Just World Belief and Ethics of Morality: When Do We Derogate the Victim?

Abstract

The current study aimed to explore the influences of moral emotions, moral ethics and perceptions of the perpetrator on the phenomenon of victim derogation. Based on the assumptions of Lerner’s Just World Belief theory (JWB; Lerner 2002) and Shweder’s “Big Three” ethics theory (Shweder et al. 1997), levels of victim derogation and avoidance in response to vignettes were analyzed as a function of moral content (ethic) and emotions, as well as good or bad outcome and perceptions of the perpetrator. Study 1 examined the influence of moral contents and outcome on moral emotions, finding that disgust is salient in violation of the divinity ethic whereas anger is salient in the autonomy ethic, and that anger is more dependent on outcome than disgust. Study 2 analyzed the influence of moral content, outcome, and perceptions of the perpetrator on victim perception. Results showed that the victim in the divinity context is perceived as more morally positive regardless of the outcome, but is avoided more. Also, negative perception of the perpetrator contributes more to positive perception of the victim in the divinity ethic than in the autonomy ethic. Perception of the victim in the autonomy ethic is affected more by outcome, and, in line with the JWB hypothesis, is derogated more when the outcome is negative. The fundamental motivation of justice was shown to be related to the link between act and outcome, but to vary by moral content. Derogation of the victim as a defense of JWB appears to be a typical reaction to injustice in the autonomy ethic, while avoidance of the victim may characterize violations of the divinity ethic.

According to Just World Belief theory (JWB; Lerner and Simmons 1966) belief in a just world, or the idea that the world is a place where people get what they deserve, is a fundamental part of justice motivation. However, this belief may lead to irrational assumptions, such as that only bad people experience bad things (Lerner and Miller 1978). Lerner and Simmons (1966) argue that, when faced with highly stimulating injustice, as in witnessing the suffering of a victim, many people defend their belief in a just world by derogating victims and finding ways to blame them for their own suffering.

However, JWB theory lacks grounding in emotion-based explanations and moral views. Skitka (2009) criticized the social justice literature for neglecting “homo moralis” and suggested that future studies should examine social justice by taking moral perspectives into account. The current paper attempts to address this issue.

In their “Big Three” theory, Shweder et al. (1997) suggested that there are three ethics of morality: autonomy, encompassing individual rights and justice; community, including duties and social norms; and divinity, consisting of sacredness, nature and protection of the natural order. The CAD Hypothesis (Rozin et al. 1999) further showed that each of these ethics is associated with one of three other-blaming emotions: disgust (elicited in divinity violations), anger (elicited in autonomy violations) and contempt (elicited in community violations). In the present study we make use of this conceptualization of multiple moralities and accompanying moral emotions to better understand the operation of the JWB phenomenon of victim derogation, by analyzing the perception of injustice in two different ethics, autonomy and divinity. In addition, we examine the effects of good or bad outcome of the action and perception of the perpetrator on victim derogation.

Victim Derogation

According to Lerner and Simmons (1966), victim devaluation is triggered in an observer by two beliefs: first, that the suffering is going to continue in some form, and second, that nothing can be done to help the victim. Other scholars have also suggested that perceptions of the victim’s innocence and suffering increase victim-blaming. Correia et al. (2001) did not find any significant influence of victim innocence on avoidance of the victim, but they suggested that the victim’s perceived suffering is threatening to belief in a just world. Hence, if the victim is perceived to suffer but has no behavioral responsibility, greater JWB threat – and consequently more victim derogation – is expected.

We expect the same factors to contribute to both victim derogation and avoidance of the victim. However, we also expect that emotions toward the victim will be important components of victim perception. According to Weiner (1980), sympathy towards a victim leads to reduced blame, and Pizarro (2000) has suggested that empathy increases vulnerability to feeling distress for the other, making it possible to use empathy as an effective moral marker. We expect that the extent to which positive emotions such as sympathy and empathy for the victim are aroused will be related to the moral content (ethic) – autonomy or divinity – of the situation.

Moral Ethics and Emotions

Shweder’s (1990; Shweder, et al. 1997) theory posits that there are three ethics underlying moral decision making that vary by culture. As briefly mentioned above, each of these ethics is associated with specific types of moral issues. (1) Ethic of autonomy refers to violations of individual rights, individualism, harm to self and others, disruption of freedom, and importance of liberty and fairness. (2) Ethic of community refers to community and hierarchy violations, duty to community, respect for authority and group, or interdependence. (3) Ethic of divinity refers to violations of rules of divinity or purity, sacredness, the natural order of things, sanctity, or degradation of the soul (Shweder et al. 1997). Further, the CAD Hypothesis (Rozin et al. 1999) suggests that violation of these moral ethics is associated with three different moral emotions: anger for autonomy violations, contempt for community violations, and disgust for divinity violations.

Lerner (2002) stated that anger and sadness are the two emotions aroused when belief in a just world is threatened. Injustice may be defined as undeserved harm, and in the “Big Three” model, harm is one of the concerns of the ethic of autonomy; and the CAD hypothesis maintains that anger is aroused in autonomy violations. Therefore, the current study hypothesizes that JWB will be elicited in violations in the ethic of autonomy more than in the ethic of divinity.

In contrast, violations in the ethic of divinity are expected to arouse disgust rather than anger. Disgust has been shown to have different effects than anger on the person. Haidt et al. (1993) found that disgust leads to stronger affective responses. Russell and Giner-Sorolla (2011) also reported that disgust renders the person more vulnerable to intuitive judgments that are less likely to be justified by concrete explanations.

These emotional reactions are also predictors of behavioral reactions; for example, the feeling of anger is more likely to motivate approach, while disgust leads to an intuitive reaction of avoidance (Haidt et al. 1997; Russell and Giner-Sorolla 2011). Avoidance activated by a feeling of disgust is labeled the disgust contamination effect (Rozin et al. 2008): the person intuitively avoids the object, whatever it may be, simply because of the feeling of disgust. According to Russell and Giner-Sorolla (2011), purity violations can also include quite abstract concerns about contamination. Where there is a violation of the divinity ethic that harms someone else, we expect that the emotion of disgust will focus the observer’s attention on condemnation of the violator rather than on the threat to JWB, and that the disgust contamination effect will motivate avoidance of rather than derogation of the victim.

Influence of Good or Bad Outcomes

Lerner proposed that, as a fundamental aspect of justice motivation, people must believe that there is an appropriate fit between what they do and what happens to them (Lerner and Simmons 1966). Taking this assumption further, some investigators have suggested that outcome has an influence on decisions (Janoff-Bulman et al. 1985). These authors discussed the motivational base of JWB theory, suggesting that outcome impacts blaming behaviors; they found that people derogate victims more when the outcomes are bad. Many studies have supported the appropriate fit hypothesis (see Lerner and Miller 1978), finding that if the outcome is good, the person is also considered to be good. In the light of these findings we included outcomes in the design of the study by varying the long-term outcome of the events that led to injustice. We expected that bad outcomes would be more threatening to JWB, which in turn would lead to more negative evaluations of the victim in the ethic of autonomy than in the ethic of divinity.

Role of Perpetrator Perception in Victim Blaming

Few studies have attempted to analyze the role of the perpetrator in victim derogation (Sizemore 2013). Mitchell et al. (2009) found that a victim received less blame when the perpetrator was blamed more, and that this effect varied according to the perceived motivation of the perpetrator. Ellard et al. (2002) suggested that when individuals are faced with a particularly heinous instance of injustice (e.g., in cases of torture), they may label the perpetrator as evil. The current study also aims to explore the influence of perpetrator evaluation on victim derogation in light of the moral content. We hypothesized that greater blame for the perpetrator would reduce victim derogation in the content of divinity as compared to autonomy.

The Current Study

This study had three main goals. The first aim was to assess the influence of type of moral content (ethic of autonomy versus divinity) on moral emotions; the second aim was to examine the effect of moral contents and perpetrator derogation on victim perception; and the third aim was to explore the influence of outcome on emotions and victim perception. Two scenarios involving harm and unfairness to the protagonist were created. One, portraying a person harmed by an accident, was designed to represent the autonomy ethic and trigger anger; the other, portraying a witness to incest, was designed to represent the divinity ethic and arouse disgust. Each story was written in two versions, one with good and the other with bad outcomes for the protagonist.

Two studies were carried out. The first study was concerned with the moral emotions elicited by the two stories, according to moral ethic and outcome, while the second study focused on evaluations of the victim and the perpetrator in each story, again according to moral ethic and outcome. The same set of scenarios was used in both studies.

Study 1

Study 1 aimed to explore the nature of participants’ emotional responses with regard to the content of the story, the victim and the perpetrator. Hypotheses were as follows:

• 1. Among emotions aroused by the stories, disgust will be greater in the incest story regardless of outcome, whereas anger will be greater in the accident story with a bad ending.

• 2. Among emotions felt toward the perpetrators, disgust will be greater in the incest story for both outcome conditions, whereas anger will be greater in the accident story with a bad ending.

• 3. A higher level of positive emotions will be reported for the victim in the incest story compared to the victim in the accident story, regardless of outcome.

Method

Participants

134 participants (93 women, 41 men) were recruited through convenience sampling (no compensation) by posting the study link on several social network platforms. The sample had a mean age of 26.35 (SD = 5.47). 33.6% reported that they were undergraduate university students, 40.6% were working, and the remainder did not specify their status.

Instrument

The questionnaire included demographic questions such as age, sex, work or school status, marital status and residence in addition to the stimulus materials, which consisted of the two stories detailed below, each with good or bad outcomes according to the condition, as well as questions about the participant’s emotional reactions to the stories. Following each story, participants responded to questions about both the perpetrator and the victim.

Construction of the Scenarios

Studies of JWB and their critiques cumulatively provide a set of standards for stimulus materials that can successfully arouse JWB effects. First, the stimulus must be emotionally engaging to elicit unconscious defensive reactions (Hafer and Bègue 2005; Lerner and Miller 1978; Lerner 2002). Second, it must be clear that the victim has no responsibility for the injustice; if the victim bears any responsibility, derogation cannot be attributed to threats to JWB. Third, there should be no ambiguity about the injustice – it should be obvious (Hafer and Bègue 2005). The accident story used in a study by Correia et al. (2012) meets all these criteria and was adapted for use in the present study to represent the autonomy ethic. Originally written in the first person, the story was translated into Turkish, using standard translation-back translation methods, and rendered in the third person. In the original version the story ended badly, but for the current study a story with a happier ending was also constructed in order to investigate the effect of story outcome. In both stories, the protagonist has a unisex name. Since Turkish pronouns are not gendered, the sex of the protagonist is unspecified. In the English translation given here, the “singular they” is used to avoid the cumbersome s/he, him/her, and his/her (Walsh 2015).

“A car ran a red light, hitting a student named Evren, who was crossing the street on a green light. On seeing Evren the driver tried to stop but it was already too late. It turned out that the driver was under the influence of alcohol.”

In the “bad outcome” condition the story ends with the following statement:

“The accident changed many things in Evren’s life since they are now paralyzed and have to use a wheelchair. It will go on like this for many years.”

In the “good outcome” condition, the ending is as follows:

“The accident changed many things in Evren’s life, including their outlook on life. They thought that, even though they fell behind in classes, going through a long period in intensive care gave them an opportunity to grasp the meaning of life. Now, Evren is writing a novel.”

This story meets the criteria listed above by portraying direct physical harm, which has been shown to arouse relatively strong automatic reactions (Cushman et al. 2006; Greene et al. 2001), and since there is no ambiguity regarding responsibility, it should threaten belief in a just world. As it involves harm and a violation of individual rights, it is expected to arouse the intuitive moral judgments related to the autonomy ethic.

A second story was constructed that would meet the above-mentioned criteria but that would evoke the intuitive judgments and emotions related to the divinity ethic rather than autonomy. On the basis of several studies (Haidt and Hersh 2001; Haidt et al. 1993, 1997; Rozin et al. 1999; Russell and Giner-Sorolla 2011), the scenario constructed for this purpose made use of violation of the incest taboo.

“Ada walked into their father’s office only to see him kissing his sister [i.e., Ada’s aunt]. When Ada told their mother about this, the parents got divorced and the mother became ill. All this changed many things in Ada’s life. Ada decided to give up school despite getting an attractive offer from a college abroad and started to work to provide treatment for their mother.”

In the “bad outcome” condition, the story ends as follows:

“After a while, Ada’s mother recovered but Ada could not succeed at any job they got, so they gave up their ideals and continued to have trust problems in personal relationships.”

In the “good outcome” condition, the ending is as follows:

“After some time, when Ada’s mother recovered, Ada decided to return to their education and accepted the offer from the college abroad. Then Ada and their mother moved to live abroad and led a comfortable life.”

In this story, a sacred taboo is violated and an unambiguously innocent victim (Ada) is harmed by it, both by witnessing the violation and by suffering interruption of his/her education. Also, the victim takes care of the mother through her illness, showing devotion to family relations (following the “natural order of things”). These elements were considered to be consistent with the ethic of divinity. No direct physical harm toward the victim was portrayed, in order to avoid automatic anger activation; rather, injustice was represented through emotional harm.

For each of the short stories, participants rated the intensity of their feelings (1) toward the situation (“what did you feel in this situation?”), (2) toward the protagonist (victim), and (3) toward the perpetrator (“what did you feel toward Ada/Evren?”/“what did you feel toward Ada’s father/the driver?”), using a scale ranging from 0 (I do not feel it at all) to 5 (I feel it extremely strongly). The lists of emotions felt toward the situation were anger, contempt, disgust, pity, pride, sadness and shame.. With regard to the victim and perpetrator, empathy and sympathy were added to this list. Disgust, anger and contempt are the other-blaming emotions discussed in the CAD hypothesis (Rozin et al. 1999), while empathy and sympathy were included to measure positive emotions (Pizarro 2000; Weiner 1980). The remaining items were used as filler items. The list of emotions was randomized for each of the stories.

Procedure

The questionnaire was administered online using Qualtrics web survey. Items were presented only after the participant indicated informed consent. No names were collected. Half the participants responded to stories with good outcomes and the other half responded to stories with bad outcomes.

Results

Emotion scores with regard to the situation were subjected to a repeated measures mixed design ANOVA (story X outcome X emotion toward situation) to test the first hypothesis, that disgust felt toward the situation would be salient in the incest story under both outcome conditions, whereas anger would be salient in the accident story with a bad outcome. Results revealed that scores for disgust in the incest story were significantly greater than in the accident story, (M = 3.41, SD = 1.83 versus M = 1.65, SD = 1.35), F_{1, 111} = 77.58, p ≤ .0001, ɳ² = .41.

Both a main effect for story (F_{1, 111} = 4.29, p = .04, ɳ2 = .04) and an interaction effect for story X outcome were found significant for anger (F_{1, 111} = 28.13, p ≤ .0001, ɳ² = .20). Anger was greater in the accident story (M = 3.79, SD = 1.85) than in the incest story (M= 3.39, SD = 1.95) the interaction results showed that outcome had a significant effect on anger scores in the accident story: the accident story with a bad ending (M = 4.49, SD = 1.62) elicited significantly more anger than the incest story with a bad ending (M = 3.11, SD = 1.97), p ≤ .0001). In the good outcome condition, participants also reported more anger in the incest story (M = 3.68, SE = 1.90) than in the accident story (M = 3.07, SE = 1.82), p ≤ .03. In the incest story, outcome had no influence on either disgust or anger, whereas in the accident story, anger was significantly higher in the bad outcome condition. The interaction effect supports the hypothesis. Figure 1 shows the interaction effect graphically.

Fig. 1

Disgust and anger scores toward the situation by outcome

To test Hypothesis 2, that disgust toward the perpetrator would be greater in the incest story, regardless of outcome, and that anger toward the perpetrator would be greater only in the bad outcome condition, repeated measure mixed design ANOVA (story X outcome X emotion toward perpetrator) was carried out. The results revealed that the only significant main effect on disgust toward the perpetrator was for incest story. Neither outcome nor the story X outcome interaction had a significant effect. The perpetrator in the incest story (M = 4.32, SD = 1.95) elicited significantly more disgust than the perpetrator in the accident story (M = 3.16, SD = 1.92), F_{1, 111} = 31.69, p ≤ .0001, ɳ² = .22. For anger toward the perpetrator, both the main effect of story (F_{1, 111} = 9.99, p = .002, ɳ² = .08) and the interaction between story and outcome were significant, F_{1, 111} = 8.70, p = .004, ɳ² = .07. The perpetrator elicited more anger in the accident story (M = 4.94, SD = 1.49) than in the incest story (M = 4.41, SD = 1.76). The interaction effect showed that the good outcome condition did not significantly affect anger toward the perpetrator, but that the bad outcome elicited significantly greater anger toward the perpetrator in the accident story (M = 5.30, SD = 1.16) than in the incest story (M = 4.26, SD = 1.88), p ≤ .0001. The interaction is shown graphically in Fig. 2.

Fig. 2

Interaction graph of anger toward the perpetrators

To test the third hypothesis, that positive emotions would be more strongly endorsed for the victim in the incest story compared to the accident story, repeated measures mixed design ANOVA (story X outcome X emotion toward victim) was conducted. Results revealed that both the main effect of story (F_{3, 333} = 12.95, p ≤ .0001, ɳ² = .10) and the interaction between story and outcome (F_{3, 333} = 2.86, p = .04, ɳ² = .03) were significant, but no main effect of outcome on positive emotions was found. Pairwise comparisons showed no significant difference between incest and accident stories in terms of empathy and sympathy for victim. However, interaction results revealed that the accident victim who experienced a bad outcome received significantly higher rates of sympathy, whereas the incest victim who experienced a good outcome received significantly higher rates of empathy (See Table 1.) This hypothesis was not supported.1

Table 1 Positive emotions toward victim by story and outcome

Study 1 Discussion

The findings were mostly supportive of the expected links between the moral emotions, anger and disgust, and the moral ethics implied by the stories, autonomy or divinity, respectively. In line with JWB, a bad outcome in the accident story was associated with more anger than in the good outcome, whereas more disgust was aroused in the incest scenario, regardless of the outcome. This finding is parallel to Russell and Giner-Sorolla’s (2011) statement that disgust is less likely than anger to be sensitive to situational and contextual cues. This suggests that outcomes may shape reactions to autonomy violations more than reactions to divinity violations.

Contrary to expectation, the incest victim did not receive any more sympathy than the accident victim. The fact that neither victim’s sympathy nor empathy scores varied by outcome also runs counter to JWB expectations. Interestingly, however, interaction results revealed that the bad outcome condition led to more sympathy for the accident victim, while the good outcome condition led to more empathy for the incest victim.

One possible explanation for the findings with regard to positive emotions is that the accident story was not fully successful in threatening participants’ JWB. This possibility is examined in the second study, where we expect participants to perceive greater injustice and greater suffering for the victim in the accident scenario. A second possibility could be cultural differences. Furnham (2003) highlighted cultural variations in JWB and showed that cultures with more property, wealth and power may have stronger JWB than those that have little; the latter may believe that the world is unjust rather than just. A similar argument from the moral psychological perspective was provided by Haidt et al. (1993), who argued that Western cultures have a more harm-based morality, in line with the autonomy ethic. As members of one of the world’s “developing” economies, Turkish participants may have a weaker tendency toward JWB threat than those in “developed” economies. In addition, Tepe et al. (2016) reported that in Turkey the ethic of divinity is the most salient moral ethic in justifying judgments regarding harmless taboo violations; this could be related to the positive emotions felt toward the victims.

Study 2

The second study, using the same scenarios as the first study, was designed to test the effects of perpetrator derogation on evaluations of a victim, taking moral contents and outcomes into account. The study had two parts: in the first part, evaluations of the victim were analyzed; and in the second part, the effects of perceptions of the perpetrator on victim perception were analyzed.

Hypotheses

• 1. Moral content (autonomy or divinity) will affect victim evaluations.

• a. Because of the disgust contamination effect, the incest victim is expected to receive higher avoidance scores.

• b. Because anger is related with derogation and is also more sensitive than disgust to outcome, the accident victim is expected to be rated as less attractive and derogated more than the incest victim, especially in the bad outcome condition.

• c. Because the incest victim is portrayed as otherwise behaving in accordance with the morality of divinity, avoidance is not expected to be associated with reduced victim attractiveness.

  • 2. More negative perception of the perpetrator will be associated with greater perceived attractiveness of the victim. However, this link is expected to be weaker for the accident victim.

Method

Participants

Participants were 203 undergraduate students (125 women,78 men), recruited from three universities in Istanbul. Their ages ranged from 18 to 44 (M = 21.02, SD = 2.66).

Instrument

The questionnaire included questions regarding age, sex, marital status and residence. The same stories used in Study 1 constituted the stimulus materials.

Dependent Variables

Victims were evaluated on seven variables, adapted from Correia et al.’s (2001) study, namely attractiveness, avoidance, behavioral responsibility, derogation, deservingness, injustice and suffering. Perpetrators were evaluated only on avoidance, behavioral responsibility and derogation.

The attractiveness scale for victims consisted of the following adjectives: fair, good, honest, mature, moral, strong, and successful. The derogation scale items for victims were: immoral, irresponsible, selfish, silly, unsuccessful, weak, and wrong.

The items on the attractiveness scale for perpetrators were: good, intelligent, moral, responsible, right, strong, and thoughtful. Derogation scale adjectives for perpetrators were bad, immoral, irresponsible, selfish, silly, weak, and wrong. These items were rated in a Likert format, with responses ranging from 1 (No, does not describe the target person) to 5 (Yes, describes the target person).

The remaining variables (avoidance, behavioral responsibility, derogation, deservingness, injustice and suffering) were measured by responses to direct questions, such as “Would you like to spend some time with a person like X in real life?”, “Was X responsible for what happened?”, or “Did X deserve what happened to them?” on a scale from 1 (No, definitely not) to 7 (Yes, definitely).

Manipulation Check

To check whether the stories actually had the expected effect on participants, after making the evaluations they were asked how much the stories bothered or disturbed them. They were also asked how confident they were in their judgments. In light of findings by Russell and Giner-Sorolla (2011) regarding the feeling of disgust, it was expected that the incest story would be associated with greater feelings of disturbance and lower levels of confidence. Likewise, activation of JWB is thought to depend on perceptions that the victim is suffering unfairly (Hafer and Bègue 2005; Lerner and Miller 1978; Lerner 2002). Questions regarding fairness, responsibility and suffering served as a check on these perceptions (see analyses in the Results section.)

Procedure

Data were collected in a classroom setting with groups of 45–60. No names were collected. Participants were briefly informed about the content of the task and they were told that the stories described real events. Deception was used to increase emotional engagement with the stories. After completing the questionnaire, participants were informed that the stories were fictional. Half the sample read stories with good outcomes and half read stories with bad outcomes. In both forms the incest story was read first. All questions and ratings with regard to the first story were completed before proceeding to the second story.

Results

Reliability and Factor Scores for Dependent Variables

Factor analyses were carried out for attractiveness, derogation, avoidance and behavioral responsibility items for victim and perpetrator in each story separately, using Principal Axis and Direct Oblimin methods. For all scales except one, all items in each scale loaded on a single factor. Only in the case of the incest victim, the attractiveness items loaded on two factors: “competency” (Cronbach alpha = .84), which included the items mature, strong and successful, and “moral attractiveness” (Cronbach alpha = .72), which included the items fair, good and moral. “Honesty” did not load on either factor and was omitted from further analyses. Reliability coefficients are shown in Table 2.

Table 2 Cronbach alpha scores for dependent variables

Manipulation Checks

Incest Story

As a confirmation that the story was perceived as a divinity violation, participants were expected to be more bothered or disturbed, and to give lower ratings of confidence in their judgments, in the incest story. As expected, repeated measures ANOVA results revealed that participants’ disturbance scores were higher in the incest story (M = 4.35, SD = 2.14) than in the accident story (M = 3.95, SD = 2.06), F_{1, 201} = 6.30, p = .01, ɳ² = .03; and confidence in judgments was significantly lower in the incest story (M = 5.32, SD = 1.54) than in the accident story (M = 5.59, SD = 1.45), F_1,201 = 5.19, p = .02, ɳ² = .03.

Accident Story

To serve as an adequate stimulus for JWB, the accident story needed to evoke perceptions that the situation was unfair, that the victim is not responsible for the situation and does not deserve to suffer, and that the victim was suffering. As expected, participants rated the accident story as more unfair than the incest story (M = 5.69, SD = 1.58 versus M = 4.75, SD = 1.82), F_{1, 157} = 41.87, p ≤ .0001, ɳ² = .21. While outcome did not have a direct effect on perception of fairness (p = .15), the interaction between story and outcome was significant, F_{1, 157} = 9.17, p = .003, ɳ² = .06. Perception of injustice varied by outcome only in the accident story, not in the incest story (bad outcome M = 6.08, SD = 1.40 versus good outcome M = 5.31, SD = 1.67). In both stories, the protagonist was rated as not having responsibility for, and not deserving the things that happened. There was no significant effect of story or outcome on responsibility or deserving. The main effect of outcome on judgments of suffering was significant, F_{1, 201} = 56.11,, p ≤ .0001, ɳ² = .22; that is, victims in both stories were seen as suffering more when the outcome was bad (M = 6.44, SD = .89 versus M = 5.25, SD = 1.31). The interaction between story and outcome was also significant; as expected, the accident victim in the bad outcome condition (M = 6.71, SD = .92) was perceived as suffering more than the incest victim in the bad outcome condition (M = 6.17, SD = 1.33), F_{1, 201} = 7.06, p = .009, ɳ² = .03.

Hypothesis 1

Evaluation of the victim will depend on the moral ethic (autonomy or divinity) portrayed in the story.

• a. The incest victim will receive higher avoidance scores. Repeated measures ANOVA showed that the incest victim was avoided significantly more (M = 2.43, SD = 1.21) than the accident victim (M = 1.73, SD = .98), F_{1, 201} = 63.73, p ≤ .0001, ɳ² = .24. There was no significant main or interaction effect of good or bad outcome on avoidance scores. This hypothesis was supported.

• b. The accident victim will be perceived as less attractive and will be derogated more than the incest victim. Repeated measures ANOVA showed that both story and outcome, as well as their interaction, significantly affected attractiveness scores. Since the attractiveness scale for the incest victim yielded two different factors, labeled “competency” and “moral attractiveness”, scores for each of these factors were treated as dependent variables. Mean attractiveness scores by story and outcome are shown in Table 3.

Table 3 Pairwise comparison of victim attractiveness scores based on the outcome condition

The main effect of story (F_{2, 402} = 39.11, p ≤ .0001, ɳ² = .16) showed that the accident victim received lower attractiveness ratings (M = 3.69, SD = .79) compared to both measures of attractiveness of the incest victim (competency M = 4.02, SD = 1.09; moral attractiveness M = 4.24, SD = .74); for the incest victim, moral attractiveness scores were also significantly higher than competence scores, p ≤ .0001.

In addition, the main effect of outcome (F_{1, 201} = 72.11, p ≤ .0001, ɳ² = .26), showed that a good outcome (M = 4.34, SD = .57) makes the victim more attractive than a bad outcome (M = 3.63, SD = .62). The interaction between story and outcome also significantly affected victim attractiveness scores, F_{2, 402} = 25.12, p ≤ .0001, ɳ² = .11. Interaction results are illustrated in Fig. 3. All pairwise comparisons were significant at p < .01, except for the difference between the accident victim’s attractiveness and the incest victim’s competence in the bad outcome condition (p = .53).

Fig. 3

Interaction graph of attractiveness scores

In parallel with the attractiveness ratings, scores on the derogation scale showed a similar pattern. The main effect of story indicated that the accident victim (M = 1.90, SD = .85) was derogated significantly more than the incest victim (M = 1.60, SD = .70), F_{1, 201} = 22.98, p ≤ .0001, ɳ² = .10. A bad outcome (M = 1.94, SD = .63) leads to significantly more victim derogation than a good outcome (M = 1.55, SD = .59), F_{1, 201} = 19.92, p ≤. 0001, ɳ² = .09. Hypothesis 1b is fully supported.

• c. Avoidance scores will not be associated with reduced victim attractiveness. Inspection of Fig. 3 shows that even though bad outcome reduced attractiveness for both victims, the incest victim’s moral attractiveness scores were reduced considerably less compared to both their competence scores and the accident victim’s attractiveness scores. Thus, as predicted, although the incest victim was avoided more than the accident victim s/he was perceived as more attractive and was derogated less. Hypothesis 1c is fully supported.

Hypothesis 2

More negative perception of the perpetrator will be associated with greater perceived attractiveness of the victim; the association will be stronger in the incest story.

Linear regression analyses were carried out, separately for each of the stories, to determine the association between perpetrator derogation and attractiveness of the victim. Perpetrator derogation significantly predicted accident victim attractiveness, β = .23, t(201) = 3.34, p = .001. Similarly, perpetrator derogation significantly predicted moral attractiveness of the incest victim (β = .34, t(201) = 5.075, p = .0001); however, it did not predict competence ratings. Regression coefficient scores for both the accident and incest victims are higher, showing that perpetrator derogation makes both victims more attractive. Results showed that the hypothesis is partially supported, but the predictive power of perpetrator derogation was higher in the incest story than in the accident story.

Study 2 Discussion

This study aimed to assess the separate and combined effects of moral ethic and outcome, as well as the effect of perceptions of the perpetrator, on perception of victims in terms of attractiveness, derogation, and avoidance. The pattern of responses regarding victim suffering, responsibility and deserving, as well as the respondent’s feeling of disturbance in response to the stories, confirmed that the stories adequately represented the two moral ethics (autonomy and divinity) and were adequate to threaten just world beliefs (JWB).

Comparisons of the accident story (autonomy ethic) with the incest story (divinity ethic) revealed that responses to the accident story followed the expectations of JWB theory: although both victims were perceived as not being responsible and as not deserving of the events that befell them, the accident victim was perceived as less attractive, and was derogated more, than the incest victim. The incest victim’s attractiveness was not unitary but fell into two categories, namely “competence” and “moral attractiveness”, both of which were rated higher than the attractiveness of the accident victim. These results bolstered the claim that JWB theory applies mainly in the moral ethic of autonomy.

Despite being rated as more attractive in both categories, the incest victim was avoided more than the accident victim, supporting the hypothesis of disgust contamination (Rozin et al. 2008).

Comparisons of stories with good endings versus those with bad endings showed that, as JWB theory would predict, victim attractiveness is greater in the case of good outcomes in both moral contexts. However, compared to the incest victim, the accident victim was rated as less attractive and was derogated more when the story ended badly, in line with JWB theory and the appropriate fit hypothesis (Janoff-Bulman et al. 1985; Lerner and Simmons 1966). On the other hand, the difference in attractiveness ratings of the accident victim in the bad outcome condition and the incest victim in the same condition was only significant for the moral attractiveness of the incest victim, not their competence. The appropriate fit hypothesis may apply only to the autonomy ethic, where the magnitude of moral wrongness is highly dependent on the degree of harm done (outcome); the divinity ethic is much less open to considerations of degree – if a rule is broken, it is broken; if purity is violated, purity no longer exists. From this perspective, it is likely that competency is salient in the autonomy ethic and moral attractiveness is salient in the divinity ethic. This could explain why moral attractiveness ratings of the incest victim did not vary by outcome but competency did vary.

Derogation of the perpetrator had a similar effect in both stories, with greater derogation of the perpetrator leading to greater attractiveness ratings for the victims, although this effect was stronger in the incest story.

To summarize, the derogation expected by JWB theory was elicited more strongly in the accident context (autonomy), particularly when the outcome was negative for the protagonist, while paradoxically the more attractive incest victim (divinity context) was avoided more. Attraction of the victim was also increased by greater derogation of the perpetrator, especially in the divinity context. This study also showed that avoidance need not be in line with negative attitudes toward the victim, since the victim in the incest scenario is perceived as more personally attractive. However, it should be noted that the higher attractiveness score of the victim in the incest story may be related with taking care of the mother through her illness. Further studies using different types of justice infractions across moral domains may provide clarity on this issue.

General Discussion

The two studies reported here aimed to investigate the effects of moral context and outcome on moral emotions, victim derogation and other aspects of perception of the victim.

The main finding of the first study was that the moral emotions, anger and disgust, were differentially aroused by the different moral contexts and outcomes. While anger was aroused by both stories, in the accident story the intensity of anger was reduced by a good outcome and increased by a bad one – as predicted by JWB theory and the appropriate fit hypothesis (Janoff-Bulman et al. 1985; Lerner and Simmons 1966). In contrast, outcome made no difference in the amount of anger aroused by the incest story.

On the other hand, disgust was higher in the incest story regardless of outcome. This result is congruent with recent literature investigating the antecedents of anger and disgust (Russell and Giner-Sorolla 2011). It is also congruent with and provides a portion of the explanation of some of the findings of the second study.

Positive emotions (empathy and sympathy) toward the two victims did not differ meaningfully, although there was some variation across different combinations of story and outcome. It is possible that participants did not distinguish reliably between empathy and sympathy. It is also possible that victims of any sort arouse empathy and sympathy regardless of the moral context or outcome. Further studies using different types of justice infractions with different outcomes may help to clarify this issue.

The second study extended the distinction between the autonomy and divinity ethics to their effects on evaluations of and attitudes toward the victims. Russell and Giner-Sorolla (2011) have pointed out that anger prompts hostile approach whereas disgust promotes avoidance. In line with this, the greater anger expressed in the bad-outcome condition of the accident story in Study 1 is congruent with the reduced attractiveness and greater derogation of the victim under the same condition in Study 2. Likewise, the greater avoidance of the incest victim in Study 2 is congruent with the disgust aroused by the incest story in Study 1 (see also Rozin et al. 2008 on disgust contamination). These complementary findings strongly support our overall hypothesis that different moral ethics produce different reactions to unjust suffering, including different other-blaming emotions (anger versus disgust) and different attitudes toward the victim (derogation versus avoidance).

Limitations and Future Directions

The present study needs to be tested cross-culturally to see whether or not the same patterns will be seen in other cultures. Both the emotions aroused and perceptions of victims in moral violations may be affected by the relative dominance of one or the other moral ethic (Haidt et al. 1993; Tepe et al. 2016). In addition, there was significant reliance on convenience sampling, especially for study 1, and additional research should be extended to more varied ages and so forth. However, random assignment to condition (good versus bad outcomes) and randomization of order of presentation of some of the items should reduce any effects of sampling method.

The scenarios in the current study were designed to provoke anger and disgust within a context of injustice. Even though the scenarios aimed to threaten justice perception while also arousing moral emotions, they may include some confounding variables. For example, the gender of perpetrator is clear in the incest scenario whereas in the accident scenario it is not specified. Also, the incest scenario portrays emotional harm whereas the accident scenario portrays physical harm.

The two studies reported here took up emotions and victim perceptions separately, with different respondents and different questions, so that we could not directly look at the relations between them. Future studies should investigate the effects of moral emotions on perceptions of both perpetrators and victims in moral violations.

The current study did not take into consideration individual differences in characteristics such as personality traits, just world belief (Rubin and Peplau 1975), disgust sensitivity (Haidt et al. 1994), moral foundations (Haidt and Graham 2007) or system justification (Jost and Banaji 1994). Niemi and Young (2016), for example, have found that individual differences in moral foundations predict quite different patterns of victim-blaming. Further studies may pursue the interaction between personal characteristics and context in responding to moral violations.

Conclusion

The present study makes two main contributions to the literature on perception of victims. First, it demonstrates clearly that just world belief theory (JWB), with its prediction of victim derogation in situations of injustice, describes processes that operate mainly in the autonomy ethic, with anger as the dominant blaming emotion, whereas in the divinity ethic, the emotion of disgust and associated victim avoidance predominate. Second, it shows that the long-range outcome of the experience of injustice exerts a strong influence on the emotions and perceptions of the observer, but that this influence is moderated by the moral context. Specifically, the combination of the autonomy ethic and bad outcome most strongly elicit the derogation of the victim predicted by JWB, whereas outcome does not affect derogation of the victim in the divinity ethic.