Introduction

Following violent incidents, victims commonly fear stigmatization and perpetrators are often assumed to be evil. Researchers in cognitive and social psychology have examined these judgments, among whom Lerner and Miller (1978) proposed justice motive theory to provide an explanation. According to justice motive theory, people hold beliefs that they live in a fair world, in which positive actions are rewarded and negative ones are punished (Lerner, 1980). In the last decades, psychologists have investigated the behavioral and cognitive consequences of BJW. Such research has distinguished a belief that one’s own world is just (personal belief in a just world; PBJW) from a belief that the world is just for people generally (general belief in a just world; GBJW). Justice motive theory suggests that the ultimate reason for GBJW is that, if the world is “just for others,” it is also likely “just for me,” and therefore, PBJW and GBJW are correlated. However, despite the correlation between PBJW and GBJW, the two dimensions of BJW have unique correlates. PBJW has been shown to be positively correlated with life satisfaction (Dalbert, 1999), self-esteem (Donat et al., 2016), self-trust as well as trust in others and the world (Cubela Adoric & Kvartuc, 2007), and investment in long-term goal-pursuit (Hafer & Rubel, 2015a). By contrast, GBJW is associated with harsh, punitive responses to perpetrators of injustice and harsh responses to the disadvantaged (Bègue & Bastounis, 2003; Sutton & Douglas, 2005; for a review see Bartholomaeus & Strelan, 2019).

Overall, BJW has important functions in people's lives, such as encouraging investment in long-term goals, the pursuit of those goals through prosocial means, and providing a sense of purpose in life (see Dalbert, 2001; Hafer, 2000; Hafer et al., 2005; Hafer & Rubel, 2015a). Therefore, when just-world beliefs are threatened by contradicting evidence, people engage in strategies for preserving rather than giving up their beliefs (Lerner, 1980; Lerner & Miller, 1978). For instance, the phenomenon of victim blame is claimed to be one of such strategies that are utilized following threats to BJW (Lerner & Miller, 1978). Importantly, situational characteristics and individual differences determine the specific strategies that people use to protect their threatened BJW (Hafer & Gosse, 2010). Prior studies in this area have focused on blame, derogation, and psychological distancing. In the present research, the understudied strategies of demonization and compensation will be added to judgments of blame and derogation to examine how people, as observers, preserve their BJW in a hypothetical victimization scenario. More specifically, the current research explores how individual differences in observers’ BJW and victims’ as well as perpetrators’ characteristics interact to predict the ways people defend against threats to just-world beliefs. According to our reasoning, BJW should have stronger effects on a strategy when victims’ and perpetrators’ characteristics suggest the availability of that strategy and/or the non-availability of other strategies.

Responses to Injustice

Past research suggests that judgments of responsibility and blame about perpetrators and victims may depend on many factors; the outcome of the incident being positive or negative (Cushman, 2008), the harm-doer’s mental state (Weiss et al., 2021), controllability of mental states (Cusimano & Goodwin, 2019), perceivers’ moral values (Niemi & Young, 2016; Tepe et al., 2020), and moral content of the violation (e.g., harm violations or purity violations; Young & Saxe, 2011). A body of research claims that reactions to actors and victims of harm can be partly explained by justice motive theory. Studies on reactions to innocent victims have shown that when observers’ BJW is threatened in such situations, multiple strategies exist to deal with the threats (Lerner, 1980). In his 1980 book, Lerner discussed several strategies (see also Hafer & Bègue, 2005; Hafer & Gosse, 2010), some of which will be discussed here.

According to justice motive theory, if the option of restoring justice by helping the victim or punishing the perpetrator is unavailable, then people are likely to reinterpret the event (Lerner & Miller, 1978). Derogating or blaming the victims are called non-rational strategies as they encompass reinterpreting what led to the suffering, in order to restore a sense of justice. By derogating the victim, one perceives the victim as an undesirable person who deserves the negative outcomes (Hafer, 2000). Alternatively, in victim blaming, one attributes responsibility to the victims of injustice (Lerner & Miller, 1978). Also, by derogating or blaming victims, observers distance themselves from the victims and secure a sense of their own safety. According to Lerner (1980), victims are sometimes derogated because they possess negative characteristics or qualities. However, when a suffering victim is likable or virtuous, derogating the victim’s character is no longer a viable option. Alternatively, people may restore their threatened BJW by blaming the victim for acting unwisely or carelessly (Lerner & Miller, 1978). Previous research has found victim derogation and victim blaming to be dissimilar BJW-maintenance strategies focusing on victims’ character and behavior, respectively (Haynes & Olson, 2006; Jones & Aronson, 1973; Karuza & Carey, 1984).

There is also a set of rational strategies which can be characterized by the recognition of injustice. Compensating or helping victims are considered rational strategies because they involve accepting the presence of injustice (Miller, 1977). People are motivated to help or support the compensation of victims to restore justice and preserve their BJW. From a justice motive theory perspective, compensating or helping victims following the observance of injustice is considered a BJW-maintenance strategy (Lerner, 1980).

Another set of strategies includes perpetrator-oriented responses such as punishment and demonization of assailants. Following Darley (1992) and based on justice motive theory, Ellard et al. (2002) found that in certain instances of injustice when maintaining BJW by restoring justice or cognitively distorting outcomes is more effortful and less available, people may label perpetrators as evil. There are reasons to assume that demonization is a BJW-maintenance strategy. First, demonizing helps maintain BJW by assigning perpetrators to a rare population of evil characters (Ellard et al., 2002). Furthermore, demonizing perpetrators may allow a person to reject the possibility that evil could be committed by almost anyone when the circumstances call for it. Finally, demonizing may further benefit BJW by increasing observers’ perceived intensity of aggression done by perpetrators, which could facilitate the justification of greater punishments (Vasturia et al., 2018). In line with the reasoning, Gromet et al. (2016) found that people condemn perpetrators more harshly if they perceive malevolent desires and hedonic behavior in them.

Previous research has shown that higher BJW is associated with higher engagement in defensive strategies (Gravelin et al., 2019; Hafer & Bègue, 2005; Strömwall et al., 2013; Vonderhaar & Carmody, 2015), suggesting that observed injustice is perceived as more threatening for people high (vs. low) in BJW, which explains why high-BJW perceivers react more sensitively to observed injustice and are more likely to engage in BJW-maintenance strategies than low-BJW perceivers.

Situational Determinants of Responses

Several principles have been suggested that help determine how people choose to protect against threats to their BJW (Hafer & Gosse, 2010). As a general rule, people prefer coping mechanisms that are more readily available and less effortful. Situational variables affect the perceived availability of various strategies and the extent to which they require cognitive effort. Thus, characteristics of victims and perpetrators can facilitate or impede the use of a specific BJW-maintaining strategy.

For derogation, the relative respectability or likability of a victim’s character can be influential (Haynes & Olson, 2006; Jones & Aronson, 1973). For example, Haynes and Olson (2006) found that participants tend to derogate a drug dealer more than a volunteer youth sports coach. Participants exposed to the suffering of a volunteer youth sports coach seemed to find it hard to derogate the victim’s character and thus resorted to blaming or compensating the victim. In the present study, we likewise expect that the respectability of a victim has an effect on victim derogation. Specifically, we expect that victim derogation as a BJW-maintenance strategy is higher when the victim is less respectable (vs. highly respectable) in the eyes of the perceivers. Thus, we expect a stronger positive association between BJW and victim derogation when the victim is not (vs. highly) respectable.

For demonization, perpetrators of crimes may have some characteristics which can increase observers’ tendency to demonize them (Ellard et al., 2002). Demonizing is rooted in the concept of evil and the predisposition to believe that evil individuals are behind evil actions (Baumeister, 1999; Darley, 1992). Thus, cultures worldwide have developed an “archetype of evil” that Baumeister (1999) referred to as the “myth of pure evil.” The stereotypical attributions about evil actions and their agents include the intentional infliction of harm and the infliction of harm merely for the pleasure of doing so. The appearance of these stereotypical cues of evilness can influence judgments about actors of harm and facilitate the use of demonization to uphold BJW. In the present research, we likewise expect that the presence of evilness cues has an effect on demonization. Specifically, we expect that perpetrator demonization as a BJW-maintenance strategy is higher when perpetrators show cues of evilness (compared to no such cues). Thus, we expect a stronger positive association between BJW and perpetrator demonization when the perpetrator shows (vs. does not show) cues of evilness.

For victim blaming, the behavioral responsibility of the victim for causing the incident, such as a careless action, is influential. In the present research, we will not focus on behavioral responsibility and cues about the availability of victim blaming.

For compensation, the possibility to help the victim and how much attempts at compensation would successfully alleviate the victim’s suffering is impactful. Likewise, we will not focus on the availability of compensation in the current research.

Given that BJW-maintenance strategies are employed to uphold the belief in a just world, it is likely that people no longer engage in any of these strategies when BJW threats have been efficiently removed by a single strategy. Following the same logic, people may engage in an unavailable strategy more when other strategies are also not available. Thus, it is plausible that the availability of a strategy influences the extent to which other strategies are used. For instance, when there are no cues about the availability of victim derogation (i.e., an unrespectful victim), other strategies may be used more. Specifically, demonization, victim blaming, and compensation may have a stronger association with BJW when the victim is respectful. Similarly, when there are no cues about the availability of demonization (i.e., an evil perpetrator), other strategies may be used more. Specifically, victim derogation, victim blaming, and compensation may have a stronger association with BJW when the perpetrator does not show cues of evilness.

Furthermore, when a perpetrator displays signs of evilness, observers may resort to perpetrator demonization for preserving their BJW and no longer derogate the victim even though the victim’s respectability was low. Similarly, when the victim is displayed as unrespectable, observers may engage in victim derogation and no longer demonize the perpetrator even though the perpetrator is presented as evil. Therefore, we expect that when the victim is presented as less respectable and the perpetrator is presented as evil, victim derogation and perpetrator demonization are negatively correlated with each other, after controlling for BJW.

Alternatively, people may use a particular strategy irrespective of the availability of any other strategies: After all, various strategies can principally be used simultaneously. For example, signs of low respectability in victims may promote victim derogation, but demonization or victim blaming might be employed to further defend the belief in a just world. More specifically, using a strategy as a response to BJW threat may not be different depending on whether or not cues influencing the availability of other strategies are present. According to this reasoning, different BJW-maintenance strategies should be unrelated to each other when the victim is presented as less respectable and the perpetrator is presented as evil.

Previous work suggests a hydraulic effect, in which using one BJW-maintenance strategy reduces the likelihood of employing another (Bordieri et al., 1983; Correia et al., 2001; Hafer & Rubel, 2015b; Haynes & Olson, 2006; Jones & Aronson, 1973; Kenrick et al., 1976; Lerner & Simmons, 1966).

The Present Study

In the current study, we assessed multiple strategies of defending BJW (derogation, blaming, demonization, and compensation) after measuring participants’ general belief in a just world and manipulating situational variables (victim’s respectability and perpetrator’s evilness) that influences the availability of different BJW-maintenance strategies. As it will be shown in the vignette below, for respectability manipulation, one of the victims was a professor (highly respectable among Iranians) and the other was a car dealer (not much respectable among Iranians). Also, for the manipulation of evilness, the perpetrator with evilness cues was calm and serious and inflicted harm intentionally, while the perpetrator without evilness cues was anxious and inflicted harm unintentionally (see Ellard et al., 2002).

To test our hypotheses, we will be conducting hierarchical multiple regression for each of the four BJW-maintenance strategies, by entering the independent variables in the first step (victim respectability, perpetrator evilness, and BJW) and the two-way interaction terms between independent variables in the second step.

For Hypothesis (1), we predict that victim derogation will be higher in the low victim respectability than in the high victim respectability condition. Then, in Hypothesis (2), we predict an interaction effect between BJW and victim respectability on victim derogation such that in the low respectability condition, BJW is more strongly (positively) associated with victim derogation than in the high respectability condition. For Hypothesis (3), we expect that perpetrator demonization will be higher in the high perpetrator evilness than in the low perpetrator evilness condition. In Hypothesis (4), we predict an interaction effect between BJW and perpetrator evilness on perpetrator demonization such that in the high evilness condition, BJW is more strongly (positively) associated with demonization than in the low evilness condition.

For Hypothesis (5), we predict an interaction effect between BJW and perpetrator evilness on victim derogation, victim blaming, and compensation such that there is a stronger positive association between BJW and victim derogation as well as victim blaming and compensation when the perpetrator does not show (vs. shows) cues of evilness. Moreover, we predict an interaction effect between BJW and victim respectability on perpetrator demonization, victim blaming, and compensation such that there is a stronger positive association between BJW and perpetrator demonization as well as victim blaming and compensation when the victim is respectable (vs. not respectable).

In addition, we test the correlation between victim derogation and perpetrator demonization when both of these strategies are available. For the unrespectable victim/evil perpetrator condition, we expect victim derogation and perpetrator demonization to be negatively correlated with each other, after controlling for BJW (Hypothesis 6).

Methods

Participants

After excluding 15 participants who failed to answer the attention check correctly (please choose “strongly agree”), we recruited 200 Iranians from social media platforms (e.g., Twitter, Telegram). A sensitivity power analysis revealed that with a sample of N = 200 and a statistical power of 0.80, we can reliably detect an effect size of 0.05 in multiple regression analysis with increase in R-square as effect size statistic. We recruited participants by passing along invitations for participation in different groups and channels. There were 86 males and 114 females between 18 and 62 years old (M = 29.49; SD = 8.076) and the majority were not married (n = 131). Out of the 200 participants, 23 had a diploma or less, 86 held a bachelor’s degree, 87 held a master’s degree, and 4 had a doctorate.

Experimental Design

This study uses a 2 (Respectability of Victim: Car Dealer vs. Professor) × 2 (Perpetrator Evilness: Evil vs. Not Evil) between-subjects design. Participants’ BJW was measured as a continuous variable.

Procedure and Measures

After providing informed consent, participants first completed the BJW scale. Then, they read the designed scenario and responded to measures for the four BJW-preserving strategies, which were presented in random order using a Latin squares design. All responses were made on 7-point scales ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). After completing the questionnaire, participants were thanked and fully debriefed. This included informing them that the victimization story presented was fictional. The contact information of the main researcher was provided in case participants had further questions. Participants did not receive any compensation.

Belief in a Just World

BJW was assessed with the Persian translation of the six-item General Belief in a Just World Scale (Dalbert et al., 1987) that was validated by (Mikani et al., 2022). With the aim of not priming participants with the topic of justice, we intertwined 16 unrelated items with the primary items (e.g., “I prefer traveling by train than planes”). Only the scores from the GBJW items were averaged showing good internal consistency (Cronbach’s α = 0.80). Higher scores indicated a stronger endorsement of the BJW. We should note that in this study, we used GBJW (not PBJW) to measure BJW because it is an aspect of BJW that is more central in responses to injustice toward others and, as a result, is more relevant to the focus of our study.

The Victimization Scenario

We randomly assigned participants to read an alleged news article from Hamshahri – a very well-known newspaper in Iran – about an aggravated robbery. Regarding the evilness manipulation, the evil perpetrator had some of the characteristics of a stereotypically evil agent that Webster and Saucier (2013, 2015) highlighted. Our manipulation of evilness was based on the pertinent literature on evilness and demonization (Baumeister, 1999; Darley, 1992; Ellard et al., 2002; Webster & Saucier, 2013). Thus, as can be seen in the scenarios, the perpetrator with evilness cues was calm and serious and inflicted harm intentionally, while the perpetrator without evilness cues was anxious and inflicted harm unintentionally.

We manipulated respectability by varying the profession of their victim. Thus, in our work, one of the victims was a professor and the other was a car dealer. While professors are highly respectable in Iran, car dealers are among the least respectable. This especially holds true for our participants who were mostly highly educated individuals. In constructing the scenario, we complied with the set of recommended standards for stimulus materials that can successfully elicit effects of high threat to BJW. First, the scenario must be emotionally engaging to arouse defensive reactions (Hafer & Bègue, 2005; Lerner, 2002; Lerner & Miller, 1978). Second, it has to be clear that the victim is not responsible for the injustice and has no control over the situation. Third, there can be no ambiguity about the injustice (Hafer & Bègue, 2005). In addition to the victim’s innocence in our scenario, the magnitude and duration of suffering and the fugitiveness of the perpetrator elicit a high threat to BJW in all conditions. We asked participants to view the text from all perspectives and think carefully about it. In all conditions, the victim had no name. Participants were told that the name and sex of the victim are redacted “for confidentiality reasons”; in actuality, we did this so that the sex of the victim would not unduly influence participants’ responses. In the English translation given here, s/he, him/her, and his/her are inevitably used which might seem fake. However, Persian pronouns are not gendered and the sex of the victim was unspecified in our story without seeming fabricated.

To increase ecological validity, we prepared an original article in a journalistic writing style matching the articles in the “incidents” section of a Hamshahri newspaper.

Participants read the following vignette:

“Last Wednesday, a terrible crime happened in Tehran. At 12:30 am, a [professor / car dealer] living in Tehran was found wounded at the corner of a street. After inspecting the evidence and hearing the victim, the detective reported that after working late and at 11:30 pm, the victim had decided to walk the 15-minutes walking distance from the university to his/her home. On the way home, the perpetrator approached him/her, put a knife on his/her right-side abdomen, and commanded in a [shaky / serious] voice to hand over his/her cell phone and watch. In the meantime, the victim’s cell phone had rung, and the perpetrator [panicked / calmly rejected the call] and then [anxiously / slowly] thrust the knife into the victim’s abdomen. The perpetrator had taken the cell phone and the watch, and [nervously / calmly] ran away. The injuries from the stabbing were very severe and the [professor / car dealer] was close to death. According to the medical report, s/he miraculously survived but will suffer from irrecoverable internal injuries. Criminal investigations are ongoing but no vital clue has been found yet.”

Next, participants completed questions measuring their endorsement of BJW-maintenance strategies, which were presented in random order using a Latin squares design.

Victim Derogation

Participants completed 5 items (Cronbach’s α = 0.75) measuring how much they derogated the victimized target’s character. These items included: “the extent to which the victim is likable in general” and “the extent to which the victim is a careless person.”

Victim Blaming

Participants completed 4 items (Cronbach’s α = 0.66) measuring how much they blamed the victimized target’s suffering for his/her behavior. The items included: “the victim’s behavior is responsible for the fact that he/she was harmed” and “the victim’s behavior is to blame for the fact that he/she was harmed.”

Perpetrator Demonization

Participants completed van Prooijen and van de Veer (2010) 5-item demonizing scale (Cronbach’s α = 0.84). The items included: “the perpetrator is only motivated to destroy everything that is benevolent.”

Compensation

Participants answered a single item concerning “apart from insurance companies, how much money would you help the victim if you were to decide from 0 to 100,000,000 Tomans.” Participants were free to specify any amount in the range.

Results

The descriptive statistics and correlation coefficients among measured variables are shown in Table 1. In our main analyses, we examined the effects of BJW (continuous variable; centered around the sample mean), victim’s respectability (+ 1 = high; -1 = low), perpetrator’s evilness (+ 1 = evil; -1 = not evil), and their interactions separately for victim derogation, perpetrator demonization, victim blaming, and compensation. As can be seen in Table 2, in the first step, the three main effect terms (BJW, victim’s respectability, and perpetrator’s evilness) were entered. Then, in the second step, the terms for the two-way interactions were added.

Table 1 Descriptive statistics and correlations among the study variables
Full size table
Table 2 Hierarchical regression of on BJW-maintenance strategies
Full size table

Victim Derogation

In the first step, the predictors (BJW, victim’s respectability, and perpetrator’s evilness) accounted for a significant portion of the variance, adjusted R2 = 0.13, F(3, 196) = 11.67, p < 0.001. Participants derogated the victim who was a car dealer more than the victim who was a university professor, M = 4.41 (SD = 0.08) versus M = 3.81 (SD = 0.08), b = 0.29, t = 4.49, p < 0.001 (Hypothesis 1). Also, BJW was positively related to victim derogation, b = 0.20, t = 3.13, p = 0.002. However, perpetrator evilness did not have a significant effect on victim derogation, M = 4.13 (SD = 0.09) versus M = 4.10 (SD = 0.09). In the second step, there was only a significant victim’s respectability × BJW interaction effect, b = 0.68, t = 3.75, p < 0.001, R-square increase = 0.06 (Hypothesis 2, 5). As can be seen in Fig. 1, simple slope analysis showed that in the low respectability conditions (i.e., victim was a car dealer), BJW was positively associated with victim derogation b = 0.47, t = 5.35, p < 0.001. In the high respectability conditions (i.e., victim was a professor), BJW was not significantly associated with victim derogation, b = 0.01, t = 0.11, p = 0.908.

Fig. 1
figure 1

Predicted values of victim derogation as a function of BJW and respectability

Full size image

Perpetrator Demonization

The predictors (victim’s respectability, perpetrator’s evilness, and BJW) in the first step accounted for a significant portion of the variance in perpetrator demonization, adjusted R2 = 0.29, F(3, 196) = 28.58, p < 0.001. As expected, the perpetrators with cues of evilness were demonized more than perpetrators without such cues, M = 4.59 (SD = 0.11) versus M = 3.46 (SD = 0.11), b = 0.45, t = 7.58, p < 0.001 (Hypothesis 3). Also, BJW was positively associated with perpetrator demonization, b = 0.31, t = 5.20, p < 0.001. However, victim’s respectability did not predict the demonization of perpetrators, M = 4.01 (SD = 0.12) versus M = 4.04 (SD = 0.12). In the second step, the perpetrator’s evilness × BJW interaction was significant, b = 0.39, p = 0.020, R-square increase = 0.01 (Hypothesis 4, 5). As can be seen in the simple slope analysis shown in Fig. 2, in both evilness conditions, BJW was positively associated with demonization. However, in the evil perpetrator conditions (i.e., perpetrator with cues of evilness), the association between BJW and demonization was stronger, b = 0.47, t = 5.35, p < 0.001 than in the non-evil perpetrator condition, b = 0.15, t = 2.14, p = 0.03.

Fig. 2
figure 2

Predicted values of perpetrator demonization as a function of BJW and evilness cues

Full size image

Victim Blaming

Together, the predictors accounted for a significant portion of variability in victim blaming scores, adjusted R2 = 0.07, F(3, 196) = 4.89, p = 0.003. Among the variables in the first step, only BJW significantly predicted victim blaming, b = 0.25, t = 3.58, p < 0.001. In the second step, none of the two-way interaction effects were significant (Hypothesis 5).

Compensation

Taken together, the predictors accounted for a significant portion of the variance in the first step, adjusted R2 = 0.06, F(3, 196) = 5.24, p = 0.002. Participants compensated the car dealer less than the professor, b = -0.18, t =—2.72, p = 0.007. Also, BJW was positively associated with compensation, b = 0.17, t = 2.54, p = 0.012. In the second step, none of the two-way interaction effects were significant (Hypothesis 5).

As for Hypothesis (6), in the unrespectable victim and evil perpetrator condition and after controlling for BJW, victim derogation and perpetrator demonization were not significantly correlated with each other, r = 0.11, p = 425.

Discussion

The present study extends previous research on justice motive theory and the situational and individual difference determinants of BJW-maintenance strategies. In the current study, we manipulated situational cues regarding the characteristics of victims and perpetrators. More specifically, we manipulated victim’s respectability and perpetrator’s evilness which led to more victim derogation (Hypothesis 1) and perpetrator demonization (Hypothesis 3), respectively. Our results show clear differences in how people react to a violent crime depending on diverse situational conditions and different levels of BJW. Consistent with previous research (Haynes & Olson, 2006) and our Hypothesis (2), we found that respectability of the victim (i.e., professor vs. car dealer) was associated with how much victim derogation was used to defend BJW. Furthermore, our results supported Hypothesis (4) suggesting that the existence of evilness cues in the perpetrator was related to how much demonization was utilized as a defensive mechanism against BJW threats. Thus, in line with prior research, BJW-maintenance strategies that were more available were preferred by perceivers of the injustice whose BJW was threatened (Hafer & Gosse, 2010). The findings also indicate that perceivers of criminal acts are responsive to implicit evilness cues, such as calmness, in determining the evilness of the perpetrator. Also, the extent to which a victim is derogated can be influenced by one’s mere profession. In contrast with what we expected (Hypothesis 5), no two-way interaction effect on victim blaming and compensation was significant. Also, the interaction effect between BJW and victim respectability on perpetrator demonization and the interaction effect between BJW and perpetrator evilness on victim derogation were not significant. These finding indicate that the availability of some strategies may not remarkably influence the use of other strategies to maintain BJW. Nevertheless, by further looking into this, we found that BJW was associated with blame and compensation in the victimization scenario where the victim was highly respectable and there were no cues of perpetrator evilness. While the interaction effects were not significant, these findings suggest that when there are no cues for specific defensive strategies, perceivers of injustice may use some defensive mechanisms even if they are not particularly available. However, more than one defensive strategy may be used by perceivers of injustice so that BJW may be further secured. Furthermore, our findings did not provide support for Hypothesis (6), further suggesting that the use of certain strategies may not notably influence the use of other strategies for defending BJW. The null findings, which were somewhat inconsistent with a number of previous research (for a review see Hafer & Rubel, 2015b), may have been caused by the limitations of this study in methodology that we will discuss later. Thus, future studies with different victimization scenarios, different samples, and fewer limitations in methodology are needed to reach conclusions.

Interestingly, a marked difference was also observed in the association between BJW and compensation of victims depending on their respectability. The difference between compensating the car dealer and the professor remained significant when BJW was controlled. It appears that when victims are perceived as not very respectable, compensation is not only an unpopular BJW-maintenance option, but it is also lower irrespective of how much one believes in a just world.

Our study had a number of limitations. First, the claim that the need to preserve BJW can affect attitudes toward actors and victims of harm may be constrained by the findings of previous research suggesting that perceivers’ moral values (Haidt, 2001; Niemi & Young, 2016), and the moral content of the wrongdoings (Young & Saxe, 2011) can influence judgments of blame and representations of causal responsibility. Future research should consider that our scenario involved a harm violation with a very harmful outcome that would motivate punishment desires. Also, future research investigating BJW-maintenance strategies could benefit from including other important predictors such as moral values (Niemi & Young, 2016), belief in pure evil (Webster & Saucier, 2013), and agreeableness (Habashi et al., 2016) to provide a more comprehensive picture of the beliefs, values, and personality factors that influence reactions to specific instances of injustice. For instance, Niemi and Young (2016) suggested that individualizing moral values may constrain the legitimacy of BJW, and binding moral values may be the true driver of victim blaming and victim derogation.

Second, we assumed that participants’ just-world beliefs were threatened equally in all conditions after reading the scenarios. It could be that the differences in various conditions result in unequal degrees of threat to BJW. In other words, the evilness of the perpetrator and low respectability of the victim may be less threatening to an individual’s BJW. Not restricted to the current study, this concern reflects a general uncertainty about manipulations in many studies in the literature of BJW. This issue can be attenuated partly by administering the modified Stroop task to participants immediately following their exposure to the scenario and before they have a chance to use defensive strategies (Hafer, 2000). Differences in latencies on the Stroop task for justice and neutral words can be investigated in order to be assured of equal degrees of threat in all conditions.

Third, compensation was measured by a single item of non-consequential and monetary help, and the victim blaming measure yielded a low alpha level. These limitations in measurements should be taken into consideration when results are interpreted, and future research may benefit from assessing these constructs with better measures. Also, we did not manipulate the ordering of items which may be crucial because people may have restored their threatened BJW before reaching the final items that were measuring BJW-maintenance strategies. However, the items regarding the BJW-maintenance strategies were randomized which, to some extent, solves the issue.

Finally, our sample mainly consisted of young and educated people. Future studies should replicate our findings in more representative samples and bigger sample sizes.

Conclusion

This study provides empirical support for the notion that harsh judgments of blame and character on victims and perpetrators partly emerge as a consequence of threats to one’s belief in a just world. To reduce the anxiety that arises as a result of living in an unjust world where victims may be innocent, many individuals endorse beliefs that confirm justice in the world and protect these beliefs when they face injustice. We indicated that preferred strategies to defend BJW depend on situational factors such as the characteristics of victims and perpetrators. Thus, when those who believe in a just world observe undeserved victims of crimes, maintaining just-world beliefs may motivate them to harshly judge victims’ character or behavior, demonize the actors of crime, or compensate victims depending on which strategy is more accessible according to available information about the incident.