1 Introduction
In welfare states, there is often a prevailing belief that welfare policies should be implemented and preserved through legislation. This legislation, in turn, should be able to reconcile conflicting institutions and interests (Åström Reference Åström1988). In this article, we are concerned with irregularised migrants – that is, inhabitants without a residency permit or a recognised right of residence – and their access to social assistance (ekonomiskt bistånd). This is a means-tested cash support programme that is administered by the municipal social services and is widely regarded as the ultimate safety net of the Swedish welfare state (Stranz et al. Reference Stranz, Karlsson and Wiklund2017). We present a longitudinal study of how the right to social assistance has changed over time, focusing on the significance of legal status for accessing this right. The study covers a four-decade period, from 1982, when the Social Services Act (SFS 2001, 453, hereinafter the SSA) was adopted by parliament, up until the election of a right-wing government in 2022.
The SSA is intended to provide support to people who have no other means of subsistence. The act does not specify who is eligible for support, but makes a distinction between short-term visits – for example, tourists – and municipal residency. Further, the SSA is framework law, meaning that it gives wide discretionary power to the municipalities. Their interpretations and reinterpretations of the law have far-reaching consequences for people’s opportunities to access support. Decisions by municipalities can in most cases be appealed to administrative courts, among which the highest court delivers judgments that guide practices and local guidelines. The relevant paragraphs of SSA have not changed much since its entry into force in the early 1980s.
A point of departure for our research is that the significance of legal status for access to the ultimate safety net can provide insights into the inclusive and exclusive mechanisms of the welfare state and social law more broadly. Based on extensive empirical research, this article therefore aims to contribute to knowledge about when and how changes in the right to social assistance happen, from the legal position of irregularised migrants. Furthermore, our aim is to explore why the changes have taken place. Methodologically, we combine the depth and detail of qualitative research with the breadth of analysing many instances, allowing for more comprehensive insights while maintaining a qualitative perspective. This implies that although we do not make quantitative claims with the present study, our research may contribute to quantitative research (see Collier Reference Collier2011). Through a qualitative analysis of 372 court judgments, twenty-eight municipal guidelines, three focus groups and six individual interviews, we identify the practices and strategies used by municipalities, courts and social workers that lead to changes, and point to the consequences of these.Footnote 1 We also highlight some methodological reflections from our research that are important for investigating changes to social rights from a historical institutionalist perspective.
All in all, we provide an understanding of the sociopolitical developments that is attentive to the function of law in this development. In a larger perspective, we provide insights on how social rights can be safeguarded, and how inclusionary social citizenship in the welfare state can be understood and reinvented, in a time of increasingly populist criticism of immigration and violent deportations of people who do not have their asylum claims recognised by migration authorities.
The normative starting point of our research is the assumption that we live in a globalised world characterised by the forced and voluntary movement of people across nation-state borders. If such a world is to be a liveable one for all, an important concern is how basic social rights increasingly may be linked to presence (Squire and Darling Reference Squire and Darling2013) rather than legal status. In this context, we are interested in the negotiations that take place around what Linda Bosniak has termed ethical territoriality (Bosniak Reference Bosniak, Fine and Ypi2007). To gain knowledge about the function of legal processes and ‘hereness’ vs. ‘residence status’ as a basis for social rights, we have studied legal practices and strategies concerning irregularised migrants’ access to rights. While we do not seek to answer the question of what a world order guided by ethical territorialism could or should look like, our findings highlight some legal mechanisms that can enable steps in this direction.
The text is organised in six steps. First, we present a literature review to frame our study and the research field to which we want to contribute (Section 2). This is followed by a section on methodological considerations including a presentation of our data and analytical strategy (Section 3). Next come theoretical concepts (Section 4). This is followed by a section where we present contextual information on the rights holders and the social rights that are the focus of our study. Here, we also present an overview of the broader sociopolitical development in terms of rights to welfare services in Sweden (Section 5). The next section constitutes the main empirical and analytical portion of the paper. Here, we trace several continuities and critical turning points that can be identified concerning the right to social assistance for irregularised migrants in Sweden over four decades (Section 6). The concluding discussion (Section 7) draws on the analysis and asks what the wider implications of our findings are for the fundamental idea of a welfare state that ensures public accountability for needs-based social rights.
2 Literature review
The nature of changes in welfare states is a broad and heterogeneous field of research, encompassing many different disciplinary approaches and theoretical perspectives. Research in this field has, for example, focused on the role of the New Right Movement (King and Waldron Reference King and Waldron1988), budget decisions by political leaders (Wilensky Reference Wilensky1974), decommodification as an indicator for changes in social rights (Åmark, Reference Åmark2005), attitudes among the public (Svallfors and Taylor-Gooby Reference Svallfors and Taylor-Gooby2012) and how the beliefs of political actors shape their decisions and the reforms they enact (e.g. Béland and Cox Reference Béland, Cox, Béland and Cox2011). Yet other studies have focused on austerity politics and neoliberalism, including the effects of such political processes on welfare distribution – and on access to social security benefits and social assistance. This latter body of research understands the late 1970s as a time when active, progressive redistribution policies had made Sweden into one of the most equal countries in the world, economically speaking (Bengtsson Reference Bengtsson2020). At this time, there was also widespread support for the belief that poverty was a result of inequality and needed to be dealt with accordingly. Writing in this period, the sociologist Walter Korpi argued that the very need for social assistance arose from the factors that create inequality between people, the same factors that form class-based socio-economic hierarchies. Thus, he argued, if inequality was addressed, poverty would eventually also be erased (see Esping-Andersen and Korpi Reference Esping-Andersen and Korpi1986).
The importance of rights in this process, as well as the gradual nationalisation of social provision and the role of institutions such as the social services, had previously been investigated by TH Marshall in his work on social citizenship (Marshall Reference Marshall, Marshall and Bottomore1950; see also King and Waldron Reference King and Waldron1988). For Marshall, the linkage between citizenship and welfare was a question of equality in society, to counteract existing inequalities and thereby improve people’s life chances. More specifically, Marshall’s idea of ‘social citizenship’ was about extending the scope of the notion of equality through civil, political and social rights (King and Waldron Reference King and Waldron1988). This meant, among other things, supporting the shift away from poor laws to welfare policies, which he believed ought to be characterised by universal welfare provisions to avoid, as far as possible, the stigmatising effect of selective and disciplinary poor law provisions. Several scholars have problematised Marshall’s argument. Desmond S King and Jeremy Waldron (Reference King and Waldron1988) argue that some economic security is a prerequisite for meaningful political citizenship, and John Crowley insightfully explains how Marshall’s treatment of the national scale as self-evident is problematic given that the nation-state is a historically situated construct. Crowley’s argument is particularly pertinent given that Marshall wrote about social citizenship in Britain at a moment of waning imperial power, when Britain was remaking itself as a nation-state (see El-Enany Reference El-Enany2020). Crucially, Marshall’s conception of social citizenship also underplays the significance of the local scale and of local residency (Crowley Reference Crowley1998).
In the Swedish context, the welfare state has gradually been reformed and weakened since its proverbial ‘golden era’ in the post-war period. The development was accelerated in the 1990s in connection with the financial crisis when inequalities began to widen and absolute poverty exploded (Larsson et al. Reference Larsson, Letell and Thörn2012). However, mass unemployment and financial hardship were not, as had been the case in the past, tackled via new investments. Rather, austerity became the central idea in finding solutions to fiscal and monetary policy issues. In conjunction with this shift in policy, the social democratic values that had guided social policy from the 1930s were gradually replaced with what journalist Kenneth Werne (Reference Werne2014) characterises as a newfangled liberal conservatism that portrayed the poor as irresponsible and unwilling to work. The solidarity-based security systems that had been developed over decades now began to be described as the cause of unemployment, mirroring similar discursive and policy shifts in other Western countries. This new policy approach resulted in several significant changes in the regulation and administration of social assistance: a national norm for social support which guides the right to social assistance was introduced in the late 1990s and has since drifted so that it has gradually come to include fewer living expenses. Meanwhile adjustments of this norm have not kept pace with price and wage developments in general (Hjort Reference Hjort, Swärd and Engelmark2012). In this way, social assistance as a social right has been hollowed out since the 1980s (Östberg Reference Östberg, Koivunen, Ojala and Holmén2021). The restructuring of the welfare state and austerity policies have led to more difficult conditions for those social workers providing rights (Jönsson Reference Jönsson2015; Kamali and Jönsson Reference Kamali and Jönsson2018).
Today, Sweden is one of the countries within the OECD where income inequality is increasing most rapidly. Between 2019 and 2021, the number of billionaires in Sweden increased from 206 to 542 (Cervenka Reference Cervenka2022). Based on the EU measure of relative poverty, which includes anyone with an income below 60 per cent of the median income, 12 per cent of the Swedish population was living in relative poverty in 2008. In 2011, the figure had risen to 14 per cent. More foreign-born than domestic-born persons live in relative poverty, and among foreign-born persons, the group born outside Europe is the largest (Therborn Reference Therborn2020). According to statistics from 2019, nearly one in five children in Sweden lives in relative poverty, but foreign-born children are disproportionately affected by poverty: while 16 per cent of children born in Sweden live in poverty, the rate comes up to 54 per cent for foreign-born children (SCB 2021).
Scholars of welfare restructuring have shown how immigrants have been scapegoated as immigration has often been perceived and construed as a reason for the shortcomings and limitations of welfare states in contemporary times (Geddes Reference Geddes2003; Hansen Reference Hansen2021). This frame of reference, which sees immigration as a burden, has served to downplay and shield broader processes contributing to inequality in Sweden. The last few decades have seen the rise of growing welfare nationalism (Barker Reference Barker2017), right-wing populism (Mulinari and Neergaard Reference Mulinari and Neergaard2014), welfare chauvinism (Norocel Reference Norocel2016), and a dramatic shift in the Swedish migration control regime (Lindberg Reference Lindberg2023). Meanwhile, other studies have focused on the effects of austerity policies on migrants (Sainsbury Reference Sainsbury2012; Scarpa and Schierup Reference Scarpa and Schierup2018; Dahlstedt and Neergaard Reference Dahlstedt and Neergaard2019; Kazemi Reference Kazemi2021), and more specifically new intersections between social and migration law (Lundberg and Kjellbom Reference Lundberg and Kjellbom2021; Nordling and Persdotter Reference Nordling and Persdotter2021).
A theoretical concept in these scholarly discussions, which also figures in our research, is the notion of a ‘deportation regime’ (Peutz and De Genova Reference Peutz and De Genova2010). For our purposes, the concept is useful to name those processes that aim to separate those deemed entitled to social assistance from those who should be excluded from such support because of their immigration status.
Andersson and Nilsson (Reference Andersson and Nilsson2011) have shown unexpectedly that social rights of migrants were strengthened during the 1990s, despite economic recession in Sweden. One reason for this, according to the authors, was the focus of the broader debate on asylum-seeking children and their rights. However, later research has shown that children’s rights, in particular the principle of the best interests of the child, have instead become an argument for deportation, when political opinion wind has turned more restrictive in terms of asylum immigration (Holmlund Reference Holmlund2021). In sum, the broader development that is made visible in previous research is that migration regulations directly targeting migrants have been rolled out in conjunction with welfare state cutbacks and market adaptation (Sainsbury Reference Sainsbury2012; Scarpa and Schierup Reference Scarpa and Schierup2018; Kazemi Reference Kazemi2021).
Another discussion in previous research, which we build upon and contribute to with this article, concerns international human rights commitments and the difficulties encountered when attempting to use these commitments to assert the rights of irregularised migrants (Goldblatt Reference Goldblatt, Langford and Young2022; Noll Reference Noll2010). One example is that human rights treaties have been shown to offer weak protection for irregularised migrant workers, even though there are policies that would allow them to take legal action against an employer (Noll Reference Noll2010; Selberg Reference Selberg2016). Similarly, there are significant barriers for irregularised migrants to access health care, another important human right. Since the early 2010s, irregularised migrants have a formal, albeit limited, right to health care in both Sweden (Björngren Cuadra Reference Björngren Cuadra2014) and Norway (Haddeland Reference Haddeland2019). In practice, however, fear of deportation as well as ignorance of the duty of confidentiality of health-care providers present obstacles to access to care. Unclear legislation also creates uncertainty among health-care professionals, which leads to arbitrary practices towards irregularised patients (Björngren Cuadra Reference Björngren Cuadra2014; Haddeland Reference Haddeland2019).
Against this backdrop, the remainder of this article seeks to make an empirical and politically relevant contribution to scholarship on trade-offs and conflicts over resource allocation and political priorities that affect access to social rights (particularly the right to social assistance) in the Swedish welfare state.
3 Methodological considerations
In line with Sven Steinmo’s historical institutionalism (Reference Steinmo2008), we view institutions as rules that structure behaviour and whose content is simultaneously shaped by agents that work in and represent these institutions. Seen from this perspective, institutions are themselves the product of human agency – as humans enact institutions, they transform them correspondingly (Farrell Reference Farrell2018). Thus, ‘institutions are not ahistorical constants’ (Farrell Reference Farrell2018, 26) and history is not a series of independent events but interdependent processes and interactive effects of multiple causal variables that often shape one another.
We follow Béland and Powell’s call for studies on complex patterns of policy development that challenge the ‘blunt binary “paradigmatic” or “incremental” characterisations that permeate much of the literature’ (2016, 136). We situate ourselves in a stream of research that questions ‘the static representation of a “frozen” welfare state landscape’ (Esping-Andersen and Korpi Reference Esping-Andersen and Korpi1986, 24) and exaggerated welfare state resilience (ibid., 132). In the following we first describe our approach to change more closely and then present our selection of empirical data and our analytical process.
3.1 To study change
To identify and understand continuities and critical turning points using several datasets, we needed to have a coherent working definition of change and an idea of how we can study it. Besides a timeframe, which for this study was set by the SSA being in force, we needed analytical tools to identify change. Our analytical approach was to focus on how rights are limited and extended through various independent yet interconnected regulations and their application, investigating national law, court judgments and municipal guidelines. These datasets are different, and they have evolved over time depending on new relations between the institutions (national law vs. municipal guidelines) and through actions and practices within each institution, as well as in relation to unforeseen changes in the outside world, which we try to capture through contextualisation.
Change can happen slowly and incrementally, and depicting change can sometimes be difficult, not least due to the interconnectedness of institutions and the resistance from those benefiting from the status quo (Steinmo Reference Steinmo2008). Therefore, we applied a ‘tracing method’ (Collier Reference Collier2011) to identify instances where dominant regulations were contested, involving a transformation of the legal content that resulted in an expansion or restriction of (formal) access to social rights.
Thus, we have explored legal practices and strategies in their institutional and historical context, highlighting for instance when the legislature adopts contradictory regulations, or when a decision-maker or judge neglects certain social conditions in their assessments or allows different legal areas to interact in new ways. At times, such pervasive processes – consisting of a series of smaller non-systemic changes or soft governance – can alter key functions of a welfare state and lead to important changes (see Altermark Reference Altermark2020). Such changes can, in turn, be contested (Hall and Thelen Reference Hall and Thelen2009), strategic (Mahoney and Thelen Reference Mahoney and Thelen2010) or happen through conversion without much scrutiny (Hacker, Pierson and Thelen Reference Hacker, Pierson and Thelen2015). As Hacker (Reference Hacker2004) asserts, change may come about through ‘drift’ – that is, changes in the operation or effect of policies that occur without significant changes being made to the structure of those policies or the agents. ‘Drift’ thus refers to situations where the change is largely outside the immediate control of the law enforcer, where rules formally remain the same but their impact changes due to external conditions (for a discussion of general drift in the Swedish context, see Palme Reference Palme and Marin2015). In the present study, agents and their actions and strategies are important to study if we are to understand drift. We ask what the judge, the municipal social service committee and the social worker do in their assessments of individuals’ need of social assistance according to the SSA. Further, we do so with an eye to the potential effects and outcomes for access to rights for irregularised migrants and how these might change over time.
Our research is inspired by legal geography in that we apply a spatial perspective to law. Legal geography is useful for conceptualising how the law contributes to the creation of different spatial units (e.g. municipal boundaries), but also how spatial and material conditions affect the law (Persdotter and Iossa Reference Persdotter and Iossa2022). The fact that we regard rights, like the law in general, as something which is constantly being created and recreated through interpretation and negotiation means that our focus is not on what is a correct – or incorrect – interpretation and application of existing law. Instead, we take an empirically oriented approach where the aim is to qualitatively analyse how different possible interpretations and applications of the law are set against each other and negotiated, and what the outcomes of these negotiations are.
To investigate the spatial aspects of changes in terms of different geographical sites, contextualisation has been an important part of our analysis. Contextualisation (Banakar Reference Banakar2011) places legal actions, institutions, traditions and texts in a particular time and socio-legal space, acknowledging that law is embedded in social relations and processes. To contextualise our empirical material and help make sense of broader sociopolitical and policy changes during the long period we wanted to cover, we also read legislative preparatory work and developed a timeline of key events in the fields of migration and social politics, including relevant legislative and policy changes. The timeline was developed during the course of the whole research project to enable an analysis of situations over time, that is, trajectories of change and to a certain extent also causal connections.
3.2 Materials for capturing and analysing change
The data compiled for the present study include court judgments, municipal guidelines and interviews. We interviewed sixteen people: ten people in focus groups and six of whom we met individually. The individuals had all previously been employed as heads of various organisations across the public sector and civil society, working with the provision of social assistance. One of the interviewees was also a former minister for social affairs in the Swedish government. All individuals were aged sixty-seven years or over, the legal retirement age in Sweden. The interviews were semi-structured and aimed to gain an understanding of developments over time, but also to contextualise the court judgments and municipal guidelines from the earlier part of the study period.
The other ten interviewees, whom we met in focus groups of two to five persons, were still employed in social services. In the focus groups we used five vignettes to discuss concrete, fictive cases where irregularised migrants came to ask for help from the municipal social service. The focus group discussions provided key insights about the importance of the local context for access to social assistance.
Municipal guidelines on the SSA were selected through archival searches in three municipalities. There are 290 municipalities in Sweden of varying sizes. We selected three municipalities which vary in size and population composition, and which all have experience of receiving migrants affected by irregularisation processes. As with the interviewees, we pseudonymised the municipalities. The largest municipality we call Harbor City, the medium-sized Rivermills, and the smaller municipality Pinetown.
While the SSA is a national law, the responsibility for administering social assistance is placed at the municipal level. As mentioned already, the SSA is also a framework law that leaves room for interpretation by local decision-makers and individual practitioners in the geographical area where the law is to be implemented – that is, the municipality. In the past two decades, more and more municipalities have adopted local guidelines, decided upon by local politicians, which contain detailed advice and instructions for the interpretation and application of the SSA and other regulations such as tenancy and migration law. We would argue that the municipal guidelines are a form of local rights enactment, and they are central to how social workers apply the SSA in their everyday work. Of importance to this study on law and change is our observation that local guidelines are drawn upon for guidance also in the lower courts, even though they do not have formal status as a source of law in the Swedish legal tradition (SOU 2007:10, 96).Footnote 2
The court judgments adopted from 2011 in lower instances were selected from a digital database at the university’s library, Karnov. The Karnov database contained all Administrative Court judgments from 2013 (förvaltningsrättsavgöranden), a selection of Administrative Court of Appeal judgments from 2011 (kammarrättsavgöranden), and all judgments of the Supreme Administrative Court (Regeringsrätten/Högsta förvaltningsdomstolen). Court judgments that were not available in Karnov were searched for manually through physical visits to archives in five different locations in southern Sweden. In the archive search we decided on a delimitation as follows: three judgments per year were to be identified in the lowest instance, the Administrative Court, two judgments per year in the Administrative Court of Appeal. The decisions were to concern the need for support to cover basic needs, for someone whose legal status was conditioned or questioned (typically because they did not have Swedish citizenship or a secure residency permit/right of residence). The selection principle was to take the first judgment we found that met the criteria, spread out over the year, starting with January, then April, September and November. A qualitative judgment had to be made concerning claimants’ legal status. This was sometimes explicitly stated and sometimes assumed to be the case if the applicant participated in migrant integration programmes. Importantly, as regards individual written decisions by social workers, we did not include these in our sampling process. However, the court judgments, taken on appeal, summarised the reasoning of the local social services committee and sometimes included copies of their original assessment and decision.
Here, it might be noted that it is rare for persons without a residence permit or right of residence to apply for – and to appeal – a decision on social assistance. Our selection strategy (which included only decisions that had been appealed) meant that we did not include decisions on social assistance that benefited the individual applicant. This means that we cannot say much about the prevalence and nature of such decisions although we have reasons to believe that they are rare.
Finding and compiling court material was a time-consuming process. One day of work in the archive could mean flipping through up to forty bound court records, each one about 300–400 pages long. In most cases these records lacked tables of contents or thematic systematisation of the decisions; they had been archived in the order in which the judgments were issued. In most verdicts, the main title, that is the legal issue that the court is assessing, does not indicate whether the person concerned has a contested right of residence, and thus would fall within the scope of this study, which meant that we had to undertake a qualitative assessment of each decision. All in all, we skimmed through tens of thousands of decisions to identify cases where migrants’ access to support from social services was at stake.
Court judgments, besides providing information about the content of the law in practice and its consequences for individuals, offer insights into factual circumstances that are made legally relevant by the courts. It was necessary to contextualise the court judgments as they were brief and often did not endeavour to account for the needs of the individuals and families seeking support. To do so, municipal guidelines and transcriptions from the focus groups and interviews were important. The interview participants reflected retrospectively on welfare-related measures in the context of irregular migration.
The analytical process was inspired by Collier’s process tracing which is a qualitative analytical tool ‘for drawing descriptive and causal inferences from diagnostic pieces of evidence – often understood as part of a temporal sequence of events or phenomena’ (Collier Reference Collier2011, 823). To organise our empirical data, we used the qualitative data analysis software NVivo. We began with a qualitative close reading of the data, starting with the court judgments and municipal guidelines and then, in the next step, interview transcripts. The analysis of the court judgments highlighted both how rights are legally constructed and how the sociocultural environment has changed over time. Focusing on the content, we worked in continuous workshops over four years, interspersed with individual qualitative analyses (Mayring Reference Mayring2021).
4 Theoretical concepts
We are interested in substantive social rights, specifically in terms of entitlement to social assistance from the social services. By this, we mean formal inclusion in the sphere of rights holders by which individuals, at least formally, enjoy rights. Since we understand law as a process – as something that is constantly being performed and shaped through practice and considerations – and since we start from the assumption that there is much room for discretion in our field of study, we occupy a kind of middle ground between formal law and actual access to rights. What we are interested in is the legal prerequisite (as these are shaped through practices and strategies) to claim rights, from the position of irregularity rather than irregularised migrants’ actual experiences of the system – although these are indeed central to understanding access (and something we have touched upon elsewhere; see, for example, Suuronen Reference Suuronen2018).
Our understanding of social rights is inspired by political theorist Hannah Arendt’s reflections on the ‘right to have rights’ which she presents in The Origins of Totalitarianism (Reference Arendt1968). We also follow scholars working with Arendt who highlight her interest in issues regarding equality (e.g. Suuronen Reference Suuronen2018). Arendt specifically discusses the position of stateless people in the aftermath of World War II in Europe, and how loss of membership (citizenship) of a political community implied a loss of entitlement to rights. This was perplexing, she reasoned, since the agreements on ‘universal human rights’ adopted by the United Nations after the war signalled that one has certain rights simply because one is human. However, as Arendt observed, the social reality proved the opposite. A human being who is only a human being (devoid of political membership) effectively loses the very qualities that allow others to see them as a fellow human being: ‘The world found nothing sacred in the abstract nakedness of being human’ (Arendt Reference Arendt1968, 299); a stateless person was now nothing but a human being, stripped of all rights. This perspective also applies to the contemporary situation of irregularised migrants in Sweden. The current government is seeking even to prohibit municipalities from paying social assistance to people who are residing in the country without residency (Tidöavtalet 2022). In this way, formal residency – ‘membership’ in Arendt’s work – is given increased importance for access to social services in present-day Sweden.
Arendt’s thinking and its relation to economic preconditions of citizenship is a different and less well-known part of her work (see Suuronen Reference Suuronen2018). A few scholars following Arendt have challenged the view that Arendt ignores issues concerning social justice. One example is James Ingram’s point that Arendtian perspectives on human rights should focus not only on the issue of whether certain norms, or institutions, are upheld, ‘but on the social and political capacities – the distribution of resources and above all power – that must underlie them’ (Ingram, 2008, 414; also see King and Waldron, Reference King and Waldron1988, 426). Another example is Ville Suuronen’s (Reference Suuronen2018) reinterpretation of Arendt’s critique against the modern (biopolitical) condition, combined with her futuristic worldview. He argues interestingly for basic income, drawing on Arendt’s reflections on the distinction between the private, the social and the political. The fulfilment of basic social rights may thereby, based on Arendt, be understood as an essential condition for political rights to be fulfilled.
Besides Arendt’s reflections on the right to have (social) rights, a central theoretical point of departure for this article is Bosniak’s notion of ‘ethical territoriality’ (Bosniak Reference Bosniak, Fine and Ypi2007). This is the principle that rights should extend to each person who is present – staying – in the geographical area of a political community, by virtue of that presence. As we will show in our analysis, in Sweden sometimes legal status (even the category ‘deportable’, see Lundberg and Spång Reference Lundberg and Spång2017) and sometimes presence have constituted a basis for social rights. Our analysis describes a tug-of-war, between the right to social assistance linked to staying on the one hand (‘hereness’), and residency (formal legal status) on the other, that has been at the centre of changes in social rights in the Swedish context.
We sympathise with the ethical territoriality approach to rights – that the right to rights should be linked to actual stay. This position is also in line with the transnational approach taken by UN human rights committees, which guide organisations, states and local institutions in their protection of social rights. A case in point is the committees monitoring the implementation of the Convention on Children’s Rights (CRC) and, respectively, the International Covenant on Economic, Social and Cultural Rights (ICESCR). The latter has explicitly stated that the ICESCR applies to everyone, including non-nationals such as refugees, asylum seekers and stateless persons, regardless of legal status and documentation (CESCR 2009, para 30; cf. CESCR 2008, para 37 which talks about the right to emergency health care; also see Langford Reference Langford2007).
In the next section, we zoom in on the Swedish context.
5 Social rights legislation over time in the Swedish context
In the present study, we start from the normative position that people who reside in the country ought to have a right to be protected by the ultimate safety net, irrespective of their immigration status. The more specific definition of the right to this safety net is a matter of interpretation and judicial negotiations. Besides individual discretion, rights are shaped by institutions and broader social processes.
Irregularised migrants applying for support from the social services live on the margins of society, often partly or wholly because of their residence status being disputed or denied. For many, such a situation also implies a constant fear of being detected and expelled from the country (see, for example, Söderman Reference Söderman2019; Lind Reference Lind2020).
Irregularised migrants staying in Sweden are a diverse group. They include individuals who are commonly referred to as undocumented – that is, people who formally lack a residence permit – but also EU citizens with unclear or questionable right of residence, and in some cases also asylum seekers and individuals who have recently received a residence permit (so-called ‘new arrivals’). The latter group was included in the study from an early stage as we observed that their status as ‘newcomers’ was used by legal practitioners to limit their access to social assistance – even though they had formally obtained a right to live in the country and access public services on the same terms as full citizens. The right to social assistance is regulated mainly in chapter 4 of the SSA, which states the following:
‘Anyone who is unable to meet their own needs or cannot get them met in any other way is entitled to support from the [municipal] Social Welfare Board for their livelihood […] The individual shall be ensured a reasonable standard of living through this support. The support shall be designed so that it strengthens the person’s ability to live an independent life’ (ch 4, Sec 1, the SSA, our emphasis).
The concept of stay or presence (vistelsebegreppet) has been central for regulating access to rights in Sweden. In the government’s legislative preparatory work – which may be interpreted by the courts and treated as a source of law – it is stated:
‘[T]he municipality where the person stays must provide the individual with support and social assistance. It is thus the municipality in which the individual is located when the need for support arises that is responsible for ensuring that the individual receives the help they need. However, this should not mean that the municipality of stay should take all the measures that may be necessary even when the individual is [a formal] resident in another municipality’ (proposition 1979/80:1, 42–43, our emphasis).
The above quotes illustrate what we have asserted above, namely that the right to support from social services plays out in a tension between the rights of stayers and formal residents (Bosniak Reference Bosniak, Fine and Ypi2007). Who is covered by the law is not made clear in the SSA. As we will show in our analysis, sometimes actual residence and sometimes legal status (also deportability, see Lundberg and Spång Reference Lundberg and Spång2017) have constituted a basis for being counted as someone with rights. As we will see, differences are also drawn between access to so-called emergency support and access to long-term support.
Something important to understand is that the right to social assistance, according to law on paper, is valid for all people residing in Sweden who are unable to provide for themselves and their families in any other way. This idea of a general needs-based core right has been a social democratic scheme ever since the welfare state was in its infancy at the beginning of the twentieth century. Between the 1930s and 1970s, when a strong social democratic state developed, the collective, in the sense of the public sector, took increasing responsibility for providing an ultimate safety net (Åmark, Reference Åmark2005). At the same time, social assistance as an institution has always contained traces of a poor law logic that distinguish between the ‘deserving poor’, considered truly in need and worthy recipients of support, and the ‘undeserving poor’ (Davidsson Reference Davidsson2015; see also Himmelfarb Reference Himmelfarb1984; Shilliam Reference Shilliam2018).
In summary, membership (being included, or belonging) in the sense of a legal right to reside on the one hand, and actual stay on the other, have had greater or lesser significance for access to rights in the Swedish context at different times and in different contexts. Membership can thus be regarded not as a legal status but as a site where inclusion and exclusion are negotiated. As King and Waldron (Reference King and Waldron1988) argue:
‘[T]o be a member is not something fixed in stone for all time, but is expandable, as new ways develop in which the state can offer various benefits that accrue from collective organization, and in which members can participate in those arrangements to make benefits available to each other’ (p. 432).
Adding to this, King and Waldron further argue that the expectation, the normative system if you will, that the welfare state will provide at least some rudimentary protection is central to the concept of basic welfare support. This open approach to the concept of membership has inspired our research.
6 What continuities and critical turning points can be identified in social rights for irregularised migrants in Sweden?
In the following, we present the changes that we have identified in our analysis of the empirical data. We discuss concrete changes in the sense of specific, observable differences before and after a given event (see Zheng Reference Zheng2022). Drawing on a range of materials, we also analyse how the different actors shaped the legal content of the right to social assistance through their practices and attempt to contextualise these practices. In what follows, we present five main conclusions concerning how changes in rights come about and may be understood, specifically in the context of a world characterised by (irregular) movement across borders and differentiated social rights. As noted, the findings may also inform broader reflections of how rights are shaped through legal institutions, as well as reflections on the potential for an inclusionary welfare state beyond the national order of things (Malkki Reference Malkki1995).
6.1 Continuities
6.1.1 Local inconsistencies and struggles
We begin with two main observations of relevance for change that apply to the entire four-decade period and that highlight concerns about how temporary and tenuous the rights of irregularised groups are. First, in most of the court cases we analysed, irregularised migrants were not granted a right to social assistance, although there are important exceptions. Second, we see a growing overlap between social and migration law; it is noticeable that questions concerning the social rights of irregularised migrants become more visible over time as the language used by municipalities in the social services guidelines and in decisions on social assistance become more linked to migration legislation. For example, we see a shift from a language of ‘refugees’ and ‘people waiting for residency’ or ‘who are in a process of establishing themselves’, to a language that mirrors the categories of migration law including, for instance, the category of ‘people who do not have authorization to stay in Sweden’ (Harbor City’s guidelines, 2017; Pinetown’s guidelines, 2013, 2017).
The temporary, uncertain and fragile nature of social rights is reflected in local inconsistencies and struggles with regard to who is a rights holder and on what basis (i.e. which principle of inclusion/exclusion should apply). Irregularised migrants’ right to social assistance tends to be determined by municipal legal practices rather than through national law. We also found that multiple actors were involved in negotiations over who can ultimately acquire the right to social assistance, and thus in shaping access to rights.
In the 1980s, when the SSA had been adopted, the issue of legal status was, as far as we can see, invisible in national law. We do not see any traces of the deportation regime in the legislation during this period. Neither in the SSA nor in the legislative preparatory work is there any wording that explicitly limits or excludes any category of rights holders from the right to social assistance. It should also be noted that until specific legislation on economic support to asylum seekers came into force in the second half of the 1980s, asylum seekers were a direct focus of social services. The principle was that everyone who lacked any other means of subsistence was included under the SSA.
Simultaneously, the right to support for people whose right to residence is insecure has continuously been questioned in legal practice. An early example is a 1986 court judgment where a Polish citizen was refused social assistance because he was not considered to have right of residence. The man had applied for support to pay a debt he had incurred during a period when he was in hiding after being refused asylum. He had borrowed money for his subsistence during this period. The court rejected the man’s claim on the grounds that the loan was taken during a period when, since he had no right to reside in Sweden, he did not have a right to support under the SSA (Administrative Court 1986).
Local inconsistencies regarding the relevant principle of inclusion or exclusion are evident throughout the period under study. An example was found in our interview with a former head of social services in one of the municipalities, Ellie. Ellie, who would later leave her position as manager to work in a civil society organisation because she could not stand the economic cutbacks, reflected on a series of measures taken in the 1990s in connection with the implementation of a more restrictive national refugee policy. The restrictions were linked to the arrival of war refugees from former Yugoslavia and meant that people who had sought protection in Sweden had their status irregularised when their applications for asylum were rejected. Ellie remembers that the children from these migrant families attended school and that the families were part of the local community. The national legal changes, however, fuelled a more restrictive approach locally.
Ellie explained during the interview that the heads of social services in Harbor City wanted to protect the families’ access to social rights such as social assistance, schooling and health care. To this end, they initiated a dialogue with the municipal government. Besides politicians, civil society organisations were also involved in the discussions. Those in favour of inclusionary local policies – that is, the heads of social services – argued that the rights they sought to protect were interdependent. ‘The whole spectrum of life needs to be provided for [both supports according to the SSA and access to health care],’ Ellie explained. Children’s rights under the CRC (1989), which Sweden had recently ratified, was another central argument in the local negotiations. Yet, the municipal political leadership was not prepared to take a stand in favour of inclusion:
‘They [the national legislature] had [in the 1990s] closed the borders and would turn people away. And then you started to see these people staying behind and being hidden. And I know this because we [the heads of social services] had an animated discussion with [among others] [Civil Society Organisation] where we discussed that we wanted the children to be allowed to stay, that the children would be allowed to keep their kindergarten places and stay in school even if their family was threatened with deportation. But it wasn’t obvious [that the families should have access to rights]. And then I remember that we met with the Municipal Chief Executive, and they [the administrative and political leadership in the municipality] left it up to each teacher to decide, and they had said all along that it was the teachers who decided. But we didn’t want it that way. The city must decide what the city thinks. We cannot have differences locally in what we think. But they never really listened to us on this.’
The quote reveals an ambiguity around who should be subject of rights, and thus points to how temporary and fragile social rights are when those concerned have a contested legal status, even if the national law did not explicitly exclude irregularised migrants at the time (and, at the time of writing, still does not). The above example also captures how a variety of actors were involved in the negotiations over rights, with one exception: the rights holders themselves. The children and their families, whose rights were at stake, had no access to these dialogues.
6.1.2 (The subordination of) international children’s rights obligations
Children’s rights are given limited space in the judgments in our empirical material on social assistance to families with irregular status. In total there are 108 decisions (of a total of 372) that concern children or families with children.Footnote 3 Yet children’s needs, and claims made by children (or on their behalf), are consistently absent in the material (with a few exceptions, one of which we discuss below). This finding on children’s rights relates primarily not to immigration status, but rather to the economically vulnerable situation of the affected families and, possibly, to children’s limited legal capacity to act.
The local guidelines issued since 1990, when the CRC came into force in Sweden, all refer to the CRC and/or the principle of the best interests of the child. We expected that this should also be reflected in the legal practices of the courts, but this is not the case. Typically, the court judgments concerning children state that the claimants are a family with minors, but in the balance of social realities and legal interpretations, the demand to seek work comes through as central to get support (Administrative Court 1993b, 1995b, 1996, 1998a). Also, in cases where the children’s needs had been explicitly declared as the reason for a parent’s inability to attend labour market activities, these needs were not considered in the legal assessment (Administrative Court 1995b). While there are exceptions (see 6.2.2), the responsibility of parents to provide for their children and the increasing emphasis on the importance of work and employment (arbetslinjen) eclipses the right of children in families to receive social assistance. In one decision, a father had informed the social services that he needed to stay at home to take care of his children who had bronchitis and to support their mother who was also ill. The man stated that he was unable to attend the classes he was expected to participate in. The absence led to the withdrawal of social assistance by the social service committee. The man appealed, arguing that due to the needs of the family he had not been able to attend. The court’s response was succinct, merely stating that his explanation for his absence was deemed ‘unacceptable’. Consequently, the court affirmed the social services’ decision to deny support, yet nothing can be read in the court judgment about the consequences for the children (Administrative Court 1993b). Another case, which illustrates the requirement to be always ready to work, concerned a mother who had been unable to attend the required training because she was a single parent with two children. In this case, the court did not take notice of these circumstances. The decision states that since she had been absent, it was also correct for the social services to refuse economic support (Administrative Court 1995b).
Paradoxically, children’s rights became even more absent in court judgments towards the end of the 1990s, after the CRC had been incorporated into the SSA (see prop 1996/97:124). The legislative reform meant that, from then on, the best interests of the child should be considered (para 1, SSA) and children should have the right to be heard and have their wishes considered in relation to their age and maturity (para 6, SSA). A possible explanation for why children’s rights were downplayed, at the time, is that another change to the SSA was also introduced, namely a sharpening of the requirement for adult applicants to actively seek work and be available to take up regular employment in order to have access to social assistance. In a new paragraph added to the SSA, it was stated that when assessing applications, the social services committee may request that a person receiving income support for a certain period participate in skills-enhancing activity assigned by the committee if the individual had not been able to find suitable job training or another activation programme (para 6c). The paragraph (which is still included in the SSA) went on to specify that if the applicant refused or failed to participate in such activity without good reason, continued social assistance may be refused or reduced (para 6d).
Consistent with previous research, we thus observed a trend throughout the entire period, particularly since the late 1990s, where children’s rights to be heard in legal proceedings and to have their interests considered have been secondary to the obligation of parents to seek employment or participate in training activities as a prerequisite for economic support (also see Leviner and Holappa Reference Leviner, Holappa, Adami, Kaldal and Aspán2023).
Simultaneously, while the courts’ legal application appears to have disadvantaged children who live in economic destitution, families living transnational lives – that is, families who might have residency in Sweden, and thus are not in a position where their right to reside is under question, but who have family members in other countries – also had their needs neglected. One indicator of this is that the requirement to be available to the labour market in order to be eligible for support makes it impossible to travel abroad to attend a relative’s funeral without forfeiting one’s eligibility to receive social assistance. Our interviewees confirmed this tendency, explaining that a few days of absence is generally considered acceptable. However, when situations require a longer absence, social assistance is typically withdrawn, regardless of whether children are affected in the families.
Our findings in relation to children confirm what previous studies have found on the weakness of international human rights treaties and that children’s voices are not heard, especially in court settings (Atria and Salgado Reference Atria, Salgado, Christodoulidis, Dukes and Goldoni2019; Noll Reference Noll2010). Our findings also indicate how contradictory regulations may be adopted simultaneously. Regardless of whether one is a Swedish citizen or newly arrived in the country, there are strong requirements to seek work. These requirements frequently eclipse other concerns, not least the concern for upholding children’s rights.
6.1.3 Considerable inconsistencies in legal sources
The third continuity that we have identified through our analysis of court judgments concerns the arbitrary legal application of the SSA in the courts. Specifically, there is considerable inconsistency in reference to legal sources. In fact, it is not uncommon for references to legal sources to be completely absent. Lower courts take precedence over higher courts, older decisions take precedence over newer ones, and different social circumstances are given widely differing importance for the outcome. We identified a great variety in the legal justifications from case to case, which is noteworthy considering the central idea in the Swedish legal system (see Sandgren 2006) of an unbiased attitude and loyalty to the legislator, predictability and consistency in the legal practice, and the use of legal sources in a specific order that is separate from social issues and politics.
These findings confirm that the content of social rights is in practice an ad hoc issue. It also means that there is always room for extensive interpretation of who should be included in the social safety net of the welfare state.
6.2 Critical turning points in social rights due to legal practices and strategies
In the previous section we outlined three consistencies in our empirical material: First, local inconsistencies in the principles of inclusion and exclusion highlight the fragile nature of social rights for those with contested legal status, and rights holders are often excluded from dialogues on policies relevant to their situation. Second, children’s needs and claims are notably absent in court decisions, reflecting the challenges faced by economically vulnerable families and children’s limited legal capacity. Third, court judgments reveal arbitrary applications of the SSA, with significant inconsistencies in referencing legal sources. We turn now to discuss three significant turning points: moments when the interpretation and application of the SSA changed considerably.
6.2.1 Limbo as a basis for social rights
One of our findings regarding changes in access to social assistance concerns individuals who had either had their asylum claims rejected or had their residency permits revoked for other reasons but were unable to travel to the deportation country. In law, the term used to describe such a position is ‘practical impediments to enforcement’. In everyday language, this situation is usually referred to as ‘limbo’ (Lundberg Reference Lundberg2020). A common reason for why a person might end up in ‘limbo’ is if the deportation country does not recognise them as a citizen (making them effectively stateless). Other circumstances that entail practical obstacles to enforcement are difficulties in establishing the person’s identity. There are also cases involving people who come from countries that will not accept them if they do not want to return voluntarily, Afghanistan being one example of such a country. Notably, the Aliens Act (SFS 2005:716) states that a residence permit is granted if there is ‘reason to assume’ that the home country does not want to receive the person (ch 12, Sec 18). In practice, however, it is immensely difficult to attain a residence permit based on this paragraph (Lundberg Reference Lundberg2020).
In three judgments from our material, there is information stating that the person – or family – in question cannot leave Sweden because they are in a limbo situation. In the courts’ legal reasoning, important distinctions are made between those who themselves claim that they cannot return, for example because of fear of violence and persecution; those who refuse to co-operate in the enforcement of their expulsion decision; and those for whom there are circumstances beyond their control hindering deportation.
The first judgment in our dataset concerning a person in limbo is from the Administrative Court of Appeal in a Swedish town in 1998 (Administrative Court of Appeal 1998). The appellant was a man who had applied for social assistance after the immigration authorities had denied him support under the Reception of Asylum Seekers and Others Act (SFS 1994:137, hereafter LMA) on the grounds that he was no longer categorised as an asylum seeker because he had been notified of a deportation order in connection with a criminal act. The border police had also, on one occasion, tried to enforce the deportation, but failed to do so. The deportation order had therefore been postponed after the man had applied to the government for mercy. In the meantime, the man was staying with his family in a Swedish municipality. The local social services committee had granted him limited social assistance for a period, but then rejected his application for continuous support.
While waiting for the man’s appeal to be processed, the social service committee granted him limited support again. The Administrative Court of Appeal therefore rejected the board’s appeal on procedural grounds, stating that there were no legal conflicts to be solved as the board had eventually granted the applicant social assistance. The Court of Appeal concluded that the lower court’s judgment did not go against the social service committee and therefore could not be appealed by the board. Even though the Administrative Court of Appeal in this case did not deal with the question of whether the man was entitled to support, the social service committee’s actions and the court judgment show that limbo, which is an external circumstance, became an important for how the right to support for persons without a residence permit is determined in legal practices among both courts and municipalities.
Individuals who cannot return to their country of origin due to ‘practical impediments to enforcement’ are still today entitled to social assistance. This was confirmed by the social workers who participated in our focus groups. They explained that there is an assumption in both the courts and the municipalities that people who cannot be deported despite their co-operation with the migration authorities are entitled to support, while those who ‘hide’ from, and do not co-operate with, the authorities are not. The fact that there are court judgments that confirm this right to support allows social workers to include irregularised inhabitants as rights holders and provide social assistance in the municipality despite the deportation order. They explained that any other option would be unreasonable: ‘There will be no life for these people …’ (individual interview with Gunnel, Pinetown); ‘They are our ultimate responsibility (yttersta ansvar)’ (focus group, Harbor City). In the interviews, we also discussed whether social workers have any contact with the migration authorities in these situations, to check if the person is in limbo. They explained that they do have contact with the border police, as the latter can confirm limbo. This is done with the consent of the individual concerned. Furthermore, contact may also be made with the Migration Board to see if a new decision has been made in connection with the case, such as a temporary residence permit (focus group, Harbor City).
To conclude, limbo cases demonstrate that the right to social rights is shaped through court judgments which may safeguard access to social rights in municipal decision-making and remain as an invisible form of guidance also in times of cutbacks. More specifically, it is the courts’ interpretations of social circumstances – in the above-mentioned case the fact that the man could not be deported – that are relevant for changes in social rights in Sweden. The case exemplifies how the courts’ interpretations of a person’s social circumstances become relevant for their access to social rights. As such, it also highlights the role of the courts in driving social change through their case law.
An underlying argument for the right to social rights in the context of the limbo problem required, through the above verdicts, a legal recognition of the impossibility of leaving Sweden. Second, the verdicts are also an affirmation of the idea of the ultimate safety net: no one, including those with a deportation order, should be completely deprived of basic welfare. However, to be a rights holder in these cases presupposes that it can be proven that deportation is practically impossible. If the plan presented by the current government is to prohibit municipalities from paying social assistance to people who are residing in the country without residency (Tidöavtalet 2022), this group will be pushed out from this ultimate safety net.
6.2.2 Extended and reduced rights in municipal guidelines after the 2016 turnaround in Swedish asylum policy
Our next finding is related to the broader context of changes in national migration policy and their effect on social rights: it concerns the relationship between international treaties and local guidelines in one of our municipalities, Harbor City, as evidenced on two occasions.
The first local change was related to a nationwide push for extended access to welfare for migrants in the 1990s, 2000s and early 2010s that engaged professional groups (particularly health-care workers, see Nielsen 2016), civil society organisations and the then UN rapporteur for the right to health, Paul Hunt, who explicitly criticised Sweden for discrimination against migrants. As different actors – Hunt, municipalities, civil society organisations – repeatedly referred to the CRC, the regulations about social rights for irregularised migrants would gradually be extended. Civil society organisations, local social services and left-wing and green politicians joined forces in this struggle for the extension of rights.
In Harbor City’s guidelines, it eventually became an explicit position of the municipality that undocumented migrant children should be treated in the same way as permanent residents. Harbor City’s policy from 2013 states that when the needs of irregularised children for support were at stake, they should be treated ‘in all respects’ (municipal guidelines, 2013) as if they were residents of the city. This meant support on the same conditions as other children living in the municipality. It was explicitly ‘recommended that the principle of children’s best interests be applied in this way so that all children living permanently in the municipality would be given a reasonable standard of living’ (municipal guidelines, 2013). If a family with children applied to social services for social assistance, this meant that the children in the family were granted support at the same level as other families (what, in the Swedish context, is called ‘the full norm’), and that the whole family should be granted support to cover rent. An essential constraint, however, was that these children’s parents would not have access to the full norm in the same way as other adults living in long-term economic deprivation in the city. Here, the guidelines confirm what we noted in the introduction: that all social rights under the welfare state have always, in some way, excluded irregular migrants. Yet the Harbor City guidelines from 2013 demonstrate an inclusionary strategy that centred on ‘hereness’ as a basis for access to the ultimate safety net, at least for children. It was, if not radical, then at least a step towards social rights inspired by thinking beyond legal status as the only basis.
In 2017, Harbor City’s policy was revised. From now on, the guidelines state, the best-interests principle ‘may mean that [migrant] children are granted support while their parents are only granted emergency support’. Support to children, according to the new guidelines, ‘can be granted up to the national norm […] support can also be granted for reasonable housing costs based on the child’s needs’ (municipal guidelines, Harbor City 2017).
As we can see, this is clearly a weaker support. In previous research (Nordling and Persdotter Reference Nordling and Persdotter2021), we have shown that this change was linked to the turnaround in migration policy in November 2015, which was also behind a decision by the Supreme Administrative Court that we will discuss below, more specifically regarding who would be entitled to social assistance and why. Besides this change in Harbor City’s guidelines, we also know from our focus groups with social workers that some organisational changes were implemented: a centralised unit was created, tasked with assessing all applications from irregularised individuals and families, and it became common to grant support on a weekly rather than a monthly basis. This streamlining, in effect, made it more difficult for irregularised applicants to obtain social assistance.
The new guidelines in Harbor City, combined with new ways of handling applications, led to a curtailment of rights at a time when more people were becoming irregularised as national refugee policy took a sharply restrictive turn (Lindberg Reference Lindberg2023).
In summary, then, the right to social rights expanded over time from the late 1990s to the middle of the 2010s. For this expansion to take place, local initiatives in favour of irregular migrants were central. Local initiatives by or on behalf of irregularised migrants evolved into a broad nationwide mobilisation campaign in which international law was one important tool. Sweden would prove to be late, compared to other countries, in adopting the strategy to restrain asylum immigration by curbing asylum seekers and undocumented immigrants’ right to work and social benefits (Andersson and Nilsson Reference Andersson and Nilsson2011). While migrant children’s rights were still strong, at least on paper, after 2015, rights were quickly withdrawn from adults. This also affected children, considering the responsibility parents are pushed to take without regard to their capacities (see 6.1.2 above on parent responsibility). We understand the changes adopted in Sweden to be directly linked to the closed borders in November 2015 and a debate that was entirely about protecting Sweden’s borders and welfare system from asylum migration (Strömbäck et al. Reference Strömbäck, Andersson and Nedlund2017). As we finish writing this article, in the autumn of 2024, several further steps have been taken in this direction.
6.2.3 Reinterpretation of the linkage between the Social Services Act and migration law
Our third finding on critical turning points concerns a ruling by the Supreme Administrative Court in 2017 (HFD 2017 ref 33). The court made a reinterpretation of the linkage between the law regulating asylum seekers’ rights to state support (LMA) and the SSA. This served to destabilise the responsibility of local social services to provide social assistance for anyone present in the municipality, and left many people in a destitute socio-economic situation (see Lundberg and Kjellbom Reference Lundberg and Kjellbom2021).
While the decision concerned social assistance, the background was an amendment in another legal area. More specifically, a reform had been made to the LMA that was intended to incentivise irregularised migrants in Sweden to ‘voluntarily’ leave the country. Previously, a limited economic support to asylum seekers who had received a rejection had been disbursed until the person left state territory (subject to certain conditions). From now on, support and accommodation would be withdrawn in the event of an asylum rejection, children and their families excluded. By the end of October 2017, 25,722 people had had their daily allowance withdrawn (Migration Agency 2019).
In the government’s legislative preparatory work for the amendment to the LMA which provides a central guide to the application of the law in courts, an explicit target group was identified as people who ‘do not intend to respect a rejection or expulsion decision’. According to the government proposal, this group was not supposed to have been ‘included in the target group of the bill [LMA]’ (prop 2015/16:146, 11). In the same legislative proposal, it was also noted that restrictions to LMA might increase the pressure on municipal social assistance. As stated in the proposal: ‘The Social Services Act also provides a right to social assistance that may be brought to the attention of the person who is not authorized to be in the country’ (prop 2015/16:146, 14). On this issue – the potential that rejected asylum seekers who had been excluded from LMA support would get access to assistance – the government called for legal guidance from the Supreme Administrative Court because such guidance was said to be lacking.
An appeal brought to the Supreme Administrative Court (SAC) heeded the government’s call and issued its ruling exactly one year after the LMA amendment came into force. The ruling concerned two children and their mother who were living in Sweden with an irregularised status. The court states explicitly that there is ‘no obligation for the municipality to provide social assistance under chapter 4, Sections 1 and 2 of the SSA, even to avoid an acute emergency, to a person whose asylum application has been rejected and who avoids deportation’ (Supreme Administrative Court 2017). The lower administrative court, which ruled in favour of the family’s claims, had previously argued that support should be provided because the family was in fact excluded from support under the LMA. The reinterpretation by the highest court of how the right to social assistance should be weighed against the interests to curb asylum immigration led to a significant narrowing of the sphere of rights holders, and the ultimate safety net of the welfare state was tied to legal residence and not presence, in Bosniak’s terms.
Participants in our focus group in Harbor City referred to the court judgment as a paradigm shift. They described that since the summer of 2017 when the verdict was made, the city had applied a much more restrictive interpretation of the right to support. Before the 2017 judgment, in a normal week they would have about 200–250 cases concerning irregularised applicants (most of whom were only considered eligible for food allowances), but after the judgment this dwindled to only about four cases each week. The only people who receive support today, the social workers explained, are those who ‘fall between the cracks’, or who ‘are in legal limbo’ (see above under 6.2.1). During the period when the municipality had a relatively high number of applications from irregularised migrant families, in 2013–17, the work of the social services was also decentralised, with one responsible board in each district. The various neighbourhoods came to function as a form of separate ‘mini municipalities’, which created differences in how assessments of needs were conducted. Nowadays there is instead a common administration for the entire municipality, which the focus group participants believe has made the assessments more similar in the different neighbourhoods, but also more restrictive. A further change is that payments to those in need are today made on a weekly basis, as opposed to monthly which was the previous practice. This means that irregularised persons in need of support now need to apply on a weekly basis. This is a time-consuming process which likely has deterred some from seeking support.
To summarise, we see from the above example that one single ruling from the Supreme Administrative Court can have a large impact on access to social rights for large numbers of inhabitants, almost overnight. We argue that the court ruled in response to the government’s call in the legislative preparatory work in a different legal area. The court, one might expect, was aware of the previous limitations of the LMA and could foresee that the right to social rights – also the children’s rights – would be significantly curtailed when it ruled that it is legally acceptable to withhold even the ultimate safety net against becoming completely destitute if it can be assumed that the applicant is avoiding deportation. The court ruling also exemplifies how the question of why people go into hiding is seldom asked – or even acknowledged – in a legal context.
In addition to the government, which had an explicit strategy to restrict access to social rights to curb asylum migration and get more rejected asylum seekers to leave the country, and the court, the municipalities and social workers were also involved in changing social rights. In Harbor City, reorganisations and changes in routines took place in the wake of the judgment. This points to a concession to the tougher asylum policy that precipitated an erosion of social rights in Sweden.
7 Concluding discussion
Our aim with the present article has been to make an empirically grounded contribution to knowledge about when and how the rights of irregularised migrants to social assistance have changed over time. Making use of process tracing (Collier Reference Collier2011), we have described some of the processes that have affected migrants’ access to social rights in Sweden, from 1982, when the SSA was adopted, up to the 2022 general election in Sweden. To trace these processes, we found it necessary to conduct qualitative content analyses of guidelines used in municipalities, combined with an analysis of a large number of court judgments, and interviews and focus groups. Our combined methodology, we believe, allows for comprehensive insights while maintaining a qualitative perspective.
Our conclusions are:
-
1. Since the 1990s, local changes to social rights have first happened in an inclusive and then in an exclusionary direction. The 2015 ‘refugee crisis’ and the subsequent government response marks a crucial turning point in this context. The example of Harbor City illustrates how various exclusionary practices (reorganising local social services and adopting new guidelines linked to a harsher national debate) can have a decisive impact on the right to social rights.
-
2. Social rights in Sweden have, to larger extent than has been known, been negotiated and changed in local sites involving several different actors. During the first few decades of SSA, there was a strong belief and national and local political ambitions to see that socio-economic inequalities could be eliminated through welfare. Yet at the same time, the national scale of social citizenship and welfare was mostly taken for granted; legal status was not even mentioned in the SSA or in its preparatory works. Presence or stay (vistelse) in the country was, and still is, fundamental for the right to have social rights. However, in practice, it soon became evident, as our study shows, that the legal protection of irregularised migrants’ rights was weak. In the interview with Ellie, a former head of social services in Harbor City, a micro-level drift of responsibility from the municipal leadership can be noted, while the heads of social services had asked for a municipality-wide policy.
-
3. Regarding children’s interests and needs, we have shown that these have been subordinated to economic cutbacks and activation requirements. While immigration status has been central to restrictions on formal access to social assistance under the SSA, even more important for restrictions seems to be the view that parents are to take responsibility for supporting their children through wage labour. This principle cuts across distinctions based on legal status. The ways in which poor children’s needs and interests became first and foremost a parenting responsibility (rather than a public concern) was particularly evident in the late 1990s, when two contradictory changes to the SSA were implemented simultaneously: the first one concerning the best interest of the child and the second one a tougher workfare-like activation policy. Our findings show that the latter requirement almost always takes precedence over the first. Indeed, legal practices carried out in the courts in the 1990s show clearly that children’s right to social rights was limited. From our study, we argue that the CRC is a weak tool for protecting children’s interests in practice (also see Noll Reference Noll2010). Moreover, we demonstrate the temporariness of children’s rights in the municipalities. Our findings in the field of children’s rights also point to the necessity of analysing contemporary legislative changes in several different areas and sites where legal practices play out, and combining different types of material, to understand change.
-
4. The courts’ application of the SSA is unpredictable. There is considerable inconsistency in their references to legal sources, and it is not uncommon that such references are scarce or even completely absent in the judgments. Very similar social circumstances are given completely differing importance for the outcome in the legal cases. Also, the justifications for decisions, legally speaking, vary from case to case. This result is unexpected, because legal certainty and a strict hierarchy of sources of law is still today an important reference point in the Swedish legal context.
-
5. For irregularised migrants, the (in)ability to comply with a deportation order has been an important circumstance for access to social rights. As we discuss in the first half of the paper, there is a salient idea that no one should be left completely outside the ultimate safety net of the Swedish welfare state (see King and Waldron Reference King and Waldron1988). In practice, however, the rigorous evidentiary requirements for proving that one is stuck in legal limbo, a more accurate description than hereness as a basis for rights, are practical attempts by the individual asylum seeker themselves to leave as such basis.
-
6. Finally, we have highlighted the politicised power of the Supreme Administrative Court. Our case description from 2017 revealed that the local conditions for irregularised migrants’ rights were affected in several ways as a direct result of the court’s decision. We argue that the court was drawn into the increasingly ardent deportation regime post-2015, as it followed the government’s statements in quickly prepared legislative preparatory works about the need for legal guidance and overlooked the formulation in the proposition that the SSA safeguards the right to a reasonable standard of living. Here, our study complements others such as Barker’s (Reference Barker2017), which seeks explanations in the internal logic of the welfare state as the basis for the rapid changes we have seen in Swedish asylum politics. However, rather than emphasising an internal and inherent logic, we highlight instead how these logics play out and the importance of single court judgments as well as local decisions at the municipal level.
What are the wider implications of these changes for the idea of a welfare state that ensures public accountability for needs-based rights and an ultimate safety net?
First, irregularised migrants have always been excluded from the right to have rights. This was observed by Arendt in her time, and our results confirm that this has been, and still is, the case in Sweden. The irregularised subject has, for obvious reasons, mostly been affected by non-rights. This is the case even though international treaties oblige Sweden to uphold a welfare system that safeguards some support for all residents. Having said this, we propose that legal status was and is not a legitimate basis for exclusion from rights according to the SSA. Instead, presence, in line with the principle of ethical territoriality, constitutes a basis to respond to social rights claims.
In many respects, the exclusions that affect irregularised migrants and migrants with precarious status also affect regularised citizens who are made to rely on social assistance as the ‘ultimate safety net’. Our findings show that access to rights is conditioned and restricted by various requirements related to the obligation to seek work.
Finally, our findings raise questions about whether it is always in the interests of vulnerable groups to make their situation visible and claim formal rights, or whether it can sometimes be advantageous to handle things informally through local (and largely invisible) legal practices. Having witnessed how a discourse of inclusion was replaced by one of suspicion, vilification and exclusion over the course of just a few years following the 2015 long summer of migration, we are weary of the risks of relying on visibility as a strategy to claim recognition and rights. How then can we, who have citizenship, professional status and access to elected representatives, act for a more inclusive policy? As citizens we may see social inclusion as a common concern both locally in municipalities and internationally through social rights commitments. As scholars, we may reveal changes in social rights empirically, using a wide range of material in the relevant context.
Acknowledgements
This research was generously supported by the Swedish Research Council for Health, Working Life and Welfare, project ‘Advanced legal practices in the welfare state. A study of displacements of the right to social assistance for undocumented persons and poor EU citizens, in three Swedish municipalities’ (2018-00458). The project has been approved by the Swedish Ethical Board (2019-01596)
Competing interests
None.
Open access
