How terrorist attacks distort public debates: a comparative study of right-wing and Islamist extremism

Völker, Teresa Article — Published Version How terrorist attacks distort public debates: a comparative study of right-wing and Islamist extremism Journal of European Public Policy Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Völker, Teresa (2024) : How terrorist attacks distort public debates: a comparative study of right-wing and Islamist extremism, Journal of European Public Policy, ISSN 1466-4429, Taylor & Francis, London, Vol. 31, Iss. 11, pp. 3487-3514, https://doi.org/10.1080/13501763.2023.2269194 This Version is available at: https://hdl.handle.net/10419/307972 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ How terrorist attacks distort public debates: a comparative study of right-wing and Islamist extremism Teresa Völker Center for Civil Society Research, WZB Berlin Social Science Center, Berlin, Germany ABSTRACT Previous research has shown how terrorist attacks attract media attention and influence public opinion and decision-makers. However, we lack a comparative assessment of the extent to which extremist ideologies matter and how they matter. Therefore, this paper compares mass media debates over extreme right and Islamist terrorist attacks. Theoretically, it innovates by linking research on discursive critical junctures and issue-specific discursive opportunity structures, emphasising the systematic differences between the two ideologies. Empirically, the study is based on an original, large-scale content analysis of mass media debates on all seven fatal attacks in Germany since 2015 (N = 9047). It combines relational quantitative content analysis with frame and network analyses. The results show how ideologies behind terrorist attack shape political reactions and the framing of the key security threat. Notably, both types of attacks provide favourable conditions for the far right, and political elites play a central role in the diffusion of far-right frames. In contrast, victims and ethnic or religious minorities have little voice in public debates. Overall, the study contributes to a better understanding of the impact of terrorist attacks on Western democracies by emphasising the impact of ideology and distorted threat perceptions in public debates. ARTICLE HISTORY Received 21 December 2022; Accepted 5 October 2023 KEYWORDS Far right; terrorism; public debates; right-wing extremism; Islamist extremism; public opinion Introduction In March 2019, a right-wing extremist shot over 50 people at two mosques in Christchurch, New Zealand. That same year, a right-wing extremist killed two © 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. CONTACT Teresa Völker [email protected] Center for Civil Society Research, WZB Berlin Social Science Center, Reichpietschufer 50, 10785 Berlin, Germany Supplemental data for this article can be accessed online at https://doi.org/10.1080/13501763.2023. 2269194 JOURNAL OF EUROPEAN PUBLIC POLICY 2024, VOL. 31, NO. 11, 3487–3514 https://doi.org/10.1080/13501763.2023.2269194 people in Halle, in Germany, after unsuccessfully trying to enter a synagogue. Four months later, in February 2020, another right-wing extremist targeting ethnic and religious minorities killed eleven people in Hanau, Germany. The same year saw Islamist attacks in Dresden, Nice, and Vienna. These recent attacks illustrate that extreme right and Islamist terrorism trigger public reactions in various ways that can drastically affect Western democracies. As Jenkins’(1974, p. 4) metaphor ‘terrorism is theatre’aptly illustrates, terrorist attacks are performed in such a way that they gain media attention and visibility in the public sphere. Without media attention, terrorist attacks remain ‘non-events’–they are invisible and cannot influence public perception or foster political action (e.g., Andrews & Caren, 2010; Koopmans, 2004). Previous research has shown that mediatised terrorist attacks affect public opinion (e.g., Agerberg & Sohlberg, 2021; Jacobs & van Spanje, 2021). The underlying ideologies of Islamist and extreme right terrorist attacks – namely Islamism and nationalism –can, according to Fukuyama (2018, p. 58), be seen as ‘two sides of the same coin.’Both oppose processes of modernisation and benefit from identity conflicts within Western societies. However, research on right-wing and Islamist extremism has remained divided. To date, there is ample research on the radical and extreme right (Berntzen, 2020; Castelli Gattinara & Froio, 2023; Ravndal, 2018) and on Islamist extremism (e.g., Della Porta et al., 2020; Giani, 2020; Ruigrok & van Atteveldt, 2007). Yet, there is rarely any research comparing Islamist and extreme right violence in Western democracies and their discursive context. Moreover, most research focuses on the effects of terrorism on the individual level (e.g., Barceló & Labzina, 2020; Bueno de Mesquita & Dickson, 2007; Godefroidt, 2022; Gould & Klor, 2010). In this paper, I provide the first comparative empirical assessment of the conditions under which extreme right and Islamist terrorist attacks influence public debates. To this end, I address the following research questions: to what extent and how do terrorist attacks influence public debates? What are the differences between public debates after extreme right and Islamist terrorist attacks? To answer these questions, I combine research on social movements, extremism and political communication. First, I examine how issue-specificdiscursive opportunity structures determine which actors and issues gain access to and influence public debates. Discursive opportunity structures refer to pre-existing values and visions around issues in the broader political culture of a country (Ferree, 2003; Koopmans & Olzak, 2004). Second, I analyse whether and how terrorist attacks affect public debates and produce what scholars have called discursive critical junctures, transforming existing political alignments and visions around issues (Della Porta et al., 2020; Larsson & Lindekilde, 2009). Combining these two perspectives, I have developed the discursive radicalisation model, serving as a theoretical 3488 T. VÖLKER heuristic to analyse how terrorist attacks may influence public debates and how radical actors may gain visibility, resonance, and public legitimacy. Empirically, this paper covers mass media debates after all seven fatal acts of political violence since 2015 in Germany. These incidents represent the key radicalisation phenomena in Western democracies: Islamist and right-wing extremism. The data was collected using an original form of a relational quantitative content analysis of newspaper coverage. Overall, the data set consists of 9047 statements by a wide range of political and societal actors. I used a mixed methods approach combining the relational quantitative content analysis with frame analysis and network analyses. Overall, the results show that both extreme right and Islamist attacks provided favourable discursive opportunities for far-right actors in Germany. In contrast, victims and ethnic or religious minorities hardly gain any voice in the public debate. The ideological motive for the terrorist attacks influenced the political interpretation of the events and public framing of the central security threat: after Islamist attacks, the majority of statements referred to Muslims as a broad outgroup and migration as the central security threat. In contrast, after extreme right attacks, the majority of statements referred to right-wing extremists as a narrow outgroup and reduced the central security threat to isolated fringe actors. In line with previous research, the results highlight that political elites played a central role in the public interpretations of these attacks and the diffusion of far-right frames. The study adds to previous research by highlighting the crucial role of the extremist ideology at stake. This comparative perspective is highly relevant for better understanding the political and democratic impact of terrorist attacks on Western democracies. The paper is structured as follows: First, I outline the theoretical framework, combining insights from research on social movements, extremism, and political communication. The second part describes the case selection, methodological approach, and data analysis. Next, I present the empirical results in three steps: first, I present quantitative analyses of the three main dimensions of discursive radicalisation –namely visibility, resonance, and public legitimacy. The second step of the analysis provides an in-depth qualitative analysis of the discursive construction of Islamist and extreme right terrorist attacks. The final step uses discourse network analyses to show the driving forces in the debate and the relationship between actors and issues. I conclude with a summary and discussion of the implications of the study. Theoretical framework The theoretical framework bridges research on social movements, extremism, and political communication. In particular, it combines two perspectives to JOURNAL OF EUROPEAN PUBLIC POLICY 3489 identify issue-specific structural opportunities within public debates and the effect of critical events on existing political conflicts. First, I integrate studies that adopt a ‘top-down’perspective and deal with discursive opportunity structures (DOS). The concept of discursive opportunity structures can be defined as ‘institutionally anchored ways of thinking that provide a gradient of relative political acceptability to specific packages of ideas’(Ferree, 2003, p. 309). DOS refer to pre-existing values and visions around issues in a country’s political culture. I follow previous research that has used this concept to examine which actors and issues gain access to and influence public debates (e.g., Koopmans, 2004; Wahlström & Törnberg, 2021). I focus on issue-specificdiscursive opportunity structures, which refer to the cultural environment that defines the resonance of certain public demands in an issue field. Although issue-specific discursive opportunity structures are relatively stable, critical events can produce ruptures and change the content and dynamics of public debates. Second, I combine this approach with a bottom-up perspective on discursive critical junctures; this refers to focal moments in which public controversies and polarisation intensify and political cleavages are transformed (Della Porta et al., 2020; Larsson & Lindekilde, 2009). In particular, acts of political violence can potentially create such discursive turns that change political alignments, identities, and visions around particular issues. As moments of uncertainty or contingency, they can lead to intense political reactions and counterreactions and open up unexpected windows of opportunity for certain actors (Della Porta et al., 2020). Public debates following critical events shape public memory and reveal discursive conflicts that can have lasting political effects. Linking the two approaches, I combine both (1) the issue-specific discursive opportunity structures over right-wing and Islamist extremism and (2) the effect of political violence on dynamics in the public sphere. I adapt Koopmans’(2004) selection criteria of discursive opportunities and develop the discursive radicalisation model. The discursive radicalisation model allows me to analyse empirically how events can provide windows of opportunity for actors to put their issues on the public agenda. This model offers a theoretical heuristic to analyse how radical actors may shape public debates after critical events such as terrorist attacks. Following Koopmans’triad of visibility-resonance-legitimacy, the model assumes a three-fold sequence, summarised in Figure 1: (1) The first step is visibility. Visibility refers to the extent to which radical actors and events such as terrorist attacks attract public attention. (2) The second step is resonance. Resonance describes the political reactions that radical actors and events provoke and how they shape discourse dynamics on contested issues. 3490 T. VÖLKER (3) The third step is public legitimacy. Public legitimacy refers to the extent to which actors and issues resonate positively and gain support in the public sphere. 1 Ultimately, I argue that high visibility, resonance, and public legitimacy of radical actors in the context of critical events can lead to incremental shifts in public debates that enhance the normalisation and reproduction of radical claims in the long-term. This model helps to shed light on the mechanisms by which critical events can provide windows of opportunity to shift public debates and normalise radical frames. Based on previous research, in the next section I derive hypotheses for each stage of the discursive radicalisation model. Visibility: the effect of terrorist attacks on public debates Previous research has shown that news values, action repertoires and the characteristics of events influence their media attention and public visibility. For instance the size of the event, its level of disruptiveness, use of unconventional and violent action repertoires or the degree of symbolic drama – influence the effects of events (Castelli Gattinara & Froio, 2023; Walgrave & Vliegenthart, 2012). Terrorist attacks, understood as ‘(…)a conspiratorial practice of calculated, demonstrative, direct violent action without legal or moral restraints […] performed for its propagandistic and psychological effects on various audiences and conflict parties’(Schmid, 2012, p. 158) are both newsworthy and disruptive and offer symbolic drama. As highly disruptive events, terrorist attacks and extremists can change public perception, generate fear and increase public demand for security and counterterrorism (Agerberg & Sohlberg, 2021; Giani, 2020). Terrorist attacks can have different Figure 1. Discursive radicalisation model. JOURNAL OF EUROPEAN PUBLIC POLICY 3491 political effects, as previous research has shown. On the one hand, terrorism can backfire on the political goals of extremist actors, by reducing their legitimacy and support (e.g., Barceló & Labzina, 2020; Gould & Klor, 2010). Such moments of crisis can have ‘rally-round-the-flag effects’when governmental actors gain support from citizens and foster securitisation (de Graaf, 2011). On the other hand, terrorist attacks can provoke disproportionate political reactions such as repressive strategies or failed counter-terrorism that question the legitimacy of the government (Bueno de Mesquita & Dickson, 2007). Discursive opportunities are a decisive factor in determining the appropriateness of political reactions and public visibility of terrorists and victims. Typically, terrorists reach a wide public audience (Schmid, 1989), while the perspective of victims tends to receive little attention. However, public attention can vary depending on who is the perpetrator and who is the victim (Knupfer & Matthes, 2021). Victims who belong to the perceived ingroup are treated more favourably than victims who belong to the perceived outgroup. For example, minority groups are over-represented as perpetrators and under-represented as victims in public debates (Dixon & Linz, 2000). This marginalisation of minority groups is a general pattern in everyday debates in Western societies and is reinforced in moments of crisis (Kroon et al., 2016). In Germany, ethnic and religious minorities, such as Muslim actors, have had limited opportunities to influence public debates on cultural and identity issues (Cinalli & Giugni, 2013; Dolezal et al., 2010). Resonance and public legitimacy of Islamist and right –wing extremism What are the differences between extreme right and Islamist terrorist attacks? Both events are based on extreme ideologies that oppose processes of modernisation and liberal democracies, and benefit from identity conflicts within Western societies (Fukuyama, 2018). 2 Islamist extremism can be defined as ‘a political ideology that strives to create a state and society in conformity with religious doctrine and Sharia law’(Precht, 2007, p. 16). Right-wing extremism is ‘a collective term for anti-democratic dispositions and attempts, that are traditionally positioned at the extreme “right”of the left-right spectre’(Backes & Jesse, 1993, p. 474). 3 Right-wing extremism is based on an authoritarian, nativist or ethnic nationalist belief system (Mudde, 2000,2019; Ravndal, 2018). When I refer to the term ‘far right’, I include both radical right actors who reject the democratic constitution and extreme right actors who fundamentally oppose the democratic constitution and aim to destroy the democratic system (Pirro, 2022). The goal of terrorist attacks is to generate fear and provoke intense political reactions that call into question the legitimacy of the political elite and the political system as a whole. A key difference is the ideals they offer as 3492 T. VÖLKER solutions to identity struggles: In the case of right-wing extremism, the collective identity is constructed around the ideal of a culturally homogenous nation state. The identity of Islamist extremists is based on the ideal of a state in conformity with Sharia law. Even though both extreme ideologies pose a threat to liberal democracies, there is a greater fear in Western societies about the compatibility of Islam with democratic values and its potential for radicalisation (e.g., Panagopoulos, 2006). After extreme right attacks, there is less of a broader debate about cultural issues or ideology such as nationalism; instead, the perpetrator is often portrayed as a ‘lunatic’and the attack is linked to mental health issues (Solheim, 2021). Right-wing terrorism is often perceived as incidents committed by socially isolated actors or ‘lone wolves’(Dreier et al., 2022;Huff& Kertzer, 2018). In contrast, outgroup hostility tends to be stronger after Islamist attacks and stronger towards Muslims and immigrants than hostility after attacks with another ideological motive and towards other societal groups such as Christians or nationalists (Godefroidt, 2022). Islamist terrorists are not perceived as a narrow outgroup of extremists but are viewed as connected to the broad community of Muslims (Ahmed & Matthes, 2017). The 9/11 attacks are the best-known example of Islamist terrorism in Western democracies that produced negative sentiments and associations between Islam, the integration of Muslims and public security (Ruigrok & van Atteveldt, 2007). A more recent example was the Charlie Hebdo attack in France 2015. The Charlie Hebdo attack led to discursive turns in different European countries (Della Porta et al., 2020). While security and freedom of expression became the dominant issues, the highest degree of polarisation concerned issues related to Islam and migration. Far-right actors successfully mobilised as an ‘entrepreneur of fear’and ‘bulwark against multiculturalism and Islamisation’by linking security issues to Islam (Della Porta et al., 2020, p. 202). Moreover, they influenced attitudes to immigration and the propensity to vote for an anti-immigration party (Jacobs & van Spanje, 2021; Solheim, 2021). This is in line with empirical research on the effects of extreme right violence. The anti-foreigner riots in the German town Hoyerswerda in 1991 intensified the asylum debate and triggered political reactions from domestic politicians. Similarly to the dynamics after Islamist attacks, public debates around these issues tended to increase the visibility of the far right (Koopmans, 2004). Moreover, after public debates on these riots, asylum seekers became an increased target of far-right violence (Koopmans, 2004, p. 384). Far-right actors have not only been successful in setting the agenda in public debates, but have also been the driving force behind the politicisation of immigration in Europe. They triggered other political actors to emphasise the issue as well (Gessler & Hunger, 2021; Hutter & Kriesi, 2019). In particular, mainstream parties and their reactions played a JOURNAL OF EUROPEAN PUBLIC POLICY 3493 crucial role in the success and normalisation of far-right claims on immigration. Even critical and negative reactions by moderate parties towards far-right actors increased their resonance and improved their opportunities to shape the direction of public debates (Muis, 2015). This shows that far-right actors benefit from any resonance in the public sphere as far as the mobilisation of far-right supporters is concerned (Bonikowski, 2017, p. 207). The success of far-right actors is linked to the national context and path dependencies of individual countries (Brause & Kinski, 2022). In Germany, far-right actors have by now a high degree of institutionalisation and strong links to party politics, as illustrated by the electoral success of the Alternative für Deutschland (AfD). Based on previous research findings, I examine the public visibility, resonance and legitimacy of radical actors, issues and events by testing the following hypotheses. First, with regard to visibility, I assume that, regardless of the ideological motive, extremists are newsworthy and attract public attention. In contrast, I expect that victims and targets of the attacks, particularly ethnic and religious minorities, hardly gain any public visibility. Based on ingroup and outgroup explanations, I assume that Islamist extremists gain more visibility in public debates than right-wing extremists. This is addressed by the first two hypotheses. Visibility H1a: Extremists gain more visibility in public debate than victims of terrorist attacks. H1b: Islamist extremists gain more visibility in public debates than right-wing extremists. Second, with regard to resonance, which refers to the thematic and political responses to these attacks, I expect that after both Islamist and extreme right attacks, right-wing parties successfully mobilise as entrepreneurs of fear (Della Porta et al., 2020) and gain more public resonance than left-wing parties. However, I assume that the thematic emphasis of the debate and public framing of the central security risk depends on the ideological motive for the event. In line with previous research (e.g., Godefroidt, 2022; Ruigrok & van Atteveldt, 2007), I expect that Islamist attacks are broadly discussed and connected to immigration and Islam as central security threats. Broad debates go beyond the attack, discuss a wide range of social issues and link the attack to a broad outgroup such as immigrants or Muslims. In contrast, I expect that extreme right attacks are narrowly discussed, focusing on right-wing extremism and the terrorists as isolated fringe actors. Narrow debates are limited to the attack, focus on issues around extremism and link the attack to a narrow outgroup such as terrorists. Resonance H2a Reactions by right-wing parties resonate more than reactions by left-wing parties after terrorist attacks. H2b After Islamist attacks there is a broad debate about immigration and Islam, while after extreme right attacks there is a narrow debate about right-wing extremism. 3494 T. VÖLKER (38.54 %) after the Islamist attacks. This may be related to the fact that the CDU was part of the government, which improves any party’s chance to shape public debates. In contrast, centre-left parties were less influential after extreme right attacks and their claims were less visible than claims by right-wing politicians. Only politicians of the Green party Bündnis 90/Die Grünen, as the classic opponent of the AfD, played a significant role in debates after Islamist attacks –22 % of the statements came from them. The results in Table 2 show that certain parties, in particular the AfD, were over-represented in public debates relative to their vote share in parliament (for details, see Table B8 in the Appendix). After Islamist and extreme right attacks, right-wing parties were the more successful crisis communicators, although extreme right attacks would have provided opportunities for leftwing parties to criticise extreme right ideologies. This confirms previous research findings that right-wing actors are successful issue entrepreneurs in moments of crisis (e.g., Della Porta et al., 2020; Krzyżanowski, 2020). What thematic debates emerged after Islamist and extreme right attacks? Figure 4 compares the thematic emphasis of debates following Islamist (red) and extreme right terrorist attacks (blue). The analyses demonstrate that the content of public debates after terrorist attacks was related to the ideological motive behind the attack. The findings confirm hypothesis (H2b) that, after Islamist attacks there is a broad debate about immigration and asylum, while after extreme right attacks there is a narrow debate about right-wing extremism. After the Islamist attacks studied here, the debate evolved around the question of how and to what extent migration and Islam may be a breeding ground for radicalisation. Islamist attacks were linked to broad outgroups such as immigrants. This linkage between security and migration played into the hands Table 2. Resonance: share of political parties as subject and object in political debates after terrorist attacks, percentages. Extreme right attacks Islamist attacks Political party Subject Object Subject Object AfD 25.74 50.79 18.54 49.25 FDP 6.05 5.82 1.95 0 CDU 18 10.58 15.37 4.48 CSU 6.75 3.17 23.17 19.4 Right-wing parties in % 56.54 70.36 59.03 73.13 SPD 16.03 10.05 13.9 13.43 Bündnis 90/Die Grünen 15.47 10.05 21.46 11.94 Die Linke 11.95 9.52 5.61 1.49 Left-wing parties in % 43.45 29.62 40.97 26.86 Total in % 100 100 100 100 N 711 189 410 67 Note: The Table shows the share of statements as percentage of all statements referring to parties. In total, there were more political statements after extreme right attacks (N = 900) than after Islamist attacks (N = 477). JOURNAL OF EUROPEAN PUBLIC POLICY 3501 of far-right actors as ‘issue owner’. In contrast, after extreme right attacks, the focus of the debate was on the perpetrator’s motives, individual radicalisation and right-wing extremism. In line with the public invisibility of the victims, many of whom were Muslims, there were hardly any debates on Islamophobia following the attacks. Moreover, there was no broader debate on issues such as nativism, national identity or Christian traditions and their relation to right-wing extremism (for further details, see Table B8 and B9 in the Appendix). This indicates that the ideological motive for the attack relates to the public interpretation of the cause of terrorism. Regarding the third step of the discursive radicalisation model, public legitimacy, the results show that there were differences between Islamist and extreme right attacks. Yet, average attitudes on most issues did not change significantly. Table 3 shows the degree of public legitimacy of issues and actors covered in the mass media. It compares the level of public support for issues and actors as object of statements one week before and one week after Islamist and extreme right attacks. The observed shift captures the change of (average) positions on issues and actors as the objects of statements (−1 stands for a negative relationship and 1 for a positive relationship) covered in the mass media. The analysis supports H3a that terrorist attacks backfire in terms of legitimacy of extremist groups. Extremist actors and events associated with both ideologies predominantly attracted negative publicity and were evaluated Figure 4. Resonance of central sub-issues, share of statements. Note: The figure compares the distribution of the sub-issues in public debates after Islamist and extreme right terrorist attacks. It includes a threshold for a minimum number of statements (n> 16). 3502 T. VÖLKER as illegitimate after the terrorist attacks. The average negative shift of positions in mass media towards extreme right actors after extreme right attacks was stronger than towards Islamist actors after Islamist attacks (see Table B12 in the Appendix for the total number of statements). Comparing the legitimacy of issues, the analyses show the continuity of a broad (negative) consensus on issues of migration, nationalism and radicalisation. Public support for issues related to migration decreased after both types of events, and statements linking security issues to failed integration gained popularity. Even though there was a generalised debate about migration and Islam instead of Islamist extremism after Islamist attacks, there were on average more positive statements referring to the Islam than negative statements. In contrast to hypothesis H3b, the average public position towards nationalism decreases to a greater extent after extreme attacks than the public legitimacy of Islam does after Islamist attacks. This indicates that terrorist attacks can disrupt general patterns in everyday debates that portray Islam more negatively than nationalism. Nevertheless, in both cases, the negative framing of migration issue was used to describe ethnic and religious minorities as a security threat and an illegitimate outgroup. 13 The framing of security threats and extreme right and Islamist extremism is analysed in more depth in the next part. Different constructions of security threats: lone wolves or collective threats How do different actors interpret the causes and consequences of terrorist attacks? To address this question, I use frame analysis to qualitatively reconstruct the interpretative frames about extreme right and Islamist terrorist attacks in greater depth. Following Benford and Snow (1988, 2000), I analyse the diagnostic frame, which refers to the cause of the attack (who is to blame) and the prognostic frame, which refers to the consequence of the attack (what should be done). The frame analysis Table 3. Legitimacy shift of actors and issues. Public legitimacy shift (average position) Islamist attacks Extreme right attacks Statements referring to Actors extreme right actors −0.18 −0.35 Islamist actors −0.28 −0.37 Statements referring to Issues Islam 0.35 −0.11 migration −0.19 −0.55 nationalism −0.19 −0.16 radicalisation 0.01 −0.06 Note: Public legitimacy shift is the change of average position from one week before the event to one week after the event (additional information in the Appendix). JOURNAL OF EUROPEAN PUBLIC POLICY 3503 shows systematic differences in the framing of extreme right and Islamist attacks in public debates. While extreme right terrorist attacks were framed as incidents by isolated individuals or ‘lone wolves’,Islamist attacks were framed as a collective threat connected to the broader group of Muslims and immigration. This difference was also evident in the prognostic framing: right-wing extremism was framed as a national threat, focusing on individual radicalisation, and Islamist extremism was framed as an imported security threat, focusing on a reorientation of refugee policy (for further details, see Table A10 and B11 in the Appendix). Different actors were engaged in a discursive competition to interpret the causes and consequences of the terrorist attack and to link them to cultural and identity questions. Organisations representing ethnic minorities highlighted the experience of discrimination against the victims from vulnerable groups and pointed to the structural problem of racism in Germany. As the previous section of this paper demonstrated, these actors did not receive a lot of discursive space to propagate their claims in the debate. After the Hanau attack, members of ethnic minority organisations compared the attacks to other acts of extreme right violence in Germany in the past to illustrate that they were not isolated cases. They criticised the fact that far-right actors such as the AfD and the ‘lone wolf’framing of extreme right terrorist were receiving too much media attention. In the following quote by the chairman of the Immigration and Integration Council (BZI), who is a member of the Green Party, this perspective is described as follows: ‘We have not forgotten Mölln, Solingen, RostockLichtenhagen and Hoyerswerda. Back then, the right-wing republicans only made it into local parliaments. With the AfD, we now have a parliamentary arm of right-wing radicalism and right-wing extremism in the Bundestag and in all state parliaments.’(Memet Kilic, 21 February 2020, Die Welt). 14 In contrast, after Islamist attacks the central security threat was attributed to migration and the lack of integration of refugees. In particular, right-wing and conservative actors demanded improved security policies and surveillance in order to fight Islamist actors and prevent potential radicalisation of migrants. The most visible far-right party AfD framed terrorists as an imported security problem caused by immigration and argued that the religion of Islam was in conflict with Western culture and democratic values. The following quotes by AfD politicians after the attacks in Berlin in 2016 and in Dresden in 2020 illustrate how far-right actors attributed the attacks to the Islam and multiculturalism and framed deportations as a solution to radicalisation. CSU politicians supported this demand of the AfD by arguing against a general ban on deportations of Syrian refugees. 3504 T. VÖLKER ‘Poggenburg claims a “direct consequence of a policy of multiculturalism at any price”, adding: “Anyone who continues to deny that Islam is directly linked to Islamist terror is complicit in every additional victim”.’(André Poggenburg, 21 December 2016, Welt) ‘All legal options must be exhausted in order to deport potential Islamist actors and violent criminals as quickly as possible. (…) There must not be a general ban on deportations’(Alice Weidel, 26 October 2020, Die Welt). This difference was also evident in the prognostic framing and the political solution proposed in response to the attacks. Right-wing extremism was framed as a national threat and Islamist extremism as an imported security threat. The debatesaftertheright-wingattacksfocusedonindividualradicalisationofextremists in Germany and right-wing extremism as a national problem. In contrast, after the Islamist attacks, the transnationalscope of radicalisation played a more important role and there was a focus on a reorientation of immigration policy. After the extreme right attacks, the key political level chosen by the actors to address theproblem (forexamplecounterterrorismmeasures)wasthedomestic level (80 %t domestic and 20 % international regulations). In contrast, after the Islamist attacks, only 51 % of actors referred to the local or domestic level and 48 % to the international or European regulatory level as being the level at which policies should be introduced (see Table B12 in the Appendix). Overall, the analysis indicates that Islamist attacks were framed as an imported threat related to immigrants threatening society as a whole, whereas extreme right attacks were framed as a threat related to isolated fringe actors threatening a part of society, namely ethnic and religious minorities. This may be explained by the increased ingroup solidarity and a sense of connection to the nation as a link to and source of protection from the dangerous other (e.g., Islamist extremism). Discourse networks Who were the most influential actors in pushing these frames and issues onto the media agenda? Discourse network analysis provides a bird’s eye view on the relationship between actors and issues in public debates (see Annex D in the Appendix for the measurement and statistics of network analysis). Figure 5shows the discourse networks after Islamist and extreme right attacks and indicates the driving forces in the debate. 15 The nodes of the directed network are the actors and the issues of the debate. 16 The connections between the nodes are the ties. The ties show the relationship between the actors and issues using colours indicating the average position: turquoise is positive (1), yellow is neutral (0), pink is negative (−1). The line width reflects the weight of the ties,defined as the number of statements towards the issue by the group of actors. JOURNAL OF EUROPEAN PUBLIC POLICY 3505 Figure 5. Actor-issue networks after. (a) Islamist attacks (top); (b) extreme right attacks (bottom). 3506 T. VÖLKER The network analysis shows that similar actor constellations emerged and dominated public debates after terrorist attacks. The degree of centrality of the nodes indicates which actors and issues played a central role in the debate. The out-degree coefficient indicates which actors had the highest number of connections and the in-degree coefficient measures which issues had the highest number of connections (see Tables D3 and D4 in the Appendix). In both cases, governmental actors and political parties had the highest out-degree and were the central actors driving these debates. They put the highest emphasis on security policies and on strengthening counter-terrorism. Regardless of the ideological motive for the event, there was a broad consensus in public debates that counterterrorism efforts should be expanded and state competencies widened. This supports findings by de Graaf (2011) that moments of crisis can foster processes of securitisation that encourage governments to expand security policies. Yet, the networks illustrate that there were different issues pushed onto the media agenda: after extreme right attacks, political actors put an emphasis on right-wing extremism, racism, antisemitism and radicalisation (the issues with the highest in-degree). The position of political parties on rightwing extremism reported in mass media was neutral on average. This was influenced by the fact that the debate centred on the claims of the AfD. Moreover, extreme right actors –and their racist, anti-Semitic and nativist claims – were at the centre of the debate. In contrast, after Islamist attacks, political parties put an emphasis on radicalisation and immigration issues and on average referred negatively to asylum issues. In contrast, they hardly ever referred to Islamist extremism. Looking at the dynamics in the debates, the network analysis shows that in both cases the debates were characterised by a low level of density and distinct actor alliances. The density coefficient indicates the share of actual connections from potential connections in the discourse network and can be interpreted as the coherence of the discourse network 17 (see Table D2 in the Appendix). In both cases, the discourse networks were characterised by a low level of density. After Islamist attacks, the discourse network was slightly more cohesive (0.085) than after extreme right attacks (0.078). This indicates that heterogenous and distinct actor alliances were engaged in a discursive competition to link the debates on the attacks to different issues. In sum, these findings demonstrate that terrorist attacks and their ideological motive had an impact on the agenda setting in public debates and reinforced existing actor alliances and power dynamics in public debates. The debates after terrorist attacks in Germany were characterised by different patterns: On the one hand, attacks that had Islamist motives were debated as an imported security threat that was connected to Islam and immigration. On the other hand, extreme right attacks were debated as a problem that was linked to a narrow outgroup of extremists. This JOURNAL OF EUROPEAN PUBLIC POLICY 3507 may be caused by the fact that victims, such as ethnic and religious minorities rarely gained any discursive space to share their perspective and influence public perception. Moreover, political elites played a central role in the public interpretation of these attacks and the diffusion of far-right frames, particularly the connection between security issues and immigration. This shows that the extent to which an ideology is historically rooted in the country seems to influence the debate about the causes and political consequences of the event. In line with my theoretical expectations, topdown mechanisms can explain actor alliances in public debates that were strongly path-dependent and reflected existing conflicts over migration, Islam and nationalism. Moreover, from a bottom-up perspective, the ideological motive for the event influenced public attention as well as the interpretation of the threat. Conclusion In this article, I have presented comparative evidence on the different ways in which extreme right and Islamist terrorist attacks influence public debates in Germany. Drawing on a new relational dataset based on a quantitative content analysis of newspapers combined with qualitative frame analysis and network analyses, I have compared mass media debates on Islamist and right-wing extremism. The results demonstrate the crucial role of ideologies for better understanding the political and democratic impact of terrorist attacks on Western democracies. The empirical results support the theoretical argument that the ideological motive for the terrorist attacks plays a central role in the public interpretation of the events in question and shapes the public framing of the security threat. The public resonance of actors and issues mirrors the issue-specific discursive opportunities in Germany. Overall, the attacks have reinforced rather than transformed existing conflicts and actor alliances in public debates. First, both types of terrorist attacks provide favourable conditions for the far right to influence the direction and content of the debate. In contrast, victims such as ethnic or religious minorities have little voice in public debates. Second, the ideological roots of the perpetrator influenced the resonance of issues and the interpretation of the central security threat and who or what is responsible for the terrorist attacks: After Islamist attacks, the majority of statements referred to the broader group of Muslims and immigration as the central security threat. In contrast, extreme right attacks were framed as incidents by ‘lone wolves’, portrayed as a narrow outgroup of isolated fringe actors. Third, the results stressed the crucial role of political elites for the public understanding of the central security threat and diffusion of far-right frames. 3508 T. VÖLKER The present study adds to previous research by highlighting the crucial role of ideology. The extent to which an ideology is historically rooted in the country plays a crucial role in the public perception of extremist actors and events. The high public resonance of far-right actors in the context of critical moments can lead to incremental shifts in public debates that enhance the normalisation of radical claims in the long-term. This is in line with previous research that showed that the success of the far right depends on favourable discursive opportunities (Castelli Gattinara & Froio, 2023; Della Porta et al., 2020). The discursive radicalisation model I have developed in this paper can serve as a tool to further analyse such mechanisms and explain how events can offer windows of opportunity for radical actors to gain public visibility, resonance, and legitimacy. As public debates are decisive for the success of radical actors and their effect on public perception and the political agenda, there is a need to further analyse the discursive impact of radical events from a comparative perspective, by comparing countries and different ideologies. This study sheds light on short-term dynamics in the mainstream media but as we live in an age of social media, it is necessary to integrate social media data and analyses of longterm trends to get a full picture of dynamics in the public sphere. In addition, further research is needed to cover a wider range of countries and cases in order to test more systematically how ideology relates to alternative factors such as the severity or location of the event. The fact that Islamist and extreme right attacks are treated differently in public debates and that victims gain no discursive space may bias public perception of security threats and have important implications for counterterrorism and the general approach to radicalisation in Germany and beyond. The reduction of the problem (of right-wing extremism) on the one hand and the expansion of the problem (of Islamism) on the other hand may induce disproportionate political reactions and prompt the formation of counterterrorism initiatives that do not solve the problem. Notes 1. Public legitimacy is not conceptualised as a legal or political term but as the publicly visible evaluation of a statement as reasonable and acceptable in public debates. It does not depend on resonance. For example, racist claims can gain a lot of resonance but a low degree of public legitimacy. 2. Ideology is defined according to Wilson as ‘a set of beliefs about the social world and how it operates, containing statements about the rightness of certain social arrangements and what action would be undertaken in the light of those statements’(Wilson, 1973, p. 91). 3. I use these ‘umbrella’definitions for comparative purposes, although there are different sub-types and ideological variation. For a comparison of the ideological features of Islamist and right-wing extremism see Annex A in the Appendix. JOURNAL OF EUROPEAN PUBLIC POLICY 3509 4. The extreme right attacks include the 2016 attack in Munich, the attacks on the politician Walter Lübcke and in Halle in 2019, and in Hanau in 2020. The Islamist attack include the attack in Berlin 2016, in Hamburg 2017 and in Dresden 2020. 5. The measures and attention to RTV vary in different countries. In Germany it includes attacks or discoveries of major arms repositories. 6. Online media platforms are not included as this study is particularly interested in the publicly visible forms of political violence in mass communication. 7. The research interest is the overall perception of public discourses, as the sum of communication processes produced in in the public sphere by multiple information providers. Therefore, the articles were not weighted. 8. This measurement is based on a quality variable –ranking between -1 (negative relationship) and 1 (positive relationship) and including 0.5 and -0.5 for a weak relationship and zero (0) as neutral. 9. Both categories do not include parties. 10. For example, after the extreme right attacks in Hanau 2020 Muslims were the central target but hardly visible in the debate. This changed in the following year when the #saytheirnames initiative drew attention to this issue and politicians and media reports referred more to individual victims. 11. This refers to the visibility of radical actors as the share of statements related to the extremists. The visibility of radical events follows a different pattern (see Figure 2). 12. In line with previous research on party politics in Germany, I classify the far-right AfD, right-wing liberal FDP and right-wing conservative CDU/CSU as right-wing parties and the social democrats SPD, green party Grüne and left party Die Linke as left-wing parties. 13. It is important to note that this legitimacy measure varies between the cases. Figure C3 in the Appendix illustrates the legitimacy shifts for each case separately. 14. Quotes have been translated by the author. 15. The analysis only includes actors with more than 9 statements. It is based on the force-directed Fruchterman Reingold layout algorithm (see Annex D, Appendix). 16. In a directed network, connections between nodes (ties) are directional. The arrow indicated the direction of the tie. 17. Modularity is the strength of division of a network into modules or clusters. Density is the intensity of interaction as the ratio of present interactions to the maximum possible interactions in a network (0 = no interactions to 1 = all possible interactions are present). Acknowledgements I would like to thank the anonymous reviewers for their excellent comments and Swen Hutter, Sophia Hunger, Lisa Zehnter, Daniel Saldivia Gonzatti, Pietro Castelli Gattinara, Manès Weisskircher, Donatella della Porta and Øyvind Bugge Solheim for their invaluable feedback. I am also grateful to all those who provided comments at the ECPR General Conference, the Political Behavior Workshop Berlin-Brandenburg, WZB Workshop ‘The Rise of Grassroots Radicalism and its Political Effects’and the Colloquium of the WZB Centre for Civil Society research. Special thanks go to the incredible team of research assistants involved in the WZB ProtestMonitoring: Mika Bauer, Antonia Kühmichel and Timon Imig. 3510 T. VÖLKER