certain immigrants, because the state actively used its policies to shape the physical and cultural composition of society. At the turn of the 20th century, fear of Asians dominated debate, resulting in racist exclusions (Huttenback, 1971; Klotz, 2013: Chapter 2). Concurrently, increasingly precise procedures effectively filtered out undesirable whites until the end of minority rule in 1994 (Klotz, 2013: Chapter 4; Peberdy, 1999). Meanwhile, Africans received less attention in these debates, as they had never been viewed as potential immigrants: the brutal apartheid system mainly added harsher methods of enforcing their exclusion (Evans, 1997; Klotz, 2013: Chapter 3; Posel, 1997).

A puzzle, however, immediately emerges. The major post-apartheid policy reform, the Refugees Act of 1998, offered unprecedented rights to non-nationals, especially foreign Africans. If an exclusionary conception of existential threat remains entrenched, as the Copenhagen School would lead us to expect, why liberalize at all? Alternatively, if foreign Africans are no longer viewed as an existential threat in a non-racial society, which the demise of apartheid would lead us to expect, why do detention and deportation remain keystones for the 2002 Immigration Act?

Potential answers drawn from the initial Copenhagen School formulation focus on the state’s speech acts. Perhaps post-apartheid leaders chose not to reiterate the dominant discourse of exclusion, in effect enabling normal parliamentary politics to proceed on issues of immigration policy. Or those leaders may have tried to invoke an existential emergency, but perhaps their speech acts failed to convince the relevant audience. Accordingly, we examine the presence or absence of state-led speech acts in order to gauge any discursive shifts in the representations of foreigners in articulation of threats. Narrowing down the amount of evidence, not least for reasons of feasibility, we concentrate on official responses at the critical moment (Dunn, 2008) of exceptionally widespread and violent anti-foreigner attacks in May 2008, supplemented with a countervailing government-sponsored anti-xenophobia campaign.

Drawing on the commonplace notion within the immigration literature of negotiated identities (see e.g. Kastoryano, 2002; Lesser, 1999), we characterize actors as either negotiating or silent about potential threats. Crucially, our use of the term ‘silence’ to characterize a discourse of the state differs significantly from others’ reference to the marginalization of alternative discourses by the state. Thus, we eschew the term’s normative connotations, in contrast to both Hansen’s (2000) post-structuralist critique of the Copenhagen School and Schimmelfennig’s (2001: 65) notion of successful rhetorical action.⁵

Test 2: Institutionalized discourses

In current debates over contemporary South African immigration policy, one answer to this puzzle about the disjuncture between historical legacies and recent reforms concentrates on bureaucracies: even if political leaders want deeper reform, enforcement agencies, notably Home Affairs and the police, routinely replicate past practices (see e.g. Vigneswaran, 2008). This argument parallels that of the so-called Paris School, noted in the middle row of Table 1, which emphasizes the distinctiveness of bureaucracies as ‘professionals in charge of the management of risk and fear’ whose purview gradually expands to include immigration issues (Bigo, 2002: 63). These bureaucracies arguably implement the everyday practices of a general Western logic of control that is more powerful than alternative societal discourses; gradual intensification also trumps any speech acts of leaders (Bigo, 2002: 64–65).⁶

Such intensification should produce consistent discourses of threat across bureaucracies, and the apartheid-era South African state certainly fits this picture, with draconian practices of racist exclusions and forced removals, administered by an authoritarian security apparatus (Cawthra, 1993). Yet why would these features intensify after the demise of white minority rule in South Africa? Remedying excessive determinism in the Paris School, Van Munster (2009: 6–7) suggests that key political moments open space for alternatives. Intergovernmental bargains play this role in Europe. For South Africa, 1994 should be such a critical moment.

To test for key political moments, we build on the complementary argument that bureaucracies routinely configure categories for data analysis and that these categories articulate positive or negative images (Anderson, 2001; Caplan and Torpey, 2001; Curtis, 2001; Sadiq, 2005; Scott, 1998). Thus, a genealogy (Vucetic, 2011) of institutional migration data can gauge the extent to which historically rooted representations of national identity endure or ebb in the post-apartheid era. Potentially, therefore, the South African case challenges Wilkinson’s (2007: 10) dichotomy of Western versus non-Western (or Third World) state–society relations. More generally, this test also assesses whether the state speaks with a unitary voice, as initially articulated by Wæver (1993: 36), or whether autonomous bureaucracies create the discursive focal point for securitization, as Bigo (2002: 71–74) suggests.⁷

Specifically, territorial control (Bigo, 2002: 74–75) would be manifest in a link between formal borders and the labels given to foreigners. Also, symbolic control over collective identity (Bigo, 2002: 76–78) would be evident in representations of national or transnational belonging used in these labels. On the basis of the content of these data categories, we characterize stable discourses about foreigners as expressing either a dichotomous distinction between insiders and outsiders or a layered view that incorporates multiple forms of outsiders. However, we do not presume that binary distinctions between foreigners and citizens are normatively less desirable (cf. Bigo, 2002: 79). Furthermore, the possibility of unstable discourses points towards a great role for societal contestation, a point we address next.

Test 3: Societal contestation

Many securitization theorists do acknowledge that societies are not empty vessels that merely take up speech acts (Doty, 1998; Hansen, 2000; McDonald, 2008; Salter, 2008a). And, given its history of exceptionally contentious parliamentary and extraparliamentary politics, as well as the deepening of democracy with the country’s first non-racial elections in 1994, South African society should manifest a high level of contestation over any discourses of threat. Yet political contestation can take an almost infinite number of forms, not all of which are necessarily salient. In particular, only claims based on identity would be logically consistent with societal security arguments. Our third test therefore requires delineation of hypotheses, thus far mostly implicit, about which actors or activities would characterize identity-based involvement in immigration policy debates.

Facing a putatively existential threat from foreigners, the most relevant actors from the Copenhagen School’s ethno-nationalist perspective should be cultural movements acting rhetorically, whether in agreement with state discourse or challenging it. However, European experiences arguably underscore the key role of rights advocates in creating an alternative civic identity that tolerates greater diversity (Jacobson, 1996; Kastoryano, 2002). Furthermore, Bigo (2002: 79) makes the counter-claim that European rights discourse is actually a ‘by-product’ of securitization, because acceptance of asylum-seekers reinforces an underlying binary distinction between citizens and foreigners. In the light of these contradictory arguments, securitization analysts should pay closer attention to rights discourse, especially the question of whether it can move the immigration issue out of an ‘emergency mode into the normal bargaining process of the political sphere’ (Buzan et al., 1998: 4).

To distinguish contending claims about the role of contestation, we compare securitization theory to the predominant theory of migration policy in democracies, which predicts interest-based forms of rights advocacy linked to political coalitions. In this alternative coalitional framework, rights advocates promote policies that benefit their organizational interests, typically leading to a political alignment with businesses that also favor liberal immigration policies (Hollifield, 1992; Tichenor, 2002).⁸ If these key interest groups disagree over policies, their opposition takes place primarily in direct negotiations with government actors or through political parties. Also, coalition theorists frequently argue that the extension of rights to immigrants precludes a return to previous levels of restriction because of political and legal factors (Hollifield, 1992, 2004; Jacobson, 1996; Joppke, 1998). Thus, coalition theory would predict a major role for political institutions, especially courts, for an outcome that might be described as desecuritization.

Another significant difference between the two theories is the process through which societal actors respond. Public challenges based on alternative discourses are consistent with securitization theory, whereas interest-based bargaining marks the instrumentalist logic of coalition theory. Therefore, we treat evidence of highly coordinated rights advocates seeking to influence bargains between the state and key economic actors, either directly or through the courts, as favoring a coalitional argument, whereas public contestation by either economic actors or rights advocates suggests a process more consistent with rhetorical action. Although businesses and unions play no explicit role in the Copenhagen School framework, we would expect to see these key economic actors make public arguments against attempts at securitization within immigration policy.

Because our aim is to characterize societal contestation, we eschew the extensive debates over measuring discourse (e.g. semiotics versus content analysis). Instead, we concentrate on which actors make what types of arguments, which we describe mainly in terms of their consistency with the securitization framework. On the basis of the divergence or convergence of discourses across key societal actors, we then summarize contestation as either weak or strong. However, the coherence of contestation (or lack thereof) does not necessarily predict policies. Therefore, we also examine claims about outcomes.

Test 4: Policy outcomes

The implications of securitization, at least for immigration policy, have thus far been based mainly on inferences from experiences in the European Union. Since immigration policy itself has multiple facets, including labor markets, political asylum, citizenship, and border controls, we focus on refugees as one aspect likely to involve societal security considerations. Specifically, regionalized procedures such as the ‘safe third country’ rule give EU member-states the ability to reject outright any asylum seeker who entered from any other member-state. Designed to prevent ‘asylum-shopping’, this rule requires that a claim be made at the first safe opportunity. That country may not be the claimant’s preferred destination. Because all EU members are ‘safe’ by definition, in essence this principle reinforces their democratic identities through a shared commitment to ‘freedom, security and justice’ (Van Munster, 2009: 2–3).

Building on this trajectory, securitization theorists have suggested that the extension of societal identity to encompass the region’s democracies as a referent object led to the creation of a shared European Union (EU) border (Huysmans, 2000). This process started with the unprecedented flows of refugees generated by the Balkans conflicts of the early 1990s. Generous yet costly protections for these refugees contributed to a budgetary crisis in Germany. In response, Germany exported the safe third country approach to its neighbors through bilateral agreements, while simultaneously pushing immigration issues up to the regional level (Post and Niemann, 2007). This formulation of a common EU approach, which broadened the borders of Germany’s societal security zone by characterizing refugees as a threat to European society as a whole, thus intensified barriers against non-EU (mainly poorer and Muslim) migrants.

Facing a similar asylum crisis since the early 2000s, South Africa has also attempted to implement the safe third country rule. This policy preference suggests the potential applicability of securitization theory. As in Europe, responses to refugees highlight tensions over rights. For instance, asylum claims jumped at least tenfold in the decade after the 1998 Refugees Act confirmed constitutional and international protections for refugees.⁹ Critics blame these new procedures for fostering false applications from economic migrants claiming to be refugees. Meanwhile, rights advocates complain that persistent procedural irregularities and outright abuses deny legitimate asylum claims. Not surprisingly, post-apartheid South African policymakers, looking for models, considered European and North American approaches.

However, directly comparing regions with dramatically different histories makes little sense. Instead, we delineate clearer claims about the dynamics of regionalization in order to test the securitization logic in the Southern Africa case. Once again, we start with the Copenhagen School, which sought to distinguish itself by stressing identity-based threats and social stability rather than prioritizing national security and strategic stability (Adamson, 2006: 181–182; Buzan et al., 1998; Wæver et al., 1993). For this analysis, political priorities are more crucial than speech acts.¹⁰ Extrapolating from the EU example, we would expect states to regionalize collective identity as the referent object of security in order to protect domestic social stability in the face of large refugee flows or a similar migration-induced crisis. Alternative policy priorities, ranging widely from other domestic political considerations to strategic concerns, serve as evidence contrary to a securitization dynamic.

A second variant treats securitization as complementary to traditional strategic approaches. In particular, Rudolph’s (2003) sequencing argument posits a relationship between strategic, economic, and societal securities. In times of high geostrategic threats, he predicts, states emphasize defense policies and economic production. Consequently, they prefer more open migration policies that accommodate demands for labor mobility. The production of wealth and rallying around a common purpose mitigate perceptions of societal threat. Inversely, decreasing threat at the geostrategic level should allow for tightening borders and flaring domestic divisions.¹¹ The link between migration and security is thus less direct: migration may affect economic prosperity, which in turn affects the capacity to invest in military power. Refugee flows, therefore, should only play a role when substantial enough to create financial concerns.

These two lines of argument enable us to untangle alternative mechanisms underpinning regionalization, which are conflated in the European case. Securitization theory emphasizes acceptance of existential threats to a broader regional collective identity, which spurs integration. In contrast, traditional strategic arguments stress externalization of national interests, where regionalization assuages geostrategic threats that include a migration component. These distinct strategies thereby point to either an expanded construction of societal security that extends beyond traditional nation-state boundaries or retention of a restricted definition based on those traditional boundaries and their corresponding notion of national interest.

Securitization in South Africa

Like many European countries, South Africa has received significant migration inflows at the same time as it has resisted discourses that embrace immigration. Anti-immigrant mobilization has also been on the rise, as in Europe. In particular, percolating xenophobia since 1994 escalated into exceptionally widespread attacks in 2008. Yet demographic trends appear stable. Although reliable statistics for South Africa are not readily available, 2008 data from its Department of Home Affairs enabled us to estimate that over 2 million migrants entered legally and illegally into a country of approximately 50 million people.¹² This estimate suggests a contemporary foreign African population in the same 8% range as apartheid-era estimates (Jeeves, 1985; Owen, 1964). Thus, no objective change in migration flows, at least at the aggregate level, can explain the flare-up. Instead, we find that the securitization framework captures reasonably well why fear of foreign Africans remains entrenched, despite the demise of apartheid. This section elaborates on the summaries in Table 2, using the qualitative metrics previously outlined for each test.

Table 2.

Societal security in South Africa.

Test	Metric	South Africa
Articulations of threat	Negotiated or silent	Silent
Articulations of threat	Negotiated or silent	State perpetuates xenophobia by not countering widespread discourse that targets foreign Africans as threat.
Institutionalized discourses	Dichotomous or layered	Dichotomous
Institutionalized discourses	Dichotomous or layered	State gradually shifts from multilayered categories of foreigners to distinguishing between citizens and all foreigners.
Societal contestation	Strong or weak	Weak
Societal contestation	Strong or weak	Societal challengers make limited inroads against state discourse of Africans as threat.
Policy outcomes	Expanded or restricted	Restricted
Policy outcomes	Expanded or restricted	State affirms national identity and rejects pan-African regionalization of society as referent object.

Test 1: Articulations of threat

Because the concept of societal security draws attention to a potential sense of emergency in reaction to domestic social conflict, outbreaks of widespread xenophobic violence in May 2008 provide an opportunity to explore the Copenhagen School’s claim that political leaders will articulate existential threats on behalf of the state and will call for emergency responses. Our sampling of discourses at that time compares South African leaders’ muted immediate reactions, which primarily treated the riots as outbursts of criminal violence, with discourses from political rivals and other branches of the state.

Initially, politicians responded to unusually widespread riots by blaming the victims and downplaying the role of street-level gangs. Despite routine media coverage of sporadic xenophobic attacks for over a decade, President Thabo Mbeki expressed surprise and shame about the violence (Mail and Guardian, 2008). Meanwhile, some of the attackers appropriated Deputy President Jacob Zuma’s signature song, Umshini Wama (‘Give Me My Machine Gun’). In response, Zuma insisted that he did not condone the attacks and claimed ‘these people are abusing this song’, but the perception remained among many urban residents that the government would ‘tell Mugabe [the president of neighboring Zimbabwe, the largest source of recent migration] to come and get his people’ (Rossouw, 2008).

Ironically, Mangosuthu Buthelezi of the Inkatha Freedom Party, who displayed the most concern, has routinely been blamed for fuelling xenophobia during his tenure as Minister of Home Affairs from 1994 until 2004. His oft-quoted statements warning that ‘millions of aliens who are pouring into South Africa’ undermined government efforts at redistribution fit into a post-apartheid discourse of economic nationalism.¹³ In 2008, he sought to dispel accusations that his followers were the driving force behind the attacks but without challenging the prevailing discourse. One supporter summed it up: Buthelezi ‘cried because the Zulus [his primary constituency] have been fingered in recent incidents, because the name of the party has been dragged through the mud and because government has taken so long to address the foreigner issue’ (Dibetle, 2008a).

Official lack of action also reflected distancing from responsibility. For instance, after two weeks the South African police had failed to quell the violence; finally, the army intervened to restore order, an unprecedented measure in the post-apartheid era (International Organization for Migration (IOM), 2009). Also, few of the attackers were prosecuted for murder or larceny (Govender, 2008). Meanwhile, the Department of Home Affairs claimed that provincial authorities should provide assistance for the thousands of displaced people, while provincial authorities pointed back to Home Affairs (Dibetle, 2008b). At times, Home Affairs and police officials even treated the situation as an opportunity to deport immigrants residing in temporary camps, regardless of their legal status (United Nations High Commissioner for Refugees (UNHCR), 2010).

Still, the demise of apartheid certainly had shifted some elements of state-led discourse. Notably, the Roll Back Xenophobia campaign, launched in December 1998 by the South African Human Rights Commission (SAHRC) along with the National Consortium on Refugee Affairs and the UNHCR, focused mainly on media coverage of immigration through educational seminars for journalists and the development of materials for dissemination. The content stressed that being a rights-respecting state required support for constitutional protections for foreigners (McDonald and Jacobs, 2005; SAHRC, 1999). Yet perpetrators of the attacks in May 2008 paid no regard to whether their neighbors had official refugee status or even South African citizenship, indicating the ineffectiveness of the campaign.

Furthermore, slow suppression of further attacks and lack of official assistance to victims of those attacks reinforced the message conveyed by the muted initial response of national leaders. In particular, the government took care not to create any appearance of offering more to foreigners than to nationals. Since township dwellers themselves cited competition for subsidized housing as one of the key issues driving the attacks, officials repeatedly stressed that emergency camps (along the lines of international relief standards) were not houses or luxurious accommodations (Joubert, 2008). In addition, most camps were disbanded after a few months, precisely to prevent their conversion into informal settlements.

Overall, these responses to an exceptional outbreak of xenophobic unrest in South Africa demonstrate the resilience of securitized discourses. Officials continued to frame foreigners as the core threat and their physical exclusion as the solution. Since the African National Congress (ANC) has remained the majority party since 1994, with no major opposition party opting to challenge xenophobia, negotiated modifications in the predominant discourse are also unlikely in the future. Still, we found evidence that supports friendly amendments to the Copenhagen School framework, as noted in the ‘Actions’ row of Table 1: silence as a form of speech act and physical actions as an alternative to speech acts can perpetuate prevalent discourses. Therefore, because its leaders do not need to reiterate a sense of existential threat, we use ‘silence’ to categorize this articulation of threat.

Taken together, these representations, based on diverse forms of verbal (e.g. Vuori, 2008) and non-verbal (e.g. Williams, 2003) modes of communication, illustrate the ontological relationship of state-as-speaker and society-as-referent-object.¹⁴ Yet speech act analysis alone, no matter how salient, is clearly insufficient for fully understanding South African discourse or policies. We have already hinted at a link between official speech acts and the bureaucratic side of the state, such as the role of the police and the creation of new commissions. In the next subsection, therefore, we focus on claims that securitization should be conceived as a process of intensification through which societal security becomes institutionalized in bureaucratic practices. Indeed, this shift broadens the debate to include the role of political institutions, including the presence or absence of democratic features of the state, which is so central to Wilkinson’s (2007) charge of Eurocentrism.

Test 2: Institutionalized discourses

Africans remain the overarching target of xenophobia in South Africa, but the dangers employed are unique to each historical era: women from Lesotho in the 1960s, Mozambican refugees in the 1980s, and undocumented Zimbabweans now. Data discourses that bureaucracies use to categorize and measure immigration can gauge such ‘threats’ across time and across institutions. For instance, South Africa has gradually shifted from a multilayered system in colonial times, under which Africans from the region and other immigrants were treated in different ways, to a simpler dichotomy in the post-apartheid era, which primarily differentiates between citizens and foreigners.¹⁵

Historically, South African treatment of foreign Africans is rooted in their role as migrant laborers. It was the mining sector rather than the government that tabulated the movement of foreign laborers in great detail, in accordance with arrangements between imperial governments (Jeeves, 1985). Therefore, state immigration data did not seek to track their flows. The one exception was Africans from the neighboring British protectorates (now Botswana, Lesotho, and Swaziland), who had generally been treated on a par with black South Africans. Until 1950, statistics referred to them collectively as ‘British’ Africans (Peberdy, 1999). Despite strictly enforced restrictions and the lack of any family unification provisions, a considerable number settled in the cities, part of a larger trend of urbanization that apartheid aimed to halt (Evans, 1997; Posel, 1997).

Meanwhile, the 1913 Immigrants Regulation Act required an ability to assimilate, in effect allowing only whites. Germans and Dutch figured prominently alongside English-speakers in the government’s attempts to bolster its desirable settler population. In the 1920s and 1930s, new quotas targeting less desirable East European Jews led the state to generate detailed statistics on birth, place of prior residence, and citizenship (Bradlow, 1978; Peberdy, 1999; Shain, 1994). Only from the 1950s onward, during the height of decolonization, did the apartheid government shift emphasis to whites as a homogenous community. Indeed, demographic survival for whites was at the heart of all apartheid policies.

Even as late as the 1960s, nationality was still fluid in South Africa. Within the British Empire, not all immigrants were automatically foreigners because many were imperial subjects. Therefore, South Africa’s declaration as a republic independent from Britain in 1961 created a novel problem: despite being long-time residents with families, British subjects of all races instantly became foreigners who needed to be identified and possibly deported. Although most foreign Africans came from tiny Lesotho, itself surrounded by South Africa, the Froneman Committee (tasked with finding a solution) warned that an ‘influx’ of these foreigners threatened to ‘swamp’ South Africa (Owen, 1964). Implementing the Committee’s recommendations, the Aliens Control Act of 1963 established the first regional border posts. And in 1965, the former protectorates were finally listed as a separate category in immigration statistics, confirming that African mobility was now regulated through nationality.

Since the 1960s, the nationality categories used by the South African state to track migration have demonstrated a basic hierarchy between foreign Africans and Africans born in South Africa (albeit denied citizenship under apartheid). Immigration statistics even directly mirrored the creation of the nominally independent Bantustans in the 1970s, as well as their subsequent reincorporation. The status of non-African foreigners – mainly whites fleeing former colonies, who were viewed as desirable immigrants because they were savvy about segregation – continued to be treated in an ad hoc fashion that facilitated their entry and integration into the 1980s (Peberdy, 1999).

The only major change in the post-apartheid era was the passage of the 1998 Refugees Act, which finally established asylum policies and procedures. This recent recognition of rights for refugees has reinforced the binary distinction between citizens and foreigners in the discourse of threat. Both politicians and the public, including perpetrators of the May 2008 attacks, routinely express concern about the fiscal impact of illegal immigrants masquerading as asylum seekers and thus overburdening social services. Yet intensification of border control in the early 1960s predates Bigo’s (2002) portrayal of a contemporary Western obsession with technologies of surveillance. Also, anti-foreigner sentiment in South Africa does not manifest a cultural definition of societal security, so notable in Europe.

This fluctuation in state-generated categories reflects predominant representations of foreigners and thus confirms that ‘the state’ can speak with a unitary voice, without denying changes over time or challenges by societal actors. Thus, the South African case refutes Wilkinson’s (2007) dichotomy between Western and non-Western understandings of the state–society relationship, because the centrality of nationality actually varies over time. Indeed, contestation is likely to be a major source of change in discourses and policies, as the next subsection confirms.

Test 3: Societal contestation

Contestation comprises many forms of competition. Juxtaposing identity-oriented and interest-based arguments untangles distinctive insights that help to explain apparent contradictions in recent South African reforms. Notably, corporatist links between economic and political spheres (Lipton and Simkins, 1993) provide economic actors such as the mining industry with the ontological standing of discourse-producing agent rather than mere audience. Also, rights advocates frequently engage in political bargaining and especially court challenges, such as those concerning new protections for refugees in South Africa (Handmaker et al., 2008). These activities demonstrate vertical competition by societal actors against overarching identities propagated by state discourses.

Since immigration is a multifaceted phenomenon that provokes many concerns beyond national identity, the absence of any economic actors in the original Copenhagen School framework is a serious limitation. However, its logic remains relevant: attention to market-based actors and rights advocates in South Africa demonstrates how the politics of coalitions and speech acts can converge around a securitization of economic interests. Crucially, major economic actors remain relatively silent in the public sphere, a legacy of apartheid-era agreements that created a dual economy. The mining conglomerates, historically the sector most reliant on foreigners, continue their arrangements for contract labor in the post-apartheid era (Crush, 2000; Crush and Tshitereke, 2001). Consequently, immigration policy is formulated on two parallel tracks, one in which business negotiates with the state, the other in the public sphere.

This bifurcation harbors a number of implications. To a large extent, the privileged position of the mining sector remained off the main policy-reform agenda in the 1990s owing to the industry’s own discursive framing of the contract system as international agreements to be honored and as a cost-effective service to the state (Chamber of Mines, cited in Crush and Tshitereke, 2001: 54–56). The industry’s partner-countries, reliant on remittances, also rarely challenged the system. This non-public track means that the mining sector does not have a big stake in public debate or parliamentary politics. Thus, in contrast to the situation in Europe, South Africans’ deeply entrenched xenophobic securitization discourse does not have to compete with an alternative pro-market discourse.

In addition, unions (legalized only in the 1980s) have played a more limited role than their European counterparts in shaping or challenging the predominant discourse. While the Congress of South African Trade Unions, an ANC ally, did challenge the two-track bifurcation, many of its proposals did not get incorporated into subsequent legislation (Crush and Tshitereke, 2001). Most notably, the National Union of Mineworkers came out early as a strong advocate for the rights of long-term residents, who were allowed to vote in the 1994 elections and to regularize their status in a series of amnesties during the late 1990s. Yet, aside from condemning the 2008 attacks as misplaced scapegoating that distracts from employer exploitation, non-mining unions have not engaged in the debate over xenophobia (Hlatshwayo, 2011).

Other forms of contestation are certainly evident, however. For instance, lawyers successfully navigated the courts to block the dismantling of camps in 2008, while some local communities managed to preclude or at least dampen violence, owing to the efforts of individuals and organizations that objected to the premise that foreigners were a threat (IOM, 2009). The attacks also generated widespread efforts at the level of civil society to set up support networks for victims dealing with issues ranging from legal and humanitarian assistance to community and government relations (Everatt, 2011). Yet these advocates remain relatively weak, partially owing to their lack of coalition partners (Klotz, 2012). As a result, longstanding state discourse and practices continue to permeate society.

Overall, notions of societal security reflect the relative strength of identity-oriented and interest-based societal groups. In South Africa, societal challengers to xenophobia remain relatively weak. The mining industry’s influence primarily outside democratic avenues also serves as a reminder that any influence of rights advocates or economic actors is structured by the institutional context. Constitutional constraints or corporatist relationships are merely a few variants of ‘democracy’ that merit greater attention in determining scope conditions. Greater attention to ways in which identities and interests become synthesized in conceptions of societal security thus offers one step towards dialogue between securitization and other theories of immigration policy. Institutional context, which is at the heart of any state–society relationship, also sheds light on competing representations of security, as illustrated next.

Test 4: Policy outcomes

Distinguishing the referent object of a putative threat enables us to compare alternative security-based arguments. Migration policy itself also has multiple facets, including labor markets, political asylum, citizenship, and border controls. Therefore, to test mechanisms potentially linking threats and policy outcomes, we focus on refugees as one aspect likely to involve security considerations. However, traditional conceptions of security are not salient: South Africa no longer perceives a strategic risk from, say, a communist invasion led by Cuban forces based in neighboring Angola. Even extending Rudolph’s (2003) sequencing logic reveals no clear connection between national security and migration policies. Significant restrictions on African mobility were lifted in 1986 as a domestic reform at the height of civil unrest and regional strife. By the mid-1990s, the level of geostrategic threat to South Africa had clearly declined, but without any corresponding shift in immigration policies.

Inconsistencies in post-apartheid policy also remain puzzling. Nelson Mandela’s new government initially sought to liberalize migration policy through the 1997 Green Paper and the Refugees Act of 1998, which established asylum for the first time. Yet the controversial 2002 Immigration Act offered only modest reforms, while retaining an emphasis on detention and deportation. South Africa also vehemently rejected the 1995 proposal by the Southern African Development Community (SADC) for a timetable toward Schengen-like free movement in the region, which extended efforts (initiated in the 1980s) to overcome structural barriers to integration.¹⁶ Instead, South Africa emphasized better administrative coordination (Oucho and Crush, 2001). Accordingly, South African bilateral policies have sought to coordinate visa requirements with neighboring countries. Home Affairs also hoped, through procedural guidelines, to prohibit asylum claims if the applicant has transited through a safe third country (Lawyers for Human Rights/African Centre for Migration & Society, 2013).

Surprisingly, the securitization logic makes some sense of these contradictions, even though South Africa, a major refugee-receiving country in Africa, has not pushed for a common regional policy. Unlike in Europe, regionalization of collective identity and externalization of national interests are two distinct potential solutions to what the South African state has characterized as a socio-economic threat posed by refugees, especially alleged false claimants from its troubled neighbor Zimbabwe. In South Africa, the region does not become the referent object for the state’s definition of societal security. Therefore, its restricted view of the region, based on a traditional notion of national interests rather than pan-Africanism, blocked European-style multilateral cooperation.

Notably, Home Affairs does not categorize Zimbabwe as a refugee-generating country, so very few victims of political violence have approved asylum status. Zimbabweans are not the only claimants to confront this hurdle, as very few parts of Africa are accepted as legitimate sources of refugees in the Home Affairs lexicon.¹⁷ However, unlike those fleeing conflict in the Congo, the Lakes region, or West Africa, most Zimbabweans cross directly into South Africa. Consequently, the safe third country rule, a key strategy that successfully targeted the main source of refugees into EU member-states, fails to diminish the primary flow of asylum seekers into South Africa.

Furthermore, if South Africa formally acknowledged the depths of political repression in the Zimbabwean conflict, its categorization of refugee-generating countries would have to expand dramatically. The consequences would be a much larger number of legitimate asylum claims in addition to a much narrower set of potential safe third countries on the continent. Given its acknowledgment of both international and constitutional legal protections for refugees, South Africa would then be required to accept more African refugees. In effect, European-style regionalization would exacerbate rather than solve its refugee ‘problem’.

Thus, the securitization logic, combined with accidents of geography, explains this Southern African trajectory, without resort to factors such as non-Western cultures or authoritarian political systems. Emphasis on ontological threats to collective identity also highlights a conundrum. South Africa’s discourse of societal security, which rejects pan-Africanism, defines foreign Africans as the primary existential threat. This restricted notion of national identity fosters foreign policies that unintentionally exacerbate the migration flows it seeks to avoid. Such surprising policy insights lead us to urge further testing of the framework’s general applicability to other regions. Given the uniqueness of each regional context, however, we caution against widespread application without sensitivity to geography. Such caveats return us to questions of scope.

Conclusion

The South African case has confirmed that ‘societal security’ can overcome Eurocentrism without excessive conceptual stretching. Bringing together key insights from each of our four tests helps to explain apparent contradictions in contemporary South African policy. Notably, reconfigured discourses of societal security in the post-apartheid period have facilitated the extension of some rights for non-citizens alongside tightened exclusions of those deemed undesirable migrants. This core feature in recent reforms would have been overlooked without the application of securitization theory. And even though South Africa did not follow the European path of regionalization, its policies still support the logic of securing existential identities.

Extending the scope of analysis would help to expunge other counterproductive elements of Eurocentrism. For example, the South African case casts doubt on the existence of any singular ‘Western’ understanding of societal security. While Europeans do appear to be securitizing immigration in response to concerns about Islam, veils in particular do not have the same salience elsewhere. Notably, the status of Muslims in South Africa is not linked to current immigration or controversies over religion. One set of Muslims is associated with the so-called Coloureds, whose ancestors arrived as slaves in the 1700s; the other set is associated with so-called Indians, whose arrival was linked to the use of indentured labor in the late 1800s. More recently, antagonism to Muslims from elsewhere in Africa, such as Somalia, is based on their nationality, not their religion.¹⁸

This one application alone cannot resolve the questions of scope that it highlights, however. Clearly, a focus solely on speech acts is too narrow, but how many other forms of action can be encompassed without losing the concept’s distinctive emphasis on discourse? Similarly, how many actors or forms of contestation can be included without losing sight of the core ontological assumption of states-as-actors? And which other theoretical perspectives should be leveraged to differentiate securitization processes from other plausible explanations for policy outcomes? Concentrating more carefully and systematically on societal security as its conceptual core can help the securitization framework retain its innovative insights into collective identity.

Keeping in mind that no theory is without weaknesses, we suggest that securitization theory merits more systematic inclusion in analyses of migration. In particular, we suggest probing additional variation among non-European democracies. For example, Japan is another ethno-nationalist country that has historically resisted the notion of immigration (Gurowitz, 1999; Tsuda and Cornelius, 2004). Our analysis of contestation in South Africa suggests that the Japanese state is unlikely to relent on its exclusionary definition of society security, in part because of the weak role of organized labor. Also, juxtaposing identity-based social groups against the economic interests of business and labor in Brazil or Argentina might explain recent shifts in official preference for select groups categorized by race, class, ethnicity, and geographic origins (Albarracin, 2004; Lesser, 1999). Furthermore, Malaysia’s recent outsourcing of its societal security enforcement to ultra-nationalist voluntary groups takes the analysis of contestation to non-governmental daily practices in immigration control (McGahan, 2009). Another transnational step would be to assess across more cases the extent to which domestic commitments to international human rights principles (Jacobson, 1996; Joppke, 1998) possibly temper discursive threats, as happened when compliance with constitutional and international law pushed South Africa to stronger protection of refugee rights.

These suggestions only hint at the full potential of securitization theory to offer useful insights beyond Europe. Indeed, many democracies are routinely ignored in both the immigration and the securitization literatures, even though the majority of population flows take place outside advanced industrial economies (Human Development Report, 2009). Despite potential pitfalls and lacunae, the societal security concept merits greater integration into comparative studies of immigration policy and immigrant incorporation. To widen their geographical gaze, we suggest that securitization theorists first be ‘granted visas’ to the diverse democracies in Latin America and Asia. Only then can we assess whether there are ‘border controls’ on how far the framework can travel.

Footnotes

Acknowledgements

Earlier versions of this article were presented at the Annual Convention of the American Political Science Association in September 2009 and at the Annual Convention of the International Studies Association in February 2009. Thanks especially to Emily Rodio for research assistance, and to Rens van Munster, Kevin McGahan, Matt Cleary, members of the ‘identity group’ at Syracuse University and participants at the 2011 National Endowment for the Humanities (NEH) Summer Seminar on Rethinking International Migration, including its organizer Roger Waldinger, for comments.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Asli Ilgit is an Assistant Professor of Political Science at Gustavus Adolphus College, with research interests in the politics of identity, immigration, foreign policy, and discourse analysis. Her work has been published in International Studies Review, Mediterranean Politics, and MERIP, and as a chapter in Psychology and Constructivism in International Relations: An Ideational Alliance, edited by Vaughn P. Shannon and Paul A. Kowert (University of Michigan Press, 2011).

Audie Klotz is a Professor of Political Science in the Maxwell School of Citizenship and Public Affairs at Syracuse University. Her interests include theories of international relations, qualitative methods, transnational activism, global migration, and identity politics, with a regional specialization on Southern Africa. Her latest book, Migration and National Identity in South Africa, 1860–2010 (Cambridge University Press, 2013), includes a macrohistorical comparison with Australia and Canada.

References

Adamson

(2006) Crossing borders: International migration and national security. International Security 31(1): 165–199.

Albarracin

(2004) Selecting Immigration in Modern Argentina: Economic, Cultural, International and Institutional Factors. Doctoral dissertation, University of Florida.

Amin

Chitala

Mandaza

(eds) (1987) SADCC: Prospects for Disengagement and Development in Southern Africa. London: Zed Books.

Anderson

(2001) Imagined Communities: Reflections on the Origin and Spread of Nationalism, 2nd edn. London: Verso.

Balzacq

(2005) The three faces of securitization: Political agency, audience and context. European Journal of International Relations 11(2): 171–201.

Balzacq

(ed.) (2011a) Securitization Theory: How Security Problems Emerge and Dissolve. New York: Routledge.

Balzacq

(2011b) Enquiries in methods: A new framework for securitization analysis. In: Balzacq

(ed.) Securitization Theory: How Security Problems Emerge and Dissolve. New York: Routledge, pp.31–53.

Bigo

(2002) Security and immigration: Toward a critique of the governmentality of unease. Alternatives 27(1): 63–92.

Boswell

(2007) Migration control in Europe after 9/11: Explaining the absence of securitization. Journal of Common Market Studies 45(3): 589–610.

10.

Bradlow

(1978) Immigration into the Union 1910–1948. Doctoral dissertation, University of Cape Town.

11.

Buzan

(1991) People, States, and Fear: An Agenda for International Security Studies in the Post-Cold War Era. Boulder, CO: Lynne Rienner.

12.

Buzan

Wæver

De Wilde

(1998) Security: A New Framework for Analysis. Boulder, CO: Lynne Rienner.

13.

Caplan

Torpey

(2001) Documenting Individual Identity: The Development of State Practices in the Modern World. Princeton, NJ: Princeton University Press.

14.

Cawthra

(1993) Policing South Africa: The SAP and the Transition from Apartheid. Cape Town: David Philip.

15.

Chebel d’Appollonia

(2012) Frontiers of Fear: Immigration and Insecurity in the United States and Europe. Ithaca, NY: Cornell University Press.

16.

Ciutâ

(2009) Security and the problem of context: A hermeneutical critique of securitization theory. Review of International Studies 35(2): 301–326.

17.

Cornelius

Takeyuki

Martin

Hollifield

(eds) (2004) Controlling Immigration: A Global Perspective. Stanford, CA: Stanford University Press.

18.

Crush

(2000) Migrations past: An historical overview of cross-border movement in Southern Africa. In: McDonald

(ed.) On Borders: Perspectives on International Migration in Southern Africa. New York: St. Martin’s Press, pp.12–24.

19.

Crush

Tshitereke

(2001) Contesting migrancy: The foreign labor debate in post-1994 South Africa. Africa Today 48(3): 49–70.

20.

Curtis

(2001) The Politics of Population: State Formation, Statistics, and the Census of Canada, 1840–1875. Toronto: University of Toronto Press.

21.

Dibetle

(2008a) Buthelezi scores brownie points. Mail and Guardian, 31 May.

22.

Dibetle

(2008b) Displaced ‘not govt’s responsibility’. Mail and Guardian, 15 August.

23.

Doty

(1998) Immigration and the politics of security. Security Studies 8(2): 71–93.

24.

Dunn

(2008) Historical representations. In: Klotz

Prakash

(eds) Qualitative Methods in International Relations. Basingstoke: Palgrave Macmillan, pp.78–92.

25.

Evans

(1997) Bureaucracy and Race: Native Administration in South Africa. Berkeley, CA: University of California Press.

26.

Everatt

(ed.) (2011) Special issue on xenophobia and civil society. Politikon: Journal of the South African Association of Political Studies 38(1): 1–189.

27.

Floyd

(2011) Can securitization theory be used in normative analysis? Towards a just securitization theory. Security Dialogue 42(4–5): 427–439.

28.

Gad

Petersen

(2011) Concepts of politics in securitization studies. Security Dialogue 42(4–5): 315–328.

29.

Govender

(2008) About 142 xenophobic cases before SA courts. Mail and Guardian, 3 June.

30.

Government of South Africa, Department of Home Affairs (2008) Annual Report, 2007–2008. Pretoria.

31.

Gurowitz

(1999) Mobilizing international norms: Domestic actors, immigrants, and the Japanese state. World Politics 51(3): 413–445.

32.

Guzzini

(2011) Securitization as a causal mechanism. Security Dialogue 42(4–5): 329–341.

33.

Handmaker

Lee

Klaaren

(eds) (2008) Advancing Refugee Protection in South Africa. New York: Berghahn.

34.

Hansen

(2000) The Little Mermaid’s silent security dilemma and the absence of gender in the Copenhagen School. Millennium 29(2): 285–306.

35.

Heard

(1974) General Elections in South Africa, 1943–1970. London: Oxford University Press.

36.

Herd

Löfgren

(2001) ‘Societal security’, Baltic states and EU integration. Cooperation and Conflict 36(3): 273–296.

37.

Hlatshwayo

(2011) Is there room for international solidarity within South African borders? COSATU’s responses to the xenophobic attacks of May 2008. Politikon: Journal of the South African Association of Political Studies 38(1): 169–189.

38.

Hollifield

(1992) Immigrants, Markets, and States: The Political Economy of Postwar Europe. Cambridge, MA: Harvard University Press.

39.

Hollifield

(2004) The emerging migration state. International Migration Review 38(3): 885–912.

40.

Human Development Report (2009) Overcoming Barriers: Human Mobility and Development. New York: Palgrave Macmillan.

41.

Huttenback

(1971) Gandhi in South Africa: British Imperialism and the Indian Question, 1860–1914. Ithaca, NY: Cornell University Press.

42.

Huysmans

(2000) The European Union and the securitization of migration. Journal of Common Market Studies 38(5): 751–777.

43.

Huysmans

(2002) Defining social constructivism in security studies: The normative dilemma of writing security. Alternatives 27: 41–62.

44.

International Organization for Migration (IOM) (2009) Toward Tolerance, Law, and Dignity: Addressing Violence Against Foreign Nationals in South Africa. Arcadia, Pretoria: IOM Regional Office for Southern Africa.

45.

Jacobson

(1996) Rights Across Borders: Immigration and the Decline of Citizenship. Baltimore, MD: Johns Hopkins University Press.

46.

Jeeves

(1985) Migrant Labour in South Africa’s Mining Economy: The Struggle for the Gold Mines’ Labour Supply 1890–1920. Kingston: McGill-Queen’s University Press.

47.

Joppke

(1998) Why liberal states accept unwanted immigration. World Politics 50(2): 266–293.

48.

Joubert

(2008) Finally, the SA govt moves into action. Mail and Guardian, 30 May.

49.

Kastoryano

(2002) Negotiating Identities: States and Immigrants in France and Germany. Princeton, NJ: Princeton University Press.

50.

Klotz

(2012) South Africa as an immigration state. Politikon: South African Journal of Political Studies 39(2): 189–208.

51.

Klotz

(2013) Migration and National Identity in South Africa, 1860–2010. New York: Cambridge University Press.

52.

Lawyers for Human Rights/African Centre for Migration & Society (2013) Policy Shifts in the South African Asylum System: Evidence and Implications. Johannesburg: African Centre for Migration & Society.

53.

Lesser

(1999) Negotiating National Identity: Immigrants, Minorities, and the Struggle for Ethnicity in Brazil. Durham, NC: Duke University Press.

54.

Lipton

Simkins

(eds) (1993) State and Market in Post-Apartheid South Africa. Johannesburg: Witwatersrand University Press.

55.

McDonald

Jacobs

(2005) (Re)writing xenophobia: Understanding press coverage of cross-border migration in Southern Africa. Journal of Contemporary African Studies 23(3): 295–325.

56.

McDonald

(2008) Securitization and the construction of security. European Journal of International Relations 14(4): 563–587.

57.

McGahan

(2009) The securitization of migration in Malaysia: Drawing lessons beyond the Copenhagen School. Paper presented at the Annual Meeting of the International Studies Association, New York, February.

58.

Mail and Guardian (Johannesburg) (2008) Mbeki says govt wasn’t warned about attacks, 4 June.

59.

Mail and Guardian (Johannesburg) (2012) Headscarves raise hackles, 20 January.

60.

Oucho

Crush

(2001) Contra free movement: South Africa and the SADC migration protocols. Africa Today 48(3): 139–158.

61.

Owen

(1964) Foreign Africans: Summary of the Report of the Froneman Committee. Report 111/64. Johannesburg: South African Institute of Race Relations.

62.

Peberdy

(1999) Selecting Immigrants: Nationalism and National Identity in South Africa’s Immigration Policies, 1910 to 1998. Doctoral dissertation, Queen’s University, Kingston, Canada.

63.

Posel

(1997) The Making of Apartheid, 1948–1961. Oxford: Clarendon.

64.

Post

Niemann

(2007) The Europeanization of German asylum policy and the ‘Germanization’ of European asylum policy: The case of the ‘safe third country’ concept. Paper presented at the 10th Biennial Conference of the European Union Studies Association, Montreal, May.

65.

Rossouw

(2008) The Zuma factor. Mail and Guardian, 24 May.

66.

Rudolph

(2003) Security and the political economy of international migration. American Political Science Review 97(4): 603–620.

67.

Sadiq

(2005) Lost in translation: The challenges of state-generated data in developing countries. In: Monroe

(ed.) Perestroika! The Raucus Rebellion in Political Science. New Haven, CT: Yale University Press, pp.181–199.

68.

Salter

(2008a) Securitization and desecuritization: A dramaturgical analysis of the Canadian Air Transport Security Authority. Journal of International Relations and Development 11(4): 321–349.

69.

Salter

(2008b) Imagining numbers: Risk, quantification, and aviation security. Security Dialogue 39(2–3): 243–266.

70.

Sartori

(1970) Concept misformation in comparative politics. American Political Science Review 64(4): 1033–1053.

71.

Schimmelfennig

(2001) The community trap: Liberal norms, rhetorical action, and the Eastern enlargement of the European Union. International Organization 55(1): 47–80.

72.

Scott

(1998) Seeing Like a State: How Certain Schemes to Improve the Human Condition have Failed. New Haven, CT: Yale University Press.

73.

Shain

(1994) The Roots of Antisemitism in South Africa. Charlottesville, VA: University of Virginia Press.

74.

South African Human Rights Commission (SAHRC) (1999) 4th Annual Report. Available at: http://www.sahrc.org.za (accessed 15 August 2009).

75.

Stritzel

(2007) Towards a theory of securitization: Copenhagen and beyond. European Journal of International Relations 13(3): 357–383.

76.

Tichenor

(2002) Dividing Lines: The Politics of Immigration Control in America. Princeton, NJ: Princeton University Press.

77.

Tjalve

(2011) Designing (de)security: European exceptionalism, Atlantic republicanism, and the ‘public sphere’. Security Dialogue 42(4–5): 441–452.

78.

Tsuda

Cornelius

(2004) Japan: Government policy, immigrant reality. In: Cornelius

Takeyuki

Martin

Hollifield

(eds) Controlling Immigration: A Global Perspective. Stanford, CA: Stanford University Press, pp.439–476.

79.

United Nations High Commissioner for Refugees (UNHCR) (2010) Global Report: South Africa. Available at: http://www.unhcr.org/4c08f2c19.html (accessed 20 July 2010).

80.

Van Munster

(2009) Securitizing Immigration: The Politics of Risk in the EU. Basingstoke: Palgrave Macmillan.

81.

Van Niekerk

Van der Waldt

Jonker

(2001) Governance, Politics, and Policy in South Africa. Cape Town: Oxford University Press.

82.

Vigneswaran

(2008) Enduring territoriality: South African immigration control. Political Geography 27(7): 783–801.

83.

Vucetic

(2011) Genealogy as a research tool in international relations. Review of International Studies 37(3): 1295–1312.

84.

Vuori

(2008) Illocutionary logic and strands of securitization: Applying the theory of securitization to the study of non-democratic political orders. European Journal of International Relations 14(1): 65–99.

85.

Wæver

(1993) Societal security: The concept. In: Wæver

Buzan

Kelstrup

Lemaitre

(eds) Identity, Migration and the New Security Agenda in Europe. London: Pinter, pp.17–40.

86.

Wæver

(2011) Politics, security, theory. Security Dialogue 42(4–5): 465–480.

87.

Wæver

Buzan

Kelstrup

Lemaitre

(eds) (1993) Identity, Migration and the New Security Agenda in Europe. London: Pinter.

88.

Watson

(2009) The Securitization of Humanitarian Migration: Digging Moats and Sinking Boats. London: Routledge.

89.

Weiner

(1992/1993) Security, stability, and international migration. International Security 17(3): 91–126.

90.

Wilkinson

(2007) The Copenhagen School on tour in Kyrgyzstan: Is securitization theory useable outside Europe? Security Dialogue 38(1): 5–25.

91.

Williams

(2003) Words, images, enemies: Securitization and international politics. International Studies Quarterly 47(4): 511–531.

92.

Williams

(2011) Securitization and the liberalism of fear. Security Dialogue 42(4–5): 453–463.