Numbers, Ignorance and Swedish Migration Policy

Introduction

In a press conference on 24 November 2015, then Swedish Prime Minister Stefan Löfven announced that Swedish migration policy was going to be temporarily adapted ‘so that more people choose to seek asylum in other countries’ (Crouch 2015). Sweden, he announced, needed ‘breathing room’ following an unprecedented increase in the number of arriving refugees. ‘It pains me that Sweden is no longer capable of receiving asylum seekers at the high level we do today. We simply cannot do any more’ (Ibid.). With this announcement, Sweden began what has been termed a ‘u-turn on asylum’ (Parusel 2016), moving from having some of the most generous policies in the EU towards those seeking protection to becoming one of the most restrictive.

Whilst much has been said and written about the larger effects of this new law (see Elsrud, Gruber & Lundberg 2021), one aspect that has thus far gained very limited attention is how the debate around the law involved a decoupling of the effects of the law from its stated aim to reduce the total number of asylum seekers. Between when the new policy was announced in November 2015 and when it was finally approved and implemented in June 2016, the number of monthly asylum arrivals had already dropped from 36,726 people to just 2,104 people (Migrationsverket 2019a), a reduction of 94%. In spite of this steep drop in arrivals, the main focus of the policy as well as the debate surrounding it remained on decreasing the number of asylum seekers – a policy goal that had arguably already been achieved. What is also noteworthy is that not only had the total number of asylum seekers already come down by the time the law was passed, but the law itself also raised a number of concerns. Documents produced through the public consultation process typical of Swedish law-making (the so-called remissrunda) revealed widespread concern from a broad range of societal actors (see Stern 2018) about whether the law would be effective in its stated goal of reducing migrant numbers. Furthermore, there were also concerns that the law would have significant negative impacts on migrants already in Sweden and on Swedish society in general.

In this article we therefore seek to investigate how the focus on reducing numbers impacted the debate around the law in its initial form and in its subsequent incarnations, and what implications this has for our understanding of migration policymaking at large. That is, we take an interest in the fact that reducing migrant numbers could be used as a policy goal without much overt specification, allowing ‘migrant numbers’ to be invoked regardless of either the actual current number of arrivals or how much impact a law was likely to have on those numbers. Thus, unlike Ruhs and Martin’s classic concept of ‘numbers vs. rights’ (Ruhs & Martin 2008), in which states make trade-offs between the number of migrants and the rights they are afforded, here it seems that the calculation was instead ‘numbers vs. everything else’. A focus on the number of migrants, in other words, became untethered from a relation to the actual number of migrants or other societal issues. Devoid of any specific meaning or definition, ‘migrant numbers’ as a political argument could therefore be sustained over time and justify any particular set of consequences. Thus, if reducing numbers becomes a way to justify policy regardless of the number of migrants arriving, how can we then understand such policies?

Such a question places this article at the confluence of two streams of literature. The first relates to the broader politics of knowledge production, and in particular, the ways in which expert knowledge becomes utilised within policymaking. In migration policymaking in particular, previous literature has examined how knowledge is both constructed (Amelung, Scheel & van Reekum 2024) and used (Boswell 2009) in policymaking. Whilst we draw on ideas from these readings, we also make use of the literature that places explicit emphasis on ‘strategic ignorance’ (McGoey 2019) to investigate how knowledge is also rejected. As such, we also take an interest in the constructions of non-knowledge. The second strand of readings relate to the politics and ‘seduction’ (Merry 2016) of numbers. That is, their ability to convey information and an air of certainty at the same time that they serve to decontextualise the world they seek to describe.

The argument presented here therefore evolves in two steps. First, we examine the role that migrant numbers played in both justifying the legal reforms and as an ongoing point of discussion. In the second step we examine the expert knowledge that policymakers had access to and how that information was consistently ignored (Stel 2016: 1402) in favor of a discourse that repeatedly centered on numbers related to migration. This focus on numbers suggests both the ability of a numbers-based justification to be politically successful as well as how the constant invocation of numbers – increasingly untethered from their social reality – makes meaningful discussion of their actual value difficult if not impossible. Instead, we argue that numbers and numeric figures could successfully be used to convey an ‘aura of scientific exactitude’ (Hansen 2015), attributing these policy changes to authority or legitimacy, despite ignoring or marginalising other available and known sources of knowledge.

Background and Methodology

During 2015, Sweden granted one the largest number of asylum claims in the EU relative to its population size (Holloway, Faurès & Leach 2020). Whilst initially being quite welcoming to those seeking asylum, the number of arrivals began to create a political sense of a need to ‘do something’. The chosen mechanism was to temporarily restrict the rights of asylum seekers for 3 years with the so-called Temporary Law, which came into force in mid-2016. Despite its construction as a temporary measure, this law was renewed in 2019 with some minor adjustments, and in 2021 it was made permanent following further changes. On the basis of our analysis of the legal reforms, the Temporary Law sought to achieve its goal through several principal measures (see also Asylkomissionen 2020; Garvik & Valenta 2021). First, it made temporary rather than permanent residence permits the main rule. Those arriving seeking protection would initially receive a permit valid for 13 months or 3 years, depending on one’s categorisation as in need of subsidiary protection or as a refugee, with the possibility for renewal in increments of 2 years. Secondly, family reunification was restricted for most arrivals (although this restriction was eased somewhat in the 2019 renewal). Finally, the law offered the possibility for permanent residency only to those who could prove that they were financially self-sufficient either through work or their own entrepreneurial activities. Whilst the self-sufficiency requirement to receive permanent residency was initially restricted only to those with protection-related permits in the 2016 and 2019 iterations, the 2021 version, which enshrined the law permanently, also extended this rule to all individuals applying for permanent residency (see Table 1 for a summary of the key aspects of each law).

Our analysis draws on an extensive number of official documents generated as part of these legal processes in 2016, 2019 and 2021. For each law amendment, we examined four primary types of documents (see Figure 1 for an overview of the lawmaking process in Sweden). First, we examined the law drafts which included initial investigations for the suggested amendments by the government. Second, the documents produced as part of the official consultation process, known as the remiss system, were studied. An important and unique feature of Swedish policymaking, the remiss system allows a wide range of organisations the right to be consulted before decisions are made (Lundberg 2013: 60). The actors consulted can include government agencies, municipalities, courts, unions and civil society organisations as well as any other actor wishing to submit written commentary. Since Swedish agencies maintain a relatively high degree of autonomy, such documents offer a degree of independence that might be less present elsewhere. Third, we examined the government’s official proposal for the law (known as the proposition), which includes not only the technical legal details of the proposed changes, but also the motivation for the changes and the expected consequences. In putting forward their propositions to parliament, the government is also expected to engage with and respond to the comments from the consultation round. Lastly, we studied the transcript of the debate in parliament over these changes. Overall, this means that our analysis relies on data from each key stage in the Swedish law-making process.

Figure 1

Two key features of the Swedish policy system make the examination of the use of expert knowledge here especially interesting and enlightening. Firstly, due to the Principle of Public Access to Official Records (Offentlighetsprincip), all documents that form part of the Swedish law-making system are publicly available. The combination of the remiss system and the principle of public access offers a rare possibility to trace and gain a broad overview of the views of the government (government proposal/proposition), those of societal actors (consultation reports/remiss) and the parliament as a whole (parliamentary debates) as the law is proposed, debated and decided upon. Secondly, a large amount of ‘expert’ information about both the content and likely effects of a proposed law are contained within the policy process through the consultation (remiss) system. Therefore, unlike in many of the contexts discussed in the literature on research and migration policymaking where how research enters the policy process remains an open question (see Boswell & Smith 2017), the expert knowledge discussed here is endogenous to the policy process. In the typical case, policymakers respond to the consultation responses (remissvar), justifying their own position, which also means that it is possible to follow more precisely how policymakers have engaged with the expert knowledge generated through the consultation process.

All official documents were coded and analysed using Nvivo software. Our analysis of these documents was focussed on examining the law on its terms. Since the stated focus of the law was the reduction of migrant arrivals, we looked at two key issues. We first examined how migrant numbers were presented and discussed in the legal reforms. Secondly, we examined what effects the law was expected to/observed to have. Examining the law over its several rounds also allowed us to look at how information about the law’s effects were considered (or not) over time.

Knowledge and Ignorance

Policy responses towards migration have been criticised for producing unintended consequences, for failing to achieve their stated objectives (Pettrachin & Hadj Abdou 2024) and for deploying tools which have been shown not to be effective (Natter & Welfens 2024). Scholars have further identified gaps between policy developments and objective evidence, a so-called evidence–policy gap (Pettrachin & Hadj Abdou 2024). In light of the above, scholars have taken an interest in how expert knowledge is utilised (or not) within migration policymaking processes (Ruhs, Tamas & Palme 2019). For instance, the work of Boswell points to the instrumental, legitimising and substantiating uses of knowledge by political actors (2009) as well as the different modes by which expert knowledge is communicated to policymakers (Boswell & Smith 2017). More broadly, studies on the utilisation of knowledge have pointed to the ways in which its use is politicised, such that even though a large amount of research is produced, it is engaged with only selectively (Natter & Welfens 2024; Ruhs et al. 2019).

Yet if politicians and policy processes are known to generate information which is not utilised, it is also important to think about the use of ‘non-knowledge’ or ‘ignorance’ as its own form of knowledge practice. This broader idea of what has been called ‘agnotology’ (Proctor & Schiebinger 2008) or ‘ignorance studies’ (Gross & McGoey 2015) focusses on the power of not knowing. Whilst unknowing may be simply the product of a complex world, it is also something that may be used intentionally to generate ‘strategic ignorance’, exerting power through (strategically) maintaining ignorance about a given topic (McGoey 2012). In its uses within the context of migration governance, ignorance has been used to indicate how states decline to collect certain kinds of information (Boswell & Badenhoop 2021) or obfuscate the nature, significance or reliability of certain kinds of information to lend undue credence to data (Scheel & Ustek-Spilda 2019). The kinds of ignorance presented here are therefore connected to practices of knowledge construction and presentation. However, following Aradau (2017), we are here not so much interested in the generation of ignorance or non-knowledge as we are how it is used productively.

There remains another dimension of this ‘ignorance’ that is not related to how knowledge is generated, but rather to how it is used, like the proverbial elephant in the room or the presence of a persona non grata at a party – something which is known but is simply actively ignored. Thus, it is useful here to also think about ‘strategic ignoring’, the ongoing act of refusing to engage with, mention or acknowledge information that is clearly known. Here, what Stel (2016: 1403) has termed ‘professed ignorance’ or ‘the things they know but pretend not to know’, thus represents not merely a strategy of avoiding acquisition or general obfuscation, but also a deliberate strategy of denial. In this, the specific focus on ignoring also highlights ‘the dynamic relationship between knowledge and ignorance’ (Borrelli 2024: 5).

The Special Role of Numbers

In the politics of knowledge production around migration, numbers have come to play an increasingly important role. In fact, the field of migration has seen an almost extraordinary emphasis on, or even craving for (see Lemberg-Pedersen & Haioty 2020), quantifiable data. Such data have become part of the ‘knowledge practices’ that have been ‘mobilized to make migration as intelligible, actionable objects of policy-making’ (Scheel, Ruppert & Ustek-Spilda 2019: 579). In such processes, discussions around migration, particularly within the public sphere, have often ended up fixating on migrant numbers rather than other aspects or attributes. One reason for this, as Scheel and Ustek-Spilda (2019: 665) suggest, is that in policy debates numerical facts are often attributed to authority ‘which imbue their producers with the aura of scientific exactitude’ (see also Hansen 2015). However, numbers do not just imbue certainty or create subjects, they also obscure complexity. Indeed, it is precisely this combination of specificity and comparability, along with a decontextualisation, that makes numbers so ‘seductive’ to use, despite the fact that numbers often ‘provide knowledge that is decontextualized, homogenized, and remote from local systems of meaning’ (Merry 2016: 3). Numbers contain within them certain understandings of the world that led to their generation, but this meaning becomes stripped out once they are presented as mere digits. In this article, our goal is not so much to challenge the knowledge that generated the numbers (although as we will note in the upcoming chapters, it was in many of our cases highly speculative) but to point to how migrant numbers, through their both inherent and constant decontextualisation, became stripped of meaning in the Swedish policy debate.

Whilst it is easy to see how numbers can be attractive to policymakers in the sense that they present a simplified version of reality, research has emphasised that numbers are not simply neutral or static facts ‘out there’ but rather ‘contingent and on-going sociopolitical constructions’ (Ostaijen & Scholten 2017: 493). Numbers are, in other words, imbued with meaning: a process that involves taking complex social realities and distilling them down into quantifiable indicators (Broome & Quirk 2015; Merry 2016). This is especially relevant as migrant numbers tend to be treated as crucial facts in migration debates (Fotopoulos & Kaimaklioti 2016), whilst there is often not a clear understanding of what these actually mean. As Anderson writes:

The proliferation of statistics around migrant numbers provides a way of quantifying what is happening, but as Anderson so aptly points out, there is typically no clear understanding of what the ‘right’ number of migrants is amongst policymakers. It is not that the numbers produced are necessarily wrong (although of course as mentioned above their collection and articulation are not neutral acts) but rather that the meaning and judgement attached to such numbers tend to be both unstable and unclearly defined.

Whilst it is difficult to neatly condense three periods of sweeping migration policy reform and their attendant hundreds of pages of documents into such a short space, we here provide three key moments in which we see both an emphasis on numbers as well as decoupling of those numbers from larger consequences through the ignoring of key information. First, we examine a persistent uncertainty regarding the slipperiness of achieving the goal of reducing migrant arrivals – both in terms of what it meant and if it could be known whether the laws could achieve it. Second, we look to the debates in parliament to examine how policymakers avoided attributing constantly referenced numbers with any meaning, producing a sort of non-knowledge about these numbers that made it difficult for politicians to define what was a large or small number of migrants. Finally, we examine how the government continuously ignored significant and widespread concern about the larger impacts that the laws would have on Swedish society and migrants themselves.

The Relationship between the Law and Migrant Numbers

Although the legal changes made through the Temporary Law were clearly focussed on reducing the number of migrant arrivals to Sweden, both official commentaries and the government itself expressed uncertainty about whether it would actually have (or already had) this effect. Given the variety of factors at play, including the actions taken by other countries in Europe, the Administrative Court of Stockholm, for instance, noted that ‘it isn’t possible to determine if the goal of the proposed legal change will be achieved’ (Förvaltningsrätten i Stockholm 2016).¹ Commenting on the proposed extension of the Temporary Law in 2019, the Administrative Court of Gothenburg also noted that ‘…no sufficient analysis has been made of the reasons why the number of asylum seekers in Sweden has decreased. No definite conclusion can be drawn about the impact the Temporary Law has, or an extension of this law would have, on the number of asylum seekers’ (Förvaltningsrätten i Göteborg 2019: 1).

Figures on the projected number of arrivals assembled by the Swedish Migration Agency did estimate that a continuation of the Temporary Law would lead to fewer arrivals compared with a return to the previous, and more welcoming, legal framework. According to their projections, an extension of the Temporary Law would lead to about 5,000 fewer arrivals in 2019, and about 11,000 fewer arrivals per year in the years 2020–2022. However, whilst the Migration Agency predicted lower numbers, such figures were given with a fairly high degree of uncertainty. For the years 2020–2023 the predictions were, for example, 16,000–28,000 arrivals per year with the continuation of the Temporary Law or 15,000–40,000 if the policy reverted back to its original form (Migrationsverket 2019b).

In the government’s own initial legal proposal in 2016, they wrote that the ‘main purpose of the government’s proposal is to decrease the number of asylum-seekers to Sweden. However, it is not possible to predict to what extent the number of asylum seekers will be affected’ (Regeringen 2016b: 39). By the time the law became extended in 2019, the government noted, yet again, that it is ‘true that it is not possible to determine with certainty to what extent each measure, alone or in combination with other measures, has affected the number of asylum seekers’ (Regeringen 2019: 29). The reduction of numbers was, as already mentioned in the introduction, characterised in some cases as giving Sweden a temporary ‘respite’ or ‘breathing space’ (andrum) to better address the reception of those who had already arrived. As the government wrote in their proposal: ‘…these measures are also necessary for Sweden to be at a minimum level in accordance with EU law and international conventions and create a breathing space for Swedish refugee reception’ (Regeringen 2016a: 44). This idea of ‘breathing space’ was often invoked, pointing to the temporary nature of the law and the large number of people seeking protection; 5 years after the introduction of the Temporary Law, the government, however, continued to insist that ‘it is not possible to determine to what extent each measure, alone or in combination with other measures, has affected the number of asylum seekers’ (Regeringen 2021: 51).

Despite relying on the idea that it was necessary to implement these laws to reduce the number of arrivals, the government persistently claimed that it could not be known whether the law would have or had had that effect. Here, they clearly relied on the strategic application of ‘unknowing’ (Gould & Stel 2022), as they maintained that it was ‘not possible to determine’ which measures have had which effects on the number of individuals arriving. Yet in continuing to promote these laws with the goal of reducing migrant numbers, the ‘ignorance’ about its effects was not operationalised for the purposes of caution, but rather for further action. By not providing analysis or evaluations on why the number of asylum seekers in Sweden had decreased, the government thus consistently declined to collect certain kinds of information (Boswell & Badenhoop 2021) and strategically maintained ignorance about the effect on migrant numbers (McGoey 2012). This is especially clear when considering that policymakers must have known before implementing the initial law that the number of arrivals had already drastically decreased. This was, for example, brought up in the Parliamentary debate in 2016, exemplified by this statement from an MP for the Center Party:

This statement sheds light on the fact that the situation in Sweden had already changed quite dramatically even before the new law was passed or implemented, and that this was known amongst policymakers. Claiming that it cannot be known to what extent the law changes would impact or had impacted the decrease in the number of arrivals, whilst knowing that the numbers had already dramatically decreased before the law could have had any effect, could thus be understood as ‘strategic ignoring’ and ‘professed ignorance’. The government, in other words, refused to engage with what was clearly known, but they ‘pretended’ not to know (Stel 2016: 1403).

The number of asylum seekers continued to be the key justification to prolong the measures in the Temporary Law in 2018 and 2020 – even as these numbers continued to decrease and despite the fact that the government expressed uncertainty about (and even denied the ability to know) how much of an effect the Temporary Law actually had on reducing the number of arrivals to Sweden. Whilst several studies pointed to an effect on migrant numbers and legal conditions as being a motivating factor behind migrants choosing particular destinations, it also remained difficult for policymakers to identify how large of an impact the legal changes in Sweden had been relative to other changes, such as the EU-Turkey Statement or the large number of European borders which remained closed (Migrationskommittén 2020).

What is a Big or Small Number?

Although migrant numbers also regularly featured in the parliamentary debates around the Temporary Law and its successors, the use of different numbers was rarely accompanied by any explicit reference as to how these should be interpreted. This means that despite ‘evidence/facts’ being presented in the form of numbers and numeric figures, a form of ignorance was produced, as these numbers were rarely tethered to a specific meaning. A prime example of this is the speech of the then Justice Minister, when presenting the new law for discussion in parliament (Riksdagen 2016: 34–41). He began by noting that ‘last year [2015] Sweden took in over 160,000 asylum seekers’ and that ‘the previous year, 2014, Sweden took in over 80,000. This means that, as many have pointed out earlier today, that our land in two years has taken in almost a quarter million asylum seekers’ (Riksdagen 2016: 34). As the speech continues, he notes how many Syrians have come to Sweden since 2011 (130,000), how many came in the last 4 months of 2015 (114,000) and of those how many were unaccompanied minors (26,000), all within the first 2 minutes of his speech. He also noted how many people had taken classes for Swedish and social studies (75,000), how many people were living in housing provided by the Swedish Migration Agency (90,000, down from a previous high of 100,000) and how many applications were being processed each month (7,000–8,000). In all, he used 24 different numbers to describe asylum seekers, yet largely left aside any discussion or mentioning of what those numbers meant.²

Whilst we can assume that the choice of statistics in the speech was designed to point to a high number of people seeking asylum, there were markedly fewer statistics used when discussing the effects of those arrivals. What this means is that whilst numbers themselves were used rhetorically to signal a problem writ large, the lack of specific examples of the effects of those numbers also means that the perceived problem (or even threat) that they were supposed to point to remained amorphous and undefined. In other words, whilst constantly evoked, the numbers were not filled with meaning, and the then Justice Minister, alongside many other parliamentarians, rather seemed to evoke the semantic and ‘seductive’ (Merry 2016) power of the statistics – and the authority they portray – whilst also leaving them there decontextualised, stripped of any particular meaning. Rather, it seems that numbers became a means to signal the use of evidence/knowledge/facts to attribute authority to the policy debate statement (Hansen 2015) rather than clarifying what the numbers actually meant.

The excessive usage of statistics combined with a lack of specificity as to their actual meaning led to some fairly confusing and somewhat peculiar exchanges in parliament as the parliamentarians tried to sort out what meaning should be attached to a particular statistic. One example of this is when, in the parliamentary discussion on the renewal of the Temporary Law in 2019, two liberal/conservative parties, both broadly in favour of reduced immigration, engaged in a debate on what exactly constituted a small number of asylum seekers. A member of parliament (MP) for the Moderate Party, the largest conservative party, responding to an MP from the Christian Democrats, begins:

In response, the MP from the Christian Democrats notes that this right to family reunification already exists in many other EU countries and is therefore unlikely to have a large effect on the number of arrivals. He further emphasised, noting broader values that ‘…I think that the possibility for family reunification plays well into our ambition as Christian Democrats to ensure that we have a society where we live with each other rather than alongside each other’. Unconvinced by this response, the member from the Moderates replied: ‘How can immigration be reduced as drastically as one would like it to be if, at the same time, people in this House repeatedly vote for more comprehensive immigration?’ (Riksdagen 2019: 79). In his final response, the Christian Democrats MP moves away from the question entirely, pointing to other priorities for the Christian Democrats such as vocational training programs, tougher welfare standards and greater clarity in asylum decisions.

What is revealing about this exchange is both the difficulty, and perhaps also avoidance, of defining what constitutes a big number, as well as the difficulty in and silence around determining what those numbers mean. What was important for the Christian Democrats was that families had the opportunity to be reunited, yet for the Moderates, the merits or demerits of family reunification were not mentioned or discussed – only the fact that the policy would lead to more immigration. Thus, two parties that are both broadly in favour of reducing the number of arrivals end up in a situation where they cannot agree on whether 8,000 migrants is a small or large number.

Numbers, despite their claim to specificity, precision, certainty and countability, here end up serving an almost opposite purpose. Whilst the two politicians do not argue about the specific number, they struggle to find a common understanding of what that number means and whether it is small or large. Notably, this conversation is not focussed on a cost–benefit analysis of 8,000 additional people, but rather a more general discussion of whether it is a small or large number – a discussion that is rendered almost meaningless absent any shared, stable criteria for assessment or an emplacement of these figures within larger societal implications. By avoiding defining what is a big or small number, not filling the invoked numbers with meaning, policymakers were able to avoid specifying what their ‘preferred number’ was, and the argument to ‘reduce numbers’ was consequently also sustained and re-used over time without specificity. In this sense, the government mobilised and exploited unknowns to avoid liability for earlier actions and created unknowns ‘to generate support for future political initiatives’ (McGoey 2019: 3).

Interestingly, it was the Social Democrats and the Moderates, the two largest parties representing the centre-left and centre-right, respectively, which relied most on numbers in their speeches before parliament in 2016. Whilst the 8 references to numbers in the speech by the Moderates pales in comparison with the aforementioned 24 used by the then Justice minister of the Social Democrats, they are both significantly more than the number used by the far right and staunchly anti-immigrant Swedish Democratic MP, who used zero (Riksdagen 2016). As discussed above, the use of figures for the Social Democrats and for the Moderates provide a way to justify the need for action without an ideological case. Echoing Broome and Quirk, the pivot to numbers ‘takes what might otherwise be highly contentious normative agendas and converts them into formats that gain credibility through rhetorical claims to neutral and technocratic assessment.’ (Broome & Quirk 2015: 813). In this, they empty ‘migrant numbers’ of their meaning, constantly reducing them to bare figures and struggling to place them into societal context. The choice of the Swedish Democrats to then not discuss migrant numbers makes sense because as a party which is always sceptical of migration, they have no need for such vague numerical abstractions. They have no problem stating their opposition openly and ideologically over time as to them, migrants are a problem to begin with, and any number is too high.

Our emphasis in these first two sections has been on the way numbers were used in the debate without any specificity or meaning, creating a situation in which any number could be evoked as a number that was ‘too high’, strategically not engaging with any specific discussion of what those numbers meant. However, in terms of the productive role of non-knowledge, we find it equally interesting to look at the way that expert knowledge around the effects of the new law was continuously ignored throughout the policy-making process. As expert knowledge about expected effects of a law is baked into the policy process in Sweden, non-engagement with this information required not just a lack of engagement, but also an ongoing ignoring. Therefore, in the next section we review both the kinds of information that policymakers were provided with and how they did (and did not) engage with it.

What Collateral Damage?

Whilst the stated aim of the Temporary Law and its legal antecedents, as we have seen in previous sections, was to reduce the number of migrants, the chosen mechanism of reducing rights for individuals already in Sweden received significant critique. Many entities involved in the consultation (remiss) process pointed to negative effects of the law on the health, welfare, labour-market participation, legal security and integration of asylum-seekers in general with emphasis on how children, families and women would be especially disadvantaged. Whilst these were some of the principal concerns, they were by no means the only ones. Consultation reports throughout the years also pointed to a variety of issues with implementation (Förvaltningsrätten i Göteborg 2016; Migrationsverket 2016, 2019b, 2020; Sveriges Advokatsamfund 2020), equality (Civil Rights Defenders 2020; Jämställdhetsmyndigheten 2020; Länsstryelsen Stockholm 2020; RFSL 2020; Röda Korset 2020; Svenska Kyrkan 2020) and the rule of law, amongst other concerns. Taken together, these consultation reports reflect a significant number of severe concerns, voiced by a wide variety of actors both inside and outside of state institutions, about the impacts that these laws would have.

Although the Temporary Law emphasised employment, official consultations on the law highlighted from the beginning that the policy would likely bring distortions to the labour market. For instance, the Swedish Public Employment Service expressed concern that the law would disincentivise arriving individuals with needed skills from getting work in their area of expertise. Rather, in the search for quick employment they would take lower-skilled jobs that could be obtained faster, pushing those newly arrived individuals with lower-skills out of the Swedish labour market entirely (Arbetsförmedlingen 2016). Furthermore, a number of organisations pointed out that linking permanent residency to employment for newly arrived individuals risked creating situations of unequal power relations in the workplace. As one union noted, unlike with labour migrants, with asylum-seekers ‘it entails an increased vulnerability and an even stronger dependency on the employer. There are also no incentives to report irregularities, as the residence permit risks being revoked if it turns out that the law’s requirements have not been complied with’ (SACO 2016). Even employers were reluctant about the role that they would play in potentially granting long-term status to newly arrived migrants. The Swedish Federation of Business Owners, Företagarna, noted that the requirements of the law ‘places an additional burden on employers who will not only bear responsibility for another person’s livelihood and safety in the workplace, but also for their right to reside in Sweden’ (Företagarna 2016). Whilst in 2015 such concerns were expressed as anticipated effects if the law were to be passed, the consultation rounds in 2019 and 2020 revealed that many of these issues had indeed come to pass.

Many of the consultation reports further pointed to both anticipated and observed health impacts of the policy changes. In both 2019 and 2020, many of the reports, including those from the Swedish Red Cross, Save the Children and the National Board of Health and Welfare, pointed to concerns about the negative psychological effects that could come from being separated from family and from living in a constantly temporary status (Rädda Barnen 2019; Röda Korset 2020; Socialstyrelsen 2019). The National Board of Health and Welfare for instance noted: ‘Increased mental illness has been observed with regards to temporary residency’ (Socialstyrelsen 2019), and furthermore, that care providers often were unwilling to begin treatment that can take many years, as conditions may worsen if the care is cancelled early – factors which become crucially important when it is not known how long someone will be allowed to stay. Reports also pointed to the particular vulnerability of children who could be affected not only by general familial uncertainty, but also by the constant possibility of separation from school and friends (Barnombudsmannen 2020; MUCF 2020; Röda Korset 2020).

Both consultation reports and the government also noted that the law would negatively impact the integration of new arrivals. The government itself wrote that: ‘The increased uncertainty and the fact that it takes longer before family reunification can take place, risks counteracting integration’ (Regeringen 2016a). Consultation reports submitted as part of the 2021 reform reiterated such observations, with a number of reports pointing to effects on integration both due to issues of family separation (Amnesty International 2020; FARR 2020; Malmö stad 2020; Röda Korset 2020) as well as effects on long-term planning and investments in both businesses and education (Region Jönköping län 2020; SKR 2020).

Yet this extensive tranche of information was largely met with silence from policymakers. In the parliamentary debate over the 2016 law, only one politician, from the Center Party, mentioned the consultative reports that had been received; 3 years later, in debates over the 2019 renewal, they were likewise only brought up by a member from the Left Party. Later, in 2021, the main discussion of the consultative process occurred in an exchange between a MP from the Christian Democrats and one from the ruling Social Democrats. In that exchange, the Social Democrats responded to criticism from the Christian Democrats that the current bill had been watered down from its original form by pointing out that these changes were done ‘partly because the consultation bodies pointed to these the issues had to be dealt with’ (Riksdagen 2021: 14). Responding to this, the Christian Democrat MP said: ‘It is very convenient for [him] to be able to hide behind the consultation responses’ (Riksdagen 2021: 14), accusing the Social Democrats of making the changes not due to the content of these issues, but rather to placate their coalition partners, the Green Party.

Even when the existence of the consultative reports was acknowledged, their content was often dismissed. In its official legislative proposal following the consultation process in 2016, the government wrote: ‘The Government therefore shares the consultation process’s assessment that the proposed temporary residency permits will have negative consequences for both individuals and for the agencies. Despite this, the government believes that it is necessary to temporarily introduce rules on temporary permits’ (Regeringen 2016a: 29). No further discussion or elaboration is offered on this point in the proposal. Although here the government recognised the content of the reports, it made no effort at all to defend or argue its position.

Not only was the information gathered as part of the consultation process actively ignored by policymakers, but there was also an effort to not gather further information. As was written in the Children’s Ombudsman’s consultation report in 2019:

Rather than conducting the kind of evaluation that had both been promised and would be typical for a policy change of this magnitude, the government actively chose not to find out what the effects of this law were. What we see here then is both an ongoing ignoring of the information they already had as well as a lack of interest in gathering further information – an intentional ‘unknowing’ around the negative effects of the law, which we see here as not incidental, but rather strategic (McGoey 2019). Even as they argue that the law is ‘necessary’ (to reduce migrant numbers) and recognise (although not in detail) its ‘negative consequences’, no attempt is made by the government to justify or explain this trade-off or even the rationale underpinning it. By not providing a justification they can avoid responsibility or critiques of their logic – because they have not offered any.

Conclusions

What emerges from our review of Swedish policy documents is a picture of immigration policy in which a focus on numbers overshadowed all other concerns. The reduction of migrant numbers became a goal not just to alleviate pressure on the reception system, but also a goal in and of itself. At the same time that the legal process made relatively few efforts to deduce the effects of different migrant numbers on society, it likewise did not give significant consideration to the effects of the law on migrants in Sweden. The discussion of migrant numbers thus became largely untethered from societal consequences.

Given that the consultation process means that expert knowledge is an inherent feature of Swedish policymaking, it creates a situation which demands some form of response to that knowledge from policymakers. However, what emerges as notable is that their response here was largely to ignore the knowledge produced. Rather than selectively or strategically using the expert knowledge produced (see Boswell 2009), policymakers here rather engaged in a sustained ignoring, a process which was supported by making the focus of the issue one of migrant numbers. This focus on numbers allowed policymakers to use the authoritative nature of numbers (see Hansen 2015) whilst also removing those numbers from their broader social meanings (see Merry 2016). Building on others who have pointed to the role of both ignorance (Gould & Stel 2022) and ignoring (Borrelli 2024), we have additionally shown how meaningful political debate over contentious issues becomes difficult in the context of a focus on decontextualised numbers.

Ultimately, this paper makes no judgement of whether Sweden’s reception system was overburdened or beyond its capacity to operate effectively or whether there were ‘too many’ or ‘too few’ individuals arriving in Sweden. Rather, what we highlight here is that Swedish policymakers struggled to make that assessment because migrant numbers were never discussed in relation to broader societal capacities or goals. If ‘few’ was never defined, there could also always be the argument that there were ‘too many’. Examining the role of numbers is especially salient in the context of migration debates, where the number of arrivals is often treated as an important social and political issue. What we reveal here is that even in the case of Sweden, a country known for its (generally) rigorous policymaking process and vigorous information gathering, the meaning of migrant numbers can become quite ethereal. As discussions of migrant numbers become all the more commonplace in both policy and societal spheres, it remains important to examine what meaning is given to migrant numbers and how information about them is used.

Funding Information

This research was supported by the Swedish Research Council for Health, Working Life and Welfare (FORTE) grant 2018-00316.

Competing Interests

The authors have no competing interests to declare.