Constructing “violence-affirming extremism”: a Swedish social problem trajectory

ABSTRACT

Violent extremism has internationally become an established and highly prioritised social problem. While there is extensive and ongoing international diffusion and coordination, there are national specificities in terms of meanings conferred to this social problem and to the processes of its establishment in the public sphere. In this article, the trajectory of the Swedish notion of “violence-affirming extremism” is studied through an analysis of the coverage of “extremism” in national media from 1995 to 2019. It is argued that the introduction of the term “violence-affirming extremism” constituted a novel social problem frame established and institutionalised during the period 2008–2019. This process was driven to a large extent by terrorist events domestically and abroad, instilling a sense of drama and urgency. Whereas the new frame merged Islamist, left-wing and right-wing extremism as part of the same phenomenon, there were also differences in terms of perceived causes and solutions. Islamist extremism received the majority of mentions and was associated with the highest proportion of control-oriented solutions, such as legislation, policing and surveillance. These types of solutions remained salient throughout the latter part of the period, while other solution types appeared to lose momentum as the problem became institutionalised.

KEYWORDS:

• Crime prevention
• framing
• radicalisation
• social problems
• Sweden
• violent extremism

Introduction

Ever since the 9/11 attacks in 2001, the world has become used to regarding terrorism as an urgent and prioritised social problem. However, the ways in which it has been understood and responded to have undergone significant changes during the subsequent two decades. The initially largely reactive and coercive responses to terrorism after 9/11, launched under the broad label “War on Terror”, were soon complemented by the articulation of a more preventive approach, initially developed in Europe (Harris-Hogan, Barrelle, and Zammit 2016). This approach, often referred to as prevention of violent extremism (PVE), was closely linked to shifts in predominant ways of understanding terrorism and political violence, in turn connected to the emerging centrality of the concepts radicalisation and violent extremism (Baker-Beall 2016).

Of course, European states also historically sought to control “extremists”, “terrorists”, and “revolutionaries” through various forms of repressive measures (e.g., Della Porta 1995). On the international level, there has been official European collaboration on counterterrorism since 1976, with the establishment of the platform TREVI (Terrorisme, Radicalisme, Extrémisme et Violence Internationale) (Coolsaet 2010). Harris-Hogan, Barrelle, and Zammit (2016) identify a shift starting in the UK with the introduction of the programme Contest in 2003. The terrorist bombings in Madrid 2004 and in London 2005 also fuelled a sense that terrorism was not only a threat from the outside but also partly “home-grown” in European states. Related to this notion was the idea that citizens “at risk” of becoming terrorists could be diverted from their path towards committing crimes of political violence, and that this could be achieved through various preventive interventions from domestic authorities towards people not yet involved in politically violent groups and activities (Heath-Kelly 2012; Jämte and Ellefsen 2020b; Martin 2014; Zedner and Ashworth 2019). Whereas the new PVE policies and interventions typically acknowledge different types of extremist motivations for political violence, several scholars have noted that authorities tend to zoom in on Muslim communities (e.g., Baker-Beall, Heath-Kelly, and Jarvis 2015). The increased focus on proactivity has also led to the engagement of public officials (such as social workers and teachers) and organisations with hitherto none or very little involvement in preventing political violence (Ellefsen 2021).

One might regard these shifts in concepts and ideas about terrorism and violent extremism as directly reflecting the appearance of a new social problem “out there”, requiring novel approaches. However, although objective conditions – such as high-profile terrorist events – may contribute to the public salience of a problem and how it becomes perceived (Andersson Malmros 2021a), such conditions do not by themselves constitute violent extremism qua social problem. As argued in classical contributions by Blumer (1971) and Spector and Kitsuse (1987), more recently summarised by e.g. Best (2017), social problems are what become defined as such in interactions between different social actors. The ideas that come to dominate in society about a social problem’s character, causes, and solutions, in turn influence how various actors respond to it. Critical analyses of the processes by which social problems become established serve to historicise taken-for-granted assumptions about their character and to trace the trajectories of their development (Ungar 1998).

While it is important to study broad international trends in counter-terrorism strategies, the national and local levels – with their varying political cultures and institutional arrangements – remain important arenas for social problem articulation and subsequent implementation of prevention strategies (Andersson Malmros 2019). The aim of this study is therefore to focus on one national arena and analyse how the contemporary understanding of violent extremism became established as a social problem, by tracing the associated problem definitions over time. The study investigates, firstly, the process of how a novel approach to politically motivated violence enters a national arena and the key dynamics of its development. Secondly, it seeks to unpack the content of the emerging social problem definitions by identifying the types of actors included in the definition and elucidating the extent to which they are discussed differently in terms of causes and proposed solutions. This serves to problematise what categories of people are considered potentially dangerous, and who are thereby more likely to become targets of proactive interventions.

The case selected for analysis is Sweden – one of the countries that early on expressed engagement in the European Union (EU) fight against radicalisation (Coolsaet 2010). At the same time, Sweden is distinguished by its Scandinavian welfare model, strong protection of civil liberties and tradition of a political culture of compromise, cooperation, and dialogue (Peterson, Thörn, and Wahlström 2018), which makes the case a useful reference point internationally. For methodological reasons, Sweden is also a practical case, since (as will be elaborated below) the new approach to violent extremism and radicalisation was accompanied by a new and previously rarely used concept: “violence-affirming extremism” (Sw. “våldsbejakande extremism”).¹ This makes the process easier to trace. The focal period of study is from 2008 to 2019, but I will contextualise this period in relation to prior social problem discourse.

The main empirical source is newspaper articles, which provide information on how social problems are negotiated in the public arena and indirectly in other arenas (such as parliamentary discussions, among public authorities, and in academia). Influential claims-makers can be identified both in news reporting and as authors of opinion pieces. Mass media also contributes to shaping the problem through the ways in which events are reported and to setting the agenda for politicians, especially during crises and during the establishment of new policy fields (Van Aelst et al. 2014; Koch-Baumgarten and Voltmer 2010). That mass media not only report on other actors’ problem definitions, but also influence policy, actually underlines the suitability of newspaper data for identifying developments of a social problem over time.

The next section will discuss previous research within the field of violent extremism, radicalisation, and PVE. After outlining the theoretical starting points of the study and methodological considerations, I will turn to a brief history of the contemporary wave of claims-making associated with violence-affirming extremism (hereafter VAE), after which follows a more detailed analysis of the dominant frames in the claims-making activities.

Previous research and the growth of a scientific sub-field

Broadly speaking, the classification of political dissent – violent or not – as extremism is not new (Backes 2009). However, there is a clear tendency of increased use of this category to understand and control political and religious militancy. This tendency is also visible in the academic literature. The first research article registered in the database Web of Science on “extremism” was published in 1959, and the first relative peak occurred in 1968 with 12 articles. Before 2000, articles on “extremism” were less than 30 per year, and after a subsequent slow rise the numbers grew sharply to 298 in 2017 and 537 in 2021. The concept of “violent extremism” only occurs as an article topic in the database once in the 1990s and only 6 of the total number of 682 publications by the end of 2021 were published before 2010.

“Radicalisation” is another notable member of this conceptual family that has received an enormous increase in research attention in the last decades (Malthaner 2017). Whereas the concept had previously mainly been associated primarily with ideological “radicalism”, its current association with political violence and (home-grown) terrorism is traced by Coolsaet (2010) to the Dutch intelligence service AVID, after the murders of Pim Fortuyn in 2002 and Theo van Gogh in 2004. Following the London bombings in 2005, the European Commission officially established countering “violent radicalisation” among the prime goals of the EU counter-terrorism policy. Later the same year, the Council of the European Union (2005) presented countering radicalisation and recruitment to terrorism as central goals of the “prevent” aspect of the EU Counter-Terrorism strategy. Violent radicalisation subsequently became a topic of several scholarly reports – both independent and linked to the EU – which connected the notion especially to “jihadist terrorism” (e.g., Alonso et al. 2008; Neumann and Rogers 2007; Ranstorp 2010).

In addition to the many studies outlining explanations and suggesting prevention strategies for violent extremism, violent radicalisation, and terrorism, there is also a growing critical literature. While parts of this literature on the one hand critically evaluate and indicate detrimental effects of specific prevention programmes (e.g., Lindekilde 2012), other parts critically analyse policy and the use of core concepts such as “radicalisation” and “extremism” (Baker-Beall 2014; Sedgwick 2010; Backes 2009). Hörnqvist and Flyghed (2012) point to two competing discourses underlying the concept of radicalisation: one of cultural conflict and one of terrorism in response to social exclusion. According to Kundnani (2012; cf. Silva 2018), the concept of radicalisation has been strongly associated with research that downplays the political context of terrorism and directs focus at theological, individual, and community factors that can purportedly be pre-emptively spotted and acted on by authorities.

Heath-Kelly (2012) similarly analyses the governing logics that the concept of radicalisation is linked to in the UK context and points to the construction of (and tension between) “at risk” and “risky” subjects: individuals who are regarded as vulnerable to radicalisation efforts, blending over into those who become threats to society. Martin (2014) builds on this analysis and highlights how counter-radicalisation work in the UK has become a governing of the unknowable, directed not at actual threats but at individuals and groups posing potential threats (which in each case has only a miniscule risk of realisation). He argues that the introduction of the term “violent extremism” implies a broadened targeting of the governing efforts: “the argument is now that extremist groups, even those committed to non-violence, theoretically and theologically provide legitimation for ideas which may be used as justifications for violence” (p. 68). Onursal and Kirkpatrick (2021) show how the labels “terrorism” and “extremism” increasingly converge from 2010 to 2017, while non-violent forms of extremism became discursively established as precedents of involvement in terrorism. However, with the exception of the latter study, few go into detail regarding the discursive process of how the (new meanings of) the concepts become established.

Several studies acknowledge the importance of mass media representations for shaping dominant understandings of terrorism, radicalisation and extremism in society (e.g. Berbers et al. 2015; Figoureux and Van Gorp 2020). A striking pattern identified in this body of research is that Islam and Muslims tend to be disproportionately regarded as associated with terrorism, whereas incidents with perpetrators associated with, e.g., the extreme right appear to receive less attention and to be less likely to be characterised as terrorism (Kearns, Betus, and Lemieux 2019; Powell 2018; Zdjelar and Davies 2021). However, importantly, characterisations of extremism/radicalisation, and people involved in it, are under ongoing negotiation and appear to vary depending on the actor quoted (Berbers et al. 2015; Figoureux and Van Gorp 2020).

Turning to Sweden, the specific national context of this study, there are a growing number of publications focusing specifically on Swedish PVE policies and their implementation. Jämte and Ellefsen (2020b) identify three distinguishing characteristics of the contemporary Swedish PVE arena: (1) interventions against individuals based on a pre-emptive logic, (2) an outsourcing of social control to agencies not normally involved in control of political activism, and (3) a pluralisation of social control, notably involving various forms of “soft repression” (Jämte and Ellefsen 2020a). Similar to research findings from the UK, Andersson (2018) argues that the Swedish discussion on terror has moved focus from the politics of violent acts to the characteristics of individual perpetrators.

There are a good number of studies of local expressions of Swedish (and Scandinavian) PVE practices. In a systematic analysis of Scandinavian municipal action plans against violent extremism, Andersson Malmros (2019) identifies a strikingly diverse number of different interpretations of radicalisation and policies often weakly grounded in local conditions. Mattsson (2018) scrutinises Swedish anti-radicalisation policies in relation to primary education, identifying a discrepancy between public actors’ perception of social and structural causes of VAE and the largely individual focus of the interventions. In a similar vein, Johansson and Arvidsson (20166) identify models for understanding and dealing with violent extremism among teachers, social workers, police and politicians. They distinguish an individualised model that labels violent extremists as villains to be met with repression and a more structural model that regards the same group as victims to be met with social support. In a comparison of public officials’ perspectives on Islamist, right-wing, and left-wing VAE, Jämte and Ellefsen (2020b) note how local public officials tend to perceive Islamists as representing the highest level of political violence while violent right-wing extremists are typically perceived as more prevalent than the other groups. Overall, extant research points to an overall tendency that PVE policies and interventions lack a clear demarcation line between lawful ideas and activities and those that are linked to violence.

Despite the burgeoning literature on PVE policies and practices, there is clearly a need to improve our understanding of the processes by which the contemporary notion of violent extremism – in Sweden, VAE – became established as a social problem. A recent contribution relating to this issue is Andersson Malmros’s (2021a) institutionalist analysis of the development of prevention of terrorism, extremism and radicalisation as an issue field 1999–2019. He especially highlights the role of critical events in this process and distinguishes between redefining events, supportive events and contentious events, depending on their impact on the issue field. However, Andersson Malmros’ study also shows the need to further elaborate on how definitions, policies, strategies and dominant explanations developed during the most intense period of social problem formation of VAE.

Theoretical approach: the framing of social problems

The present analysis departs from the tradition of constructionist perspectives on social problems associated with early criticisms of sociological approaches that take the social problems under study for granted. Herbert Blumer (1971) argued that:

a social problem exists primarily in terms of how it is defined and conceived in a society instead of being an objective condition with a definitive objective makeup. […] The societal definition gives the social problem its nature, lays out how it is to be approached, and shapes what is done about it. Alongside these decisive influences, the so-called objective existence or makeup of the social problem is very secondary indeed.

This perspective was developed by Spector and Kitsuse (1987) who likewise argued that social problems should be studied as “claims-making” activities of various actors who try to establish and/or modify what should generally be considered social problems. Some of the early constructivist work on social problems attempted to identify the discrete stages of their “natural history”, such as their (1) emergence and (2) legitimation, (3) the mobilisation of action around the problem, (4) the formulation of action plans and (5) the implementation of these plans (Blumer 1971; cf. Best 2017). Acknowledging that social problems definitions develop over time, but not independently of prior understandings and practices related to the problem, Ungar (1998) developed the notion of a ”social problem trajectory”. He argued that the directions in which a social problem is likely to develop in the future are limited, albeit not completely fixed, by it accumulated past.

Others have been interested in the arenas in which social problems are negotiated, established and dealt with, such as mass media, political bodies and public institutions. Hilgartner and Bosk (1988) argued that “the number of social problems is determined, not by the number of harmful or dangerous situations and conditions facing society, but by the carrying capacity of public institutions” (p. 61). Given that this “carrying capacity” is limited, the social problem game at any given moment is a zero-sum game in the sense that if a new social problem becomes acknowledged, others will have to (at least temporarily) receive less attention. Among the factors that according to Hilgartner and Bosk determine the success of a social problem in an arena are: the arena’s organisational characteristics; the level of drama by which a problem is presented; the problem’s novelty and conversely the saturation of similar attention to the problem that eventually tends to lead to its decline; the cultural and political resonance of the problem; and the character of the feedback between different arenas.

To conceptualise how claims-makers interpret and present the character of a social problem they are concerned with, several studies within the field of social problems draw on the notion of framing (Best 2017; Loseke 2011). Various versions of this concept have been developed to analyse the meaning-making activities by, in particular, mass media and social movements (Benford and Snow 2000; Gamson 1992; Gamson and Modigliani 1989; in turn based on Goffman 1974). Benford and Snow (2000) made a widely used distinction between three core aspects of framing: (1) diagnostic framing – identifying the character and the causes of a problem, including who/what is regarded as responsible, (2) prognostic framing – the solutions to the problem, and (3) motivational framing – what it is about the problem that makes action necessary and urgent. The below analysis will mainly focus on the diagnostic and prognostic aspects of framing, whereas the topic of establishing and maintaining urgency will be brought up in the broader discussion.

The success of a collective action frame is typically regarded as hinging on its resonance, which Benford and Snow (2000) break down into two aspects: its credibility and salience. Whereas the former concerns the inner consistency of the frame, its empirical credibility and the personal credibility of its proponents, the salience aspect includes the degree to which the frame fits with the audiences’ personal experiences and its narrative fidelity, i.e. its resonance with dominant ideologies and myths in a culture. Different framings of an issue compete in framing contests in the public arena, as part of the broader social problem game. In their often strategic framing activities, actors in a social problem game may more or less successfully subject the frame to various processes: frame bridging – linking previously unconnected issues; frame amplification – increasing the resonance of a frame by invigorating an existing value of belief; frame extension – extending an existing frame to cover new phenomena or areas; and frame transformation – in other ways changing the content of a frame.

In terms of prognostic framing of VAE, the various solutions proposed can be analysed according to four overarching logics of crime prevention that Sahlin (2000) identifies by intersecting two overriding values – individuals’ welfare vs. social order – with the object of the prevention effort – individual behaviours or the social institutions that form preconditions for preventing crime. As illustrated in Figure 1, the combinations of values and objects result in the four logics: socialisation, control, changing structures/environment, and increasing efficacy of control institutions. These four logics can be used as a reference point to typologise different preventive approaches to VAE, and to draw parallels with broader trends in crime prevention.

Figure 1. Four logics of crime prevention (adopted from Sahlin 2000)

Research methods

To acquire an overview of the process in which the social problem of “violence-affirming extremism” takes shape, background information was taken from official documents, including notes from committees, parliamentary debates, policy documents and official statements from various institutions in relation to governmental reports. However, the analysis focuses qualitative and quantitative analysis of newspaper content in order to identify the development of social problem frames in the public sphere. As noted in the introduction, mass media representations of a problem are closely intertwined in policy making processes in complex ways; mass media provides information about key events and (public) positions of more powerful claims-makers while the influence of these events and claims on policy is enhanced by the media reporting itself (Koch-Baumgarten and Voltmer 2010). Conversely, dominant mass media frames lend legitimacy to some parties in framing contests in other arenas and contribute to developing public conceptions of the character of social problems (McCombs and Valenzuela 2021). Clearly, some processes in the establishment of a social problem, such as various negotiations and discussions outside the public realm, may not be accurately represented in mass media. However, newspaper coverage nevertheless provides an overview of positions and actions taken by several different actors over time, which is difficult to achieve with other kinds of material (Baumgarten and Grauel 2009).

Developments in mass media framing of “violence-affirming extremism” were studied through two samples of Swedish news articles, drawn from the database Retriever. First, all articles from the 25 years between 1995 and 2019 containing all derivatives of the word “extremism” (search term: “extremis*”) were sampled from the major Swedish newspapers Aftonbladet, Dagens Industri, Dagens Nyheter, Expressen, Göteborgsposten, and Svenska Dagbladet (henceforth S1). Part of the reason for choosing the starting point 1995 was pragmatic, since from this year onwards all the above newspapers were available in full text in the database. However, the point in time also coincides with what Lööw (2017) identifies as a period during which some of the characteristics of the contemporary “violent extremism” and “violence-affirming extremism” frames begin to develop. This sample contains 21,130 news articles and due to its size it was not manually coded but only subjected to automated text analysis.

In order to create a more manageable sample for a closer manual analysis of social problem framing, I also made two overlapping samples (S2 and S3) of articles only containing derivatives of the specific term “violence-affirming extremism” (search term: “våldsbejakande extremis*”). The term had its first printed news media appearance in 2008, which therefore marks the starting point for this sample. Since the use of the term was relatively infrequent before 2015 (see Figure 2), and since it was deemed particularly important to adequately capture the initial period of the public life of the term VAE, S2 included all newspaper articles from 2008 to 2014 in the Retriever database using this term, altogether 308 articles. Use of the term violence-affirming extremism exploded in 2015 and I therefore chose to create a sample – S3 – containing only the newspapers Aftonbladet (social democrat) and Dagens Nyheter (liberal) during the decade from 1 January 2010 to 31 December 2019 (neither Aftonbladet nor Dagens Nyheter mentioned the term before 2010). S3 contained in total 576 articles after excluding false positives.

The S2 and S3 samples were initially coded using a coding scheme containing a number of open questions primarily concerning diagnostic and prognostic aspects of the social problem frame, as well as what actors were cited in the articles. After an inductive analysis of the responses, the open questions for S3 were coded in a second round using a set of dummy variables based on the most salient frames identified in the prior inductive analysis. The coding work was shared by the author and a research assistant.

For the automated text analysis of S1, the articles of the sample were grouped in files containing 6 month segments and imported into the software Orange, which provides tools for word count analyses and collocation analysis. The coded S3 dataset was imported into the statistical software SPSS 25 for descriptive quantitative analysis.

The following analysis will begin with outlining the process of establishing the contemporary notion of VAE in Sweden. Subsequently, I will turn to an analysis of this development in mass media based on quantitative analysis of the material. Starting with the relative distribution between right-wing, left-wing and Islamist terrorism, the predominant diagnostic and prognostic social problem frames will then be analysed.

The social problem of “violence-affirming extremism” 2008-2019 and its prehistory

During the 20^th century, controlling and categorising threats to state security in Sweden involved a number of different concepts with slightly different connotations. This included “activities hostile to the state”, as well as “politically unreliable” and “politically encumbered” actors. According to Lööw (2017), Swedish policies in this area have been largely shaped by the work of the Committee Concerning Subversive Activities in the 1930s. Examples of new policies introduced at this time were laws against uniformed and paramilitary organisations. The dominant approach by the Swedish state has been to nevertheless allow organisations considered subversive but to put individuals and groups in this category under strict surveillance and to exclude them from mainstream politics. The 1930s Committee also identified the school as a potential arena for interventions.

However, during the main part of the 20^th century the task of controlling subversive activities in Sweden and political violence almost entirely belonged to the police and the military. Linked to peaks in extreme right and militant antifascist mobilisation in the 1990s, Lööw identifies the beginnings of a shift in perspectives leading to more socially preventive approaches. At this time, this was mainly visible in terms of exit support for activists of the extreme right and the problem came increasingly to be framed as a youth problem. Lööw argues that from the 1990s onwards there has been a tension between the (traditional) security perspective and one more oriented towards social prevention – in Sahlin’s (2000) terms, a shift from control-oriented approaches towards including a socialisation logic. Paradoxically, this came at a time when, according to Sahlin, control-oriented approaches to crime prevention in general were taking over from socialisation-oriented approaches.

By compiling the newspaper article texts from Sample 1, it is possible to build a timeline of the use of the key concepts in the contemporary Swedish media discourse related to extremism (see Figure 3). Whereas “extremism” and “terrorism” are both used with varying frequency already in the beginning of the studied period, rising in frequency in the 2000s, the terms “radicalisation” and “violence-affirming extremism” (VAE) are hardly used at all in Swedish mass media until the 2010s. The shift to using VAE might seem like a superficial change in relation to previous terms, including “violent extremism” (Sw. “våldsam extremism”), which did occur already in the 1990s (although much less frequently). However, the neologism is significant on several levels. First, the term marks the beginning of a period of dramatically increased prominence of political extremism and political violence as interrelated social problems. Second, the introduction of the term VAE in Sweden also coincides with influences from the contemporary European discourse about radicalisation and the search for means of individual-level prevention (cf. e.g., Coolsaet 2010; Kundnani 2012). Third, even though the definitions of VAE vary, it arguably denotes not only a property of individuals and groups actually using violence for political purposes but also the mere expression of positive attitudes towards such violence (see also Andersson 2018).

The first newspaper incidence of the term VAE in the material occurred in 2008 in the context of a report about a presentation by the researcher Magnus Ranstorp on preventing terrorism and radicalisation in Europe.² Ranstorp was at the time affiliated to the Swedish Defence University, and had a background at St Andrews University in Scotland, a node for the international expertise in terrorism studies. At this point in time, the established term within the security police and other authorities still appears to have been the term “violent extremism”, as indicated by the name of the report Violent Political Extremism (Korsell et al. 2009) (Sw. “Våldsam politisk extremism”) released in 2009, written by academics in collaboration with the security police and exclusively focusing on left- and right-wing violence. In the documented Swedish Parliamentary debates, the term VAE made an appearance 2009 in connection with a question concerning a scientific review – by Ranstorp and Santos (2009), – about prevention of VAE, which had been ordered by the Government.

The delivered review was a report only on the risks associated with Islamist extremism in a part of Sweden’s third-largest city, Malmö, and the Social democrat interpellator questioned its narrow focus. The introduction of the term VAE thereby coincided with the addition of Islamism to the two ideological categories previously associated with political violence in Sweden. In other words, while violence motivated by extremist ideas was not new as a social problem in Sweden, by also including Islamist violence the new term represented a frame extension (Benford and Snow 2000). The contemporary EU-level discourse on countering (especially jihadist) radicalisation (cf. Baker-Beall 2016) presumably influenced this frame extension through various channels, but most visibly through Ranstorp, who had himself participated in shaping this discourse as a member of the European Commission’s expert group on violent radicalisation (Alonso et al. 2008).

In 2010, the prominence of the VAE concept increased in the Swedish public discussion. This was largely connected to the media coverage of an attack in May against the Swedish artist Lars Vilks during a lecture in Uppsala (linked to his recent depictions of the prophet Mohammed as a dog), and an unsuccessful attempt at a terror attack in Stockholm in December. At this moment, the problem of VAE – and the specific term “violence-affirming” – was picked up by editorial writers and other actors besides experts and politicians. On December 1, the Swedish Security police released a report on “Violence-affirming Islamist extremism in Sweden” (Säkerhetspolisen 2010). The far-right party, the Sweden Democrats, called for a Parliamentary debate about Islamist extremism, and in response, the Parliament decided to organise a debate also including right- and left-wing political extremism.³

After an initially predominant use of VAE to establish the social problem of a growing risk of Islamist violence in Sweden, right-wing VAE gradually received a more prominent position in the news framing of VAE during 2011. The turn began already in early 2011, but was sustained by the discussions ensuing on the Utøya terrorist attacks in Norway. In September, the Swedish EU commissioner Cecilia Malmström founded another important player in the international social problem game of violent extremism: the Radicalisation Awareness Network (RAN) on the EU level (cf. Melhuish and Heath-Kelly 2021).

During 2011, the Swedish Government worked on a national action plan against VAE, which was presented in December. In it, three types of VAE in contemporary Sweden were identified: “the autonomous left, the white supremacy milieu and Islamist extremism” (Swedish Government 2011, 9), and on the title page these categories were represented by three images: a burning barricade at a riot in connection with the EU Summit in Gothenburg 2001, the back of someone’s head where the letters “SS” are shaved, and an image from the failed terrorist attack in Stockholm the previous year. These images are followed by a fourth blank image with a question mark. However, no attempts were made to explore any other categories than the three mentioned, and although the report mentions “violence in the name of religion” (p. 13), no religions other than Islam were mentioned as connected to extremism and violence.

In 2012, mass media coverage on the issue went down somewhat, seemingly in the absence of major events driving news reporting. One theme that received some attention was the London Olympic games, which were discussed in relation to possible terrorist threats. The Swedish Government appointed the committee “Investigation of more effective work to prevent violence-affirming extremism”, which received relatively little media attention until the official report When We Care (SOU 2013, 81) was presented in December 2013. The report received public criticism from former members of the committee itself concerning the marginalised role of ideology in its diagnostic framing of VAE. This, the critics argued, was an inheritance from the mission statement by the Government (specifying the aim of designing a joint strategy for all three identified forms of VAE). Others criticised the report for its too broad definition of VAE and for recommendations that they feared would lead to increased control of young people’s political engagement. In the aftermath of the report’s publication, mainly journalists and experts engaged in critical discussion about consequences of various ways of defining the problem (diagnostic framing), whereas politicians cited in the newspapers almost exclusively talked about what must be done (prognostic framing).

In 2013, a Swedish Military Academy report was published about “foreign fighters”. The problem frame of domestic VAE was extended to encompass this phenomenon through the risk of people radicalising abroad and aquiring competences that could be used in domestic terror acts. However, “foreign fighters” as a distinct subcategory of violent extremism, did not attract major mass media attention until 2014, when it quickly became almost exclusively an issue about people travelling to fight for the newly founded Islamic State (IS). Whereas much of the early discussion of this issue focused on how to prevent people from travelling to fight for IS (e.g. whether to legislate against “terror journeys”), the discussion would later turn to how the society should act towards “returnees” – primarily meaning people who had taken part in wars in Syria and Iraq for the Islamic State. This corresponds to the framing of “returnees” in the contemporary EU policy discourse (Baker-Beall 2019). Whereas some Swedish municipalities took early initiatives to support and reintegrate those returning from war zones abroad, others criticised this approach and instead demanded stricter sanctions.

In 2014, the right-wing coalition government appointed the former Social Democrat minister, Mona Sahlin, as the first National Coordinator against VAE. 2014 was a year of Parliamentary elections and extremist threats against politicians became a salient issue. New legislation against “terror journeys” was proposed by the right-wing government coalition and the subsequently elected left-wing government coalition continued the work on this legislation.

The mass media discourse on VAE in 2015 was dominated by three acts of terror: the attack against the magazine Charlie Hebdo offices in Paris, a lone-actor attack (killing three individuals) at a school in Trollhättan, Sweden, and the November 13 attacks in Paris that included the concert hall Bataclan. The number of newspaper mentions of VAE increased dramatically compared to the previous year. In Sweden, a number of national-level interventions against VAE were reported, including the institution of so-called municipal “knowledge houses” (resource centres for local authorities coming in contact with VAE) and the founding of the Segerstedt Institute to promote research and dissemination of knowledge on prevention of racist groups and violent ideologies. The National Coordinator against VAE furthermore released the Conversation Compass⁴ – a website to provide knowledge about VAE and resources for practitioners to facilitate dialogue with young people running the risk of becoming radicalised. The Conversation compass led to controversies in mass media, partly because of its branding of a number of specific political organisations as “violence-affirming extremists”.

In the autumn of 2015, connected to the establishment of many new housing facilities to accommodate the high influx of asylum seekers, such facilities were increasingly subject to arson attacks (Törnberg et al. Forthcoming). Some voices were raised in mass media in favour of labelling these attacks VAE or even terrorism. However, most news reporting on such attacks did not use the “terrorism” label.

In 2016, attention to VAE was fuelled by media reports of terror attacks in Brussels and subsequently in Nice. One of the perpetrators from the Brussels attacks had links to Sweden’s third largest city, Malmö, and the National Coordinator against VAE used this to sharply criticise the Malmö municipality for not having a formal action plan against VAE. Not long after, the National Coordinator against VAE – Mona Sahlin – was forced to resign after a corruption scandal.⁵ The National Coordinator was replaced (twice) the same year, and the turbulence may have contributed to the growing criticism directed against the institution, which mainly concerned the fact that it was not already from the start subordinated to any of the already existing national authorities.

In 2017, much media reporting centred on the April terror attack on Drottninggatan in Stockholm, in which 5 people were killed. In the ensuing discussion about prevention, one strand concerned the need for various forms of situational crime prevention (e.g. better road blocks for pedestrian streets), as well as demands from the police for increased surveillance capacities. Terror attacks in Great Britain (Westminister, Manchester and London) also received Swedish mass media attention during the year. While all these were understood as the result of Islamist extremism, during the autumn the same year the media discussion on VAE shifted towards Nazi groups. This was primarily linked to activities of the Nordic Resistance Movement (NMR) and its demonstration in Gothenburg during the city’s annual book fair. Parts of the subsequent discussion concerned the degree to which Neo-Nazis’ right to demonstrate could be restricted and whether the police had been too lenient in relation to far-right protesters’ incitement to racial hatred (e.g. displaying certain Nordic runes with neo-Nazi connotations).

In the beginning of 2018, the National Coordinator against VAE was replaced by Center against Violence-Affirming Extremism (CVE; the official English name is the less literal translation: the Swedish Center for Preventing Violent Extremism), which became part of the Swedish National Council for Crime Prevention (BRÅ). A substantial share of the news reporting in 2018 concerned the trial of the perpetrator from the Stockholm terror attack the previous year and commentaries on the anniversary of the attack. 2018 was also election year and several news articles and opinion pieces were published in which political parties outlined their approach to crime prevention and the police. In these, VAE was in many instances dealt with in the same context as gang violence. This frame bridging was accomplished by subsuming both into the broader problem complex of crime-ridden areas in cities. In connection with the parliamentary election, there were also reports of extreme right threats against politicians and mobilisations against the extreme right, in particular the neo-Nazi organisation NMR.

In June, the security police released a new report on growing Salafi Islamism in Swedish towns and cities (Ranstorp et al. 2018). The report accused several civil society organisations of VAE connections and named a large number of individuals allegedly associated with the Salafi VAE environment. The report was followed by a large number of articles about the Swedish state not having taken Salafi extremism seriously, while notably few criticised the report’s extensive labelling of unconvicted individuals (implicitly or explicitly) as potentially dangerous extremists.

In early 2019, a salient theme was women and children associated with IS imprisoned abroad, and what to do with those who are allowed to return to Sweden. The previously discussed theme of lacking knowledge among municipal authorities received renewed attention, as well as the legal (and moral) challenges regarding how to deal with children of IS members. Over the year, attention was also directed towards right-wing VAE (instigated by the March 15 Christchurch attacks in New Zealand), as well as some coverage of left-wing VAE (triggered by the presence of a person with an Anti-Fascist Action t-shirt on a poster for the Left Party). However, Islamist VAE received more overall attention, partly driven by reporting on the state decision to deport six allegedly violence-affirming Islamist Imams (based on intelligence withheld from the public).

Overall patterns

A word count of the corpus S1 confirms the developments narratively outlined above. Figures 4a-c show the added frequencies of some predominant keywords for the respective categories in Swedish articles on extremism. It is notable that since this corpus contains not only articles on VAE but “extremism” in general, there are some discrepancies to the above narrative, which focused on VAE in particular. In the broader corpus concerned with “extremism”, Islamist extremism is less overrepresented compared to right-wing extremism, except for the large peak mainly connected to the 2017 Drottninggatan terrorist attack. This is particularly evident during the peak in coverage on the political right in 2014 – an election year – which indicates intensified discussions about the parliamentary far-right Sweden Democrat party. At the same time, 2014 had considerably less coverage on VAE compared to subsequent years, indicating that the parliamentary far right was broadly considered “extremist” but not of the “violence-affirming” kind. The left is consistently underrepresented in the “extremism” corpus, and the y-axis in figure 4(c) is therefore only 1/10 in relation to Figure 4 a and b. Figure 4(c) showing a general tendency over the time period to talk proportionately less and less about Communism as the major extremist threat from the left and more about “left-wing extremism” in general.

Based on the coding of articles in sample S3 (which all included the term VAE), one can furthermore identify how explicit mentions of the three forms of extremism intersect in Swedish public discourse (articles not mentioning any particular form of VAE – 29% of the sample – are excluded). As shown in Figure 5, Islamist VAE is not only by far the most common category of VAE in the sample; it is also more often mentioned on its own than the other two VAE categories taken together. In contrast, both right-wing and left-wing VAE (especially the latter) are rarely mentioned on their own but most frequently in articles concerning all three types of VAE. This indicates that while concern with Islamist VAE is driving the social problem discourse, the other two forms are mainly mentioned in the context of dealing with these phenomena as part of the same social problem (or, in some cases, mentioned as criticism of such lumping).

Interestingly, Islamism co-occurring with right-wing extremism is the most common dyad, by a wide margin. These patterns also generally correspond to the findings of Jämte and Ellefsen (2020b) that these two groups are regarded by Swedish public officials as the most pressing parts of the social problem of VAE, whereas the left to a large extent is included in the discourse out of the need to create symmetry between a focus on the political left and right. However, as noted above, no other religious motivations than Islam have at any point been brought to provide a similar symmetry with regards to religion.

There are some notable dynamics characterising the development outlined above. On the whole, the studied period represents a seemingly complete cycle of the establishment of a social problem, in line with the “natural history” pattern of social problems identified by e.g. Blumer (1971). The security police and actors linked to the Swedish Military Academy acted as domestic social problem entrepreneurs starting to define the problem in the late 2000s, under clear influence from the social problem framing on the EU level. However, the discourse did not take off until some spectacular events that amplified the social problem frame occurred in 2010. The subsequently installed National Coordinator against Violence-Affirming Extremism then started to call for both national and local action plans and to produce information material defining the problem. The establishment of the permanent Center against Violence-Affirming Extremism as part of the Swedish National Council for Crime Prevention signifies that the problem achieved a level of relative permanency in Swedish public administration.

What is striking concerning the beginning of the attention cycle when VAE is established in Sweden, is that there was initially hardly any discussion about this being a new concept, different from “violent extremism” used in prior decades to designate actors among the far right and the radical left-libertarian movement. VAE meant a frame extension, both in terms of including Islamists as well as adding “affirming” to include not only those who act violently or plan to do so, but more broadly those who approve of the use of violence. This in turn implied a widening of the net of social control towards a larger group of people (Cohen 1985). However, politicians, social problem entrepreneurs from the Military Academy, and the security police hardly ever discussed definitions in public. The idea of a coherent phenomenon causing extremist violence, emanating in analogous ways from Muslims, the political left and the political right, was initially largely taken for granted. A significant debate about this occurred only after a controversial official state report on VAE was released in late 2013.

It is also clear that the development was highly event-driven, with terrorist attacks, both domestic and European, creating the necessary drama to become salient in public discourse, creating the attention peaks visible in Figures 2–4 above. Compared to other social problem candidates during the period, VAE had, in a sense, the “advantage” of recurrent events fuelling the social problem discourse with drama and novelty and prevented it from becoming saturated, to use the terms of Hilgartner and Bosk (1988). Throughout the process, international events and discourses made an impact on the Swedish social problem game (see also Andersson Malmros 2021a). However, the ways in which these events and discourses were connected to Swedish social problem work appear to have been largely shaped by on the one hand the social problem trajectory (Ungar 1998) inherited from understandings and practices concerning left- and right-wing violence, and on the other hand, a relatively limited number of high-profile social problem entrepreneurs. As we saw in 2016, an international terrorist attack was used by the national coordinator to push for conformity from the Malmö municipality in producing a specific type of municipal action plan. In framing terminology, this is an example of how international events were used for motivational framing purposes, by reminding the public about the urgency of taking action. This sense of urgency was also fuelled by public agencies releasing new reports mapping the problem.

Figure 2. Mentions of the term “violence affirming” (Sw. “våldsbejakande”) in articles also mentioning “extremism” in major Swedish newspapers (S1) per 6 month period, from 1995 to 2019.

Figure 3. Mentions of derivatives of the terms extremism, radicalisation, violence affirming and terrorism in major Swedish newspapers (S1) per 6 month period, from 1995 to 2019.

Figure 4. (a) Frequencies of words in corpus S1 associated with Islamism. (b) Frequencies of words in corpus S1 associated with right-wing extremism. (c) Frequencies of words in corpus S1 associated with left-wing extremism.

Figure 5. Proportions of articles explicitly mentioning each of the the three types of VAE, alone or in combination S3.

In the subsequent sections, I will scrutinise in more detail how two central aspects of the problem definition were specified during the period: the perceived causes of, and proposed solutions for, VAE. In other words, I will develop the content of the diagnostic and prognostic framing of the problem.

Diagnostic mass media framing of VAE

Although VAE is, in a sense, itself a kind of diagnosis about the cause of violence perpetrated for religious and political reasons, it is in public discourse typically conflated with the violent acts, or actors, themselves. In the case of Islamist VAE, high-profile terrorist acts such as suicide bombing dominate, whereas right-wing extremists are more often associated with individual hate crimes, and online harassment, with the Utøya attack as a notable exception towards mass-violence. The left is more typically associated with harassment of politicians and journalists and doing property damage.

Overall, the causes of VAE, as framed in Swedish public discourse can be divided into four broader categories: culture and structure; ideational factors; relational factors; and individual vulnerabilities. Culture and structure here refers to, on the one hand, the broader context of individuals engaged in VAE. It can be subdivided into factors concerning material factors such as poverty and segregation (and associated lack of integration in broader society) and, on the other hand (sub)cultures and traditions that are regarded as conducive to VAE. The latter is separate from ideational factors, which include references to ideology and religion, and, as a distinct category in the analysis, propaganda (typically connected to online arenas). Relational factors include recruitment and social pressure, revenge against other groups, and various forms of backlash effects of actions (or inaction) by authorities. What is notably absent from the public discourse in this connection are references to escalating group processes, leading to radicalisation within groups (cf. Malthaner 2017).

Lastly, individual vulnerabilities, including on the one hand, individual needs and problems (typically “vulnerable youth”) and on the other hand, an individual history of (non-political) criminality. “Terror travel” – e.g. travelling to Syria or Iraq to join Al Qaeda or Daesh – forms a special category, since it is on the one hand often regarded as an expression of VAE abroad but at the same time is a frequently cited cause of domestic VAE (see also below). The sample of articles mentioning VAE (S3) were all coded in relation to mentioned causes (when several causes were mentioned, several codes were used). The results of this analysis are presented in Figure 6, in which the percentages represent the prevalence of each type of cause among the articles mentioning each specific category of VAE.

Figure 6. Mentioned causes of VAE proportional to all articles about each subcategory.

Overall, ideational factors, and specifically ideology/religion, are the generally most cited causes. Comparing typical causes across the categories of VAE, Islamism and the Right are most strongly associated with both individual and structural deficiencies and vulnerabilities. Individuals in these two “camps” appear often to be seen as being pushed by external circumstances and at the same time pulled in by propaganda and the ideology/religion itself. The issue of recruitment and social pressure is by far most salient in relation to Islamist VAE, primarily associated with so-called “terror travel”. Journeys to fight for IS in Iraq or Syria were themselves also highlighted as a possible cause for future Islamist terrorism, leading to radicalisation and improved skills in weapons and demolitions. (This is almost entirely depicted as an Islamist phenomenon; the small percentages of articles on the left and right also mentioning this cause are only an artefact of articles mentioning “terror journeys” in a discussion where all three types of VAE are mentioned.) It is also striking that segregation/poverty is a rather common theme specifically in relation to Islamist VAE. Lastly, specific causes were not always mentioned and some were not very precise. In proportion to the number of articles on each type of VAE, they are most often unclear in relation to left-wing VAE.

Prognostic mass media framing of VAE

Overall, a larger proportion of articles in sample S3 bring up prognostic frames – what to do about VAE – compared to how many mention specific causes. This signifies a general focus in the media debate on how to take care of the problem rather than understanding its causes. In order to clarify broader patterns in prognostic framing, the suggested solutions to VAE can be grouped into the four broader categories of crime prevention suggested by Sahlin (2000): control, socialisation, changing structures/environment, and making control institutions more effective (see Figure 7). Three sub-categories are subsumed under the umbrella of control-oriented interventions: legislation and penal sanctions (including also calls for increased policing and deportation); increased surveillance, intelligence gathering and control of information flows; and various forms of situational crime prevention, including Crime Prevention Through Environmental Design (CPTED).

Figure 7. Mentioned prognostic frames in relation to VAE proportional to all articles about each subcategory.

Interventions aimed as socialisation are either assigned to the sub-category of social support to risk groups (by e.g. social services and schools) or various forms of exit programmes. Some other categories are closer to Sahlin’s notion of more broadly changing structures/environment, including: reducing social inequalities and/or segregation; civil society dialogue; and everyday behaviour by ordinary citizens (such as “taking the discussion” with professed extremists). Lastly, several suggestions concern the meta-level of making crime preventing institutions more effective; setting up action plans, increasing collaboration between authorities; increasing and spreading knowledge about the problem, and interventions involving the international level.

Figure 7 indicates a tendency in the public discourse to be generally more specific about prognostic framing in relation to Islamist VAE, with an emphasis of control-oriented strategies such as new legislation, more control resources and increased surveillance. Much of the discussions in this regard concern novel legislation and surveillance to prevent “terror travel”. It is also in the context of preventing “terror travel” that articles discuss the need for increasing collaboration – and exchange of (surveillance) information – between authorities (also overrepresented in relation to Islamist VAE). Demands for instituting exit support are also common in relation to Islamist VAE, largely related to the fact that exit support was previously only available for leaving organised crime and far-right groups.

Extreme right VAE only stands out in relation to the civic responsibility of all citizens to “take the discussion” with people expressing neo-Nazi views. This perspective is more limited in relation to Islamism, presumably because Islamists are largely regarded as isolated from broader groups of “ordinary” citizens, and that this kind of intervention is instead mediated through civil society dialogue, that is, talking to approachable religious and civil society leaders about addressing VAE in their communities.

The observation that adherents of Islamist and right-wing VAE are more often regarded as vulnerable in the public diagnostic framing is also reflected in their overrepresentation in calls for interventions involving social support by schools and social services. However, only a very small proportion of articles on either type of VAE propose interventions targeting inequalities, segregation or any other kind of primary prevention (Brantingham and Faust 1976).

The prognostic social problem framing can be further clarified by regarding its trajectory over time (see Figure 8). Few solutions were publicly proposed early on in the process. However, tellingly, calls for increasing knowledge stand out somewhat in early years. Numbers of proposed solutions increase sharply (with the number of articles on VAE) in 2015, a year with several Islamist terrorist attacks in Europe. Calls for legislative changes and interventions associated with policing initially predominate and remain comparatively high in all subsequent years. Demands for more surveillance has a stable relatively high frequency from 2015 onwards. These patterns indicate a trajectory characterised by a perceived need for continuously increasing control-oriented interventions. In contrast, calls for action plans and collaboration between authorities peak in 2016 after which they go down (along with calls for more knowledge about the problem). Presumably, the successive institutionalisation of the problem on many levels in the Swedish bureaucracy (Andersson Malmros 2021a) also means less focus on prevention efforts focusing on making institutions more effective. Solutions that are already implemented do not need much further public discussion. Except for a peak in discussion about exit support in 2016, socialisation-oriented prevention received clearly less support than control-oriented prevention during the latter half of the period.

Figure 8. Mentioned prognostic frames in absolute numbers of articles in sample S3, per year.

Conclusions

The development of “violence-affirming extremism” as a social problem in Swedish public discourse generally followed international trends and was partly driven by high-profile terrorist events in other countries. However, the above analysis demonstrates how the framing of VAE was also shaped by the inherited national social problem trajectory of “violent extremism”. The threat of Islamist political violence was introduced as a domestic social problem by “piggybacking” (Loseke 2011) on an established problem frame concerning left- and right-wing extremist violence, illustrated by the initial reluctance in the Swedish Parliament to discuss Islamist VAE as a separate problem. Along with extending the prior frame, by including Islamist political violence, came the seemingly innocuous terminological shift from “violent” to “violence-affirming”. It nevertheless signified a more pronounced pre-emptive logic by encompassing also (ostensibly) extremist groups and individuals that support (“affirm”) violent means without necessarily resorting to them.

Considering the apparent ease with which VAE achieved and maintained broad recognition as a prioritised social problem in Sweden, we must remind ourselves that many troubling conditions never acquire a comparable status (or only do so temporarily). A crucial factor behind VAE becoming a salient social problem in Sweden during the 2010s was a series of high-profile terrorist events, domestic and in other Western countries, which provided drama and a recurrent sense of novelty. This made it possible for the problem to secure its place among the most important social problems year after year.

However, the analysis also shows that the terrorist events cannot by themselves explain the development and establishment of the social problem VAE. For example, the terrorist attacks on 9/11 2001, although causing a spike in the number of mentions of “terrorism” in Swedish public discourse, did not bring about the huge surge in attention to radicalisation and VAE that later occurred during the 2010s. A number of more geographically proximate events did play a more direct role and a number of social problem entrepreneurs were crucial in shaping the initial diagnostic and prognostic frames. The subsequent institutionalisation of VAE as a core social problem for the Swedish state – through action plans, local coordinators and finally the establishment of a new national authority – was a collaborative effort by several actors in different parts of the state bureaucracy together with the mass media.

Concerning the content of the social problem definitions of VAE, the analysis revealed a tension between treating it as a unitary phenomenon and as three disparate types of extremist actors and practices. When discussed separately, they first of all received very different levels of public attention. Islamist VAE quantitatively dominated the public discourse, whereas left- and right-wing VAE were primarily mentioned in discussions concerning all three categories. This illustrates that, despite problem framings of Islamist VAE initially piggybacking on left- and right-wing VAE, the main thrust in the VAE social problem trajectory appears to have come from concerns about Islamist terrorism and Islamist extremism in general. Thereby, a whole set of policies and practices were introduced in relation to left- and right-wing VAE that presumably would not have come about without the frame bridging with Islamist VAE. At the same time, the ostensibly impartial overall framing of VAE conferred legitimacy to the expansion of the PVE arena, which more claims-makers would arguably have regarded as highly questionable if the sole expressed ambition had been to prevent Islamist radicalisation.

The analysis also demonstrated clear differences in terms of both perceived causes and proposed solutions to Islamist, left-wing and right-wing VAE. While causes of left-wing VAE were largely considered unclear, the main perceived cause of Islamist and right-wing VAE was clearly the ideology/religion as such, and associated propaganda disseminated online. Direct recruitment into extremist activities was a common concern connected specifically to Islamist terrorism, but had little resonance with the other two milieus. While both individual and structural vulnerabilities are mentioned as partial causes, especially in relation to Islamist VAE, only rarely do suggested solutions in the material focus on changing structural conditions such as addressing segregation or inequalities. It is individuals who are the main targets of interventions, through control efforts (including legislation, policing and surveillance) and socialisation (including exit support). Articles focusing on Islamist VAE most frequently mentioned control-oriented solutions, which could indicate that (suspected) Islamist extremists are the main intended targets when such solutions are discussed. Regarded over time, it is striking how proposals concerning making institutions more effective tend to wane after a peak in salience around the mid-2010s. Meanwhile, control-oriented prognostic frames remained salient until the end of the studied period. The latter tendency could be interpreted as a trajectory towards further expanding the scope and intensity of surveillance and coercive control. From a civil liberties perspective, this is a source of concern.

Lastly, since the narrative in this article is largely based on newspaper data, it is possible that aspects of the analysis would have turned out different if it had relied on other sources, such as televised news, parliamentary documents or access to the inner workings of the Swedish security police. A closer scrutiny of such alternative sources would provide a valuable complement to create a more complete picture of the social problem trajectory. One must also acknowledge the discrepancy between the official framings of the problem identified here and how they are translated to various levels of the state and municipal bureaucracy and interpreted by the “front-line” practitioners assigned to deal with the problem (Andersson Malmros 2021b; Jämte and Ellefsen 2020b). This remains an important topic for future study, along with the need for further cross-national comparisons of social problem trajectories relating to violent extremism and radicalisation.

Acknowledgments

I would particularly like to thank Elvira Johansson, who coded parts of newspaper data used in the analysis. Also thanks to Jan Jämte, Magnus Wennerhag, Nils Gustafsson, the anonymous reviewers and the journal editor for advice and helpful comments on previous versions of the manuscript.

Disclosure statement

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

Additional information

Funding

This work was supported by Swedish Research Council for Health, Working life and Welfare (Forte) under Grant number 2016-00925

Notes on contributors

Mattias Wahlström

MattiasWahlström is Associate Professor of sociology at the Department of Sociology and Work Science, University of Gothenburg. His research concerns various aspects of social movements, political protests, political violence and social control.

Notes

1. Andersson (2018) also maintains that the Swedish notion of “våldsbejakande extremism” is unique and not directly translatable to “violent extremism”. He suggests that the meaning of the term “våldsbejakande” lies somewhere between “violence approving” and “violence affirming”. In line with Jämte and Ellefsen (2020b) and Andersson Malmros (2021a), this study sticks to the latter translation.

2. Note that this chronology differs from that of Andersson Malmros (2021a), who dates the introduction of VAE to a report by the Swedish Security Police, published in 2010.

3. As shown in a quantitative analysis by Aziz (2021), the Sweden Democrats were the most frequent users of the term “extremism” in Swedish parliamentary debates.

4. Previously available at URL: www.samtalskompassen.samordnarenmotextremism.se

5. She had issued a false certificate to help one of her personal bodyguards to get a bank loan.