University of Amsterdam Graduate School of Communication
Research Master’s programme in Communication Science
Investigating the Effects of Hate Speech Prosecution of an
Anti-Immigration Politician on the Propensity to Vote for His Party
Using Social Identity Theory
Miriam Tribastone 11368551
Master’s Thesis
Supervisor: Dr. Joost van Spanje Word count: 7,441
Abstract
Hate speech prosecution of anti-immigration politicians represents a controversial case in democracy because free speech is restricted to protect minority rights. Little is known,
however, about the consequences. A study by van Spanje and de Vreese (2015) observed that legal action against Wilders increased the electoral support for his party, the PVV, especially among voters who are moderately opposed to multiculturalism. The present study built upon and extended this by investigating under which conditions the decision to prosecute an anti-immigration politician for hate speech affects the propensity to vote for the corresponding party. Starting from hate speech’s intergroup nature, this research applied social identity theory to this field for the first time; it posited that hate speech prosecution could be perceived as a threat by certain voters and that this could affect the electoral support for the accused’s party. The moderating roles of group identification, relative deprivation, and attitude towards multiculturalism, and the mediation role of anger among assimilationists were hypothesized. An online experiment where participants (N = 414) were randomly exposed to one of two Facebook news posts about the decision to (a) prosecute or (b) not prosecute a politician for hate speech was conducted in Italy. Findings revealed that voter group identities did not moderate the effect of hate speech prosecution of an anti-immigration politician on the propensity to vote for his party. This research showed that the decision not to prosecute an anti-immigration politician for hate speech led to more anger than the decision to prosecute among those citizens who are moderately opposed to and those who are in favour of
multiculturalism. In line with previous studies, the decision not to prosecute could be perceived by these citizens as a threat to their expectations for democracy. Theoretical and practical implications for democracy are discussed.
Keywords: anti-immigration parties, hate speech prosecution, social identity theory, electoral utilities, multiculturalism.
Investigating the Effects of Hate Speech Prosecution of an Anti-Immigration Politician on the Propensity to Vote for His Party Using Social Identity Theory
Research on anti-immigration parties—those parties that stress immigration issues in electoral campaigns and aim to reduce the immigration influx to their countries, such as the Freedom Party (PVV) in the Netherlands and the League in Italy (Fennema, 1997; van Spanje, 2011)—has identified several factors that could determine their electoral success or failure(e.g., van der Brug, Fennema, & Tillie, 2005). Previous studies demonstrated that electoral support for these parties is influenced by voters’ ideological predispositions and preferences for restrictive immigration policies (e.g., van der Brug & Fennema, 2003), and their perceived legitimacy and effectiveness (Bos & van der Brug, 2010). Another factor that has gained attention is the legal actions taken against anti-immigration parties. Among these legal measures, hate speech prosecution has caused public debates in Western European countries where anti-immigration politicians have been prosecuted. A notorious example is the case of Geert Wilders, the PVV leader, who was convicted for hate speech against Moroccans in the Netherlands at a rally in 2014 (Vrielink, 2016). After asking his supporters if they wanted more or fewer Moroccans and they replied with fewer, he promised to arrange it (Vrielink, 2016).
Hate speech is defined as the expression, promotion, and incitement to harm,
discrimination, intolerance, and hostility based upon the target’s identification with a racial, ethnic, religious, gender, or national group (UNESCO, 2015). Since 1965, hate speech prosecution has been one of the legal measures that countries can take to eradicate
discrimination, based on the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD), as adopted by the United Nations. Legal foundations of hate speech prosecution are present at the national, European, and international levels (Fennema,
2000). Nevertheless, hate speech prosecution remains controversial because of the conflict between hate speech and freedom of expression in the democratic context (Fennema, 2000).
Research on the effects of hate speech prosecution of anti-immigration politicians is limited and little is known about the consequences of this legal action. Previous studies have distinguished between the intended effect, minority group protection, and unintended effects, such as decreased democratic consensus (Rekker & van Spanje, 2019; van Spanje & de Vreese, 2014; Wichgers, Jacobs, & van Spanje, 2019). Among the unintended effects, van Spanje and de Vreese (2015) observed that, in the case of Wilders, the decision to pursue hate speech prosecution bolstered the probability of voting for his party across the board and increased vote intention among those citizens who were moderately in favor of minorities’ assimilation into the national culture.
This study builds upon and extends the work by van Spanje and de Vreese (2015) by further investigating under which conditions the decision to prosecute an anti-immigration politician for hate speech boosts the propensity to vote for his party. Propensity to vote
concerns the empirical observation of the electoral utility, which voters derive from voting for a party (i.e., voter’s assessment of the utility to vote for a party, on which the choice is based; van der Eijk, van der Brug, Kroh, & Franklin, 2006). In contrast with, for instance, party choice, this measure provides information about the preferred party as well as the parties that were considered (van der Eijk et al., 2006). Moreover, propensity to vote was chosen because it is considered appropriate for multi-party systems such as in Italy and the Netherlands (van der Eijk et al., 2006). The research question underlying this study is the following
RQ: What are the effects of the decision to prosecute an anti-immigration politician
for hate speech on the propensity to vote for his party?
To answer the research question and contribute to the field, social identity theory is applied to hate speech prosecution of an anti-immigration politician for the first time. This
theory suits the controversial topic based on its intergroup nature, as per the UNESCO (2015) definition of hate speech. Moreover, Minkenberg (2006) highlighted that parties’ legal
repression could consolidate social identities as an unintended effect. Thus, it is argued that if the decision to prosecute a politician for hate speech is considered a social identity threat, this could help to explain the electoral effects as a way to restore a positive social identity.
The academic relevance of this study is fourfold. First, building upon previous findings (e.g., van Spanje & de Vreese, 2015) that hate speech prosecution’s effects tend to vary between citizens, the influencing factors at the citizen level are further investigated. By testing the moderating roles of group identification, perceived relative deprivation, and attitudes towards multiculturalism as well as the mediating role of anger among those who oppose multiculturalism (i.e., assimilationists) on the effects of hate speech prosecution on the propensity to vote for the politician’s party, a contribution to the literature is made. Secondly, this experimental study complements previous studies that have mainly used
non-experimental designs (e.g., van Spanje & de Vreese, 2015). This study tests the effects of the decision to prosecute versus the decision not to prosecute an anti-immigration politician for hate speech on the propensity to vote for his party. Thirdly, whereas previous studies on the effects of hate speech prosecution have focused on exposure to traditional news media such as TV newscast (Wichgers et al., 2019), this study considers social media for the first time, based on their recognized role as a source of (political) information in the literature (e.g., Gil de Zúñiga, Jung, & Valenzuela, 2012; Hermida, Fletcher, Korell, & Logan, 2012). Finally, this study analyzes hate speech prosecution in Italy: This country has not been considered thus far and represents a compelling case because, in comparison with the Netherlands, an anti-immigrant party (i.e., the League) is currently in the governing coalition. The
investigation of the Italian case could cross-validate the evidence found in the Netherlands by van Spanje and de Vreese (2015).
This study has notable societal relevance since it tackles the intended-unintended effects paradox of hate speech prosecution and contributes to the understanding of citizen-specific factors leading to the unintended effects. This study aims to provide evidence that could help researchers and practitioners understand hate speech prosecution and its effects.
Theoretical Framework
Social Identity Theory
Social identity theory has been widely applied to several contexts by different disciplines to explain intergroup phenomena and its academic relevance has increased in the age of multicultural societies (Brown, 2000).
People define themselves based on peculiarities, traits, roles, and characteristics in relation to the multiple social groups to which they belong (Deaux, Reid, Mizrahi, & Kathleen, 1995). The theory, formulated by Tajfel and Turner (1986), focuses on social identity in the intergroup context. According to the theory, individuals derive their social identity from group membership. Social identity is the part of the self-concept concerning the cognitive, evaluative, and affective components of group membership (Tajfel & Turner, 1986). The salience of group memberships varies as a function of the importance individuals’ attach to the groups and circumstances (Tajfel, 1978). A group is a collection of individuals that define themselves and are defined by others as group members; membership implies an emotional attachment to the group and a shared consensus about its evaluation (Tajfel & Turner, 1979). When individuals are categorized as members of another social group or in situations of intergroup conflicts, citizens tend to behave as members of their group, rather than as individuals (Tajfel & Turner, 1986). The individual will compare groups and
Turner, 1986). Indeed, social identity theory is “a theory of in-group love” (Brown, 2000, p.763).
Based on the intergroup nature of hate speech (UNESCO, 2015) and on the tension it creates between the reinforcement of an identity and the distinction from another (Waldron, 2012), social identity theory represents an appropriate framework to explain the controversial case of hate speech prosecution as a social identity threat (i.e., actions or communication that could undermine the value of group membership; Grant & Brown, 1995). In line with
Wichgers et al. (2019), hate speech prosecution could be perceived as a threat to the values, norms, beliefs, and practices of a larger group in society than the supporters of the prosecuted politician (i.e., those who identify with the person on trial).
The Moderating Role of Group Identification
Group identification is a central concept in social identity theory since it helps
individuals define their self-concept and understand their place in society; this is done through the mechanism of social categorization, which organizes individuals into groups (Tajfel & Turner, 1979). Group identification refers to the cognitive and affective attachment to a certain group in society (Hartmann & Tanis, 2013; Voci, 2006). When a group is salient, the individual’s self-concept will depend on the valence of this group membership (Voci, 2006). In the presence of a group threat, group membership will increase in its salience and the perceived differences between the ingroup and other groups in society will also increase (Voci, 2006).
Research on social identity theory has highlighted the relevance of group identification in predicting perceptions of and actions taken to re-establish a positive social identity (Doosje, Spears, & Ellemers, 2002). The study by Doosje et al. (2002) observed how the degree of identification influences group support in intergroup threatening situations; in particular, low identifiers are more prudent in supporting their group than high identifiers. The moderating
role of group identification was also observed in voting behavior (Ben-Bassat & Dahan, 2012; Fowler & Kam, 2007). In an attempt to improve the position of their group, citizens who highly identify with their group tend to participate more in politics and express a vote
preference for an ingroup member as compared to a member of another group (Ben-Bassat & Dahan, 2012; Fowler & Kam, 2007).
Based on Waldron’s (2012) argument that hate speech reinforces the ingroup identification and the discussed literature, hate speech prosecution could be perceived as a social identity threat by those who identify with the politician’s group, and this can affect the electoral support for the politician’s party as a means to restoring a positive social identity. The effect of hate speech prosecution on party support will vary between citizens who do not identify, those who to some extent identify, and those who strongly identify with the ingroup. Therefore, it is expected that group identification will moderate the effects of the decision (not) to prosecute an anti-immigration politician for hate speech on the propensity to vote for his party.
H1: The decision to prosecute an anti-immigration politician will lead to a higher propensity to vote for his party than the decision not to prosecute him, and this effect will be stronger as citizens identify more with the in-group.
This study explores the role of religious social identity, which is considered one of the most collectivist and therefore most suitable for the investigation of social identity theory (Deaux et al., 1995). The identification with Catholics is considered as it is a majoritarian group in Italy; 80% of Italians identify as Christians and specifically Catholics, and 78% of Christians in Italy declare that they are proud of this identity (Pew Research Center, 2018). In addition, this group is comparable to the group targeted by hate speech in this study (i.e., the Muslim minority). Group identification based on religion aligns with the idea on trial (rather
than the person or party on trial) since Islam is considered incompatible with Italian culture and values by half of Italians (Pew Research Center, 2018).
The Moderating Role of Perceived Relative Deprivation
The concept of relative deprivation is considered an extension of social identity theory (Walker & Pettigrew, 1984), however, it is not mutually exclusive of group identification. Relative deprivation is defined as the difference between the valuable resources an individual feels entitled to and the ones actually possessed (Tajfel, 1978). Since this is a perception based on relative standards, it is never objective (Crosby, 1976). The concept has an affective component (i.e., the feeling of displeasure, resentment, and vulnerability) and a cognitive component (i.e., the perceived violation of expectations; Grant & Brown, 1995; Mummendey, Kessler, Klink, & Mielke, 1999). In the intergroup context, relative deprivation arises from a negative social comparison: When individuals compare themselves to others and the outcome of this comparison is negative, this can undermine their social identity (Grant & Brown, 1995; Tajfel, 1978).
Despite recognizing the relevance of this concept, early literature on relative deprivation has failed to conceptualize and address it consistently, especially regarding the intergroup and intragroup levels of comparison (Walker & Pettigrew, 1984). A recent conceptualization clarified this issue and argued that perceived relative deprivation arises from the comparison between the individual as a member of the deprived people in society and the generalized others (see Elchardus & Spruyt, 2012; Elchardus & Spruyt, 2016). This study follows this conceptualization.
Recent studies highlighted the moderating role of perceived relative deprivation and found that citizens who feel relatively deprived tend to see society as divided into groups and subsequently embrace populism (Elchardus & Spruyt, 2016; Hameleers et al., 2017). Indeed, relative deprivation is a strategy and persuasive mechanism used to cope with vulnerability in
society; this is done by pinpointing and blaming another generalized group such as
immigrants, who are perceived as those who “steal” the resources the individual feels entitled to (Elchardus & Spruyt, 2012; Jetten, Mols, & Postmes, 2015). Deprived citizens tend to prefer right-wing, authoritarian, and ethnocentric political views of society (Elchardus & Spruyt, 2012), and tend to be attracted to populist discourse and parties (Elchardus & Spruyt, 2016).
Citizens who feel deprived could therefore perceive hate speech prosecution of an anti-immigration politician as a threat to their vulnerable group, and this could affect the electoral support for the party in an effort to improve their condition. Thus, the effect of hate speech prosecution on electoral support for the politician’s party will vary between citizens who do not feel deprived, those feeling deprived to some extent, and those feeling greatly deprived. In line with previous research, it is expected that perceived relative deprivation could moderate the relationship between hate speech prosecution of a politician and the propensity to vote for his party.
H2: The decision to prosecute an anti-immigration politician for hate speech will lead to a higher propensity to vote for his anti-immigration party than the decision not to prosecute him, and this effect will be stronger as citizens feel more relatively deprived.
The Moderating Role of Attitude Towards Multiculturalism
Hate speech prosecution of anti-immigration politicians frequently concerns the assimilationist view of minorities in society. Indeed in the case of anti-immigration
politicians, hate speech is often public criticism towards the multiculturalist ideal (van Spanje & de Vreese, 2015) based on the perceived incompatibility of the national culture and
foreigners’ culture (Fennema, 1997).
Attitude towards multiculturalism, defined as the extent to which citizens agree that ethnic minorities and foreigners should integrate into the country’s culture, is an established
determinant factor in this context1 (van Spanje & de Vreese, 2014). Indeed, previous studies on hate speech prosecution have shown that the strength of the effects varies in function of the attitude towards multiculturalism (Rekker & van Spanje, 2019; van Spanje & de Vreese, 2015; Wichgers et al., 2019).
Whereas van Spanje and de Vreese (2015) observed a general electoral lift-off for the politician’s party after the decision to prosecute him, the effects on vote intention were not significant among assimilationists and multiculturalists. This is likely because many
individuals in the first group already supported the party before the decision to prosecute the politician for hate speech, while most of the second group would never vote for this party, regardless. In contrast, a strong increase in vote intention for the party of the anti-immigration politician was observed among those citizens who were moderate assimilationists (van Spanje & de Vreese, 2015).
Based on the extant literature, it can be assumed that the decision to prosecute could be perceived by citizens who agree with the political idea on trial as a threat to their identity, ideals, and practices and that, in coping with the threat, the electoral support for the
politician’s party is influenced. It is expected that the effect of hate speech prosecution on the propensity to vote for the politician’s party will vary between citizens who strongly reject multiculturalism, those who moderately reject it, and those who support multiculturalism. Thus, it is hypothesized that attitudes towards multiculturalism moderate the effect of hate speech prosecution of an anti-immigration politician on the propensity to vote for his party:
H3: The decision to prosecute an anti-immigration politician for hate speech will lead to a higher propensity to vote for his anti-immigration party than the decision not to prosecute, and this effect will be stronger for citizens who moderately oppose multiculturalism than citizens who support it and citizens who oppose
The Mediating Role of Anger Among Assimilationists
The mediating role of emotions is well established in communication science and other social disciplines (e.g., Lecheler, Schuck, & de Vreese, 2013). Research has highlighted that emotional reactions can influence information processing and decision making
(Valentino, Brader, Groenendyk, Gregorowicz, & Hutchings, 2011). Among emotions, anger is relevant to this study because of its behavioral implications. Anger is defined as the
outcome of a perceived offense by a recognized external agent to an individual or the individual’s group (Arpan & Nabi, 2011). Anger also arises from a perceived threat and implies a sense of certainty and control over the situation (Lerner & Keltner, 2001; Valentino et al., 2011).
One of the main characteristics of this emotion is the consequential action tendency to restore a comfortable state (Arpan & Nabi, 2011; Lerner & Keltner, 2001). Angry individuals tend to make optimistic choices (Lerner & Keltner, 2001) and are motivated to reach a
positive outcome (Valentino et al., 2011). Anger is considered a functional emotional reaction (Lecheler et al., 2013); in the political context, it has been found to be a powerful mobilizer (Valentino et al., 2011; Valentino, Gregorowicz, & Groenendyk, 2009). When citizens feel angry, they tend to participate more in elections, even engaging in costly activities they might not otherwise (Valentino et al., 2011).
In the controversial case of hate speech prosecution, it is hypothesized that individuals who agree with the idea on trial could experience anger in reference to the situation. Among assimilationists, it is argued that the decision to prosecute an anti-immigration politician for hate speech could arouse anger, which can then influence the electoral support for the politician’s party.
H4: Anger mediates the effect of the decision to prosecute an anti-immigration politician for hate speech on the propensity to vote for his party among citizens with a negative attitude towards multiculturalism.
The theoretical framework and hypotheses are presented in the model in Figure 1.
Figure 1. Theoretical model.
Method
The Case Under Investigation
This study examines the effects of news exposure on the decision (not) to prosecute an anti-immigration politician for hate speech on the propensity to vote for his party. Social media and, in particular, Facebook is considered because it is the most popular social medium in Italy with 34 million users (Mazzoleni & Bracciale, 2018). In addition, Facebook is the most used social medium to retrieve news: 64% of Italians use Facebook to inform themselves, with 50% retrieving news on a daily base (Pew Research Center, 2018).
Among the news media outlets present on Facebook, Rai News, an Italian all-news public service channel, was chosen; their official account on this platform is “RaiNews.it.” At the time of data collection, Rai News was reported as the most trusted news organization, with a perceived central placement on a left-right scale by its audience and citizens who have heard
of it, and the main news source, with 69% of Italians who turn to it for the news at least once a week (Pew Research Center, 2018). Thus, it was considered suitable for this study.
Regarding the anti-immigration party, the League was chosen because it is a different case as compared to newer Dutch anti-immigration parties, such as the PVV and the Forum for Democracy (FvD), that have been considered in previous studies on the topic (e.g., van Spanje & de Vreese, 2015; Wichgers et al., 2019). Indeed, the League is an established anti-immigration party that was already in power in 1994, 2001, and 2008 (Calossi & Cicchi, 2018). After a transformation from a party with a strong regional identity (i.e., the Northern League) to one with a firm national character (i.e., the League), the League’s votes
quadrupled, with unprecedent results in Central and Southern Italy in the 2018 General Elections (Cerruto & Raniolo, 2018). Currently, it is governing alongside the Five Stars Movement (Calossi & Cicchi, 2018). The investigation of the League case could cross-validate the effects found in the Netherlands.
Following Wichgers et al. (2019), the second man of the party, Giancarlo Giorgetti, instead of the party leader, Matteo Salvini, was investigated to reduce the influence of other potential factors, such as strong pre-existing opinions or extensive news coverage about him. Salvini, along with other party members, has previously been legally targeted for hate speech against Roma and immigrants, while Giorgetti has never been accused of hate speech. In addition, he is suitable for the manipulation because, despite his established influence in the party, he has attracted only limited media attention. Thus, participants are likely to perceive the situation about him as credible.
Muslims were selected as the target group of the hate speech because they are often criticized by the League and, at the time of data collection, no high-profile legal controversy concerning hate speech has involved them as a group targeted by this party. This decision
could exclude the potential influence of other factors, such as the media coverage on the cases of hate speech against Roma and immigrants.
Experimental Design
An online survey-embedded experiment with a between-subject design was conducted. The between-subject variable, decision to prosecute, had two categories, decision to prosecute vs. decision not to prosecute. Participants were randomly assigned to one of these two
experimental groups. Since the context under investigation is Italy, the entire experiment was programmed in Qualtrics in Italian. Data were collected2 between December 6th-16th, 2018.
Participants
A convenience sample of 414 Italians eligible to vote took part in the online experiment. Participants were primarily recruited from researcher’s personal network through private messages and public posts on social media (e.g. Facebook, Twitter, and WhatsApp). To ensure diversity in the sample, participants were encouraged to share the link with family and friends. The sample (M age = 32.57, SD age = 13.75, ranging from 18 to 75 years, 52.2 % female, 7.3% no high school degree, 42% high school degree, 50.7% university degree) was mainly from Southern Italy (25.7 % North Italy, 7% Central Italy) and slightly left-leaning (on an 11-point left-right scale, M = 4.52, SD = 2.60). No compensation was offered for the participation in the study.
Procedure
Participants accessed the study in Qualtrics with an anonymous link. They were first presented with the factsheet and asked to give their informed consent. The formal study began with questions regarding the moderators (i.e., group identification, relative deprivation, and attitude towards multiculturalism). Then, participants were exposed to three posts. The first two were buffer posts by Rai News about the weather and cohousing; these related to buffer
questions that were included to veil the goal of the study. These two topics were selected because they are both emotionally neutral and newsworthy. Participants were randomly assigned to one of two stimuli. To ensure that participants would read the posts, a minimum reading time of eight seconds per post was set. In the third part of the study, the mediator (i.e., anger), the dependent variable, the manipulation checks, and demographics were measured. On the final page, participants were thanked and debriefed (full questionnaire in Appendix A). The experiment required approximately 10 minutes to complete.
Stimuli
Participants were exposed to three Facebook posts by RaiNews.it, of which two were buffer materials and one was the stimulus. All three original posts3 were retrieved from the official Facebook account of Rai News. To increase ecological validity, only necessary changes to ensure comparability and the success of the study were made in Photoshop (e.g., temporal clues were removed and posts were edited to say that they were published one hour earlier). The independent variable, the decision to prosecute an anti-immigration politician for hate speech, was manipulated in a Facebook post by Rai News as follows—one version of the post reported that the Public Prosecutor decided to prosecute Giorgetti for hate speech against Muslims, while the other version reported that the Public Prosecutor decided not to
prosecute Giorgetti for hate speech against Muslims (see Appendix B). After an analysis of
the channel’s content, the title and the caption were re-created in the experimental materials to be as similar as possible to real posts: This not only implied the use of the same font and size, but also the use of Italian legal terms that are used in media coverage. To ensure that all participants would understand the message regardless of their educational level, the main message of the title was rephrased in two other ways in the caption following the practice of several news media organizations operating on Facebook (e.g., quoting the Public
the reference to the party, the reactions, shares, and comments to the posts) were held constant.
Measures
Propensity to vote. The dependent variable was the propensity to vote for the League.
To assess it, the measure by van der Eijk, van der Brug, Kroh, & Franklin (2006) was used, which asked participants to indicate on a scale from 1 to 10 the probability that they would ever vote for the League (MLeague = 3.20, SDLeague = 3), and for four other main parties in the Italian system (i.e., Five Stars Movement, Democratic Party, Forward Italy, and Brothers of Italy).
Group identification. The scale by Leach et al. (2008) that was chosen for group
identification distinguishes between two dimensions (i.e., ingroup-related self-investment and ingroup-related self-definition), which are composed respectively of solidarity, satisfaction, centrality, and individual self-stereotyping and ingroup homogeneity. For each of these five components, one item from their original scale was adapted (e.g., I feel a bond with Catholics - full operationalization in Appendix A). The answer scale ranged from (1) not at all to (7) very much. Identification with Catholics was computed as the average of the five scores (Cronbach’s α = .89, M = 3.69, SD = 1.75).
Perceived relative deprivation. The 3-item scale by Hameleers, Bos, & de Vreese
(2017) was chosen to gauge relative deprivation. Participants were asked how much they agree with the items (e.g. if we need anything from the government, ordinary people like us always have to wait longer than others - full operationalization in Appendix A), using a 7-point scale from completely disagree to completely agree. This perception score was calculated as the mean of the three items (Cronbach’s α = .76, M = 3.91, SD = 1.39).
Attitude towards multiculturalism. To measure participants’ attitude towards
multiculturalism, the 4-item scale from Rekker & van Spanje (2019) was used (e.g.,
immigrants and ethnic minorities must fully adapt to Italian culture - full operationalization in Appendix A). The 7-point answer scale ranged from completely disagree to completely agree. To compute the score, two items were first reversed and then the four scores were averaged. A high score on this new variable meant a positive attitude towards multiculturalism
(Cronbach’s α = .624, M = 4.63, SD = 1.30).
Anger. This emotion was measured using three items from Mackie, Devos, & Smith
(2000). Participants were asked to report if they felt furious, displeased, and angry on a scale from (1) not at all to (7) very much. Scores were then averaged to create a new variable (Cronbach’s α = .79, M = 3.22, SD = 1.49). These items were presented together with other discrete emotions, such as enthusiasm, taken from Lecheler, Schuck, & de Vreese (2013).
Manipulation Checks
To test if the manipulation worked as intended, four questions asked about the politician, his party, the probability that he will be prosecuted for hate speech, and about the Public Prosecutor’s decision. Of the sample, 91.5% recognized Giorgetti and 66.9%
recognized his League membership. Participants in the condition where he was not prosecuted (M = 5.88, SD = 1.39) reported a higher probability that the politician would not be
prosecuted compared to participants in the prosecute condition (M = 4.98, SD =1.81). This mean difference, -0.905, t(332) = - 5.352, p < .001, 95% CI [-1.24, -0.57] was significant; therefore, it can be concluded that the manipulation worked. Regarding the decision to prosecute the politician, only 54.1% of participants in the prosecute condition indicated the decision to prosecute, whereas 71.2% of participants in the not prosecute condition indicated the decision not to prosecute.
Randomization Check
There were no significant effects between the prosecute condition (N = 208) and the not prosecute condition (N = 206) for age, t(402) = - .80, p = .425, gender, χ2 (1, N = 412) = .09, education, χ2 (7, N = 412) = 8.50, macro-region, χ2 (2, N = 412) = 1.18, or possession of a Facebook account, χ2 (1, N = 412) = .00. A significant effect between condition was found for left-right ideological position, t(409) = 2.12, p = .034, with more right-wing participants in the prosecute condition (M = 4.79, SD = 2.70) than in the not prosecute condition (M = 4.25, SD = 2.49), therefore ideological position will be included as a covariate in the analysis.
Data Analysis
To answer the research question and test the hypotheses, indirect effect analyses using SPSS PROCESS macro by Hayes (2013) with 95% bias-corrected bootstrap confidence intervals based on 10,000 bootstrap samples were conducted. Model 1 was used to test H1, H2, and H3 for moderation, while model 8 was used to test H4 for moderated mediation. The two experimental groups were dummy coded (1 = prosecute, 0 = not prosecute) and used as the criterion variable, while the dependent variable is the propensity to vote for the League. Predictors and moderators were automatically mean centred by PROCESS5. Simple slopes were examined for high (i.e., one SD above the mean), moderate (i.e., the mean), and low (i.e., one SD below the mean) values of the moderators. Left-right ideological orientation was included as covariate.
Results
All results are presented in Table 1 and Table 2.
Hypothesis Testing
H1. The first hypothesis assumed that the decision to prosecute would lead to a higher
would be stronger as the identification with Catholics increased. In contrast with expectations, there was no moderating effect of group identification between the decision to prosecute the politician and the propensity to vote for his party, b = -0.12, t = -0.89, p =.375, CI [-0.39, 0.15]. This means that as voters identified more with the ingroup, they were not significantly more susceptible to the effect of the decision to prosecute on their propensity to vote than for the effect of the decision not to prosecute. Therefore, H1 was rejected6.
H2. The second hypothesis concerned the moderating effect of perceived relative
deprivation between the decision to prosecute the politician for hate speech and the propensity to vote for his party. Results did not yield a significant interaction effect of the decision to prosecute the politician and perceived relative deprivation on propensity to vote, b = 0.10, t = 0.59, p = .556, CI [-0.24, 0.44]. That is, the decision to prosecute did not lead to a higher propensity to vote for the politician’s party than the decision not to prosecute, and this effect was not stronger as citizens felt more relatively deprived. Based on these findings, H2 could not be accepted7.
H3. The third hypothesis stated that participants exposed to the decision to prosecute a
politician for hate speech would report a higher propensity to vote for his party than participants exposed to the decision not to prosecute, and this effect would be stronger for moderate assimilationists than for assimilationists and multiculturalists. The interaction effect of hate speech prosecution and attitude towards multiculturalism on the propension to vote for the League, however, did not yield a significant result, b = 0.10, t = 0.57, p = .572, CI [-0.24, 0.43]. In other words, the effect of the decision to prosecute the anti-immigration politician for hate speech on the propensity to vote for his party was not stronger among moderate assimilationists. Thus, H3 was rejected8.
H4. The fourth hypothesis regarded anger as a mediator between the decision to
party among those citizens with a negative attitude towards multiculturalism. The data revealed that the effect of hate speech prosecution on anger was conditional on the attitude towards multiculturalism, b = -0.23, t = -1.97, p = .049, CI [-0.47, -0.00]. As shown in both Figure 2 and Table 2, the moderated effect was significant among moderate assimilationists (i.e., the group with a mean score), b = -0.31, t = -2.10, p = .036, CI [-0.59, -0.02], and multiculturalists (i.e., the group with a score of one SD above the mean), b = -0.62, t = -2.93, p = .004, CI [-1.03, -0.20]. In other words, for each additional point on the 7-point anger scale, the decision not to prosecute the politician for hate speech increased anger by 0.31 among moderate assimilationists and 0.62 among multiculturalists. This means that, among individuals with a (moderately) positive attitude towards multiculturalism, being exposed to news about the decision not to prosecute an anti-immigration politician led to more anger than being exposed to news about the decision to prosecute. Overall, the predictors explained 4% of the variance in anger. Contradicting H4, anger did not mediate the effect of hate speech prosecution on the propensity to vote for the politician’s party among those with a negative attitude towards multiculturalism, index of moderated mediation = –0.00, SE = .02, CI [–0.06, 0.04]. Therefore, H4 could not be confirmed9.
Table 1.
Predictors of Propensity to Vote for the League
Propensity to vote for the League Model 1 b (SE) Model 2 b (SE) Model 3 b (SE) Model 4 b (SE) Constant 0.37 (0.22) 0.01 (0.20) 0.50 (0.23)* 0.48 (0.37) Decision to prosecute -0.33 (0.22) -0.30 (0.23) - 0.30 (0.22) - 0.30 (0.23) Group identification 0.38 (0.07)*** Decision x Group identification -0.12 (0.14) Relative Deprivation 0.19 (0.09)* Decision x Relative Deprivation 0.10 (0.17) Attitude towards multiculturalism - 0.56 (0.10)*** - 0.56(0.10)*** Decision x Attitude towards multiculturalism 0.10 (0.17) 0.10 (0.17) Anger 0.01(0.8) Ideological orientation (control) 0.63 (0.05)*** 0.70 (0.05)*** 0.60 (0.06)*** 0.60 (0.06)*** R2 .43 .39 .43 .43 F 72.99*** 61.24*** 79.39*** 63.53***
Note. N = 414. Unstandardized coefficients with standard errors between parentheses.
+
Figure 2. Interaction effect of hate speech prosecution and attitude towards multiculturalism on anger. Table 2. Predictors of Anger Anger Model 4 b (SE) Constant 3.46 (0.17)***
Decision to prosecute the politician -0.31 (0.15)*
Attitude towards multiculturalism 0.07 (0.07)
Decision x Attitude towards multiculturalism -0.24 (0.12)* Assimilationists Moderates Multiculturalists 0.01 (0.21) -0.31 (0.14)* -0.62 (0.21)**
Ideological orientation (control) -0.06 (.03)
R2 .04
F 4.31**
Note. N = 414. Unstandardized coefficients with standard errors between parentheses.
+
Figure 3. Model with unstandardized coefficients Note. +p < .10; *p < .05; **p < .01; ***p < .001
Discussion
Hate speech prosecution represents a controversial case in liberal democracies because of the tension between freedom of expression and minority protection (Fennema, 2000).
Research on the effects of this legal action against anti-immigration politicians distinguishes between the intended protection of minorities and the unintended consequences, such as decreased democratic support (e.g., van Spanje & de Vreese, 2014) and the increased electoral appeal of the prosecuted politician’s party among those who agree with the idea on trial (van Spanje & de Vreese, 2015). Building upon and extending the work of van Spanje and de Vreese (2015), this study examined under which conditions exposure to hate speech prosecution of an anti-immigration politician affects the propensity to vote for his party. A novel contribution, social identity theory was applied to this field for the first time. It was posited that, since hate speech is an intergroup phenomenon (UNESCO, 2015), the
prosecution of an anti-immigration politician for hate speech could be perceived as an ingroup threat by citizens who identify with the politician’s group, and this could affect the propensity to vote for his party as a way to cope with this threat. The moderated effects of group
identification, perceived relative deprivation, and attitude towards multiculturalism, and the moderated mediated effect of anger among assimilationists were tested in this study.
Furthermore, the Italian context was investigated by considering the case of the anti-immigration party the League.
Discussion of the Results
Findings revealed that, in contrast with the hypothesized mechanisms, the effects of the decision to prosecute a politician for hate speech on the propensity to vote for his party were not moderated by group identification, perceived relative deprivation, or attitude towards multiculturalism10. This means that hate speech prosecution does not seem to be perceived as an ingroup threat by those citizens who identify with the politician’s groups (i.e., those who identify with Catholics, those feeling relatively deprived, and those citizens who argue moderately in favor of cultural unity), and electoral support for the politician’s party is not used as a means to restore the threatened social identity. In contrast with van Spanje and de Vreese (2015), this study did not observe the effect of hate speech prosecution on electoral support among moderate assimilationists. A potential reason could be the different
methodologies used to investigate the phenomenon. The panel study conducted by van Spanje and de Vreese (2015) before and after the court decision to prosecute Wilders for hate speech in 2009 had a higher external validity but less control over the potential influence of other factors, such as media attention that proved to play an important role in this phenomenon (see Jacobs and van Spanje, 2019), than this experiment. Opposite of expectations and existing literature on the mobilizing effect of anger (e.g. Valentino et al., 2011), a mediating role of anger between hate speech prosecution and the propensity to vote for the League among assimilationists was not found. Findings, however, did reveal a significant effect of hate speech prosecution on anger, and this effect strengthened as the attitude towards
multiculturalism became more positive. Among moderate opponents and proponents of multiculturalism society, the decision not to prosecute a politician for hate speech led to more anger than the decision to prosecute. Considering the subtle manipulation, the effect size is
notable among multiculturalists; this statistic indicated that about 9% of anger was explained, and this did not change when participants who failed the manipulation checks were excluded from the analysis. This finding is in line with previous studies on the effects of hate speech prosecution, which showed that the effects are conditional to citizen’s level of agreement with the criticism expressed by the politician and, specifically, that the decision not to prosecute an anti-immigration politician for hate speech decreases the satisfaction, evaluation, and
confidence in the democratic system among multiculturalists (van Spanje & de Vreese, 2014; Wichgers et al., 2019). A possible interpretation of the effect of hate speech prosecution on anger under the influence of attitude towards multiculturalism is based on citizens’ broken expectations and demands to the functioning of democracy. In the case of moderate
assimilationists and multiculturalists, the decision not to prosecute the politicians could lead to anger because of the failed defense of minorities, which is considered important for democracy. The interpretation of these effects as a threat to citizens’ expectations for democracy is also supported by the literature on emotions. Research has seen anger as a response to a perceived offense or threat by a known source (Arpan & Nabi, 2011). Indeed, anger could be a form of discontent towards authorities for not meeting citizens’ demands.
Limitations and Future Research
When interpreting the findings of this study, it is important to consider some limitations. First, this research investigated the effects of the decision to prosecute; therefore, it is uncertain if the findings would be the same when the prosecution itself is the event at stake. Second, the sample used was not representative of Italian voters. Even though special
attention was paid towards recruiting a large and diverse sample, participants were primarily younger, from Southern Italy, slightly left-leaning, and had higher levels of education. Other groups of voters (e.g., citizens with lower levels of education), were underrepresented in the sample and this could have been a reason why electoral effects were not found. Thus, further
investigation of these effects using a representative sample is recommended. Another element to consider when interpreting the findings of the present investigation is the experimental design, which represents both a strength and a limitation. On the one hand, this experiment complemented findings of previous studies (e.g., van Spanje and de Vreese, 2015) that have primarily used non-experimental designs and went beyond them by investigating the role of emotions. This method has a high internal validity due to the control over the experimental situation, which adds confidence to the evidence found in this experiment. On the other hand, although efforts were made to increase external validity (e.g., existing Facebook posts were manipulated following established media practices), generalizability to a real-world context could be limited. In contrast with this experiment, in real life citizens could be exposed to extensive media coverage on this topic (Vrielink, 2016), which could play a role in this phenomenon. To advance this field, future studies should tackle the effect of hate speech prosecution of an anti-immigration politician on democracy in more natural circumstances. Moreover, given that the evidence among multiculturalists found both in this study and similar previous studies by van Spanje and de Vreese (2014) and Wichgers and colleagues (2019) cannot provide answers about how long the effects last, it is urged that the design of future studies allow for testing the duration of the effects. Indeed, to understand the effects of hate speech prosecution, it is fundamental to know if the effects last or disappear quickly. Considering the challenges implied by this kind of research on this topic, especially when it comes to the investigation of emotions (e.g., scarce availability of secondary data), the investigation of the effects of hate speech prosecution using panel surveys would be highly appropriate. Panels are used in unpredictable events (e.g., Boomgaarden & de Vreese, 2007) such as hate speech prosecution, present high external validity, and allow for the evaluation of the effects’ duration (Bryman, 2012).
These limitations notwithstanding, this research has important implications. This study was one of, if not the first to link emotions and, specifically, anger to hate speech prosecution. This intriguing result, that those who do not agree with the accused’s idea showed
significantly more anger when he is not prosecuted than when he is, provides a key
contribution to the understanding of hate speech prosecution effects. In addition, the evidence found for anger validates previous studies’ findings that the effects of hate speech prosecution are moderated by the extent to which citizens disagree with the accused’s statement, given that this conclusion is similar across different countries, anti-immigration parties, and research methodologies.
The practical implications of this research are highly relevant for democracy. Public debates, authorities, academics, anti-immigration parties, and even ethnic minorities could benefit from the findings that the decision not to prosecute an anti-immigration politician for hate speech could negatively affect democracy by making angry a significant group of citizens who do not agree with the idea on trial and demand minority protection from authorities.
Conclusion
All in all, hate speech prosecution of an anti-immigration politician could not affect the propensity to vote for his party, but it does have an important impact on emotions, specifically anger. It is posited that, when this legal action is not undertaken, a significant number of citizens could perceive that their demands for functioning democracy are not heard by authorities. While the understanding of hate speech prosecution effects is far from
complete, this study provided evidence that the decision not to prosecute an anti-immigration politician for hate speech could be dangerous for democracy.
References
Arpan, L. M., & Nabi, R. L. (2011). Exploring anger in the hostile media process: Effects on news preferences and source evaluation. Journalism and Mass Communication
Quarterly, 88(1), 5–22. doi:10.1177/107769901108800101
Ben-Bassat, A., & Dahan, M. (2012). Social identity and voting behavior. Public Choice, 151(1–2), 193–214. doi:10.1007/s11127-010-9742-2
Boomgaarden, H. G., & de Vreese, C. H. (2007). Dramatic real-world events and public opinion dynamics: Media coverage and its impact on public reactions to an assassination. International Journal of Public Opinion Research, 19(3), 354–366.
Bos, L., & van der Brug, W. (2010). Public images of leaders of anti-immigration parties: Perceptions of legitimacy and effectiveness. Party Politics, 16(6), 777–799.
doi:10.1177/1354068809346004
Brown, R. (2000). Social identity theory: Past achievements, current problems and future challenges. European Journal of Social Psychology, 30(6), 745–778. doi:10.1002/1099-0992(200011/12)30:6<745::AID-EJSP24>3.0.CO;2-O
Bryman, A. (2012). Social Research Methods (4th edition). Oxford, England: Oxford University Press
Calossi, E., & Cicchi, L. (2018). The Italian party system’s three functional arenas after the 2018 election: The tsunami after the earthquake. Journal of Modern Italian Studies, 23(4), 437–459. doi:10.1080/1354571X.2018.1500215
Cerruto, M., & Raniolo, F. (2018). From exchange to voice. Voting in southern Italy. Journal of Modern Italian Studies, 23(4), 418–436. doi:10.1080/1354571X.2018.1500214 Crosby, F. (1976). A model of egoistical relative deprivation. Psychological Review, 83(2),
85–113. doi:10.1037/0033-295X.83.2.85
Journal of Personality and Social Psychology, 68(2), 280–291. doi:10.1007/JHEP04(2014)104
Doosje, B., Spears, R., & Ellemers, N. (2002). Social identity as both cause and effect: The development of group identification in response to anticipated and actual changes in the intergroup status hierarchy. British Journal of Political Science, 41, 57–76.
Elchardus, M., & Spruyt, B. (2012). The contemporary contradictions of egalitarianism: An empirical analysis of the relationship between the old and new left/right alignments. European Political Science Review, 4(2), 217–239. doi:10.1017/S1755773911000178 Elchardus, M., & Spruyt, B. (2016). Populism, persistent republicanism and declinism: An empirical analysis of populism as a thin ideology. Government and Opposition, 51(1), 111–133. doi:10.1017/gov.2014.27
Fennema, M. (1997). Some conceptual issues and anti-immigrant parties in Western Europe. Party Politics, 3(4), 473–492.
Fennema, M. (2000). Legal repression of extreme-right parties and racial discrimination. In R. Koopmans & P. Statham (Eds.), Challenging Immigration and Ethnic Relations Politics (pp. 119–144). Oxford: Oxford University Press.
Field, A. (2013). Discovering Statistics Using SPSS (4th edition). London: SAGE Publications.
Fowler, J., & Kam, C. (2007). Beyond the self: Social identity, altruism, and political
participation. Social Science Research Network Working Paper Series, 69(3), 813–827. Gil de Zúñiga, H., Jung, N., & Valenzuela, S. (2012). Social media use for news and
individuals’ social capital, civic engagement and political participation. Journal of Computer-Mediated Communication, 17(3), 319–336.
doi:10.1111/j.1083-6101.2012.01574.x
relative deprivation and threats to social identity. Social Psychology Quarterly, 58(3), 195–212.
Hameleers, M., Bos, L., & de Vreese, C. H. (2017). The appeal of media populism: The media preferences of citizens with populist attitudes. Mass Communication and Society, 20(4), 481–504. doi:10.1080/15205436.2017.1291817
Hartmann, T., & Tanis, M. (2013). Examining the hostile media effect as an intergroup phenomenon: The role of intergroup identification and status. Journal of
Communication, 75(S1), 409–410.
Hayes, A. F. (2013). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York City, NY: Guilford Press.
Hermida, A., Fletcher, F., Korell, D., & Logan, D. (2012). Share, like, recommend: Decoding the social media news consumer. Journalism Studies, 13(5–6), 815–824.
doi:10.1080/1461670X.2012.664430
ICERD (1965). United Nations International Convention on the Elimination of Racial Discrimination. Retrieved from http://legal.un.org/avl/ha/cerd/cerd.html
Jacobs, L., & van Spanje, J. (2019). Prosecuted, but popular? Hate speech prosecution of anti-immigrant politicians in the news and electoral support. Manuscript in preparation. Jetten, J., Mols, F., & Postmes, T. (2015). Relative deprivation and relative wealth enhances
anti-immigrant sentiments: The V-curve re-examined. PLoS ONE.
Leach, C. W., van Zomeren, M., Zebel, S., Vliek, M. L. W., Pennekamp, S. F., Doosje, B., … Spears, R. (2008). Group-level self-definition and self-investment: A hierarchical
(multicomponent) model of in-group identification. Journal of Personality and Social Psychology, 95(1), 144–165. doi:10.1037/0022-3514.95.1.144
Lecheler, S., Schuck, A. R. T., & de Vreese, C. H. (2013). Dealing with feelings: Positive and negative discrete emotions as mediators of news framing effects. Communications,
38(2), 189–209. doi:10.1515/commun-2013-0011
Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality and Social Psychology, 81(1), 146-159. doi:10.1037/0022-3514.81.1.146
Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining offensive action tendencies in an intergroup context. Journal of Personality and Social
Psychology, 79(4), 602–616. doi:10.1037/0022-3514.79.4.602
Mazzoleni, G., & Bracciale, R. (2018). Socially mediated populism: The communicative strategies of political leaders on Facebook. Palgrave Communications, 4(50), 2-10. doi:10.1057/s41599-018-0104-x
Minkenberg, M. (2006). Repression and reaction: Militant democracy and the radical right in Germany and France. Patterns of Prejudice, 40(1), 25–44.
doi:10.1080/00313220500482662
Mummendey, A., Kessler, T., Klink, A., & Mielke, R. (1999). Strategies to cope with negative social identity: Predictions by social identity theory and relative deprivation theory. Journal of Personality and Social Psychology, 76(2), 229–245.
Pew Research Center (2018, May). Being Christian in Western Europe. Religious identity. Retrieved from http://www.pewforum.org/2018/05/29/religious-identity/
Pew Research Center (2018, May). In Western Europe, public attitudes toward news media more divided by populist views than left-right ideology. Retrieved from
http://www.journalism.org/2018/05/14/in-western-europe-public-attitudes-toward-news-media-more-divided-by-populist-views-than-left-right-ideology/
Rekker, R., & van Spanje, J. (2019). How hate speech prosecution of politicians undermines support for the legal system and democracy. Manuscript in preparation.
Tajfel, H. (1978). Social categorization, social identity and social comparison. In H. Tajfel (Ed.), Differentiation Between Social Groups: Studies in the Social Psychology of
Intergroup Relations (pp.60-76). Oxford, England: Academic Press.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The Social Psychology of Intergroup Relations (pp. 33–47). Pacific Grove, CA: Brooks/Cole Publishing Company.
Tafjel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. Psychology of Intergroup Relations, 7-24.
UNESCO (2015). Countering Online Hate Speech. Retrieved from https://unesdoc.unesco.org/ark:/48223/pf0000233231
Valentino, N. A., Brader, T., Groenendyk, E. W., Gregorowicz, K., & Hutchings, V. L. (2011). Election night’s alright for fighting: The role of emotions in political
participation. Journal of Politics, 73(1), 156–170. doi:10.1017/S0022381610000939 Valentino, N. A., Gregorowicz, K., & Groenendyk, E. W. (2009). Efficacy, emotions and the
habit of participation. Political Behavior, 31(3), 307–330. doi:10.1007/s11109-008-9076-7
van der Brug, W., & Fennema, M. (2003). Protest or mainstream ? How the European anti-immigrant parties developed into two separate groups by 1999. European Journal of Political Research, 42, 55–76.
van der Brug, W., Fennema, M., & Tillie, J. (2005). Why some anti-immigrant parties fail and others succeed a two-step model of aggregate electoral support. Comparative Political Studies, 38(5), 537–573. doi:10.1177/0010414004273928
van der Eijk, C., van der Brug, W., Kroh, M., & Franklin, M. (2006). Rethinking the dependent variable in voting behavior: On the measurement and analysis of electoral utilities. Electoral Studies, 25(3), 424–447. doi:10.1016/j.electstud.2005.06.012
van Spanje, J. (2011). The wrong and the right: A comparative analysis of “anti-immigration” and “far right” parties. Government and Opposition, 46(3), 293–320.
doi:j.1477-7053.2011.01340.x
van Spanje, J., & de Vreese, C. (2015). The good, the bad and the voter: The impact of hate speech prosecution of a politician on electoral support for his party. Party Politics, 21(1), 115–130. doi:10.1177/1354068812472553
van Spanje, J., & de Vreese, C. (2014). The way democracy works: The impact of hate speech prosecution of a politician on citizens’ satisfaction with democratic performance.
International Journal of Public Opinion Research, 26(4), 501–516. doi:10.1093/ijpor/edt039
Voci, A. (2006). The link between identification and in-group favouritism: Effects of threat to social identity and trust-related emotions. British Journal of Social Psychology, 45(2), 265–284. doi:10.1348/014466605X52245
Vrielink, J. (2016). Do we want more or fewer prosecutions of opinions: The Geert Wilders trial 2.0. Netherlands Journal of Legal Philosophy, 45(2), 2–11.
doi:10.5553/NJLP/.000053
Waldron, J. (2012). The Harm in Hate Speech. Cambridge & London: Harvard University Press.
Walker, I., & Pettigrew, T. F. (1984). Relative deprivation theory: An overview and conceptual critique. British Journal of Social Psychology, 23(4), 301–310. doi:10.1111/j.2044-8309.1984.tb00645.x
Wichgers, L., Jacobs, L., & van Spanje, J. (2019). To sue or not to sue? How prosecuting anti-immigrant politicians influences citizens’ support for the democratic system. Manuscript in preparation.
Footnotes
1
Moreover, this variable is measured in the standard Dutch national election studies (van Spanje & de Vreese, 2014).
2
The study was approved by the Ethics Review Board of the Faculty of Social and
Behavioural Sciences prior to data collection. In addition, the experiment was suitable for completion through a smartphone, tablet, or computer.
3
The original post about Giorgetti concerned his comment about the political alliance between the League, Brothers of Italy, and Forward Italy in the 2018 General Elections.
4
Deleting two items from the scale for attitude towards multiculturalism scale led to a higher reliability (Cronbach’s α = .71). However, since this improvement is not substantial, it is arguably more valid to use the complete scale rather than only two items that were slightly more reliable. Furthermore, it is important to note that participants were grouped into multiculturalists, moderates, and assimilationists in a similar way to Rekker & van Spanje (2019), who used z-scores. In this way, a sufficient sample size and statistical power are ensured among groups.
5
This operation is recommended by Field (2013) to enhance the interpretability of the data, as well as the heteroscedasticity-consistent standard errors option as to avoid concerns about heteroscedasticity. Analyses showed that findings did not substantially change when these operations are included. Among the other advantages of using PROCESS for the analysis of moderation and conditional effects is that it computes the interaction terms automatically with a precision superior to the human one (Field, 2013).
6
Concerning H1, results were consistent in the sample with participants who answered the manipulation checks correctly (see Appendix C).
7
Concerning H2, in the sample with participants who answered the manipulation checks correctly, an analysis reported a marginal effect of perceived relative deprivation on propensity to vote (see Appendix C). All other results were consistent.
8
Concerning H3, results were consistent in the sample with participants who answered the manipulation checks correctly (see Appendix C).
9
Concerning H4, in the sample with participants who answered the manipulation checks correctly, attitude towards multiculturalism significantly predicted anger. All other results were consistent (see Appendix C). It is important to note that the effect of hate speech prosecution was only observed on anger; analyses conducted on fear, hope, and enthusiasm did not report significant effects.
10
In line with previous research on anti-immigration parties (e.g. van der Brug & Fennema, 2003), group identification, perceived relative deprivation, and a negative attitude towards multiculturalism significantly predicted the propensity to vote for the anti-immigration party.
Appendix A Questionnaire
Factsheet
Dear participant,
We would like to invite you to participate in a research study to be conducted under the auspices of the Graduate School of Communication, a part of the University of Amsterdam. The study for which we are requesting your cooperation is about Italian politics. In the online survey, several questions will be asked and some Facebook posts will be displayed.
Only Italian citizens eligible to vote may participate in this study. Participants must be at least 18 years old.
The study will take about 10 minutes. Your participation is voluntary.
As this research is being carried out under the responsibility of the ASCoR, University of Amsterdam, we can guarantee that:
1) Your anonymity will be safeguarded, and that your personal information will not be passed on to third (non-scientific) parties under any conditions, unless you first give your explicit permission for this.
2) You can refuse to participate in the research or cut short your participation without giving a reason for doing so. You also have up to 7 days after participating to withdraw your
permission to allow your answers or data to be used in the research.
3) Participating in the research will not entail your being subjected to any appreciable risk or discomfort and you will not be exposed to any explicitly offensive material.
4) No later than three months after the conclusion of the research, we will be able to provide you with a research report that explains the general results of the research. Participants wishing to receive this report must make this known by email to Miriam Tribastone, miriam.tribastone@student.uva.nl.
For more information about the research, you are welcome to contact the responsible researcher, Dr. Joost van Spanje, ASCoR, University of Amsterdam, Nieuwe Achtergracht 166, Postbus 15791, 1001 NG Amsterdam, +31 (0)20 525 4827, at any time. Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing ASCoR, at the following address: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl. Any complaints or comments will be treated in the strictest confidence.
We hope that we have provided you with sufficient information.
We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate.
Kind regards,
Joost van Spanje and Miriam Tribastone
Informed Consent
I hereby declare that I have been informed in a clear manner about the nature and method of the research, as described in the invitation for this study.
I agree, fully and voluntarily, to participate in this research study. With this, I retain the right to withdraw my consent, without having to give a reason for doing so. I am aware that I may halt my participation in the experiment at any time.
If my research results are used in scientific publications or are made public in another way, this will be done such a way that my anonymity is completely safeguarded. My personal data will not be passed on to third parties without my express permission.
If I wish to receive more information about the research, either now or in future, I can contact the responsible researcher, Dr. Joost van Spanje. Should I have any complaints about this research, I can contact the designated member of the Ethics Committee representing the ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl.
√ I understand the text presented above, and I agree to participate in the research study. In this first part, we would like to ask you a few questions about your opinions. We encourage you to be as sincere as possible since there is no right or wrong answer.
Perceived Relative Deprivation
To what extent do you agree with the following statements on a scale from 1 to 7, where 1 means completely disagree and 7 means completely agree?
If we need anything from the government, ordinary people like us always have to wait longer than others.
I never received what I in fact deserved.
It’s always other people who profit from all kinds of benefits. 1 completely disagree – 7 completely agree
Buffer Questions
Crime Economic situation Unemployment Terrorism Government debt Immigration Health National security The education system Pensions
The environment, climate and energy issues Other ______
Don’t know
You may have heard the idea that the world's climate is changing due to increases in
temperature over the past 100 years. What is your personal opinion on this? Do you think the world's climate is changing?
1 definitely changing - 7 definitely not changing
Some people believe that the population is getting older. How concerned are you personally about a possible increase in the proportion of people in Italy aged 65 and over? On a scale from 1 to 7, where 1 means not at all and 7 means very much, how much would you say you are concerned?
1 not at all – 7 very much.
___________________________________________________________________________
Ideological Orientation
In politics, people sometimes talk about left and right. Where would you place yourself on a scale from 0 to 10, where 0 means the left and 10 means the right?
0 left – 10 right
Group Identification
On a scale from 1 to 7, where 1 means not at all and 7 very much, to what extend do you identify with Catholics?
I feel a bond with Catholics I am glad to be Catholic
The fact that I am Catholic is an important part of my identity I have a lot in common with the average Catholic person Catholic people are very similar to each other
1 not at all – 7 very much
Attitude Toward Multiculturalism
To what extent do you agree with the following statements on a scale from 1 to 7, where 1 means completely disagree and 7 means completely agree?
Immigrants and ethnic minorities should be able to live in Italy while retaining all customs and traditions of their own culture.
Immigrants and ethnic minorities must fully adapt to Italian culture.
Immigrants and ethnic minorities should be able to freely profess their religion in Italy.
Immigrants and ethnic minorities must speak Italian in public. 1 completely disagree – 7 completely agree
___________________________________________________________________________
In the next pages, three Facebook posts will be shown. Please read them carefully, as there will be a few questions about their content afterwards. We also would like to inform you that we set a minimum reading time of 8 seconds per post; after it, you will be able to continue the survey.
Buffer Posts
English translation
The arrival of cold air from Russia on Italy
