‘We need the money’: how welfare anxiety justifies penal and social reforms in immigration debate

ABSTRACT

The response of welfare states to immigration is a topic of concern for scholars of welfare and scholars of crime and society. Yet, debates about immigration, crime and welfare in academic literature often take place in isolation. This paper advances both literatures by studying the mechanisms that drive both welfare chauvinism and penal nationalism in relationship to each other. Rather than assuming an a priori and abstract negative construction of immigration in political debates, this study draws on a large-scale Critical Discourse Analysis of political and media debates in Denmark to show how immigration is constructed in ways that justify welfare and penal policy reforms. The analysis shows that both policy shifts – towards a two-tier exclusionary welfare system, and towards penal nationalism that reserves harsher punishment for immigrants – are driven by a discourse that constructs immigrants as threatening welfare first, and only secondarily as threatening personal safety. Based on the empirical data, the article argues that anxieties about the impact of immigration on the integrity of the welfare state is what drives and justifies the construction of a two-tiered welfare and criminal justice system.

KEYWORDS:

• Immigration
• welfare
• criminal justice
• discourse analysis
• policy

Introduction

In 2018, the center-right Danish government passed the Ghetto Package (Ghettopakke) – a legislative initiative that included a combination of social policy and criminal justice measures, aiming to disperse residential concentrations of low-income non-Western minorities and foster their social integration. As stipulated in the legislation, residential areas across the country were designated as ‘ghettos’ based on the ethnic composition and social-economic background of their residents. The meaning of such categorisation was significant, as the Ghetto Package called for the demolition or selling of a certain percentage of public housing units in the ‘ghettos’ and the subsequent relocation of their residents. In addition, the plan called for increased punishment for crimes committed within the ‘ghetto’ areas, and the removal of families from their public housing if one family member was convicted of certain crimes. Furthermore, the legislature stipulated that part of the funding for the Ghetto Package would come from benefit cuts for immigrants. This policy reform, while being more comprehensive and momentous than most others, was not singular and in some ways was not entirely exceptional. Since the turn of the century, and particularly since the 2015 refugee reception crisis (Rea et al. 2019), Denmark’s politics have become increasingly hostile to non-Western immigrants. This hostility is manifested in various legal and policy initiatives, some of which touch on core elements of the welfare state and its criminal justice system. As a classic Nordic welfare state with generous redistributive social policies and rehabilitative criminal justice policies, Denmark is an illuminating case to study whether – and, if so, how – welfare states evolve in response to immigration and growing diversity.

Scholarly debates on European welfare and criminal justice ask whether support for redistributive and rehabilitative social policies would wane in the face of immigration and diversity. Some scholars of welfare and criminal justice argue that increased diversity and immigration would lead to a weakening of the welfare state and a tougher penal system, in accordance with the United States experience (cf. Alesina and Glaeser 2004; Muncie 2008; Smith 2012). Other scholars within these related yet distinct literatures argue that the classic welfare state does not morph into an entirely different model of welfare regime. Instead, the European welfare state evolves into a two-tier system in which welfare privileges and rehabilitative policies are reserved for insiders of the welfare state (i.e. the ethnic majority citizen group), whereas the harsher and less generous treatment is kept for those who are conceived of as outsiders – ethnic minority and immigrant populations. Indeed, both sets of literatures find evidence of the emergence of dual welfare and penal regimes (Abrahamson 2019; Barker 2017; Jørgensen and Thomsen 2016; Larsen 2011).

The literatures on welfare and criminal justice share similar internal debates. Yet, by separating social-economic policies from criminal justice policies, we miss important aspects of the welfare state’s response to immigration. Often, if we turn our gaze to the intricate ways in which these two fields operate in tandem, we find legislation that mixes social policies with criminal justice policies, substitute one for the other and merge them both practically and symbolically. This paper extends Barker’s (2017) argument – that concerns for the solvency of the welfare state drive penal nationalism – by examining the mechanisms that drive this relationship. To make that argument, I constructed a large empirical dataset, which I used to analyse the communicative discourse (Schmidt 2008) around immigration, where government, civil society and private actors engaged in a debate over the most appropriate response to growing immigration.

The analysis shows how the negative construction of immigrants focused primarily on the threat they pose to the economy and the welfare system. Crime control and personal safety were secondary concerns that were often woven into welfare concerns. Abstracting from the case of Denmark, this article shows that the creation of two-tier welfare and criminal justice systems is driven by an immigration discourse that views immigrants as threatening welfare first, and only secondarily as threatening personal safety. This finding speaks to the centrality of welfare protectionism in welfare states’ public and political discourses and to the central role it plays in shaping public opinion and policy initiatives on immigration.

In what follows, I review the literatures on welfare chauvinism and welfare antagonism and connect them to the literature on penal nationalism. I describe the Danish case, outline my methodological approach and analyse political and media discourses on immigration in relation to broader discourses on welfare solvency and crime control. I will show how these discourses were meshed together, built on each other and ultimately justified welfare and penal policy reforms.

Immigration and the welfare state: understanding changes to social and penal policies

Welfare states are composed of social (more or less) redistributive welfare policies, as well as criminal justice polices. As Wacquant (2009), drawing on Bourdieu, put it, both institutions compose the 'left hand' (social policies) and the 'right hand' (disciplining forces) of the state. Both policy realms are entrusted – independently and in relationship with each other – with the management of the population and the successful reproduction and preservation of the state. Importantly, both institutions are also responsible for managing marginal populations. They do so by following the same logic of operation that lies in the principles of either the exclusion or inclusion of marginal populations (Beckett and Western 2001). Indeed, welfare and criminal justice literatures use similar terms – albeit with few cross references – to debate the shifting landscapes of welfare and criminal justice policies in the context of increased immigration and diversity.

Students of the welfare state have been trying to explain why the US has not developed a comprehensive welfare state as most other European states have done. One strand of explanation focused on the racial and ethnic diversity in the US, in comparison with the relative homogeneity of European countries. Alesina and Glaeser’s influential book (2004) promoted the thesis that greater ethnic and racial diversity leads to declining support for redistributive welfare policies. They, like others, empirically demonstrated this relationship mostly, but not exclusively, in the American context (Burgoon and Rooduijn 2021; Gilens 1999; Luttmer 2001; Quadagno 1994; Roberts 1997; Wacquant 2009). According to this literature, the key to explaining a decline in support for welfare in diverse settings is the greater tendency to demonise the poor as ‘lazy’ and ‘undeserving’ when the public perceives them to belong to an out-group. That, in turn, provides fertile grounds for attacks on the welfare state.

However, when studying this hypothesis in the European context, a slightly different theory emerged. While the tendency to stigmatise the out-group has also been recorded in Europe (e.g. Balch and Balabanova 2016; Larsen and Dejgaard 2013; Larsen 2011; Meeusen and Jacobs 2017), and is, too, based on stereotypical thinking about the out-group (Hjorth 2016), it does not necessarily lead to a denunciation of the welfare state and its redistributive principles. Instead, continental European states, and especially the Scandinavian social-democratic states, adjust their eligibility criteria so as to preserve welfare benefits but reserve them primarily for the ethnic majority population. Instead of welfare antagonism, we see the emergence of a two-tier system, sometimes called ‘welfare chauvinism’ (Brady and Finnigan 2014; Careja, Emmenegger, and Kvist 2015; Hjorth 2016; Jørgensen and Thomsen 2016), ‘welfare dualism’ (Larsen 2011) or ‘welfare nationalism’ (Barker 2013, 2017). Under this new regime, the principles of welfare universalism are still preserved (as opposed to retracting the universalist mandate and resorting to means-tested welfare policy), but their contours are changed to exclude migrant and ethnic minority populations.

The literature on penal policies is also divided along similar lines. Some scholars (e.g. Muncie 2008; Smith 2012; Wacquant 2010) claim that there is at least some evidence of the diffusion of American-style harsh penal policies. They claim that in both continents, often by direct influence or by inspiration, we see an antagonist response to diversity that leads to exceedingly harsh penal policies. Other scholars, such as Barker (2017) and Shammas (2017), contend that what we see instead is the creation of a two-tier penal system, in which rehabilitative measures are gradually being reserved for national insiders, whereas the punitive harsh policies are directed at foreign and minority populations. Vanessa Barker (2017) argues specifically that welfare nationalism leads to penal nationalism. The two-tier welfare system is reflected onto the penal system, where – similar to the distribution of welfare entitlement – the state’s coercive power is hinged on the citizenship status of the individual. It is central to Barker’s argument that state violence in welfare nations is not a consequence of the hollowing out of welfare, which leaves the criminal justice system as the sole arm of the state to manage the poor (as, for example, Wacquant (2009) argued for the US). Instead, in Europe, state violence is used to protect and reproduce the welfare state rather than dismantle or replace it. The protection of the welfare state is done on behalf of the ‘legitimate’ state members and against the out-group, which is suspected to be unfairly overburdening the system to the point of break. In Barker’s words, ‘[w]elfare nationalism drives penal nationalism, in which perceived outsiders, non-members, and especially noncitizens are subject to increased controls in the name of national interests, including the preservation of national identity and social security’ (Barker 2017, 134). While other studies have focused on the role of ethnonationalism in immigration debates (e.g. Thompson 2022), this article pays close attention to the role penal nationalism and especially welfare nationalism play in shaping states’ responses to immigration.

Denmark presents a particularly relevant case to study this phenomenon. It is a classic social-democratic state with high degrees of redistributive universal benefits and high levels of de-commodification (Esping-Andersen 1998), which provides personal social services (Sipilä 1997). While social care is a universal right enjoyed by all citizens, the state also invests a significant amount in targeted programs that treat marginalised populations. This arm of the state prioritises rehabilitative approaches both in welfare and in criminal justice institutions. According to Pratt (2008), the lenient and rehabilitative approach to punishment in Scandinavia is rooted in the states’ highly egalitarian social structure, culture and values, as well as their homogeneous origin and their developed social welfare systems that reduce criminogenic tendencies. Denmark thus presents us with an intriguing case study: it remains a rather homogeneous state – non-Western immigrants and their descendants were 9% of the population in 2021 (Statistics Denmark 2021)¹ – with high levels of trust and cohesion. Still, concerns about immigration became a leading issue in the press and in election campaigns. The fact that immigrants and minorities comprise a rather small share of the population of Denmark serves to highlight the political nature of the construction of immigrants as a threat that justifies the dualisation of welfare and criminal justice policies.

As public opinion shifted (Hercowitz-Amir and Raijman 2020), the legislator responded with several policies that aim to curb immigration and limit immigrants’ access to welfare benefits. Thereby it effectively rebuilt a dual welfare and criminal justice structure. One such policy that contributed to the dualisation of welfare and criminal justice systems is the aforementioned Ghetto Package. Another legislative enactment increased the punishment for individuals who commit a crime while being associated with a known gang. While the law targeted all gang members, regardless of their ethnic background, the terminology in the title of the law used a word that pointed to minority-affiliated gangs. The Danish language has two words for ‘gang’: ‘bande’ refers to a minority-affiliated gang, and ‘rocker’ refers to an ethnic Danish-affiliated gang. Notably, the legislation changed from being called Rockerloven (Rockers’ Law) in 1996 to being rebranded as Bandepakke (Bande Package) in the early 2000s. The third iteration of the Bandepakke, passed in 2017, included a financing bill that was designed to pay for the initiative through cuts to a number of universal and needs-based welfare benefits for immigrants and asylum seekers. Put together, both symbolically and materially, the new Gang Package of 2017 had an adverse effect on migrant and minority populations. Additionally, in the context of an overall increased emphasis on securitisation in Danish prisons, inmates who do not hold Danish citizenship are barred from participating in any rehabilitative services inside prison or after release (Damsa 2021). Most recently, the Danish government has reached an agreement to transfer foreign-national prisoners, who are scheduled to be deported at the end of their prison term, to a prison in Kosovo. The justification for this agreement was not focused on cost-saving measures. Instead, the government justified the plan based on the need to avoid the over-crowding of Danish prisons (Justice Ministry 2022). Thus, it is the welfare of (Danish national) prisoners that justified the creation of a two-tier penal system.

Despite the close ties between welfare policies and criminal justice policies, especially in social-democratic states, and albeit similar debates in the literatures, little dialogue exists between scholars of both fields. This study contributes to both literatures by showing the mechanisms that drive these policy changes. My analysis demonstrates the centrality of welfare concerns to the dualisation of both social and penal policies. Moreover, while most previous studies take the negative construction of immigrants as an a priori fact that leads to welfare states’ restructuring (e.g. Brady and Finnigan 2014; Careja, Emmenegger, and Kvist 2015), or test presupposed theoretically conceived attitudes towards immigrants (e.g. Hjorth 2016; Larsen 2011), I provide an in-depth empirical analysis of migration debates. This allows me to demonstrate the active intertwining of welfare and safety concerns in the debates. This, in turn, helps us to explain welfare states’ shifting policy landscape.

Methods

To analyse the evolving discourse on immigration and its relationship with welfare and criminal justice discourses, I conducted Critical Discourse Analysis (CDA) on a dataset of 535 articles that were published in four national news sources (online and printed) in the years 2015–2018, as well as 12 agenda-setting speeches of Danish prime ministers that were delivered in the years 2015–2020. A constructivist method, CDA engages in an inter-textual analysis of discourses in their historical and cultural contexts (Wodak 1996). It allows us to explain and interpret the dynamic relationships between different discourses, their evolution over time and their relationship with power in society (Van Dijk 2003; Wodak 1996; Wodak and Meyer 2009). My empirical analysis is inspired by the tradition of the Discourse Historical Approach (DHA) in CDA (Wodak 2009). While my theoretical inspiration draws on DHA, I restricted my sociological analysis to deconstructing a specific sample of media and political debates, analysing the themes (topoi) of public debates connecting welfare with immigration and criminal justice policies, their content, the broader socio-historical political context and specific discursive statements by government and media actors in Denmark (cf. Reisigl and Wodak 2001).

Media is a key site for such an analysis. Media outlets are central players in shaping political understanding, and in forming the ideas that become ‘common sense’ and ultimately govern politics (Yuval-Davis et al. 2017). It is also central in shaping internal boundaries between migrant populations and the majority society through the creation and dissemination of dominant narratives that form policy debates (Georgiou 2020) and have shown to impact public perception on immigration (e.g. Blinder and Jeannet 2018; Meltzer et al. 2021; Igartua and Cheng 2009). By complementing media discourse analysis with political discourse analysis, I was able to examine the tensions as well as points of convergence between the news media discourse and hegemonic political discourse. The diversity of the news sources together with the analysis of political speeches allowed me to gain a comprehensive understanding of the multidimensional construction of immigration discourse in relation to different social settings and various other discourses. To account for the dynamic nature of discourse, I conducted both historical-diachronic discourse analysis and a comparative-synchronic analysis of the data (Carvalho 2008), which enabled me to capture both the complexity of immigration discourse at each critical moment, as well as its temporal evolution.

Data and sample

The time frame of the analysis starts in 2015. This was an election year in Denmark when the anti-immigration far right Danskfolkeparti (Danish People’s Party) received a record high share of the votes. Shortly thereafter, the Syrian civil war led to a massive wave of refugees seeking shelter in Europe. The analysis, then, starts with this watershed year and follows the evolution of the discourse on immigration until the end of 2018 in the news media, and until January 2020 in prime ministers’ speeches.

News media

In order to create a dataset that is both comprehensive and representative of the evolving discourse on immigration in the news throughout the years, and at the same time is feasible for in-depth qualitative analysis, I chose a strategy of sampling that focused on critical discourse moments (Carvalho 2008; Gamson 1992). I selected the critical discourse moments based on their theoretical significance, after mapping the news for relevant events that are likely to lead to heightened coverage of the topic of immigration (Carvalho 2008). In order to identify the exact sampling dates for each of the critical discourse moments, I conducted multiple exploratory searches, which allowed me to narrow down the dates to those that contained the most extensive and varied coverage of each theoretically relevant moment.

The critical discourse moments that were included in the sample are:

1. The passage of a new integration benefit scheme that entailed limited access and reduction of benefits for refugees, and immediately thereafter, the beginning of the Syrian refugees crisis in Denmark (2–13 September 2015).

2. Coverage of the New Year’s Eve events in Cologne, in which women reported sexual violence and theft apparently by young men of North African origin (7–12 January 2016).²

3. The passage of the controversial Jewellery Law, which notoriously sanctioned the police to confiscate refugees’ valuables, including jewellery, in order to offset the state’s expenses on refugee care. In addition, this legislation included benefit reductions and various restrictions on refugees’ residence and family reunification rights (27 January–2 February 2016).

4. The passage of the expansion of criteria in the reduced integration benefit scheme, to include Danish citizens as well as foreign nationals (28 June–4 July 2016).

5. Approval of the Gang Laws Package (Bandepakke III) (24–30 March 2017).

6. Presentation of the Ghetto Package (Ghettopakke) to the public and the beginning of political negotiations (2–8 March 2018).

7. The passage of the finance act agreement, which included a ‘paradigm shift’ in the area of immigration. The shift entailed greater emphasis on repatriation rather than the integration of refugees (29 November–5 December 2018).

To construct a representative sample, I selected the two most read broadsheet newspapers (Politiken and Jyllands-Posten), the most visited online news site (Danish Radio) and one of the most read tabloids (Ekstra Bladet). The sampled news sources vary in their format, content and values. While the broadsheet and public news source emphasise traditional journalistic values, such as objectivity and fairness, the tabloid newspaper focuses on sensational stories (Skovsgaard 2014; Hansen 2016). The different newspapers are aligned with a variety of political positions, thus allowing us to capture a fuller image of the emerging news media discourse on immigration. Politiken attracts more left-leaning readers and Jyllands-Posten attracts more right-wing readers (Hjarvard 2007). Not surprisingly, their editorial lines similarly align to the left (Politiken) or the right (Jyllands-Posten) (Hjarvard 2007). Ekstra Bladet’s readership is predominantly low-income (Ekstra Bladet 2016) and is the most politically diverse among Danish tabloid readers (Hjarvard 2007). In addition, I sampled the online textual coverage of Danish Radio (DR), the oldest and largest electronic media platform in Denmark. This news source is public and, as such, it is the only Danish news source that is freely available online. Consequently, it is the most accessed news source in Denmark with 37% of the population visiting it to obtain news (Schrøder, Blach-Ørsten, and Eberholst 2019; Farkas and Neumayer 2020). Politically, its editorial line is considered balanced with fair representation of both political camps (Albæk, Hopmann, and de Vreese 2010).

Within these parameters – selected critical discourse moments and selected data sources – I searched on Infomedia, an online archive, for articles that discuss the topic of immigration in Denmark, using inflections of the Danish words for ‘immigration’, ‘migrants’ and ‘refugees’ in the search phrase. Because of the extremely high volume of articles, I limited the search to front-page coverage and opinion pieces in print newspapers, thus focusing on the most influential agenda-setting coverage. To limit the number of articles from DR, I chose to analyse the articles that are at least 450 words, thus focusing on the most substantial and in-depth coverage.³ After manually cleaning the data and excluding irrelevant articles, the search process yielded a total of 535 articles. The majority of the articles from print media were opinion pieces, and only 10% were front-page articles (of them, some were news articles and some were front-page opinion pieces). Since the different years were characterised by varying degrees of coverage volume, the final results in this paper are presented as a percentage of the overall articles that were analysed (either in total or for each year).

Political speeches

In addition to news media sources, I included all 12 prime ministers’ speeches that were given ritually at the annual opening of the parliament (October) and for the occasion of the New Year, from January 2015 until January 2020. I selected these speeches for analysis because of their agenda-setting function; these ceremonial speeches are effective at highlighting the priorities of the government, and setting the political tone for contemplating its previous and forthcoming legislative initiatives. Since there were three election cycles within this time frame, the speeches in the sample were delivered by three different prime ministers: Helle Thorning-Schmidt from the Social Democrats (SD); Lars Løkke Rasmussen from Venstre (a center-right party); and Mette Frederiksen from SD. The speeches were all transcribed and translated into English by the Danish Ministry of State and are available online at its website.

Analysis of data

The first stage of the analysis was to read a selection of the texts and identify the main theoretically relevant themes in the immigration discourse. I then constructed a codebook that contained the different codes, a clear explanation of their meanings and application, and examples from the dataset. The collection and coding of news data were conducted by the author and a Danish native-speaking research assistant. To achieve reliability, we coded the first dozens of articles in parallel and discussed our coding selections, until we reached a comprehensive understanding of the application of the codes. In subsequent reliability checks we achieved high levels of coding similarity. The prime ministers’ speeches were coded by the article’s author alone, using the same codebook and the same procedures. The third step included an in-depth CDA analysis of all the articles and speeches with relevant codes. I performed the analysis using the qualitative software package Nvivo 12 Plus. All quotes from news articles are my translation from Danish; all quotes from prime ministers’ speeches are based on the state’s official translations of the speeches.

In the next section, I analyse media and political discourses around the topics of immigration, welfare and criminality to illustrate the ways in which legislative initiatives in these fields were mobilised and justified in the media and in politics.

Findings

The Danish media discourse of immigration was characterised by a stark duality. On the one hand, media articles provided overall slightly greater coverage to pro-immigrant frames: 57% of all the articles in the sample contained a positive frame regarding migration or at least a sense of obligation to accepting immigrants. Still, it contained a large share of anti-immigrant frames (50% of all the articles).⁴ Note that some articles presented opposing arguments without taking a clear stand about their validity; hence, the overlap in numbers. In contrast to the more balanced media discourse on immigration, the political discourse was much harsher in its judgment and tone. Virtually all of the political speeches included multiple negative references to immigration, framing migration as a driver of social and cultural harm.

Considering temporality in media discourse, we can see that the share of positive and negative coverage of immigration flipped with time. If, in 2015, 64% of the coverage included positive references to immigration and 44% of the coverage was negative, by 2018 the trend had reversed, with 41% of all articles including positive references and 58% of them being negative. It is possible that after the initial moral shock elicited by the major wave of war-stricken Syrian refugees in 2015, the media discourse of immigration shifted and became more negative (Greussing and Boomgaarden 2017 find a similar pattern).

The overwhelming majority of the media articles that included any positive framing of immigration were inward looking, focusing on the Danish character as a benevolent nation that is doing well to accept refugees and immigrants. However, the theme of Denmark’s benevolence was also used – by politicians and media alike – to justify the rejection of immigrants and refugees. This argument often posited that Denmark has done above and beyond for refugees, and that it needs to ensure its own national integrity and prosperity as well. Often, this rhetorical tool was used to justify the harshest anti-immigrant policies in media debates.

We can find another instance of such ambivalence in the political discourse on immigration. Indeed, 36% of the speeches included a reference to immigrants as capable of integrating well and called to invest more in integration efforts. Yet, this endorsement of immigrants was always made in the context of a broader argument that constructs the ‘good’ immigrant, who is willing and able to integrate, in sharp contrast to the ‘bad’ immigrant, who fails to do so. In that way, any such positive reference to immigrants’ potential to integrate was used as an opportunity to reprimand ‘the other’ immigrants, who fail to integrate and positively contribute to society.

The idea of immigrants as a threatening ‘other’ in the media has been prevalent throughout the entire period of this study and increased over time. In 2015, 37% of the articles included the notion of immigrants as a threat; by 2018, this narrative was present in half of all articles (51%). This finding is in line with the results of multiple other media studies (Eberl et al. 2018). The threatening spectre of immigrants was heavily filtered through the lens of welfare protectionism and crime prevention. Below, I will examine the relative salience of – and interconnectivity between – narratives that portray immigration as a threat to the integrity of welfare systems, and immigration as a threat to public safety and law enforcement.

Constructing immigrants as a threat: crime prevention and welfare anxiety

When looking more closely at the construction of immigration as a threat, the data shows that the fear of immigrants unduly overwhelming the welfare system (welfare anxiety) has been much stronger and more consistent than the notion of immigrants as potential criminals (interestingly, very few articles or speeches made any connection between immigrants and a threat of terrorism). While 17% of all articles in the sample contained the notions of immigrants as threatening the integrity of the welfare system, only 7% of the articles included references to immigrants as posing risk to law and order. Likewise, in prime ministers’ speeches, the focus on the need to protect welfare in the face of immigration was a constant feature, much more dominant than the theme of criminal threat, and that, too, has intensified since October 2017.

From a temporal perspective, the evolution of media coverage of immigration in relation to welfare anxiety and crime concerns is striking. Throughout the entire duration of the study, welfare anxiety was consistently stronger than concerns for personal safety and crime control in media coverage. However, both themes significantly increased over time. In 2015, at the height of the refugee reception crisis, these negative themes of welfare anxiety (appearing in 10% of the 2015 coverage) and criminal threat (in merely 1% of the 2015 coverage) were rather weak. As the acute stage of the Syrian refugee crisis ended, and with it the intensive coverage of refugees’ distress ceased, the theme of welfare anxiety increased to 21% of all media coverage in 2018, and the theme of criminal threat increased to 15% of the 2018 media coverage.

Migration as a threat to the solvency of the welfare state

The theme of welfare anxiety, a dominant concern in media and political discourses, was composed of two distinct, if interrelated, discursive elements: (i) constructing the topic as a matter of fairness in welfare distribution and (ii) constructing welfare as a zero-sum game. Both discursive elements can make sense only if we remember that the understanding of welfare in the Nordic context is universal and encompasses many more state services than in liberal regimes.

The first discursive element – of fairness – built on the notion that immigrants take more than they contribute to the welfare state, and that this is simply unfair, or even dishonest. Drawing on, and constructing, an idea of social otherness, this discursive element often suggested that because of immigrants’ cultural or social background, they do not engage in full and honest employment, choosing instead to rely on welfare benefits. We can see this notion expressed, for example, in a 2016 commentary: ‘Even if you believe that we are good at getting people into jobs and education, it does not change the fact that refugees, immigrants and foreigners, in general, pull more out of Danish society than they contribute financially’ (Politiken, 30 January 2016). Another writer lamented the ‘reluctance on the part of social workers and municipalities to actually close the coffers for those who show no willingness to contribute’ (Jyllands-Posten, 2 March 2018), and another columnist bemoaned: ‘With our trust, we are almost powerless in the face of masses of immigrants who perceive our welfare services of various kinds as a buffet table. Why order something when help comes all by itself, absolutely unconditionally?’ (Jyllands-Posten, 2 July 2016). This discursive element, then, justified reduction of immigrants’ welfare benefits and eligibility criteria in order to restore the balance in a lopsided welfare system where immigrants take more than they give.

The second discursive element of welfare anxiety constructed welfare as a zero-sum game, in which immigrants who use welfare services are directly taking away from services that would otherwise benefit the ethnic majority population. In an op-ed in Politiken, Mette Frederiksen, at that point still in the opposition, explained:

I want to fight for our model of society. I will not leave other vulnerable groups in the lurch due to lack of integration. … And I want to be sure that I can fund our welfare society in the future. Otherwise, I will fail in my task as chairperson of the Social Democrats. (Politiken 30 January 2016)

Similarly, Henrik Sass Larsen, also a politician with SD, said that ‘we do not want to sacrifice the welfare state in the name of humanism’. He warned that Denmark must protect the underprivileged in Danish society as it is they who reside in neighbourhoods with the largest concentration of refugees, where welfare services (broadly defined to include day care, public housing and unions) are stretched thin. ‘This means’, he concluded, ‘that those who already are under hard strain will find themselves in an even harder situation’ (Danish Radio 2 February 2016). These and other speakers posited that accepting many immigrants and refugees, and extending them the same level of welfare support, would directly and immediately hurt the level of services that the rest of the Danish population would receive.

Importantly, this particular argument was not limited to welfare services for the lower classes alone. Writers and politicians spoke about the expenses and resources that the state has to devote to the management of immigrants as being directly (and quite literally) at the expense of the majority population of all economic backgrounds. When talking about welfare services, speakers in the media and political discourses emphasised institutions such as public hospitals, cancer treatment, childcare, elderly care, public infrastructure and care for the homeless and disabled – institutions that benefit Danes of all social-economic backgrounds. PM Løkke expressed this sentiment in a speech given in the context of the Ghetto Package debate: ‘If more people provide for themselves, then we will ease the pressure on our welfare system. And we need the money. For safe and secure eldercare. Better cancer treatment. More research. Better infrastructure’ (PM Løkke, January 2018 speech). Here, the prime minister made a direct relationship between funding for minorities’ welfare and the ability to fund infrastructure and resources that benefit the general Danish society. In a similar manner, he presented the resources that the police have to devote to areas with a large concentration of minority populations as directly taking away resources from ethnic majority-neighbourhoods: ‘If we eliminate the hotbeds for gangs, the police can make an effort in other places, and create safety and security in the streets and in everyday life’ (PM Løkke, January 2018 speech). This last point leads us to the second topic that shaped the public discourse on immigration: namely, crime control and personal safety.

Immigration and the threat of criminality

Despite several legislative initiatives where immigration served as a backdrop for penal policy reforms, the discourse regarding immigrants’ alleged criminal behaviour was not prominent in the news and political discourses. In media debates, the topic of criminality was marginal in comparison with the discussion of welfare protectionism, as we saw earlier in the paper. What is most striking is that a large share of the articles that did mention the threat of criminality simultaneously also referenced the threat to welfare integrity. This combination of welfare anxiety together with public safety anxiety appeared in 36% of all the articles that contained any reference to immigrants’ criminality. In comparison, this overlap constitutes only 14% of those articles that discussed welfare anxiety. For example, in early 2016, a columnist had warned: ‘If in doubt, look out the window, visit a ghetto … or look at crime statistics, especially the dangerous ones, dominated by migrants and … youths, high on machismo and social welfare’ (Jyllands-Posten, 29 January 2016). This quote is an example of a common rhetoric strategy that folded together the stereotypes of immigrants as threats to public safety and to welfare integrity. Another newspaper commentary made this point in saying: ‘“persons of ethnic origin other than Danish” who have obtained Danish citizenship are to a much greater extent … dependent on public support, are more criminal and less likely to support universal liberties’. Thus, he concluded:

Instead of letting criminals, the incapacitated and welfare tourists stay in the country, we here in Denmark should rather let ourselves be inspired by more well-functioning countries like Canada, Australia and Switzerland, where high demands are clearly placed on newcomers. (Jyllands-Posten, 11 September 2015)

Similarly, a year later, a publicist asserted:

I have nothing against asylum seekers or refugees, but asylum means protection or refuge from persecution due to race, religion, nationality, minority status and politics – not public support for the rest of my life with ‘villa, Volvo and doggy (vovse)’, or the practice of crime. The old Danish proverb: ‘Follow costume or leave the country’ is still valid. (Jyllands-Posten, 1 February 2016)

This last comment is a clear illustration of the fairness and (dis)honesty discursive pattern of welfare anxiety intertwined with the risk of immigrants being criminals in disguise, abusing Denmark’s generosity and good will. This data shows that media discourse on immigration tended to mix welfare anxiety with concerns for personal safety as a way to justify restrictive immigration, welfare and criminal justice policies.

We can see similar patterns in political discourse. The topic of law and order and the promise to protect citizens from criminal behaviour, historically dominant in political discourse in neoliberal regimes such as in the US, was mostly absent from prime ministers’ speeches in Denmark. It did not feature in any prime minister’s speech until October 2017, several months following the passage of the Gang Laws Package. It is noteworthy that this topic was absent during the height of the media and political debate on the refugee crisis (2015–2016). The relative marginality of the issue of crime control in the context of immigration discourse was also evident in the language and terminology that prime ministers employed in their speeches. While welfare protectionism was a powerful rallying cry awash with emotional and strong words, speakers who addressed issues of crime control used a much more nuanced and subdued language. Such was the case, for example, in speeches by PM Løkke, who used tough and domineering terminology when he talked about ‘tearing down’, ‘demolishing’, ‘punishing the parents’ and ‘taking full control’ to address alleged welfare abuse by immigrants. In contrast, the action that Løkke professed to take against criminals was much less graphically and dramatically put; he simply stated that the government shall hold criminals accountable (PM Løkke, October 2017 speech), and said only that ‘hard actions against criminals’ shall be taken (PM Løkke, October 2018 speech).

Interestingly, in the speech delivered in October 2017, PM Løkke, a member of a center-right party, tied criminality to marginality and stressed the need to address social issues that are at the core of criminal behaviour. In addition to promising to be ‘hard on crime’, he professed to take proactive welfare-oriented preventive measures, starting at a young age, in order to effectively curb criminality. In that sense, the discussion in Denmark is an example of Nordic Exceptionalism (Pratt 2008).

However, the expressed relationship between welfare and criminality went beyond understanding the root causes of criminal behaviour, to constructing a discourse that pins criminal behaviour and welfare dependency on the behaviour of non-Western immigrants and their descendants. This active connection between the three discursive themes was particularly dominant during the discussion of the aforementioned Ghetto Package. For example, PM Løkke tied minorities’ criminal behaviour with welfare dependency and cultural self-isolation in the ‘ghettos’ when he said: ‘Throughout the country, cracks have appeared in the map of Denmark. … It is a matter of a counter-culture that has developed from an environment where far too many live on transfer income … where more are criminals’ (PM Løkke, October 2017 speech). Likewise, PM Frederiksen tied together welfare – broadly defined – and criminal behaviour in stating that ‘failed integration has left traces in schools, in residential areas, in workplaces. And perhaps most evidently in crime statistics … Far too many of those who have come here, commit crimes. Or their sons do’ (PM Frederiksen, October 2019 speech). This statement suggests that immigration takes a toll on the ability of the welfare state to deliver high-quality social services and ensure the personal safety of its residents.

To conclude, the empirical data from both media and political debates shows that immigrants’ criminality was rarely a concern that stood on its own. In fact, initially it occupied only a fraction of the public discourse on immigration, and later it was most commonly discussed in connection with other social, economic and cultural anxieties around immigration and diversity. In contrast, the concern that welfare institutions will suffer as a result of increased immigration was a powerful independent anxiety that mobilised and justified anti-immigrant legislation in the fields of both welfare and criminal justice. Counter-intuitively, these findings suggest that it is mostly welfare protectionism that justified harsher penalties for crimes committed by immigrants, and not – as we would expect – a discourse that mobilises personal safety concerns.

Discussion

This article extends the debate in the literatures of welfare, criminal justice and immigration by tracking the active and dynamic process of constructing immigrants as a threat in the context of welfare and crime control. It provides the necessary empirical data to understand the mechanisms through which penal and welfare dualisation takes place and demonstrates the centrality of welfare anxieties in driving and justifying anti-immigrant policies in public discourses. This article contributes to these literatures by broadening the lens of investigation to include all three interrelated institutions, debates and policies. By expanding their investigative lens, both criminologists and welfare researchers are poised to gain a fuller and more nuanced understanding of welfare states’ responses to immigration.

Wacquant (2009) positions welfare and criminal justice regimes as complimentary institutions charged with the management of the poor and marginalised populations, where marginalised populations are subject to various degrees of care and control (Brydolf-Horwitz and Beckett 2021). The Ghetto Package is an apt example of this kind of continuum: it is a legislative initiative encompassing an assortment of policy amendments that have an impact primarily on low-income minority populations through criminal justice reforms (e.g. longer sentences for crimes committed in the ‘ghettos’) and welfare reforms (e.g. mandatory childcare for toddlers of the target group). Both types of reform include significant amounts of investment on rehabilitation and service provision for marginalised populations, yet both are also characterised by high levels of coercion, surveillance and control.

Critically, as this study shows, both types of intervention – through welfare and criminal justice institutions – have been justified with a discourse that portrays immigrants and minority populations as a threat. However, unlike in the US, where law and order have been a central theme in justifying harsher social control (Beckett and Francis 2020); here, we find that the focus has been primarily on the social threat that immigrants allegedly pose to the welfare state. Instead of focusing on immigrants as posing a threat to individual safety, both media and political discourses have focused on immigrants as posing a threat to the integrity of social institutions. Thus, the personal safety frame is foreshadowed by a much stronger social stability frame that centres on the solvency of welfare institutions.

Moreover, the article shows that it is important to understand and analyse the culturally specific definition of welfare. While welfare in the US refers only to remedial programs that serve the poor and disadvantaged and is often administered based on means-tested criteria, the data shows that in the Danish context welfare is understood in a much broader sense. Welfare is not solely an arm of the state that supports the poor and marginalised – although it certainly does that, too. Rather, welfare is all the public institutions and programs that the state administers to improve the wellbeing of all its citizens, regardless of their economic background. This broad understanding of welfare helps to explain the strong support it enjoys in the Nordic states. It also helps to explain why it has become an effective foil for the construction of immigrants as a threat. In a social-democratic regime, where welfare is a universal provider of services, welfare institutions and welfare spending enjoy tremendous support. Portraying immigrants as a potential out-group that unduly burdens this central institute at the expense of the majority group enjoys a strong resonance and can then be easily translated to a call to fortify the welfare state – rather than to dismantle it – against the out-group population. This can explain some of the mechanisms that are at the heart of the ‘welfare chauvinism’ theory. It also explains why this argument’s resonance was so strong that it countervailed the personal safety argument, even in cases when the media and politicians were discussing immigrants’ alleged criminality.

Indeed, both media and political discourses in Denmark conceptualised welfare as a zero-sum game that requires everyone’s equal and honest contribution. According to this dominant discursive strategy, if one group – namely, immigrants – do not pull their equal weight or are in need of additional resources because of hardship or lack of skills, they inevitably hurt native Danes, who would consequently receive lower standards of services. This idea of fairness in participation and a zero-sum game attitude towards welfare was often extended to and blended with the discourse on immigrants’ alleged criminality. Crime control and personal safety were secondary concerns that were often intertwined with welfare concerns and did not stand on their own. This is a key finding since the perception of immigrants as a threat to the social-economic system has been shown to lead to increased support for policies that limit immigrants’ access to welfare benefits (Hercowitz-Amir and Raijman 2020).

Given the discursive focus on immigrants’ threat to welfare integrity, there is little surprise that a process of dualisation of the welfare state occurs. What is more surprising, however, is the fact that although the legislator passed multiple policies that produce a de facto two-tier penal system, the presence of the stereotype of immigrants as criminals was not as widespread as we would expect it to be. Therefore, to the extent that we find a dualisation of the welfare system and a nationalisation of criminal justice policies, both policy developments were driven by an immigration discourse that views immigrants as threatening welfare first, and only secondarily as threatening personal safety. This speaks to the centrality of welfare protectionism in the public discourses of welfare states, and to the central role it plays in shaping public opinion and policy initiatives regarding immigration.

Acknowledgements

The author wishes to thank Jakob Johan Demant, Malgorzata Kurjanska and two anonymous reviewers for their helpful comments in the process of writing this article. The author wishes to extend special thanks to Josefine Dahl Hansen for her excellent research assistanship.

Disclosure statement

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

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement [grant number 793650].

Notes

1 Immigrants alone, without their Danish-born descendants, are only 6% of the population.

2 What may appear to be strangely delayed sampling dates is because the German national media did not treat it as a newsworthy story until 4 January 2016 (Boulila and Carri 2017).

3 The search phrase for all the newspapers was: ((heading:debat* OR heading:kronik OR heading:‘ledende artikel’) OR pagenumbers:1) AND (*migra* OR indvandr* OR flygtning*) AND ('danmark' OR dansk*). The search phrase for DR was: (*migra* OR indvandr* OR flygtning*) AND ('danmark' OR dansk*).

4 Positive frames included the themes: immigrants are beneficial for society; Denmark has a moral or legal obligation to accept immigrants; Danes are charitable and welcoming of immigrants; integration is possible. Negative frames included the themes: immigrants are a threat (including: threat to the economy by taking away jobs from Danish citizens or overwhelming the welfare system, criminal threat, terrorist threat, threat to Danish culture and values and a threat to gender equality); immigrants behave dishonestly; and immigrants are hard to integrate.