Bordering through destitution: the case of social assistance to irregularised migrants in Malmö, Sweden

ABSTRACT

In this article, we bring attention to the local-level administration of social services as a site of bordering. Specifically, we focus on the provision of social assistance (i.e. a means-tested cash support program, ekonomiskt bistånd) for irregularised migrants. Based on a close comparative reading of the City of Malmö’s 2013 and 2017 guidelines on social assistance, we analyse how entitlements to social assistance have been redefined and restricted following the 2015 so-called refugee crisis and the subsequent overhaul of Swedish asylum policy. Prior to this ‘crisis’, in 2013, the City of Malmö became the first Swedish municipality to extend access to social assistance to irregularised migrants. In 2017, the guidelines were revised with the expressed aim to discourage irregularised migrants from remaining in the country. We see this as a shift from a needs-based approach to one that, instead, sees social policy as subordinate to the goals of immigration enforcement. Further, we conceptualize this as a shift towards a type of indirect internalized bordering measure that so far has received relatively little scholarly attention in the Nordic context, namely self deportation measures that aim to deter immigration and encourage ‘voluntary return’ by restricting access to public services and welfare benefits for (irregularised) migrants. Finally, we argue that the overall specialization, juridification and standardization of social service provision supports the shift in question, providing a convenient justification for restricting entitlements to irregularised migrants.

KEYWORDS:

• Immigration
• social policy
• social administration
• social welfare
• children’s rights

Introduction

Critical migration scholars have called attention to the nexus of migration and social policy, and the growing importance of indirect and everyday practices of bordering in immigrant-receiving welfare-states in Europe and North America (Varsanyi 2010; Tervonen, Pellander, and Yuval-Davis 2018; Yuval-Davis, Wemyss, and Cassidy 2019). The paradigmatic example is the infamous ‘hostile environment’ policies, first implemented by the UK government in 2012, intended to make it progressively harder for irregularised migrants in the UK to live, work and access services, including public healthcare (Coddington, Conlon, and Martin 2020).

Everyday bordering typically relies on local-level policies and practices of exclusion (Persdotter 2019). In this article, we focus on municipal-level social service provision as a site of bordering. More specifically, we consider the provision of social assistance (i.e. a means-tested cash support program, ekonomiskt bistånd) for irregularised migrants residing in Malmö, Sweden.¹ Our aim is to examine the presence and configuration of local-level bordering practices. Specifically, we analyse changes in municipal policies and practices pertaining to the welfare entitlements of irregularised migrants following the 2015 spike in asylum-seekers arriving in Sweden and the associated reversal of asylum-policy. Based on a close reading of the City of Malmö’s 2013 and 2017 guidelines on social assistance, we consider whether and how entitlements to social assistance have been redefined and restricted over the last several years.² In doing so, we ask whether we are seeing the emergence of local-level ‘self-deportation policies’: that is, policies that aim to encourage ‘voluntary return’ by restricting access to services and welfare benefits for migrants (Park 2019).

While much previous research on the nexus of migration and social policy in the Swedish context has tended to focus on national-level policies (see Sager, Holgersson, and Öberg 2016) we highlight the municipal level as an important one for the development of indirect and everyday bordering practices. In Sweden, social assistance is administered by the municipalities, giving local policy guidelines central importance in this context. In 2013, the City of Malmö, adopted guidelines that extended partial access to mean-tested social assistance to irregularised migrants. In the wake of the 2015 ‘crisis’, the issue of irregularised migrants’ entitlements to social assistance became intensely politicized, and the policy guidelines in question were thus revised in 2017. We argue that although municipal social policy is still more inclusive of irregularised migrants than national-level policies, access to social assistance has been partially restricted as part of a strategy to encourage ‘self-deportation’. In particular, we point to three developments – the specialization of services, the concomitant standardization of practices, and the juridification of discourse (Bergmark 2014; Ponnert and Johansson 2018) – that jointly contribute to the exclusion of irregularised migrants from municipal social services.

The article is structured as follows: After giving an overview of policy changes post-2015, we introduce the concept of ‘self-deportation’. We then present the regulations on social assistance, specifically in relation to irregularised migrants, and connect this to the processes of specialization, juridification and standardization. In the analysis, we compare and analyse the Malmö guidelines for social assistance. We conclude by discussing the broader implications and relevance of our findings.

Policy changes in the aftermath of the events of 2015

The social entitlements of irregularsied migrants is a contentious issue in Sweden today. In November 2015, the Social Democratic-Green minority government made a decision to effectively close the border to prospective asylum-seekers and abandon the country’s theretofore relatively generous asylum-policies in favour of ones that just about met the EU’s minimum standards for the treatment of asylum-seekers. The decision was announced amidst increasing alarm over the supposedly detrimental effects of refugee-immigration on the solvency of the welfare state (Barker 2018), and was expressly justified by the Prime Minister, Stefan Löfven, as a means to create ‘breathing space’ (‘andrum’) for the Swedish authorities (Regeringskansliet 2015). In this section, we highlight some of the implications of these developments for irregularised migrants, a heterogenous categorization that, in the Swedish context, mainly been used to refer to recently rejected asylum-seekers, but that also includes individuals who have resided and/or worked in Sweden for years, as well as EU residents without right of residence.

The policy reversal was accompanied by a significant discursive shift in the migration policy debate, and a renewed emphasis on repatriating rejected asylum-seekers to their countries of origin (see Lindberg 2019). As part and parcel of this new policy approach, the parliament also revised the Reception of Asylum Seekers’ Act (Lag om mottagande av asylsökande m fl., henceforth LMA), cutting entitlements to state-funded accommodation and economic support for rejected asylum seekers. Later, in 2018, the Prime Minister also expressed an intention to put an end to the possibility for municipalities to give social assistance to irregularised migrants (Lönnaeus 2018). Altogether, these developments marked a turning point in the Swedish approach to asylum seekers, and a break with what now appears as a short-lived trend of expanding irregularised migrants’ entitlements to public services.³ At the time of writing, in the summer of 2020, the government has not (yet) moved to restrict entitlements to social assistance for irregularised migrants. However, the issue is still very much alive in the public and political debate (see Mikkelsen 2020).

The post-2015 asylum policy overhaul entailed the introduction of mandatory ID checks on the Danish side of the Denmark–Sweden border to prevent would-be asylum-seekers from being able to enter into Sweden. This drastically reduced the number of arriving asylum-seekers (Migrationsverket 2016). The parliament also approved a bill (Prop 2015/16:174) intended to restrict the possibilities for asylum-seekers to obtain residency status in Sweden, and instructed the Border Police to ramp up deportations. While the policy changes were introduced as a temporary measure, they remain in effect almost five years later, in 2020. As a direct result of these changes, it has become more difficult for asylum-seekers to find protection in Sweden. Some estimate that the number of irregularised migrants increased by at least 30,000 between 2015 and 2017, a doubling of the numbers since 2011 (TT 2017). Many live in the metropolitan cities – Stockholm, Göteborg, Malmö – where access to social networks and job opportunities is better than in rural municipalities (Holgersson 2016).

The aforementioned 2016 amendment to the LMA further contributed to irregularising rejected asylum-seekers. This can be seen as exemplary of a ‘minimum rights approach’ towards irregularised migrants, intended to pressure people into complying with deportation orders (Lindberg 2019). In Sweden, irregularised migrants have no rights to take up formal employment and only minimal entitlements to social assistance and other welfare services, with the exception of basic education for children and health care that cannot be deferred. The amendment to the LMA, combined with the overall reversal of migration policy, has likely funnelled people into homelessness and deepened their poverty. It has also increased the pressure on the municipal social services: With no entitlements to state-financed economic support under LMA, seeking municipal emergency social assistance is one of the last available resorts for rejected asylum seekers (Kjellbom and Lundberg 2018).

Coinciding in time with the closing of the border to asylum-seekers, the government put out a policy package aimed to ‘reduce vulnerability and begging in Sweden’ by discouraging the immigration of so-called ‘vulnerable EU citizens’.⁴ In this context, the municipalities were instructed to take a restrictive approach towards this group so as to not encourage their immigration and settlement (Barker 2017; Persdotter 2019). At the time, there were about 5,000 vulnerable EU citizens without right of residence in Sweden (SOU 2016, 6). While vulnerable EU citizens’ experience many of the same exclusions as rejected asylum seekers, they cannot as easily be expelled or deported. Elsewhere, Lind and Persdotter (2017) have referred to this as the ‘differential deportability’ of vulnerable EU citizens and rejected asylum seekers, and it is key to understanding the government approach to vulnerable EU citizens. Unable to enact restrictions on the entry of EU citizens, but nevertheless intent on deterring ‘poverty migration, the different levels of government (especially the municipal one) has turned to micro-managing conditions of stay for vulnerable EU citizens, including their access to public services. As we discuss below, the social services have also tended to differentiate between rejected asylum seekers and EU citizens when assessing claims for social assistance, but the different categorizations also overlap.

Contextualizing indirect bordering and ‘self-deportation’ policies

Immigration policy (i.e. policy related to the admission, entry, exclusion, and expulsion of non-citizens from state territory) is commonly assumed to take place at the outer borders of the state. Critical border and migration studies scholars have made the case that we need to understand ‘the border’ not as a fixed point or line in space or time but rather as a dynamic process that plays out in a multiplicity of spaces, involving a variety of different agents – outside and inside of state territory (Van Houtum and Van Naerssen 2002). Practices of bordering are not simply about conditions of entry but also – and, for our purposes, more importantly – about conditions of stay (Anderson 2017). In this article, we focus on social service provision as a site of bordering, and in doing so, we explore the emergence of a type of indirect bordering measure that aim to deter immigration and encourage ‘voluntary return’ by restricting and reducing access to public services and welfare benefits for migrants. Such measures can be conceived of as ‘self-deportation policies’, that seek to ‘make individuals into agents of the state’s goal of their removal by making their lives unbearable’ (Park, 2019, 1882).

This type of policies are best conceptualized as indirect policies in that they do not always have as their explicit and expressed purpose to regulate immigration (cf. discussion in Park 2019). Indeed, ‘minimum rights approaches’ are sometimes justified on the basis of budgetary considerations, or in relation to broader social policy goals. Such policies also tend to involve a variety of different government actors not directly or primarily responsible for immigration enforcement. In the case of ‘minimum rights approaches’, which are situated within the nexus of migration and social policy, social service providers might, for example, come to assume a significant role as collaborators in border enforcement efforts (Nordling 2017).

Notably, self-deportation policies are also characteristically sub-national policies that rely on the distinctive powers of local governments and authorities to intervene in the everyday lives of people (see Valverde 2010). As Ien Ang (2006, 32) notes, a city ‘cannot, generally, declare the closure of its borders’. Thus, municipalities that are unable to enact direct restrictions on entry sometimes find other means of controlling the mobilities of ‘unwanted subjects’, citizens and non-citizens alike (for a comprehensive review of such policies, see Persdotter 2019, 90–116). As we discuss in more detail below, it is crucial to engage with the municipal level to understand the mechanisms and effects of ‘minimum rights approaches’, and how to challenge such approaches. Municipal and other sub-national mobility control policies need not be exclusionary – there are many examples of ‘sanctuary cities’ that have adopted policies to protect irregularised migrants against immigration law enforcement (Nordling, Sager, and Söderman 2017). As we will explain, the Malmö case displays both inclusionary and exclusionary tendencies.

Local social service provision as a site of bordering – the case of social assistance

The question of how to address the public service needs of residents with irregular migration statuses has been described as an ‘intractable policy problem’ (Spencer 2018), one that sits uneasily at the intersections of border enforcement and social policy. Typically, national law guarantees only minimal welfare entitlements to those who are residing in a country without official authorization. Yet, the complete exclusion of irregularised migrants from public services is frequently contested as being morally indefensible, inconsistent with international human rights norms, and practically untenable (see Bosniak 2007).

Social assistance is a temporary and means-tested form of support, meant to cover basic needs, such as food and housing (Panican and Ulmestig 2016). It constitutes the ultimate ‘safety net’ of the Swedish welfare system.⁵ Entitlement to social assistance is regulated in the Social Services Act (henceforth SoL). This is a framework law (ramlag), meaning that it contains few detailed regulations; instead, it defines the guiding values and overall principles of the (public) social services. Claims for social assistance are supposed to be assessed on a case-by-case basis. While there are national-level recommendations, responsibility for administrating social assistance is placed at a municipal level. Municipalities typically make their own guidelines that are meant to provide direction to frontline social workers. Such guidelines are decided upon by local politicians. Social assistance is also financed directly by the municipalities,⁶ and previous research shows that many frontline social workers are under pressure from their managers to be restrictive when assessing claims for social assistance (Hjort 2012).

SoL does not contain any explicit mention of citizenship; the key category used to define who is entitled to assistance and services is instead that of ‘residency’ (bosättning). Crucially, the ambiguity of this category presents an opening for variously irregularised subjects to claim social rights at the local level (Nordling 2017). Indeed, SoL includes a clause (2 ch. 1§) on ‘local responsibility’, which specifies that ‘the municipality is ultimately responsible for making sure that those who are present (vistas) in that municipality have access to the support and assistance that they need’ (our translation and emphasis). This is referred to as the ‘principle of ultimate responsibility’, and it could be said to represent a mechanism of basic humanitarian protection and care. It makes it incumbent on the social service administration of a given municipality to address the emergency needs of those who are geographically present within their jurisdiction if there is no other authority responsible, and regardless of their citizenship status (Kjellbom and Lundberg 2018; Cuadra and Staaf 2014). Support might be given in the form of money for food and other urgent expenses, or in the form of a return trip to the applicant’s municipality or country of regularized residency.

In 2013, the City of Malmö became the first municipality in Sweden to adopt local guidelines on social assistance that explicitly addressed irregularised migrants. Similar guidelines have later been adopted in a number of other Swedish municipalities, including Stockholm and Norrköping. In 2017, Swedish Supreme Administrative Court ruled in a case concerning the obligation for municipalities to give support to irregularised migrants (HFD 2017 ref. 33; Lundberg & Kjellbom, this issue). According to the judgment, municipalities do not have any obligation to give (regular) social assistance to the group. However, they may choose to expand their responsibilities (as is done in some municipalities with reference to the UN Convention on the Rights of the Child, UN-CRC). In the case of emergency social assistance, claims are still to be assessed individually. The court judgment has been interpreted by some as a justification for limiting entitlements to social assistance for irregularised migrants (Rydhagen 2017). However, the issue is still subject to much debate (see Mikkelsen 2020).

In addition to formal restrictions, there are also practical barriers to access. For example, some irregularised migrants might avoid contacting the Social Services out of fear of deportation (see Jolly & Lind, this issue). On at least one occasion (in November 2016), the Swedish Border Police have contacted the Social Service Administration in Malmö and requested the addresses of irregularised clients. The event in question caused debate about the role and professional ethics of social workers, and became subject to a review by the Parliamentary Ombudsman.⁷ However, it remains the case that irregularised migrants who receive support from the social services risk being caught by the Border Police. In other words, social assistance is both associated with indirect bordering practices, regulating and restricting access to social services and enforcing conditions of destitution and with direct forms of border policing (e.g. being detained and deported).

Juridification, specialization and standardization

In analysing social assistance as a (potential) site of indirect bordering, we draw on three analytical concepts from sociolegal studies – specialization, juridification and standardidation. We believe that these concepts aid our understanding of bureaucratic management at the intersection of border enforcement and social policy.

Specialization refers to the overall tendency of the Swedish social services to organize the provision of services into separate divisions, corresponding to separate categories of service users, rather than taking an holistic approach (Bergmark 2014). A tendency towards specialization can be identified in the efforts to elaborate a common approach to irregularised migrants. Instead of being addressed among other service users, the Social Services search for a specialized and streamlined approach to irregularised migrants, conceieved as a separate category of service users. This kind of specialization can further be linked to two other tendencies – juridification and standardization.

Juridification can be described as the increasing reliance on legal discourse and mechanisms to resolve conflicts in society. Ponnert and Johansson (2018) make a distinction between the two legal dimensions of juridification and standardization. In their conceptualization, juridification is characterized by an increased legal regulation and/or ‘discursive institutional faith in legal regulations and legal norms as a means to improve practice’ (Ponnert and Johansson 2018, 2026). Standardization, on the other hand, implies a formalization of practice with an intention to be coherent and systematic. For example, guidelines or assessment frameworks are tools that are based on pre-defined categories of needs instead of a rights based ideal. Such tools are not necessarily legally binding, but can be understood as a form of ‘soft law’ that sometimes has greater impact in practice than ‘hard law’ (ibid.).

Crucially, the processes of specialization, juridification and standardization are not wedded to any one political agenda. In the field of migrant justice and rights, juridification can support an agenda of inclusion as well as one of exclusion. For the purposes of this article, however, we are mainly concerned with the mobilization of legal discourse to justify exclusionary measures and restrictions of the rights of migrants. Seeing these processes in the light of theories on self-deportation, we highlight how a specialization and standardization of social work practice through changes in organization and routines, as well as a juridification of language, may be used as a seemingly neutral way to implement austere and restrictive politics.

Method and delimitations: studying bordering practices at the municipal level

Geographically, the study focuses on a single city – Malmö. Situated at the proverbial ‘gateway to and from Europe’, Malmö is typically the first city of arrival in Sweden for migrants coming from continental Europe. It is also a city with a comparatively high unemployment rate, and a high percentage of households reliant on social assistance. During the last couple of years, the cost of social assistance in Malmö has become an increasingly politicized topic in local politics (Nordling 2017).

In regards to irregular migration, Malmö is interesting as a case study because it is assumed to have an unsually large population of irregularised migrants. Migration-related activism has also given visibility to the issue of irregular migration (Sager 2016). The authorities’ responses to this have been ambivalent. On the one hand, Malmö was the first municipality in Sweden to formally extend access to economic support to irregularised migrants in 2013. On the other hand, the extension of social services has been accompanied by an intensified search for irregularised migrants through controls in streets and at workplaces. The presence of irregularised migrants in the city is also frequently linked to a variety of security issues (Schclarek Mulinari 2020). Hence, Malmö is a city where dimensions of partial inclusion/exclusion and internal border policing are particularly visible.

The article is primarily based on a close reading and analysis of the main policy document that regulates the provision of social assistance by the social services: the City of Malmö’s guidelines on social assistance. In particular, we compare and contrast the guidelines adopted in 2013, with the ones introduced in 2017. As we explain above, these guidelines are intended to provide directions to social work practitioners and can be expected to greatly influence the implementation of social policy at the local-level. Our analysis is inspired by methods of discourse analysis (see Sahlin 1999). Specifically, we focus on content-related changes and discursive shifts between the two documents.

The article is further informed by our previous PhD research projects on social work with irregularised migrants and ‘vulnerable EU citizens’, both of which were based on interviews with local policy makers and practitioners (see Nordling 2017; Persdotter 2019). In order to complement and update this material, we have carried out two additional interviews as well as multiple informal conversations with voluntary-sector service providers and activists in the Malmö asylum rights movement.

Analysis: local bordering and the administration of social assistance

As mentioned already, the adoption of the Malmö City Guidelines on Social Assistance in 2013 articulated a right (albeit a partial and provisional one) for irregularised migrants to access means-tested emergency social assistance (Malmö stad 2013). Crucially, this marked the first time in recent Swedish history that irregularised migrants were explicitly recognized in municipal social policy as a (potentially) rights-bearing category of de facto residents. Although national news media described the decision as being ‘controversial’ (Sverin 2013), Nordling’s (2017) research has shown that the rights of irregularised migrants was not, in fact, a particularly contentious issue in municipal politics at the time. Rather, local politicians saw the decision as clarifying the applicable law. The guidelines were only covering partial and emergency needs and should therefore not be understood as radically altering the conditions of inclusion/exclusion from citizenship.

In 2017, a new set of guidelines on social assistance were decided upon by the Malmö City Council. The official press release described the new guidelines as ‘more restrictive’, and as being centred on the principle that rejected asylum-seekers should leave the country (Malmö stad 2017a). The impetus to revise the guidelines came in part from the aforementioned 2017 decision of the Supreme Administrative Court, which established that municipalities do not have any obligation to provide regular social assistance to irregularised migrants. The revision also reflected the post-2015 turn asylum-policy and discourse, and the government’s emphasis on repatriating irregularised migrants.

In the following sub-section, we present a detailed reading of the 2013 and 2017 guidelines. We observe that a significant shift in discourse and practice has taken place – even if no changes have been made in SoL. In particular, we highlight four significant shifts (see below). Taken together, these are indicative of a turn away from a needs-based approach towards one that bears some – if not all – of the characteristics of self-deportation policies.

Changing rationalities: from needs to self-deportation

First, when we read the 2013 and 2017 guidelines together we notice an overall shift in language: the 2017 guidelines are framed in a legalistic language, using formulations that are close to the original law text. While some of these changes were surely intended as clarifications of how the right to social assistance should be interpreted, we can see a shift in the language that signals a different standpoint compared to the earlier guidelines. We take this as an example of the re-framing of public, political and professional ethical issues in legal terms, and thus as a sign of juridification. This shift is accompanied by a re-categorization of irregularised migrants: whereas the 2013 guidelines used the term ‘undocumented’, a term frequently used by Swedish authorities, the 2017 guidelines instead use the phrasing ‘persons without permission to stay in Sweden’. The 2013 guidelines clearly stated who was to be considered as ‘undocumented’:

An undocumented person is in Sweden without permission. The group includes persons who have received an order of expulsion, persons with an expired or redrawn residence permit and persons who have never applied for a residence permit. Also EU citizens without a right of residence and their family members belong to this group. (Malmö stad 2013, 13)

In the 2017 guidelines, there is instead a paragraph that makes it clear that there must be an investigation of the migratory legal status as part of the assessment.

When applying for assistance pursuant to Chapter 4. 1§ SoL, an investigation of the applicant’s migratory status must be conducted as law as different rules apply depending on the applicant’s status and circumstances. (City of Malmö stad 2017b, 15)

In other words, the new guidelines underline the unauthorized status of irregularised migrants. This indicates a discursive shift towards a more restrictive language. In the new guidelines, the phrasing is also shorter and formulated in a more juridical language.

Second, and most importantly for our purposes, the revised 2017 guidelines are characterized by a stronger emphasis on (potential) conflicts between the provision of social assistance to irregularised migrants and the aims of border enforcement. For example, the 2017 guidelines underline that

Persons without permission to stay in Sweden lack legal residence in the country and shall therefore leave Sweden. It is their responsibility (‘det åligger dem’) to follow Swedish law and decisions made by another authority. (Malmö stad 2017b, 15)

As a comparison, the 2013 guidelines emphasize that ‘undocumented migrants’ should be individually assessed ‘in the same way as other persons residing in the municipality’, thus highlighting their status as de facto residents:

What types of support and help are considered as acute shall, in the same way as other persons subject to the regulation, be assessed through individual assessment. There are no special regulations for undocumented migrants; they should be treated in the same way as other persons residing in the municipality and applying for support from the social services. (Malmö stad 2013, 13)

While the 2013 guidelines emphasizes the needs of ‘undocumented migrants’, the 2017 guidelines instead stresses their individual responsibility and failure to comply with their deportation orders and/or other decisions from the Migration Board. We interpret this as a ‘responsibilisation’ of the individual migrant (cf. Luibhéid 2013).

Third, the 2017 guidelines place a relatively stronger emphasis on the short-term nature of emergency social assistance. The 2013 guidelines stress the principle of ultimate responsibility, emphasizing that ‘the ultimate responsibility of the municipality remains as long as the person resides in the municipality and does not get her/his support met in another way’ (Malmö stad 2013, 13). Hence, they allow for an interpretation of ‘emergency’ as a lasting, long-term situation. The 2017 guidelines does not include any mention of the municipality’s ultimate responsibility. Instead, they separate irregularised migrants from other groups, stating that the ‘unauthroised’ applicant may only qualify for emergency social assistance under extenuating circumstances (synnerliga skäl):

The fundamental rule for adults is to only grant emergency support when there are extenuating reasons to do so. The responsibility of the Social Services is therefore limited to support and aid in emergency situations when the person is residing in the municipality (2a kap. 2§ SoL). The concept of emergency situations above all concern situations that arise unexpectedly and unforeseen. Emergency support is assessed for a short period. When applying for support in emergency situations the applicant shall attend personally. (Malmö stad 2017b, 15)

The 2017 guidelines hence focus on the applicant’s temporariness, stating that the person should leave the country and that the support should be assessed for a short period of time. In the 2017 guidelines, the temporary status of applicants is treated both as a legal and organizational issue. It is underlined that the applicant must ‘follow Swedish law’ (juridification), and that the assessment should be made for short periods of time and that claims therefore needs to be reassessed more often. This emphasis on the temporariness of irregularised migrants residency makes it more difficult for the individual to plan forward in daily life. At the same time, it reinforces the idea that irregularised migrants should only be staying in Sweden for a short period of time. It can therefore be understood as an encouragement to self-deportation. Furthermore, the change in discourse – towards a juridified language – give less leeway for social workers to interpret the Social Services Act in ways that are favourable for irregularised migrants.

Fourth and finally, the 2017 guidelines redefine the best interest of the child. Here a rephrasing has been made concerning how to apply the principle of the best interest of the child. This is of special interest, as the 2013 guidelines were aimed to strengthen the child perspective (Malmö stad 2013). In relation to irregularised migrants (in the text: undocumented migrants), the 2013 guidelines stated that

In the assessments of children’s need of support and help, the principle can be used in a way that in every sense treats children as if they were residents in the city of Malmo. This means social assistance according to the full norm and right to assistance for general costs in the same way as children resident in the municipality. In order to give children residing in our municipality reasonable living conditions it is recommended that the principle is used in this way. (Malmö stad 2013, 13).

In 2013, the municipality recommended that children should receive the national norm for social assistance. It was also recommended that families with children should get support in regard to housing. The 2017 guidelines no longer state that this interpretation is recommended, but that the principle might still be applied in this way:

The Social Services have a special responsibility for children, especially children belonging to vulnerable groups. Children shall, when deemed appropriate, be able to be granted support to a reasonable standard of living. Based on the principle of the Best Interest of the Child this might mean that children are granted social assistance while their parents are only granted emergency support. Assistance to children might be granted up to full norm [riksnorm]. Considering the Best Interest of the Child, assistance might also be granted to a reasonable cost for housing that is based on the child’s need. (Malmö stad 2017b, 16)

In 2017, the formulation that ‘the principle can be used in a way that in every sense treats children as if they were residents in the city of Malmö’ was removed. This is a language change that indicates less encouragement to focus on children’s interests before migration control. The removing of recommendations also gives social workers less direct tools in their work with irregularised children. This indicates less standardization in a context where standardization of routines (i.e. always giving ‘full norm’ to children) would be advantageous for irregularised migrants. Together with a harsher approach to irregularised migrants in general, this may give children less access to social assistance. The tendencies to encourage self-deportation are therefore not only directed to adults.

Changes in organization and practice

In addition to changing formulations analysed above, the 2017 revision of the guidelines has been followed by two noteworthy organizational and practical changes in the hands-on administration of social assistance to irregularised migrants. In this section, we turn from the document analysis to consider some of these changes.

In 2018, the City of Malmö opened a specialized unit, tasked with processing applications made by irregularised migrants. Before the opening of this unit, migrants were expected to apply with the social services office in their neighbourhood. Now, they are all referred to the same one unit. In theory, this centralization and specialization could be beneficial to irregularised migrant applicants. By focusing exclusively on one group, social workers might gain a deeper understanding, both of the applicable regulations, and of the specific needs and challenges experienced by the group in question. However, according to the voluntary-sector service providers we spoke with during our research, the establishment of this specialized unit has primarily enabled a streamlining of the assessment process and a greater degree of control and oversight, with case officers now requiring more extensive documentation to ascertain the economic and migration status of the application (personal communication). This serves as yet another example of how the social services in Malmö are shifting from a comprehensive assessment of individual needs towards a category specific and group-based assessment procedure.

Crucially, the establishment of a specialized unit has also contributed to a heightened sense of unsafety among irregularised migrants who depend on social assistance. At the time of writing, the office where they have to file and renew their applications in person is only open for two hours at a time, two days a week. This creates a practical barrier for applicants to access support (cf. discussion on the control of migrant’s time in Misje, this issue). The border police in southern Sweden is widely known to have carried out internal border controls outside health care facilities and schools, and it is not difficult to imagine that the existence of a specialized office for irregularised migrants would allow for easier monitoring and surveillance. There are reasons to believe that the requirement to apply at a specific office (during a set time-span) creates anxiety and fear among migrants to such an extent that some might avoid seeking assistance altogether.

Another tendency that has been noted among voluntary-sector service providers is that the distinction between former asylum seekers and vulnerable EU citizens is drawn differently post 2015. Neither version of the guidelines (2013/2017) distinguish explicitly between former asylum seekers and vulnerable EU citizens. It used to be the case that while vulnerable EU citizens typically received emergency social assistance in the form of a bus-ticket to their home countries, rejected asylum seekers could count on receiving cash support (Nordling 2017; Persdotter 2019).⁸ Through interviews with voluntary-sector service providers, we have learned that it is now more common to give bus-tickets also to former asylum seekers, especially ones from the Balkans.⁹ This direct way to encourage former asylum seekers to leave the country can be understood as a tendency to restrict the access to social assistance also for families who have lived in Malmö for many years. The two categories of ‘undocumented migrants’ that are actualized in relation to the social services thus seem to be given more similar treatment than before, and the distinction between residing on a long term and a short-term basis seems to be blurred. Furthermore, the practice of giving out bus-tickets can be conceived of as a bordering measure, or as a technology of (self-)deportation. Indeed, it bears questioning why one of the only forms of social assistance that is given to some irregularised migrants is one that effectively functions to repatriate them.

Conclusions

As we are finishing this article, in the summer of 2020, a parliamentary committee has just put forward a comprehensive migration policy package that proposes to make core aspects of the 2016 temporary law on asylum permanent, and that centres on a logic of deterrence (SOU 2020, 54). As journalist Petter Larsson (2020) puts it, ‘the core idea [behind the policy package] is to make life more gruelling for those who are already here in order to deter and make it more difficult for others to seek asylum’. At the same time, the City of Malmö is again revising the guidelines on social assistance, renewing the formulations from 2017. This then seems like a good moment to return to the question we posed at the outset of this article.

The purpose of this article was to explore municipal-level social service provision as a site of internalized and indirect bordering. Our analysis was guided by a conviction that is important to consider the relationship between the central- and local state policy when examining indirect and internalized bordering practices. Much previous research has argued that local-level policy is, at least potentially, more open and inclusionary towards irregularised migrants, not least because local authorities tend to regard irregularised migrants as de facto residents. Our analysis suggests a more complex relationship between national-level discourse and municipal policy and practice. It also highlights the administration of social assistance as an arena where citizenship and residency status are contested and negotiated. Municipalities can adopt policies that extend access to public services and protect irregularised migrants against border enforcement – but they can also mobilize their specific powers to enact ‘minimum rights-‘ and ‘self-deportation policies’.

Malmö became the first Swedish municipality (in 2013) to adopt social policy guidelines that explicitly recognized a right for irregularised migrants to access to social assistance. Following the events of 2015 and amidst calls to align municipal social policy with state migration policy so as to not obstruct the goals of the latter, the guidelines were revised. The 2017 policy guidelines are not entirely exclusionary. There are still openings for irregularised migrants to access social assistance. However, they do signal a shift towards a more restrictive approach. An overall standardization and juridification of the guidelines has meant that there is now less leeway for social workers to interpret the relevant regulations in ways that are favourable for their irregularised clients. At the same time, a number of organizational changes (e.g. the creation of a specialized unit for processing applications from ‘undocumented migrants’) have raised practical barriers to accessing services. Altogether, this has contributed to the creation of a more ‘hostile environment’ for irregularised migrants in Malmö. While our analysis is limited to policy-on-paper and overarching organizational changes, it would be valuable to further investigate how these changes have affected day-to-day social work practice.

We argue that the 2017 revised guidelines can be understood as part and parcel of an overall policy trend that treats destitution as a means to encourage self-deportation. Crucially, the changes in municipal policy and practice that have taken place since 2015 have not required any major political or legal reform. Instead, they have been implemented through a careful re-writing of the guidelines on social assistance, and through shifts in administrative practice. In particular, they have relied on an incremental and continuously ongoing specialization, standardization and juridification of social work practice. In contrast to the UK hostile environment policies, which have been designed and implemented in a top-down manner (from the national- to the local-level), the bordering practices in Malmö have not been explicitly or formally dictated from above, but are nevertheless consistent with a general discursive shift in the migration policy debate. As well, these new bordering practices need to be interpreted as a local response to changes in national-level asylum-policy that have resulted in an increasing reliance among irregularised migrants on local support systems. Crucially, the fact that the changes in question have been enacted at the level of practice means that there is little room for demanding accountability, or for political contestation.

Altogether, our findings highlight the need for researchers within the field of border studies to pay careful attention to sub-national policies and practices. They also point to a number of wider issues of relevance to social policy and social work research. The ‘meeting and meshing’ (Barker 2017) of national-level migration policy and municipal social policy needs to be carefully monitored and scrutinized in terms of how it affects the conditions and rights situation of irregularised migrants and exacerbates social inequalities – but also in terms of its implications for social policy more broadly. For one thing, the strategic withdrawal of public services for migrants increases the burden on the voluntary sector. There is also a risk that the instrumentalisation of social policy for the purposes of border enforcement ends up eroding key principles of the social democratic welfare state by making social services less of a social right and more of a tool of disciplinary coercion and control.

Acknowledgments

This research was supported by the Swedish Research Council for Health, Working Life and Welfare (FORTE), Grant ID: 2018-00458. The authors thank Andrea Iossa, Jacob Lind, Anna Lundberg, Hanna Scott, Emma Söderman, and two anonymous reviewers for their helpful comments.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Forskningsrådet om Hälsa, Arbetsliv och Välfärd [2018-00458].

Notes

1. Unless the discussion requires legal specificity, we use the term ‘irregularised migrants’ as a catch-all term to refer to all who reside in Sweden without (formal) permission, including groups conventionally labelled as undocumented migrants (i.e. rejected asylum-seekers) as well as ‘vulnerable EU citizens’. Our understanding of the term ‘irregularity’ is informed by De Genova’s work on ‘the legal production of migrant illegality’ and deportability (De Genova 2002), which sees ‘irregularity’ (like ‘illegality’) as a product of law and policy.

2. At the time of writing, in the late summer of 2020, the Malmö City Council is processing a proposal for a revised set of guidelines. The sections that concern the rights of irregularised migrants to access social assistance are virtually the same as in the 2017 guidelines.

3. In the years before 2015, the rightlessness of irregularised migrants was partially challenged (Nielsen 2016; Nordling, 2017). In 2013, undocumented migrants were granted a right to health care ‘that cannot be deferred’, and undocumented children were given access to education80,029 Ch. 2§ and 3§. With the renewed emphasis on repatriating irregularised migrants, such rights are now increasingly being called into question. In particular, the rights of vulnerable EU citizens to access both health care and education have been highly contested (Lind & Persdotter,2017; Persdotter,2019).

4. ‘Vulnerable’ EU citizens is an established term in the Swedish context, used by the authorities to refer to citizens of other EU Member States who are staying in Sweden without a right of residence and in situations of extreme poverty and marginality, a majority of whom are Roma from Bulgaria or Romania. We wish to stress that we do not accept the terms as naming objectively existing groups of people but that we have chosen to use the term deployed by government officials and policy makers – as a label.

5. Social assistance is defined in the law as a last resort for individuals ‘if they cannot meet their own needs, or needs could not be met otherwise’ (Ch 4 § 1 SoL) This implies that an individual in need can receive social assistance from the municipality only if no other entity can be deemed responsible.

6. Most other social security schemes (e.g. employment insurance) are administered by the central state.

7. The Parliamentary Ombudsmen (JO) found that the social services are obliged to give information to the Border Police when they are asked about specific individuals, but criticized the municipality for preventing social workers from contacting their clients to inform them about the request from the Border Police (see Lundberg, Gruber, and Righard 2018).

8. This practice is supported by the guidelines on social service entitlements of ‘vulnerable EU citizens’ issued by the National Board of Health and Welfare in 2017 (Socialstyrelsen 2017, 35).

9. Interestingly, this is a sub-group that has previously been constructed as more ‘integrated’ and therefore more ‘deserving’ of support than other more newly arrived and less established groups of irregularised migrants (see Sager 2011).