Already half a century ago, Lerner and Simmons (1966) observed that, when confronted with an innocent victim who cannot be helped, people often tend to perceive the victim’s misfortune as deserved. The proposed motive for this response is people’s desire to hold and protect their belief that the world is inherently just, and that people generally get what they deserve in life (Lerner 1980). Over the next few decades, this Belief in a Just World (BJW) has received a lot of scholarly attention, with a strong focus on the defending mechanism of blaming victims for their own fate (for overviews, see Hafer and Rubel 2015; Hafer and Bègue 2005). The present study aims to contribute to the literature by investigating whether confrontation with an innocent victim belonging to a distinct outgroup (versus the ingroup) yields different effects on victim blaming.

Belief in a Just World

Lerner and Miller (1978) stated that people have a need to believe that they live in a world where people generally get what they deserve. According to these authors, this BJW is beneficial and the primary reason for that is that the BJW makes the physical and social environment predictable and as such, it enables individuals to commit to long-term goals (see Hafer and Rubel 2015, for a recent overview for the functions of BJW). However, the confrontation with an innocent victim (of e.g. a crime or disaster), who is suffering through little fault of her/his own, threatens the BJW. Since the BJW serves advantageous functions, people are very reluctant to give up these beliefs, and attempt to preserve them. An important coping mechanism to preserve the BJW when confronted with an innocent victim is to blame these victims for their misfortune (Furnham 2003; Hafer and Bègue 2005; Hafer and Rubel 2015).

Moreover, the theory proposes that threat to BJW intensifies when the victim is more similar to oneself (Lerner 1980; Lerner and Miller 1978). Empirical evidence for this assumption has recently been provided by Correia et al. (2007), who demonstrated on the basis of Hafer’s (2000a) emotional modified Stroop task that an ingroup victim threatens individuals’ BJW more than an outgroup victim. Because victim similarity influences the level of threat to the BJW, also the coping mechanisms to deal with this threat may become more prevalent as similarity between observer and the victim increases. Indeed, various studies using different operationalizations of similarity, such as mere perceived similarity with the victim (Lerner and Agar 1972; Novak and Lerner 1968), and university membership (Bal and van den Bos 2012) have demonstrated this effect. However, other studies revealed that (out- versus in-)group membership, although a clear signal of (dis)similarity, had no effect on the blaming of innocent victims (e.g. Aguiar et al. 2008; Sebby and Johnston 2012) or occasionally even a reversed effect, revealing that victims belonging to an outgroup were more blamed compared to ingroup victims (e.g. Halabi et al. 2015). We therefore propose that a mere focus on similarity to explain potentially different effects on in- versus outgroup victim blaming may be too limited. In particular, we propose to also consider BJW as a system justification ideology, which, as we will argue below, leads to the possibility that general outgroup attitudes linked to system justification ideologies may also play a role in the effects of victims’ group membership on victim blaming.

Belief in a Just World as a System Justification Ideology

The system justification theory claims that people often “are motivated to justify and rationalize the way things are” (Jost and Hunyady 2005, p. 260). Many specific ideologies have been labeled as ‘system justifying’; including Right-wing Authoritarianism (RWA), Social Dominance Orientation (SDO), and also BJW (see Jost and Hunyady 2005). However, in contrast to RWA and SDO, very few empirical studies have explicitly investigated BJW from this perspective, despite its potential to provide further insight in both the antecedents and consequences of individual differences in BJW.

According to van der Toorn and Jost (2014), the need to manage epistemic (un)certainty is a core antecedent motivation to adopt system justification ideologies. In the case of BJW, the idea that people get what they deserve and deserve what they get is indeed a very powerful “tool” to reduce uncertainty. Hence, when confronted with someone who experiences a tragic event, believing that the victim was somehow responsible, rather than considering the randomness of the situation, allows people to retain a view of the world as logical, predictable and orderly. In line with this idea, Bal and van den Bos (2012) have demonstrated that situationally induced uncertainty versus certainty leads to a more negative evaluation of innocent victims in order to maintain their BJW. There are, however, also substantial inter-individual differences in people’s need for such epistemic security. This dispositional aspect of epistemic security needs is traditionally captured in the literature by the need for closure (NFC) construct (Kruglanski and Webster 1996; for a recent overview see, Roets et al. 2015). Individual differences in NFC, or the desire for “an answer on a given topic, any answer, … compared to confusion and ambiguity” (Kruglanski 1990, p. 337) are measured with a scale that includes five facets: need for order, need for predictability, (need for) decisiveness, intolerance with ambiguity and closed-mindedness (Webster and Kruglanski 1994; Roets and Van Hiel 2007). Several studies have empirically confirmed Jost and Hunyady’s (2005) assertion that NFC underlies system justification ideologies, such as SDO (e.g. Roets et al. 2012) and RWA (e.g. De keersmaecker et al. 2017b, see Roets and Van Hiel 2011a for an overview). Hence, NFC may likely also be a source of individual differences in BJW. Yet, a thorough empirical investigation of this hypothesized link is lacking in the literature.

System justification ideologies also have a number of potential consequences, including the derogation of disadvantaged groups (Jost et al. 2004). Indeed, even long before Jost and colleagues grouped these ideologies under the label of system justification, plenty of studies have shown robust associations with various forms of prejudice. Most prominent in this extensive body of research are studies linking RWA and SDO to prejudice (see Duckitt 2001, and Sibley and Duckitt 2008 for a meta-analysis). RWA and SDO are both ‘group-oriented’, referring to compliance and protection of group norms (Altemeyer 1981) and acceptance of inequality between groups (Pratto et al. 1994), respectively. In contrast, BJW is much more focused on the individual and it may therefore be not surprising that BJW research has almost exclusively focused on attitudes towards individuals (i.e., victims). Indeed, examination of the potential relationship between individual differences in BJW and attitudes towards groups (i.e., prejudice) is still rare and has not yet revealed straightforward evidence (see Bizer et al. 2012; Liao et al. 2016 versus Lima-Nunes et al. 2013). However, despite the lack of a group-orientation, we argue that BJW could be relevant to prejudice. Although BJW may have little bearing on blatant derogation of outgroups, one may expect BJW to be relevant to aspects of “modern forms” of prejudice that tap into denial of discrimination and resentment of special favors to disadvantaged groups. Indeed, one could expect people high in BJW to be more defiant in accepting that there could be social injustices merely based on group membership in their “just world”. So, if BJW is associated with prejudice and especially denial of group injustices, this may also have a substantial influence on how people high in BJW judge innocent victims that belong to an outgroup. Indeed, although outgroup victims are less similar, and hence less threatening to BJW, they may nevertheless be more blamed for their misfortune, based on negative outgroup attitudes.

Study 1

The first aim of the present research is to provide more empirical evidence for the theoretical status of BJW as a system justifying ideology. Therefore, first, we examine in two studies (Study 1A and 1B) whether BJW is related with 1) a core dispositional antecedent of system justification ideologies, i.e., NFC, 2) other well-known system justification ideologies, i.e., RWA and SDO, and 3) a core consequence of system justification ideologies, i.e., prejudice against disadvantaged groups.

Study 1A

Method

Participants

The sample consisted of 220 undergraduate non-immigrant psychology students from a Belgium university (81.4% women, Mage = 20.06 years, SDage = 3.53) who volunteered to participate in a class assignment.

Measures

The study was conducted online and participants responded to measures in the order presented below. All measures were rated on 7-point Likert scales anchored by 1 (strongly disagree) and 7 (strongly agree), except for the NFC scale which is traditionally rated on a 6-point Likert scale. Table 1 provides the Means, Standard Deviations and Cronbach’s alphas of all measures. These scales have been recently used in recent studies with similar populations (e.g. Harber et al. 2015; De keersmaecker et al. 2017a).

Table 1 Variable means, standard deviations (SD), Cronbach’s α and correlations of Study 1A
Full size table

Need for Closure

The 15-item short version (Roets and Van Hiel 2011b) of the revised NFC scale (Roets and Van Hiel 2007; original by Webster and Kruglanski 1994) was administered. An item example is: ‘I don’t like situations that are uncertain’.

Right-Wing Authoritarianism

RWA was measured with an 11-item version of Altemeyer’s (1981) RWA scale (see e.g. Roets and Van Hiel 2006). An item example is: ‘Obedience and respect for authority are the most important virtues children should learn’.

Social Dominance Orientation

Participants responded to the 16-item SDO scale of Ho et al. (2012). An item example is: ‘Some groups of people are just more worthy than others’.

Belief Just World

Dalbert et al. (1987) 6-item General Belief in a Just World scale (see also Dalbert 1999) and Lipkus’ (1991) 7-item Global Belief in a Just World scale were administered. Two items with near-identical wording present in both scales were only administered once. Preliminary analyses showed that both scales were highly correlated (r = .82) and yielded virtually identical effects.Footnote 1 For parsimony reasons, we therefore combined the measures into a single scale to optimize the measure of this core construct. Hence, the combined scale consists of 11 items. An item example is: ‘I basically feel that the world is a fair place’.

Subtle Racism

Participants responded to a 12-item scale, adapted from Pettigrew and Meertens (1995; see Van Hiel and Mervielde 2005) to measure racism. An item example is: ‘Immigrants should not push themselves in places where they know they will be discriminated’.

Results

In line with our expectations, the correlations presented in Table 1 show that BJW, as well as the endorsement of the two other system justification ideologies RWA and SDO were positively related to NFC. Moreover, BJW showed substantial positive correlations with RWA and SDO. Furthemore, NFC and all three ideologies were positively related to subtle prejudice.

Study 1B

Method

Participants

A heterogeneous non-immigrant sample of 127 adults who identified themselves as being heterosexual (45.7% women, Mage = 47.7 years, SDage = 19.79) completed the survey. These respondents were recruited within the social network of a master student via social media platforms and email.

Measures

The study was conducted online and participants responded to same measures in the same order as in Study 1A, with the two additional measures of prejudice at the end of the questionnaire:

Modern Racism

Participants responded to the modern racism 7-item scale by McConahay et al. (1981). An item example is: ‘Discrimination against immigrants is no longer a problem in Belgium’.

Modern Prejudice against Homosexuals

Swim et al. (1995) 8-item modern sexism measure was adapted into a modern prejudice against homosexuals measure. An item example is: ‘Discrimination against homosexuals is no longer a problem in Belgium’.

Results

Similar to Study 1A, the correlations, presented in Table 2, show that BJW, as well as the endorsement of the two other system justification ideologies RWA and SDO were positively related to NFC. Again, BJW also showed substantial positive correlations with those system justification ideologies. BJW also showed a significant positive association with modern forms of prejudice (racial and sexual orientation), but here the relationship with subtle racism was not significant. RWA and SDO showed significant correlations with the racial prejudice measures, but were not significantly related to prejudice against gays.

Table 2 Variable means, standard deviations (SD), Cronbach’s α and correlations of Study 1B
Full size table

Discussion Study 1A & 1B

The correlations obtained in Study 1A and 1B offer support for the theoretical status of BJW as a system justification ideology. First, similar to the two other system justification ideologies RWA and SDO, BJW was positively related to their theoretical, common antecedent: NFC. The magnitude of the relationship between BJW and NFC was in between the relationships of RWA and SDO with NFC, although closer to that of SDO.Footnote 2 Further evidence for the claim that BJW belongs to the circle of the system justification ideologies is provided by the observation that BJW, although not a group-oriented construct as RWA and SDO are, still showed substantial positive correlations with those system justification ideologies. Finally, similar to other justification ideologies and despite not being group-oriented, BJW was related to the hypothesized outcome variables of prejudice against immigrants and homosexuals. In line with our expectations, BJW was especially strongly related to modern prejudice, which most explicitly captures peoples’ propensity to deny discrimination and to reject the special favors and demands of minorities (see McConahay et al. 1981).

Study 2

Having provided empirical support that the BJW is positively related to NFC, other system justification ideologies and prejudice towards disadvantaged groups, we return to the phenomenon of victim blaming. Similar to the work of Aguiar et al. (2008), Halabi et al. (2015) and Sebby and Johnston (2012), in Study 2, we aimed to advance our understanding of the influence of victims’ group membership (outgroup versus ingroup) on victim blaming. In particular, the BJW theory (Lerner 1980; Lerner and Miller 1978) suggests that majority members will blame minority members less than fellow majority members because of lower similarity with the outgroup victim. However, given that BJW is also a system justifying ideology, and such ideologies are generally associated with ingroup favoritism and outgroup derogation by the advantaged group, it may also be possible that outgroup members are treated more negatively overall, including being blamed more strongly for their misfortune in situations where they are innocent victims. This reasoning would be consistent with the findings of Halabi et al. (2015). A third possibility is that group membership has little or no “net effect” on victim blaming, if the opposing processes of “similarity” and “outgroup derogation” negate each other (see Aguiar et al. 2008). The aim of Study 2 is to deepen our insight into the role of group membership on victim blaming by examining the validity of these different hypotheses.

Method

Participants

A total of 139 adult participants, who were recruited within the social network of a master student via social media platforms and e-mail, completed the study. Three participants with an immigrant background and two participants who failed to correctly answer the control questions (see below) were omitted from the sample, yielding an effective sample size of N = 134 (70.1% women, Mage = 37.13, SDage = 20.04).

Procedure

The experiment was conducted online using customized software. After completing the BJW scale, participants were randomly assigned to either the immigrant or non-immigrant condition. In both conditions, participants were presented with eight scenarios in which an innocent person befalls a misfortune (e.g. a robbery). The victim’s group membership was manipulated by giving the victim either a typical majority group name, or a typical immigrant (i.e., Muslim) name. To avoid that participants in the non-immigrant condition considered the victim as an outgroup member merely based on their sex, in both conditions, the victims’ and participants’ sex was matched. Immediately after reading each individual scenario, participants responded to measures of victim blaming about that event, and then moved to the next scenario.

Measures

All measures were rated on 7-point Likert scales anchored by 1 (strongly disagree) and 7 (strongly agree).

Control Questions

To protect data quality, two response reliability checks were spread throughout the study (see also Oppenheimer et al. 2009). These response reliability checks asked: ‘Please, check number [‘five’/‘six’] for this control question’.

Belief Just World

Participants responded to the same BJW measure as in Study 1 (M = 3.35, SD = .93, Cronbach’s α = .87).

Victim Blaming

For each of the eight scenarios, participants rated four statements to assess victim blaming. Averaging the 32 items yielded a reliable overall victim blaming scale (M = 3.20, SD = .65, Cronbach’s α = .85). An item example is: ‘The victim could have avoided this situation’.

Results and Discussion

Hierarchical regression analysis was conducted to examine the effect of victim group membership, participants’ BJW and their interaction on victim blaming. Centered BJW and group membership were entered in the first step and the centered interaction term in the second step (see Aiken et al. 1991).

In line with previous research (e.g. Hafer 2000b), the analysis yielded a significantly positive relationship between individuals’ level of BJW and victim blaming (β = .19, p = .026). Group membership showed a marginally significant effect on victim blaming (β = .16, p = .058), with the ingroup victim being blamed slightly more (M = 3.31, SD = .70) than the outgroup victim (M = 3.09, SD = .56). No significant interaction emerged between BJW and group membership on victim blaming (β = −.03, p = .761).

The marginally significant relationship between victims’ group membership and victim blaming in this experiment suggests the possibility that the tendency to blame the more similar victims from the ingroup because they are a greater threat to BJW is more pronounced than the tendency to blame victims from the outgroup as part of a more general derogation of the outgroup. To investigate the stability of the current research findings, we conducted a third study.

Study 3

To examine the generalizability of the results from Study 2 to different minority groups, group membership in Study 3 was based on a different criterion: sexual orientation (heterosexual versus homosexual). Furthermore, as a response to Hafer and Bègue’s (2005) call to use high impact stimuli to induce threat to peoples’ BJW, rather than presenting the participants with written cases, we used a video testimony of a allegedly real victim, and we increased the sample size to heighten statistical power.

Method

Participants

To reach a power of .80 to detect the effect of group membership of Study 2 (β = .16), 304 participants are needed. Given that BJW was threatened with a video testimony of a white male, we recruited 325 white male participants on Amazon Mechanical Turk, which has been demonstrated to be an appropriate method of recruiting subjects (e.g. Paolacci et al. 2010; Rand 2012). Participants received $ 0.50 in return for participation. Since homosexuality was the outgroup criterion in Study 3, individuals how identified themselves as being bisexual or homosexual were omitted from the analysis, resulting in a total of 287 white male heterosexual participants who completed the study. Forty participants were omitted from the sample because they failed to correctly answer the control questions (see below), yielding an effective sample size of N = 247 (Mage = 39.78, SDage = 13.01).

Procedure

The experiment was conducted online. After completing the BJW scale, participants were randomly assigned to either the homosexual or heterosexual condition. In both conditions, participants were presented with a video of Peter, holding sheets of paper in front of a webcam (see Fig. 1), telling the story of him having been robbed and beaten up, with lasting damage to his inner ear (i.e., tinnitus). The videos in the heterosexual and the homosexual condition were identical except for four different cues in the video that explicitly referred to the victim’s sexual orientation, e.g.: ‘I texted my boyfriend John that I loved him’ verus ‘I texted my girlfriend Charlotte that I loved her’. After the video, participants responded to measures of victim blaming.

Fig. 1
figure 1

Image of the video to induce threat to BJW in Study 3

Full size image

Measures

Control Questions

As in Study 2, two response reliability checks were implemented in the study. These response reliability checks asked: ‘Please, check number [‘five’/‘two’] for this control question’. Furthermore, participants were presented with three simple statements regarding the basic content of the video and were asked to indicate whether the statements are false or true. The statements were: ‘Peter has been mugged’, ‘Peter has hearing damage’, and ‘Peter’s last message was that he hates this world’ (all statements are true). Participants who failed to correctly complete the control questions were excluded.

Belief Just World

To assess BJW, we used the same measure as in Study 1 (M = 4.06, SD = 1.42, Cronbach’s α = .96).

Victim Blaming

To assess victim blaming, we used the same six item measure as in Hafer (2000b, Study 1), rated on an 7-point Likert scale anchored by 1 (not at all) and 7 (very much). An item example is: ‘To what extend is Peter’s misfortune a result of the type of person he is’. (M = 2.23, SD = 1.32, Cronbach’s α = .94).

Results and Discussion

Hierarchical regression analysis was conducted to examine the effects of the victim’s sexual orientation, participants’ BJW and their interaction on victim blaming. Similar to Study 2, centered BJW and the victim’s group membership were entered in the first step and the centered interaction term in second step.

The results revealed a positive significant relationship between BJW and victim blaming (β = .14, p = .026). More interestingly, a significant effect of victim’s group membership emerged. Specifically, participants blamed the ingroup victim slightly more than the outgroup victim (β = .13, p = .048; Mheterosexual = 2.40, SDheterosexual = .12, Mhomosexual = 2.06, SDhomosexual = .12). Given the more substantial sample, this effect reached the conventional threshold for significance, although its size was similar to that in Study 2. Finally, as in Study 2, victim’s group membership and BJW did not interact on victim blaming (β = .06, p = .306).

General Discussion

The main goal of the present research was to examine the role of victim’s group membership (out- versus ingroup) on victim blaming. We argued that, in addition to the generally assumed effect of similarity on increased victim blaming as a strategy to protect the BJW (Lerner and Miller 1978), also outgroup status could increase victim blaming as part of a more general negative outgroup attitude associated with system justification ideologies (Jost and Banaji 1994).

BJW as a System Justification Ideology

Studies 1A and 1B aimed to contribute to the literature by examining the theoretical status of BJW as a system justification ideology. In line with our expectations, BJW was related with other system justification ideologies and was associated with a core antecedent and consequence of system justification ideologies. In particular, the results from Study 1A and 1B demonstrated that although BJW has clearly a different focus than the “group-oriented” system justification ideologies RWA and SDO, all these ideologies were moderately strongly interrelated. Furthermore, BJW was associated with dispositional NFC, an established antecedent of other system justification ideologies, such as RWA and SDO (see e.g., Jost and Hunyady 2005; Roets and Van Hiel 2011a). Moreover, the relationship between NFC and BJW found in Study 1A and 1B also provides empirical support for Lerner and Miller’s (1978) claim that BJW serves the need to make the world predictable, stable and orderly. Remarkably, although the thesis that BJW is adopted to serve epistemic needs was a core assumption in the early development of the BJW theory, it has received very little scholarly attention before. In a rare exception, Bal and van den Bos (2012) already found that emotional/personal uncertainty is linked to more victim blaming (a seminal consequence of BJW), but to the best of our knowledge, the present study is the first to empirically demonstrate the link between dispositional measures of BJW and NFC (cf. the desire for predictability and order, and aversion for ambiguity), as hypothesized in the original writings by Lerner and Miller.

Finally, individuals’ level of BJW predicted a core consequence of system justification ideologies, i.e., prejudice against disadvantaged groups. Noteworthy and in line with our expectations, modern prejudice, which is characterized with denial of discrimination, antagonism toward minority groups’ demands and the resentment of special favors, appealed more to individuals’ BJW compared to subtle prejudice, which is focused on traditional values, cultural differences and affect toward minorities. Indeed, in the mind of a person with strong beliefs that the world is a just place, there are no such unjust things like discrimination and hence specific groups do not have the right for special favors or to make demands. On the other hand, belief in a just world does not have such direct implications for ideas about cultural differences, the role of traditional values or the absence of positive emotions towards outgroups.

Victim Blaming and Group Membership

After providing evidence that BJW is related to NFC, other system justification ideologies and prejudice towards disadvantaged groups in Study 1A and 1B, Study 2 and Study 3 investigated whether victim’s out- versus ingroup membership plays a role in one of the central issues of BJW literature: the degree to which people blame innocent victims for their misfortune in order to maintain a BJW.

Traditionally, the BJW literature has argued that higher similarity with the victim elicits a higher threat to an individual’s BJW, and therefore more similar victims are more likely to be blamed for their misfortune. Moreover, if similarity is determined by group membership, the effect can be expected to be even stronger given that BJW theory holds that beliefs about the world being a just place is especially important for “our own world” but less relevant to the world of others (Lerner 1980; Lerner and Miller 1978). Hence, primarily victims of the ingroup would truly threaten BJW and as a consequence, be blamed for their own misfortune.

However, recent studies that investigated similarity in terms of in- versus outgroup membership yielded findings contradicting this assumption, showing either no effect of group-based similarity (Aguiar et al. 2008), or even a negative effect (Halabi et al. 2015) on victim blaming. We argued that our understanding of BJW as a system justifying ideology may help explaining these findings. In particular, the system justification theory holds that the endorsement of system justification ideologies is associated with ingroup favoritism and outgroup derogation among members of advantaged groups. Therefore, although outgroup victims are less similar, negative attitudes towards their group might still make them more likely to be blamed for their misfortune.

Study 2 and Study 3 showed that, in addition to a general effect of dispositional BJW on victim blaming, a small effect of group membership emerged. Importantly, the direction of this effect indicated that ingroup victims (which are more similar) were blamed to a slightly greater degree than outgroup victims. Given the smaller sample size in Study 2 compared to Study 3, this effect was only marginal significant in Study 2 whereas it reached the threshold of significance in Study 3. This finding hence provides support for the “similarity hypothesis” of Lerner and Miller (1978) and is at odds with recent findings that outgroup victims are more blamed than ingroup victims (see Halabi et al. 2015). So how might we explain the inconsistent effect on victim blaming when (dis)similarity is operationalized as group membership? We propose that both the process of blaming primarily ingroup victims in an attempt to maintain the belief in a just world, and the process of blaming outgroup victims due to prejudice play a role, and that the net outcome may depend on the relative strength of these processes. This reasonening is in line with Aguiar et al. (2008) who already argued that ingroup victims are derogated for their misfortune because of a threat to BJW, whereas outgroup victims are derogated as a result of prejudice. Future research may seek to test this dual-process explanation by investigating how specific contextual factors may determine the relative strength of these simultaneous but opposite processes. An interesting factor that might influence the extent to which victims of an outgroup are blamed is the climate between the groups. In particular, in situations where the intergroup climate is characterized with severe conflict, hostility towards and derogation of the outgroup is likely to have a greater influence on increased outgroup victim blaming than the reversed effect of lower BJW threat due to dissimilarity. This seems a sensible perspective to explain the different results of the present experiments and the study by Halabi et al. (2015). In particular, the latter authors obtained their finding that outgroup members are more blamed compared to ingroup members in the context of Israeli Jews and Arabs, which is clearly a context of severe intergroup conflict. When the overall intergroup conflict and hostility are lower but still substantial, as can be assumed to be the case in the study of Aguiar et al. (2008) who used Roma people (labelled gypsies in the original study) as the targeted outgroup, similarity and prejudice effects on victim blaming may cancel out each other. Finally, when intergroup hostility and derogation are still weaker, the effects of similarity may be relatively more powerful, as the present findings suggest could be the case for groups based on sexual orientation and migration/religious background. An interesting avenue for future research would be to systematically test this perspective by assessing, within the same situation, how victim blaming towards members of different outgroups varies depending on intergroup hostility towards each of these groups.

Conclusion

In the present research, we provided support for the theoretical status of BJW as a system justification ideology by demonstrating its association with other system justification ideologies (RWA and SDO), a common theoretical antecedent (NFC), and a common consequence (prejudice against disadvantaged groups). Subsequently we showed in two experiments that, although the negative attitudes towards outgroups associated with BJW as a system justifying ideology may curb the influence of group-based (dis)similarity on victim blaming, ingroup members were still blamed more than outgroup victims. We argue that differences in relative strength between opposing “similarity” and “prejudice” processes across different outgroups may explain the inconsistent findings in the literature and provide a promising basis for further research on the assumed influence of similarity in victim blaming behavior.