Chapter 6 Free settlement: the maintenance of a free-settlement reception in South Africa

Legal framework and registration procedures

The 1998 Refugee Act set out a national-level approach for the first time – at least in terms of procedural requirements – towards the initial reception of refugees.³ Since its inception, the Act has received international praise and is regularly held up by UNHCR as a model of how to receive refugees on the African continent.⁴ Some academics and commentators have gone so far as to suggest that the Act is a beacon of progressive African-centric legal frameworks, moving refugee law (particularly the refugee definition) beyond the European-centric 1951 Refugee Convention (Smith, 2003). For example, the Act incorporates the 1969 OAU Refugee Convention 1(2) definition and includes an expanded definition of non-refoulement. In addition, the existence of a nationally run and individualised RSD procedure combined with a strong judiciary brings further credibility to these assertions (Johnson, 2015). Although, as explored later, this idea of an example of progressive national refugee governance was significantly weakened in 2020 with the entering into force of the Refugees Amendment Act 33 of 2008 (RAA 2008) in 2020, which itself triggered the coming into force of the Refugees Amendment Act 12 of 2011 (RAA 2011) and of the Refugees Amendment Act 11 of 2017 (RAA 2017). These acts and their accompanying regulations reduce access to the asylum system and deny asylum-seekers rights that they formerly held (Ziegler, 2020).

The final component of the legal framework relating to refugee reception is the National Constitution of South Africa. This sets the legal requirements for all national legislation and the key underlying values that should be applied when interpreting legislation (Klinck, 2009). The Bill of Rights specifically grants rights to all people in the state and affirms ‘the democratic values of human dignity, equality and freedom’.⁵ By viewing the Refugee Act through this constitutional lens, it ‘points towards an interpretive approach which gives effect to South Africa’s constitutional and international human rights commitments’ (Klinck, 2009:655). Civil society confirmed this constitutional focus in relation to refugee’s rights, with a human rights lawyer noting, ‘in our submissions we start with constitutional law. This gets more traction’.⁶

In terms of the procedural implementation of the national legal framework at initial registration, the process starts (at least theoretically) when an asylum-seeker makes her/himself known at the point of entry and requests asylum. At this point they receives an asylum transit visa, which allows them to proceed to one of the RROs to deliver the application in person (Vigneswaran, 2008). Alternatively, many forced migrants make their way to urban spaces without being detected via the porous borders. They then decide to either make themselves known to the authorities or remain hidden.⁷

The 1998 Refugee Act establishes the RSD procedure; consequently, once the claim has been lodged, refugees will undergo an RSD interview at an RRO.⁸ Throughout the registration process, refugees receive a temporary asylum-seeker permit from an RRO, which regularises their stay and allows free movement within the state until the claim is decided (Vigneswaran, 2008). Regular trips back to the RROs are needed to renew the permits, which are usually granted and extended from one to six months.

The initial reception at the point of registration

The RSD process and the administrative issues and procedures that stem from it dominate the initial registration period. Beyond these procedures, the approach by the government and UNHCR towards refugees at this initial stage of reception can be considered non-interventionist. In contrast to the provisions provided in refugee camps, there is no guarantee of immediate assistance or provision of services by the state or UNHCR. Consequently, even at this initial point of reception, there is very limited access to the global refugee regime in urban areas.⁹ Indeed, at its core, beyond procedures aimed at legal registration, state-run reception in South Africa is purposefully hands-off. Nevertheless, the non-action of the state and UN agency within this free-settlement approach to reception does create its own form of welcome. This is to say, it allows refugees to move around the territory and in doing so creates the room for refugees to find political space and other forms of reception at the local and sub-local level.

As confirmed in interviews, and by numerous Department of Home Affairs officials in public events with civil society and academia, South Africa does not have an integration policy – rather it has a policy of ‘self-integration’.¹⁰ In many respects, this national approach to reception creates many freedoms (or at least a sense of freedom) for refugees in urban spaces. For example, refugees are technically permitted to settle anywhere. Yet as discussed below, the lack of assistance during registration and beyond, coupled with the ability of substantial numbers of migrants to move freely in these spaces, results in this form of ‘non-rooted’ movement being continually contested by material and ideational factors within state bodies. Chief among these are democratic pressures. Many in civil society see this central element of the national approach to reception as a major failing:

I think one of the biggest gaps or failures by the state [is] that there is no – as far as I can understand – from a national level, an active and clear coordinated coherent policy, to achieve some kind of social integration.¹¹

Due to this policy of non-intervention, refugees generally have to rely on finding reception and ultimately acceptance at the local and sub-local level within local networks and communities, rather than through state structures. In urban areas, refugees ‘join the ranks of the urban poor and other migrants (citizens and non-citizens alike)’ (Sarkar, 2017:1).¹² These self-adopted approaches to protection and acceptance are nevertheless de facto solutions and exist outside of the national legal framework and the global refugee regime.¹³ This is the same for refugees with official documentation and others living in these spaces more illicitly.

From this preliminary analysis of the registration stage, two levels of reception emerge. At the national and international level, the government’s dominant free-settlement approach (and by extension the global refugee regime) permits refugees a good deal of autonomy. Thus, due to a broadly liberal legal refugee framework that grants key norms of the global regime, refugees are free to move and settle with local communities in urban spaces. Nonetheless (as examined further below) with the requirement to regularly visit RROs to renew permits, constraints on geographical space emerge immediately for refugees during this initial reception.

At the local or sub-local level, refugees are forced to locate forms of localised reception. This is due to the lack of assistance from higher levels of the state, although avoiding state and UN agency structures is often also an active choice. These local forms of reception retain a precarious quality, with the risk of exploitation by state actors (such as immigration officials, police) and other actors (such as a hostile local community) ever present.¹⁴

Both levels of reception open up some access to the political realm, although this is constantly contested at both the local and national level. As underscored throughout this chapter and the next, there are several key reasons for this restricted access to the political space, including the multitude of exclusion mechanisms at the local and national level that prevent access to services.¹⁵ Another reason is the quality of temporariness attached to refugee status at the national level. According to civil society activists, the idea that refugee reception is intrinsically short term is deeply entrenched within elements of the Department of Home Affairs:

[There is] a dogmatic view of refugee status [as] inherently temporary. Refugees remain permanent guests on the territory.¹⁶

As a result, refugees face real difficulties accessing pathways to obtaining permanent status in South Africa (Landau, 2011; Landau et al., 2018): indeed ‘you can’t become a permanent resident … you can get this idea of long-term residence but not permanent, you cannot become a citizen’.¹⁷ Finally, due to the numerous exclusion mechanisms, refugees in cities in South Africa are regularly confined to specific areas of the urban space. Indeed, they have little choice but to live in enclaves of the city, such as informal or illegal settlements and townships (Landau and Freemantle, 2017; Chekero, 2023). Civil society activists and leaders certainly took this view, with a prominent human rights lawyer in Cape Town noting that these enclaves are ‘pretty much … the only place that asylum-seekers and refugees can get reasonable accommodation – where they can afford to live is in poor communities’.¹⁸

Material factor: contemporary movements into South Africa

The character and make-up of cross-border movement into post-apartheid South Africa has been one of the factors that has enabled the maintenance of a free-settlement reception approach. Specifically, the state has not experienced the levels of influx of refugees seen in other states on the continent, where refugee camps have been set up to respond to a large-scale emergency. Certainly, in Africa, the adoption of encampment policies remains a popular approach to receiving large numbers of refugees simultaneously (Crisp, 2003; Schmidt, 2003).

During deliberations surrounding the drafting of the 1998 Refugee Act, refugee camps were considered as the model for South Africa, but it was agreed that they would only be deployed in times of mass influx (Jenkins and de la Hunt, 2011).¹⁹ This option for setting up camps and restricting movement in times of emergency or mass influx is common in national refugee legislation on the continent (Maple, 2016). As examined in Chapter 2, long-term refugee camps often start life as a short-term humanitarian response to a mass influx of refugees. Over time, with return deemed unrealistic, such emergency camps often then transform into protracted encampment situations. Moreover, once this new form of reception is in place, the official or dominant state reception policy often switches to upholding encampment as the model for all future refugees (regardless of the nature and make-up of the movement).²⁰

The last forty years have witnessed numerous examples of a mass exodus in Africa. This includes (but is not limited to) the hundreds of thousands leaving for Sudan and Somalia during the Ethiopian famine of 1983–5, the hundreds of thousands of Hutu refugees fleeing Burundi in 1993 (ICG, 1999), and the millions fleeing South Sudan in 2016. At the sub-regional level, however, Southern Africa has not seen the levels of mass expulsion or exodus observed in neighbouring sub-regions, such as central and eastern Africa. For this reason, it is debatable – with the possible exception of the arrival of refugees from Mozambique during the 1980s – whether South Africa has experienced what would amount to a contemporary understanding of a mass influx of refugees (Jenkins and de la Hunt, 2011).

The movement of refugees and other forced migrants into South Africa has naturally fluctuated since the adoption of the global refugee regime into the national legal framework (Segatti, 2011). Nonetheless, patterns of refugee movement, while certainly increasing substantially at times, have not turned into urgent mass movements unfolding over short periods of time as seen elsewhere on the continent. This is due to a number of key reasons: (1) South Africa’s geographical location; (2) its willingness to grant freedom of movement, coupled with perceived economic opportunities; (3) the political and economic reasons fuelling the large exodus of Zimbabweans over the last twenty years to neighbouring countries; and (4) the lack of any major incidences of unrest caused by genocide, war or extreme climatic events in immediate neighbouring states since the late 1990s.

This analysis is borne out in the limited data available on asylum applications in South Africa, which shows that the majority of refugees in the state (if we set aside Zimbabwe for the moment) come from the Horn of Africa, the Great Lakes and a small but significant number from South Asian countries (World Bank, 2018a).²¹ The available data suggests that refugees who reach South Africa regularly travel long distances and pass through numerous countries before reaching the territory (World Bank, 2018a). For these reasons, the movement of refugees and forced migrations into South Africa has effectively been held in check since the introduction of the global refugee regime and its core norms into the state. That is to say, refugee movement into the state has been curbed by external geopolitical factors in the sub-region and continent. As a result, there has not been (at least until recently) a pressing need to consider an alternative reception approach aimed at substantially managing or controlling refugee movement into the interior, and by extension into the urban space.

It would be disingenuous to say, however, that patterns of refugee and forced migration movement into South Africa have not altered considerably over the last twenty years. Since the early 2000s, the country has witnessed the in-migration of hundreds of thousands of Zimbabweans (Betts, 2013a, 2014; Thebe, 2017). For example, persons applying for asylum in South Africa rose from 4,860 in 2001 to 364,638 in 2009 (Landau et al., 2011). In 2009, South Africa received the most asylum applications globally (at over 300,000) (Carciotto, 2018). Deemed as a crisis, these numbers elicited a reaction, with the state offering around 200,000 Zimbabweans the option of applying for a four-year Zimbabwean Dispensation Permit (ZDP) (Carciotto, 2018). Asylum figures increased further in the following years, with UNHCR reporting that Zimbabweans accounted for over half of the 778,600 new asylum applications in South Africa between 2008 and 2012 (UNHCR, 2015a).²²

Yet, broadly speaking these movements over the past two decades, reflecting the long-term economic upheaval and political unrest in Zimbabwe, have been sporadic (if substantial), rather than urgent and sudden as seen in neighbouring countries when large populations flee violence or civil war. For example, by 2018 a human rights lawyer based in Musina (a major border post between South Africa and Zimbabwe) observed that the movement of migrants from Zimbabwe had almost stopped:

The majority of Zimbabweans aren’t coming through because there is a new government there, so there is a feeling and understanding that the political situation is becoming stable.²³

Yet, by 2019, with little sign of the economy improving under the new government in Zimbabwe the numbers had again increased. This continuing fluctuation in refugee and migrant flows between the two countries over the past twenty years aptly illustrates how the types of migration events triggered by the situation in Zimbabwe typically result in slower patterns and lower numbers of movement (while still considerable) than those consequent to sudden outbreaks of civil unrest.²⁴

These dispersed movements into South Africa from Zimbabwe have undeniably put huge pressure onto the national asylum system (Amit, 2011). Nevertheless, ultimately these movements have not threatened the maintenance of the free-settlement approach, with high-level ministerial discussions about the adoption of camp policies remaining in the background. This is particularly important combined with the practical reality, and the political perception, that urban spaces such as Johannesburg and Cape Town are equipped to be able to absorb large numbers of new migrants. Simply put, at least up until 2011, the movement of refugees and other migrants into urban spaces was generally not seen as a destabilising threat to either cities or to the nation.

In addition, direct security concerns relating to the cross-border movement from Zimbabwe are not deemed as an urgent priority due to the make-up of the incoming population. Most forced migrants coming into South Africa from its immediate neighbour have been due to individual persecution based on political beliefs and increasingly also for a mixture of socio-economic reasons. This was the impression given by a human rights lawyer in Cape Town: ‘while we saw persecution and beatings in 2007, [it] is not what we see now, it’s more of an economic issue in Zimbabwe’.²⁵ As a consequence, there has been little risk of armed elements infiltrating the forced migrant population. Thus, during the early 2000s, there were no urgent calls to contain the whole population of incomers at the border.

In summary, the geopolitical situation in Southern Africa (and further afield on the continent) has acted as a restraint on the movement of refugees into South Africa. Large amounts of unchecked movement, particularly into urban spaces, is always likely to elicit a national reaction. Until very recently, the make-up and patterns of movement into South Africa have not necessitated additional externally imposed control. Indeed, even the number of migrants and refugees from Zimbabwe, while considerable, have been steady rather than urgent and sudden.

Nevertheless, over the last few decades, with numbers in urban spaces continuing to increase in South Africa, the government has felt the need to react in stronger terms to regulate this movement. Indeed, the numbers have continued to grow to levels now deemed a threat to the stability of the nation (Amit, 2015). Over the last twelve years, refugees and other forced migrants in South Africa are being seen almost exclusively through a security lens, with the urban space being slowly restricted as a site of reception.

Ideational and institutional factors: the lack of international involvement in the initial stage of refugee reception in South Africa

In 1996, South Africa became a member state of the global refugee regime and, unlike many of its neighbours who had adopted encampment reception policies around that time, it elected not to place any restrictions on the right to freedom of movement. In maintaining this commitment over subsequent years, the state has permitted increasing numbers of refugees and asylum-seekers to access densely populated urban spaces. As a result, the state has had to establish, fund and staff numerous RROs in major urban hubs to deal with the various bureaucratic procedures surrounding the registration of increasing numbers of asylum applications.

States in Africa regularly rely on the global refugee regime (via its key actor, UNHCR) to assist with capacity issues relating to the hosting of refugees. Yet membership of the global regime has not been matched by an increased involvement of UN agencies in South Africa. Indeed, the empirical evidence reveals a strained relationship between the state and UNHCR, which has resulted in a lack of substantive involvement by UN agencies at this initial stage of reception. In particular, since the late 2000s, UNHCR has largely been reduced to a capacity building and public education role.²⁶ At one level of analysis, as noted by a former employee at the Department of Home Affairs, this is a logical step:

In a place like South Africa where we have a functioning democratic government or functioning government, then the organisation’s role … so they take a step back not dealing with half of the issues – that is not their role in an urban situation.²⁷

The assumption that the agency does not have a role in urban spaces in South Africa, as held by elements within the Department of Home Affairs (and indeed by some high-ranking officials in UNHCR South Africa), will be analysed further in Chapter 7. The notion that UNHCR takes a step back and that a democratic host state should be running (and be entirely responsible for) its own reception policy, including an individualised RSD procedure, is rational and consistent with international norms (Maple et al., 2023). However, it is equally apparent that this reduced role for UNHCR was not a decision made entirely by the UN agency itself. Rather, as confirmed by a UNHCR officer, it was a decision by the South African government, steered by the ANC’s habitual insistence on keeping UN agencies and other international organisations very much at arm’s length.²⁸

The reluctance by the state to permit international agencies a more prominent role within the territory stems from historical ideational structures in place since the old apartheid regime. Indeed, there remains a general sense of mistrust by political elites inside the ruling political settlement concerning international organisations, particularly those within the UN structure. This has its roots in the perception of inactivity and lack of support during the ANC’s struggles against the former apartheid government.²⁹ As a result, since the dismantling of the old regime, there has been a general unwillingness to listen to advice or accept assistance on the territory from UN agencies.

By not permitting UNHCR and other UN bodies greater influence within the national refugee framework, it means that any shift of refugee reception policy (to, say, an encampment policy) could potentially fall entirely on the state to fund and implement. This contrasts sharply with neighbouring states who already house refugees in refugee camps, as they generally see these populations as the responsibility of the international community. For South Africa to switch from the current system (where refugees are expected to fend for themselves as soon as they cross the border) to an encampment policy (where refugees are often entirely reliant on aid), would require a great deal more state investment. It would also entail a far greater presence and influence of UN agencies on the territory, which is something that the political elites in South Africa, at least in recent history, seem unwilling to accept.

In the context of post-apartheid South Africa, and considering the average length of displacement seen globally today, the current free-settlement approach to refugee reception remains less of a strain on national resources than an alternative encampment policy.³⁰ This is a powerful argument when other factors are taken into consideration, such as (1) the lack of engagement by the state in refugees’ lives beyond the registration process at the RROs, and (2) a current national registration system that is failing to function properly, in large part due to a lack of adequate funding and resources by the government.³¹ This lack of engagement and resources would be harder to justify within the distinct spatial and visual boundaries of the refugee camp. In sharp contrast to this, refugees and asylum-seekers in the urban space regularly merge (and/or ‘disappear’) into larger poor neighbourhoods.

This stance by South Africa towards UN institutions brings mixed responses from UNHCR. There is understandable gratitude from within UNHCR for the state’s reception approach and its refusal, to date, to switch to an encampment policy. Yet, the attitude at the national level towards the presence of UN agencies on the territory has resulted in what can best be described as a tense working relationship between the state and UNHCR. Consequently, there is marked frustration within UNHCR South Africa and UNHCR Southern Africa about the way the state has kept international organisations at arm’s length. A prominent human rights lawyer in Pretoria suggested that the agency had effectively been ‘shut out’ by the government.³²

As examined below, the recent regressive changes in refugee reception policy in the urban space in South Africa have drawn little in the way of public discourse by UNHCR.³³ This diplomatic approach is in part understandable. As Forsythe (2001:34) notes, the agency regularly has the delicate task of engaging with the political process of influencing governments to make appropriate choices, without running the risk of being ‘charged with political interference in the domestic affairs of states’.³⁴ Equally, the contemporary historical background and the ideational factors that this fraught history has generated within sections of the ruling political settlement in South Africa adds additional layers of complexity for the UN agency to navigate.

Many within civil society and academia nevertheless feel that the recent ‘backward’ shift in relation to the reception of refugees in South Africa needs a stronger reaction from the agency. Below is a representative response from an NGO official when asked about UNHCR in South Africa:

UNHCR … has been fairly disappointing and whenever we have asked them to take a stand – a stronger position with the government, the response that we received is that ‘we are a guest of the government, we are here to assist the government and we cannot take on an aggressive approach’ – which you know, their job is to protect refugees and they do not seem to be doing a fantastic job of that.³⁵

Finally, state engagement with the 2018 Global Compact on Refugees had hinted at a change in approach, with the Department of Home Affairs appearing to show a renewed willingness to engage with UNHCR. For example, the state made a pledge at the Global Refugee Forum in 2019, to reduce the number of outstanding national asylum claims. From this, the ‘Backlog Project’ was created, with Home Affairs and UNHCR signing an agreement in March 2021 (PMG, 2021), which committed the department to spend USD 2.6 million and UNHCR USD 7 million on the project (UNHCR, 2021a). However, as of 2023, little government money had been spent. In turn, the restrictive amendments to the 1998 Refugee Act which came into force after the pledges were made severely undermine these international commitments (Khan and Rayner, 2020).

In conclusion, the combination of historical and ideational factors that have kept the UN at arm’s length in-country, and the continued institutional contestation between government departments and UNHCR, has left the agency little to do in terms of refugee reception in South Africa.³⁶ This ‘shutting out’ of the agency has left key informants remarking that it is ‘unclear at the moment the role UNHCR is playing, if any’.³⁷ This complex and strained relationship re-emphasises the dominant role the government has in setting and implementing policy on refugee and forced migrant reception. Thus capacity concerns, combined with limited interest in increasing the involvement of UN agencies on the territory, create powerful motivations for maintaining the self-settlement approach (rather than a more resource-heavy and visible camp-based approach).

Ideational factors: the process of nation-building

In analysing the maintenance of a free-settlement approach to refugee reception in South Africa, this final section concentrates on key ideational factors that permeate throughout government departments at the national level. These include the ideational importance of the National Constitution, comparisons between creating refugee camps and the previous actions of the apartheid regime, and the continued (if strained) relevance of pan-Africanism. Grouped together, these factors can be understood as part of the process of nation-building and of the ANC’s vision of post-apartheid South Africa.

Taking these factors in turn, the National Constitution in South Africa retains notable standing amongst government officials, with its commitment to human rights and regional cooperation. As Landau and Segatti (2009) note, the preamble explicitly promises that ‘South Africa belongs to all who live in it’. By making no reference to citizenship status, the Constitution grants rights for all persons on the territory. Indeed, as previously noted, the Bill of Rights has been a powerful tool for advocates of refugee rights, including the right to work. Thus, any attempt by the government to implement refugee camps, given the severe restriction on human rights that this would entail, would sit in stark contrast to the legal and ideational aspirations committed to at the beginning of post-apartheid South Africa. For example, in relation to the current proposal for processing centres at the border, civil society actors were quick to mention the constitutional fight that would ensue.

The constitution talks about freedom of movement for everyone … it will end up being litigation I’m sure. In terms of the constitution, they will have to argue why … you are limiting a right for everyone – a constitutional right.³⁸

Complications and contestation do nevertheless arise when attempting to implement particular obligations contained within the Constitution (such as access to health) for refugees and other migrants. This is because many guaranteed rights are not accessible to most nationals (Misago, 2016). Equally, as made evident by the high-level discussions over the border processing centres, an ideational commitment to granting key rights to asylum-seekers and refugees is far from ubiquitous within national-level state entities that deal with refugees. As discussed next, contestation between conflicting ideational approaches regularly occurs within the government.

Turning to the second ideational factor, there are some within the Department of Home Affairs who see uncomfortable parallels between the creation and use of refugee camps and the old apartheid regime. Specifically, the idea of moving African brothers and sisters into confined spaces is reminiscent of the legacy of segregation during the apartheid era. Many ANC members were placed in ‘homelands’ which were areas designated for black South Africans and organised along ethnic lines during the first half of the last century, as well as forced relocation to these areas in the 1960s from other areas of the state.³⁹ Thus, many feel that any move towards encampment would be a dangerous backwards step.

Interviewees noted, however, that these views were not universal within government departments. There are many key officials in the government, particularly in the Department of Home Affairs, who were moved into ‘resettlement’ or ‘relocation’ camps in the homelands during the apartheid regime. Their attitude now is that if they went through it and survived, ‘why can’t others?’ Thus, discrepancies and contradictory approaches exist at the institutional level, with security and stability-focused views held in the Department of Home Affairs conflicting with more historical and normative-based sentiments. High-level officials concerned by the perceived insecurity and instability caused by the increase in the movement of people into urban areas are behind recent pushes for camp-based reception policies for refugees at the border.

The final ideational factor is the role of pan-Africanism in the development and maintenance of the current reception policies in South Africa. Traditionally states in Africa were generous in hosting refugees, with neighbours fleeing colonial oppression readily welcomed due to feelings of solidarity. This approach was common on the continent until the 1990s when a broad shift to democracy occurred. Since then, feelings of pan-Africanism have decreased significantly, in part because refugee movements are no longer due to violence stemming from liberation struggles from colonial rule and white minority government repression (Rutinwa, 1999). South Africa has broadly followed these trends, with pan-Africanism slowly being replaced with nationalistic and xenophobic views within state bodies and the wider voting population (Palmary, 2002; Misago, 2016; Misago, et al., 2010). As explained via the ‘democracy-asylum’ nexus, the trend for democratic structures to accompany increased anti-migrant feelings among local populations (particularly in urban conurbations where there is competition for scarce social goods such as housing and employment) has meant reception policies have become markedly less generous (Crisp, 2000). Indeed, the increase of refugee and other migrant movements into South Africa has created negative democratic loops, whereby the state is politically motivated to move away from ideals based on solidarity and pan-Africanism to ones framing migrants as a ‘problem’.⁴⁰

Nevertheless, in post-apartheid South Africa, there have been some instances when the national approach has bucked this regional trend,⁴¹ and officials within international agencies still regard South Africa as broadly welcoming to its African sisters and brothers who arrive as refugees from neighbouring states. To some commentators, the state has even gone beyond its commitments under the global refugee regime by stretching the key regime norm of non-refoulement to include forced migrants from Zimbabwe who may not strictly fall under the 1951 Refugee Convention or the 1969 OAU Refugee Convention refugee definitions (Betts and Kaytaz, 2009; Betts, 2013a).

For example, as introduced above, in 2009, the state offered around 200,000 Zimbabweans who were living in South Africa the option of applying for a four-year exemption permit: the ZDP. Since then, the permits have been extended – broadly speaking – every four years, with the original permits replaced by the Zimbabwean Special Dispensation Permit (ZSP) and then the Zimbabwean Exemption Permit (ZEP) (Maple, 2023). During interviews with civil society actors, the topic of these temporary visas gave rise to the rare occasions when the Department of Home Affairs was praised. Many (albeit reluctantly) conceded that these policies were a ‘good thing’ in relation to the ongoing protection of forced migrants.

A form of legal protection offered by a host state is certainly better than no protection at all. Indeed, without the permits, individuals who refused to leave South Africa would be forced to live deeply precarious lives without legal status. Consequently, to some observers, the ZDPs show how the state is prepared to sometimes stretch its welcome to include broad groups of migrants from neighbouring states (Betts, 2013a).⁴² An alternative view, however, is that rather than this being an example of a state ‘stretching’ key refugee regime norms, the permits are an attempt at ‘regime shifting’ (Carciotto, 2018): through this process, South Africa has effectively shifted refugees from one mobility regime (the national refugee regime) to another (the national immigration regime). As a result, the migrants lose the potential human rights attached to the global refugee regime (Carciotto, 2018; Maple, 2023).

Regardless of how one views the exact motives of the state, this alternative hospitality towards forced migrants from Zimbabwe is, however, still a form of extended reception and thus remains conditional and temporary in nature. The permits are inherently short term, meaning that when they expire (every three to four years), migrants will face renewed anxiety about their continued presence on the territory.⁴³ For many, the renewal in 2017 was the third time the temporary permits had been renewed without the opportunity of alternative pathways to accessing permanent residency.⁴⁴ Thus, this prolonged form of reception is framed by some as ‘frozen futures’ (Daily Maverick, 2020a).⁴⁵

In conclusion, within the government, ideational importance is placed on being a beacon for human rights and democratic values on the continent. This framing stems from the post-apartheid nation-building, reconciliation and unity processes instigated by the ANC (Abrahams, 2016), the aim of which was to create a cosmopolitan and inclusive democracy (Polzer and Segatti, 2011). Within this context, the suggestion of containing refugees and forced migrants in refugee camps naturally evokes painful comparisons with the previous apartheid regime. Yet equally, these are far from universally shared beliefs within national-level structures. Indeed, almost paradoxically, these ideational factors are regularly contested by opposing democratic pressures within the government systems. The pressures against are motivated by a voting public, opposition political parties and media, who are now moving away from the ideals based on pan-Africanism and universal human rights in the face of rising poverty and inequality (Segatti, 2011). Certainly, the profound contradiction of pan-Africanism in South Africa today is the growth of xenophobic violence (African Arguments, 2015).

The first half of this chapter has brought together analysis on why South Africa chooses a free-settlement approach to the reception of refugees and why this has been maintained, even though the refugee camp remains the dominant reception approach in the sub-region. Nevertheless, an examination of key material, ideational and ideational factors has also highlighted ongoing tensions and contestation at the national level in relation to how refugees are received. The next section develops the analysis further by investigating major shifts in policy seen during the initial registration period of reception, since 2011.

Material and institutional factors affecting the shift in refugee policy

This section investigates the interconnected material and institutional factors that have created a situation whereby shifts in policy towards a more restrictive approach to refugee reception have been made possible. These include material concerns over capacity, institutional contestation between different governmental departments at the national level and increased security concerns that arise with increased movement on the territory. Each element represents a legitimate concern for the state, but also intersects with other factors (discussed below), including the increased securitisation of all immigrant movement in South Africa and overarching concerns about in/stability. When combined, they accentuate a dysfunctional national refugee reception policy and, in turn, a constantly shifting reception for refugees on the ground in South Africa.

Firstly, on a material level, the Department of Home Affairs has consistently struggled with capacity issues (in terms of resources and funding) when responding to the needs and demands of asylum-seekers and refugees during registration in South Africa. A senior barrister in Cape Town, when talking about the registration process, observed: ‘it’s very hit or miss and I think that’s got to do with … capacity and management of Home Affairs … it’s a capacity issue’.⁴⁸ The exact number of forced migrants in South Africa is unknown and regularly contested (Mthembu-Salter et al., 2014). Nevertheless, it is evident that the population of all types of forced migrants is at a sizeable level.⁴⁹ In addition, since the early 2000s, the number of asylum-seeker claims in South Africa has increased markedly. The sheer quantity of asylum applications, in conjunction with a general lack of government resources at the state and local level, and poor management within the Department of Home Affairs, has produced extensive (and now infamous) backlogs in asylum claims.⁵⁰ As observed by a civil society leader:

The problem is the department [Home Affairs] is so far behind. There [is] such a large backlog, people are waiting for years and years for the adjudication. Therefore the system is broken down at the moment.⁵¹

The lack of capacity and training also contributes to poor quality decisions and corruption (Mfubu, 2018).⁵² A manager of an implementing partner of UNHCR confirmed: ‘look, there is corruption in the system … even the Department of Home Affairs acknowledges that’.⁵³ In fact, many key members of civil society see the registration procedures as entirely broken. Interviews with civil society and refugee groups noted numerous examples of a dysfunctional system. These included vast delays for applicants, a general lack of information given by officials, lost paperwork and people needing to continually reapply for status or claims due to bureaucratic errors or failings. These kinds of delays result in most applicants waiting on decisions well beyond the designated response time of six months, with some waiting years or decades before a decision is made.

Stemming from the capacity issues within the Department of Home Affairs, there are stark differences in the implementation of refugee policy at the RROs. These include variations in information given to asylum-seekers and refugees by officials and differences in understanding relating to correct documentation and procedure. A human rights lawyer suggested there is ‘no standard operating procedure at Home Affairs’.⁵⁴ The same lawyer observed that ‘people will go down there and be refused, but you don’t get anything in writing as to why they were refused. Largely up to the manager at the time it seems like’.⁵⁵ A refugee community leader explained that because of the constant alterations to the policy on documentation, they must update their clients daily via WhatsApp:

The biggest issue is lack of information on the ground … coming from the Department of Home Affairs … because the policy of documentation is changing every day … when you meet one official and they say ‘x’ and then you meet another Home Affairs official and they say ‘y’.⁵⁶

This experience of idiosyncratic and arbitrary decision-making, and conflicting information, highlights how reception in the urban space is imbued with a sense of temporariness for refugees. As examined further in the next chapter, the reception offered to refugees reflects ongoing processes of negotiation and renegotiation between various key actors. During these ‘negotiations’, different factors interact, reinforce and/or contest with each other to create a given response, with refugees remaining as temporary guests within these volatile political spaces.

Secondly, connected to this capacity issue is the increased institutional friction between the national refugee framework and the national immigration framework. Since the introduction of the global refugee regime into national law in South Africa, tensions and contestation have existed between the two systems.⁵⁷ These pressures have never been satisfactorily resolved, indeed they have increased in the last decade, with increasing numbers of migrants using the national refugee framework to regularise their stay on the territory.

This national refugee/migration nexus was made explicit with the introduction of the Immigration Act (No. 11) 2002 (‘2002 Immigration Act’) and its accompanying regulations. These legal instruments created a restrictive immigration regime that assisted highly skilled immigrants but closed immigration to most – if not all – ‘low skilled’ workers (Johnson, 2015).⁵⁸ A glaring imbalance was therefore created between the ‘restrictive immigration framework and the liberal refugee protection framework’ (Johnson and Carciotto, 2018). Indeed, a former manager in the Department of Home Affairs believes the current failings in the system can, at least in part, be put down to a sub-regional ‘migrant issue’ rather than a ‘refugee problem’:

It’s a need to regulate mixed migration … many of them don’t necessarily have asylum claims but they want to be documented … and because the asylum permit gives them the right to work, it is the easiest way to be documented. The alternative is a convoluted visa application process that you probably wouldn’t qualify for anyway.⁵⁹

In a sub-region with substantial numbers of all forms of cross-border movement, the huge disparity between the different governance frameworks (for example, in terms of access at the point of initial registration) has left certain groups of migrants with little option. Thus, many migrants and forced migrants use the refugee national framework as an access point to the interior (Moyo and Zanker, 2020). Nevertheless, the narrative used to explain these issues, as illustrated in the above quote, remains problematic, albeit popular. For example, when asked about the current problems with the asylum system, a high-ranking official in UNHCR South Africa observed:

The regime in South Africa has been very good. Regrettably this situation has been abused by many people.⁶⁰

This prevalent explanation engages with the issues but frames the migrant her/himself as the source of the issue, rather than the underlying material, institutional and ideational problems within the two migration systems.

The third material factor that has created a shift in policy relates to security issues. Since the early 2000s, the increasing numbers of refugees and forced migrants coming into the state, coupled with a non-functioning registration process and the granting of freedom of movement and the right to work to all asylum-seekers, have inevitably heightened national security concerns. These include direct and indirect security issues, for example, fears over who exactly is coming into the state and applying for asylum. One civil society leader commented: ‘anyone can apply for asylum, we don’t know who they are, you don’t need passports or papers’.⁶¹ In addition, indirect security issues also emerge in urban spaces due to insecurity caused by large numbers of new cross-border migrants moving into cities with scarce resources (WEF, 2017).

Legitimate security and capacity concerns in South Africa stem in large part from the two entirely dysfunctional national frameworks related to migrant reception. Yet, irrespective of these facts, as the numbers of migrants continue to increase in urban spaces, the arrival and movement of all international migrants are beginning to be seen predominantly through a security lens.⁶²

Ideational factor affecting the shift in refugee policy: the increased securitisation of refugees in South Africa

In combination with the salient material and institutional factors set out above, there has been a marked ideational shift within key elements of the Department of Home Affairs and other branches of the government on how to approach the reception of refugees.⁶³ Specifically, the language of security is now filtering into all aspects of national policy surrounding refugee matters. This has culminated in the ruling party, and major opposition political parties in South Africa, essentially framing all refugee and African migrant movement as a security issue.

Post-apartheid South Africa continues to try and distance itself from the errors of the past, with a commitment to democracy and universal human rights. Nevertheless, the present government still takes several cues in relation to foreign policy, and in particular immigration matters, from the former apartheid regime. As Musuva (2015) and Vale (2002) both examine, the current national discourse on security (as set out below) can be traced to the former apartheid regime with its emphasis on control and surveillance. Both authors suggest this is still seen as the best policy option in dealing with the migration ‘problem’ (Musuva, 2015).

The current approach to refugees by the Department of Home Affairs has been built from this powerful historical legacy. Yet, the increased movement of refugees into South Africa, in combination with the factors set out in the previous section, has intensified and developed this security lens further still. The department, over the last twelve years, has adopted an increasingly reductive discourse in relation to the asylum-seeker and refugee ‘problem’. Building on the narrative of ‘people abusing the system’ set out above, this discourse goes further by essentially conceptualising all ‘individuals in the asylum system as illegitimate claimants without protection needs’ (Johnson and Carciotto, 2018:169).

This framing is repeated in the 2016 Green Paper on International Migration (DHA, 2016a). Sections of the policy paper that refer to refugees and asylum-seekers are continually framed by the statistic that 90 per cent of asylum-seekers who apply for asylum do not qualify, the inference being that most asylum-seekers are economic migrants or criminals infiltrating the system. Indeed, there are continued references to criminal syndicates from Africa and Asia coming into the country, with no supporting statistics or evidence.⁶⁴ As a result of these ‘bogus’ claims, ‘real’ refugees are being stopped from gaining access to protection.⁶⁵

The adoption of this cognitive framework in the Green Paper and other policy documents also implies a consensus within the political space regarding these issues.⁶⁶ Accordingly, there is no room for debate over the ‘truths’ being asserted – even though the empirical evidence set out above shows ongoing contestation taking place inside the government over the correct form of reception given to refugees. This ‘fallacy of division’ results in statements such as ‘massive abuses within the system mean “genuine refugees” are unable to gain access to protection’ being made as fundamental truths.⁶⁷

These ‘truths’ have either informed policy or have merged with policy to the point where many key informants suggest there is now a deliberate policy within Home Affairs to automatically reject asylum applications at the first stage. As a human rights lawyer suggested, the high rejection rates are now reinforcing and confirming this ‘truth’.⁶⁸ This ‘manufacture of illegality’ (Essed and Wesenbeek, 2004:68) via high levels of rejection of asylum claims and the construction of forced migrants as illegal migrants, can be understood as an attempt by the Department of Home Affairs to detach refugees from global and national refugee frameworks, and in doing so shrink the asylum space.

This reframing of refugees as illegal migrants or criminals has spread well beyond key elements of the government. Indeed, the ruling political settlement has almost universal approval for a securitisation-style approach to immigration:

Here there isn’t that kind of friction around migration. So there isn’t a political party whose constituency advocates for the migrant … so even with many of the amendments they just rubber-stamp it to run through Parliament because nobody even in opposition looks at it quite critically.⁶⁹

The data presented suggests that the motivation for the increasingly widespread use of a security lens stems from the belief that the number of refugees and international migrants in urban spaces is reaching unstable levels. As set out above, the current ruling political settlement essentially inherited a security lens for understanding migration from the old apartheid regime. Since the 1990s however, the motivation for using the lens has shifted from apparent direct security concerns to a broader preoccupation around maintaining stability. Analysis in this chapter has continued to emphasise how the state has become increasingly concerned with the perceived instability that migrant movement is bringing into the urban space. This instability emerges from concerns relating to capacity and insecurity (for example, strains on services and labour markets), as well as more democratic fears about the growing tensions within the voting public.⁷⁰

Thus, the ‘problem’ of refugees and low skilled migrants, particularly in urban spaces, can be understood conceptually as instability. This in turn suggests that securitisation is a localised process, which is prone to change and realignment over time (Donnelly, 2017). Motivations for adopting the lens will evolve given the specific context and contemporary factors. However, regardless of the reason for the increased securitisation of refugee movement, the result is typically the same for the refugee: a progressively hostile reception and a shrinking asylum space.

Exclusion from the urban space

The previous subsections have demonstrated that a combination of material and institutional factors as well as ideational and discursive approaches based on security and stability concerns, are producing key policy changes. These in turn are fundamentally reframing refugee reception in South Africa. As changes in policy and practice increase and work their way into the national legal framework, a new form of conditional reception is emerging which includes increased checks and restraints on movement – particularly in relation to access to urban spaces. Indeed, the reception of refugees in South Africa is slowly shifting away from a free-settlement policy towards something resembling encampment.

One of the key provisions of the 1998 Refugee Act was the creation of Refugee Reception Offices (RROs) (Khan and Lee, 2018). These have traditionally been in large urban centres such as Johannesburg, Pretoria, Cape Town, Durban and Port Elizabeth (Moyo and Botha, 2022). By granting freedom of movement to refugees, and by extension access to the urban space, the primary point of contact between refugees and the state during the registration and post-registration stages occurs at these RROs (Johnson and Carciotto, 2018). Certainly, the centres, which are run by the Department of Home Affairs, are the main entry point for accessing the refugee regime, regularising refugees’ status and legitimising their presence in an urban space.

The renewal time frames for asylum permits were created when the drafters of the legislature ‘envisioned that the application process would finally be adjudicated within 180 days of the application being made’ (Khan and Lee, 2018:1269). In reality, the asylum process (and by extension the registration stage of reception) can often take years, with refugees still expected to visit an RRO every one to six months to renew their temporary permits. As such, the urban RROs are ‘essential to the functioning of the system and for accessing the protection it affords’ (Khan and Lee, 2018:1271). Therefore, any restrictions on accessing the RROs – particularly in urban areas – fundamentally alters the spatial dimensions of refugee reception and the ability of refugees to pursue their own personal and economic aims.

Since 2011, the Department of Home Affairs has closed the RROs in Cape Town, Johannesburg and Port Elizabeth. Moreover, between 2011 and 2020 only three RROs were operating correctly, with others either closed or partly closed (Moyo, Sebba and Zanker, 2021). This occurred despite there being a need for offices in large urban centres. Judicial court orders from the high courts and the Supreme Court of Appeals demanding the reopening of centres have regularly been ignored. Indeed, even a Constitutional Court order to reopen the centres went unheeded (Crush et al., 2017).⁷¹ The result of the closures has meant that asylum-seekers are forced to choose between moving closer to the remaining RROs (far from key urban areas), allowing their documentation to expire or repeatedly travelling long distances to register and receive assistance.⁷²

Where you’re closing offices left, right and centre, you’re funnelling people into having an out of dated permit … if I have to travel up to Durban, to get my permit every three to six months … [is your] employer going to allow you to, to take that much leave given it’s every three months?⁷³

The policy of closing key urban centres ultimately means refugees’ ability to freely move around the interior is seriously curtailed. Specifically, due to the often overwhelming practical and material implications of repeated long journeys to the remaining centres, access to specific key urban spaces has been severely reduced. This is seen by civil society leaders as a deliberate attempt to keep refugees out of ‘crowded’ cities like Johannesburg and Cape Town. As one civil society leader noted, ‘this was an attempt to limit freedom of movement. Reducing access to the interior … [based on a] security argument to keep people at the borders’.⁷⁴ In addition, a human rights advisor commented, ‘there was no real reason to close the Johannesburg office, it was operating quite fine … the only reason to really close it was to fit into this plan’.⁷⁵

Finally, this reframing through ideational shifts and changes in policy and practice is ultimately feeding up to the legislative branch of the government. New amendments to the national legal refugee framework (as introduced above), which came into force in January 2020, will restrict the asylum space further and reduce access to the urban space (Ziegler, 2020). For example, the new proposed asylum-seeker registration forms ask questions about asylum-seekers’ bank accounts, wage slips and employment history. This information should be immaterial to RSD proceedings. In fact, these questions are more suited to visa applications for labour migration and fit with broader themes of national interest. Furthermore, the 2017 Refugee Amendment Bill amends Section 1 of the Refugee Act to change the definition of an ‘asylum-seeker permit’ to include the term ‘visa’ in order ‘to align it with the Immigration Act 13 of 2002, as amended’.⁷⁶

The Amendment Bill also removes the right to seek and attain employment for asylum-seekers. Key informants saw these amendments as ‘a direct response to get lower numbers of asylum-seekers’,⁷⁷ or the shrinking of the asylum space. These amendments will especially affect asylum-seekers at the point of registration. With no state-run assistance programmes and now no legal way of earning money, these individuals will need to rely on the informal economy, local networks, civil society and faith-based organisations to locate immediate shelter and basic necessities.

Leading figures in civil society see the ultimate objective of these incremental changes (or as one leader coined it, ‘the deterioration of asylum’) as the removal of asylum-seekers from the urban space to the border areas.⁷⁸ Indeed, this overarching change to the spatial dynamics of reception in South Africa is explicitly set out in the Green Paper, with the proposed creation of processing centres at the border. Under these proposals, asylum-seekers would be processed at the point of entry in shorter timeframes, with refugees then given access to the interior (and granted key norms within the regime).⁷⁹ If implemented correctly, the processing centres are therefore unlikely to break international or national law. Yet, as noted by a prominent human rights lawyer, in proposing this approach, the Department of Home Affairs is presuming a level of efficiency they have not been able to deliver in the last thirty years.⁸⁰ Without additional investment by the state and the international community, there is a high risk that all refugees under this proposal could remain indefinitely in encampment-like situations or informal townships near the border.