Putting the ‘‘I’’ and ‘‘Us’’ in Justice: Derogatory and Benevolent Reactions Toward Innocent Victims in Self-Focused and Other-Focused Individuals

Reactions toward innocent victims can range from harsh derogatory reac-tionstogreatefforttoalleviatethevictims’ ill fates.Usinginsightsfromresearchonjust-world theory and perspective taking, the current paper investigates both negative and positive reactions toward innocent victims. Speciﬁcally, we propose that self-focused versus other-focused motives can evoke derogatory versus more benevolent reactions, respectively, toward innocent victims. By manipulating self-focus versus other-focus, we indeed show in two studies that a self-focus enhanced indirect victim blaming and derogation and decreased helping of innocent victims. Furthermore, when participants were focused on another person these effects attenuated. Taken together, these ﬁndings extendpreviousstudiesonjust-worldtheoryandshowthatbothblamingandhelpingcanbeviablestrategiestodealwithunjustsituations.


Introduction
Imagine that, while having your breakfast on a Saturday morning, you read in the newspaper that a fellow student or colleague of yours was involved in a severe accident. That is, while cycling home after an evening on the town, he got hit by a car on an unsafe crossing. In the accident, he suffered severe head trauma. Furthermore, because he has to go through intensive rehabilitation to re-obtain all his cognitive skills and his insurance will not pay for this he probably has to give up his dream of becoming a successful researcher. The university set up a fundraiser for his rehabilitation and you can help. What would you do? Would you spend time raising money for the victim? Or would you think ''why was he so careless; everybody knows how dangerous that crossing is''? Or maybe even ''he probably had one too many beers with his friends and did not pay attention at the crossing''?
When people are confronted with victims, whether it be individual cases or mass suffering, they often have to decide how to react. In the current research, we look at key variables that can influence the construal of such events, making either blaming or helping a more likely option to deal with the confrontation with an innocent victim. That is, building on and extending earlier theorizing, we propose that innocent victims can arouse empathic feelings in an observer, because they have been treated unjustly, but that victims may also evoke fear in people of a similar ill fate bestowing on them and, as such, elicit feelings of threat (Gilligan, 1977;Lerner, 1980;Lerner, Miller, & Holmes, 1976). We put forward that these different reactions to the unjust event could be evoked by differences in focus people can have (e.g., Baumeister & Leary, 1995;Deci & Ryan, 2000;Van Lange, 1999). In particular, we examine whether people react differently toward victims of injustice than people when they are self-focused as opposed to other-focused.

Self-Focused Versus Other-Focused Orientations
The idea that humans are self-focused is pervasive in many fields of science, such as economics and law (see, e.g., Walster, Berscheid & Walster, 1973). According to this perspective, people are self-interested rational beings who weigh costs and benefits in hopefully rational manners to arrive at a decision. And indeed, in many situations people are self-interested. This homo economicus view of mankind, however, cannot explain all human behavior and every decision people make. That is, people sometimes make decisions that are clearly not self-interested, but that are more focused on others and on gaining a positive outcome for others, at times even at personal costs (Batson, 1998). Feeling valued and respected by other people and important groups fulfill core human needs (e.g., Baumeister & Leary, 1995). This other-oriented perspective is substantiated by the facts that people are social animals (Aronson, 1972;De Waal, 1996) and that humans live in groups where they need social contact and depend on others for certain goods and services (Baumeister & Leary, 1995;Martin, 1999). This homo socialis view of mankind complements the well-known homo economicus view and underscores the other-oriented focus of human nature.
In line with the proposed differences between self-orientations and otherorientations, many influential social psychological theories make a distinction between self-focused and other-focused needs or values (e.g., Baumeister & Leary, 1995;Deci & Ryan, 2000;Fiske, Cuddy, & Glick, 2007;Van Lange, Otten, De Bruin, & Joireman, 1997). Most of these theories assume that people have a predisposition to be either more self-or other-focused and that situational influences can shift people's focus toward stronger self-orientations or other-orientations (e.g., Brucks & Van Lange, 2007;Van Lange, 1999). We argue that an important field in which these different focuses could influence decisions people make is in the field of social justice, and, more specifically, in reacting to victims of injustice.

Just-World Theory
Relevant for the current purposes is just-world theory (Lerner, 1977(Lerner, , 1980, a seminal theory that aims to explain the paradoxical negative reactions people sometimes have toward innocent victims of injustice. In this theory it is argued that people have a fundamental need to believe that the world is a just place where people get what they deserve; a place in which good things happen to good people and bad things happen to bad people. This belief in a just world (BJW) serves important functions for individuals as it helps them to navigate through the social world. Importantly, it enables people to strive for long-term goals Hafer, 2000a). That is, if people can have faith that everyone gets what they deserve and, hence, that their efforts will pay off in the end, they are able to deal with the uncertainty that is associated with investing in long-term goals . Hence, the BJW is a belief that people want to uphold.
When people observe an innocent victim, they are confronted with the fact that the world in reality is not just. That is why innocent victims pose a threat to people's just-world belief. Victim blaming and derogation have been shown to relieve a BJW threat and uphold the BJW because the victims become deserving of their ill fate (e.g., Hafer, 2000b;Hafer & Bègue, 2005). Just-world theory can also explain why victim blaming and derogation is enhanced when the victims are similar to the observer (Bal & Van den Bos, 2010;Correia, Vala, & Aguiar, 2001Hafer, 2000b). According to just-world theory, these counterintuitive findings arise, because a similar victim (as opposed to a dissimilar victim) will pose a greater BJW threat and heighten the fear that a similar ill fate might bestow on the observer. Hence, both innocent victims and similar victims will evoke stronger reactions to cope with the BJW threat (Lerner, 1980;Hafer, 2000b).
In the real world, however, we do not only see derogatory reactions when people are confronted with innocent suffering. In contrast, people sometimes go to great lengths to help or support those in need, for instance by donating money and time to alleviate the victims' suffering or taking a stand against injustice. These benevolent reactions were part of just-world theory at its introduction (Lerner, 1980;Lerner & Simmons, 1966). That is, Lerner (1980) proposed restitution as a possible strategy to resolve BJW threats. Yet, since the introduction of just-world theory, most research has focused on the negative reactions toward innocent victims (for an overview, see Hafer & Bègue, 2005). Benevolent reactions have only received minor attention in research on just-world theory, with a few notable exceptions (see, e.g., Bègue, Charmoillaux, Cochet, Cury, & De Suremain, 2008;DePalma, Madey, Tillman, & Wheeler, 1999;Hafer & Gosse, 2010;Kogut, 2011). These studies have focused largely on individual differences in just-world beliefs to explain differences in helping and did not manipulate BJW threats nor did they focus on underlying processes that drive reactions to these threats.
In the current line of research, we aim to complement the existing body of literature by focusing on key social psychological moderators that can be hypothesized to influence the negative as well as positive reactions toward victims of injustice. That is, we investigate whether differences in construal of the unjust event in either a more self-focused or other-focused way can make blaming of innocent victims or an effort to help these victims a more likely strategy to resolve the injustice. In the following section, we will further explain why we think these different focuses could influence reactions toward innocent victims by discussing earlier related research.

Self-Focus Versus Other-Focus in Reactions Toward Innocent Victims
Research on perspective taking lends support for the idea that a self-focus versus an other-focus can evoke diverging interpretations of a specific event. That is, studies have shown that in perspective taking it is important to distinguish between placing yourself in the other person's shoes (imagine-self) and a more detached imagining how the other feels (imagine-other; e.g., Batson, Early, & Salvarini, 1997;Lamm, Batson, & Decety, 2007). In the study conducted by Batson et al. (1997), participants in the imagine-self condition were asked to imagine how a similar situation would be for them (self-focus), whereas they imagined how the other must feel in the imagine-other condition (other-focus). While the former enhanced feelings of distress primarily, the latter also enhanced feelings of empathic concern. Here we argue that these feelings of distress (when self-focused) and empathy (when other-focused) may result in very different reactions toward victims of injustice.
Specifically, we propose that a self-focus enhances distress of a similar fate happening to the observer (i.e., a BJW threat), resulting in enhanced derogatory reactions to resolve this threat. Furthermore, we put forward that an other-focus enhances empathic feelings, resulting in more benevolent reactions (cf. Lerner et al., 1976). Hence, when other-focused people may not interpret the confrontation with the innocent victim as a BJW threat and, hence, reactions may not strongly be affected by variations in BJW threat. How the distinction between self-focused and other-focused construals of events relates to people's benevolent and derogatory reactions toward innocent victims of injustice will be investigated in the current research.
Some studies on derogatory reactions toward innocent victims assumed, but did not test directly, that a self-focus or an other-focus could underlie victim blaming. Specifically, the role of the self in negative reactions toward innocent victims has received some attention in research on self-regulation and the BJW. For example, Loseman and Van den Bos (2012) argued that a victim enhances aversive thoughts and emotions, because he/she poses a self-threat to the observer. Hence, they suggested that people will be primarily focused on the consequences of a BJW threat for the self when blaming an innocent victim. Other studies showed that mimicking reduced victim blaming, regardless of whether people mimicked the person who they later learned was victimized or a different person. In this line of research Stel, Van den Bos, and Bal (2012) suggested (but did not test) that mimicking might put people in a general other-oriented mindset. In the current line of research, we will directly manipulate a self-focus and an other-focus and examine the effects of these manipulations on both victim blaming and victim helping.
Taken together, the studies reviewed are in line with our proposed line of reasoning and lend some indirect support for the proposition that a self-focus might enhance and an other-focus might reduce derogatory reactions toward innocent victims of injustice. However, this assumption has not been tested directly. Moreover, these studies did not examine whether a self-focus and an other-focus are related toward helping of innocent victims, as our line of reasoning predicts. Hence, we contribute to the existing body of research in at least two important ways.

The Current Research
In two studies, we set out to investigate the influence of adopting a self-versus an other-focus on the diverging behavioral reactions toward innocent victims of injustice. We propose that a self-focus will enhance derogatory reactions toward innocent victims and suppress helping of the victims involved. We further propose that an other-focus will decrease derogatory reactions toward innocent victims and enhance helping of the victims involved. Consistent with earlier studies (e.g., , Hafer, 2000b, Loseman & Van den Bos, 2012Van den Bos & Maas, 2009;Van Prooijen & Van den Bos, 2009), we expect to find these effects especially following a high BJW threat. In these instances, people have a clear need to restore their BJW, whereas under a low BJW threat, reactions can be less pronounced (Hafer, 2000b).
In the two studies we induced a self-focus or an other-focus and investigated blaming (Study 1) and helping reactions (Study 2) toward innocent victims of injustice. We also included a neutral control condition in both studies. In this condition we did not manipulate focus. By comparing the results from this control condition to the self-focus and other-focus conditions, we gain insight into how people generally react toward innocent victims and thus how we should interpret the effects obtained in the self-and other-focused conditions.

Study 1
In Study 1 we focused on negative reactions toward innocent victims and examined whether a self-focus as opposed to an other-focus would enhance victim blaming and derogation. We manipulated focus by asking participants to think back to and describe a situation in which they were either self-focused or other-focused. Following this manipulation, participants read a newspaper article about an accident. To induce a high or low BJW threat, we manipulated whether the crime was distal or proximal (Bal & Van den Bos, 2010). According to Miller (1978, p. 1031), ''as events become closer to [people's] world (…), the concern over injustices increases greatly as does the need to explain or make sense of the event.'' In earlier studies, social proximity has been manipulated to induce a high and low BJW threat successfully (e.g., Bal & Van den Bos, 2010, Correia et al., 2007. Physical proximity, we believe, works in a similar way and has been used as a manipulation of BJW threat successfully in previous research (Bal & Van den Bos, 2012, Study 3). Subsequently, we measured victim blaming and derogation in a subtle manner that may be reflective of how people ascribe responsibility to the victim in the real world. That is, we asked participants to estimate how much alcohol they thought the victim had drunk before the accident (Girasek, Gielen & Smith, 2002).
We predicted that a self-focus would lead to more blaming reactions toward the victim (i.e., more alcohol consumed by the victim) than an other-focus. This effect was expected to occur especially under a high BJW threat, so when participants were confronted with a proximal as opposed to a distal victim. To establish whether people normally react from a self-or an other-focused perspective when confronted with an unjust situation, we also included a neutral control condition in Study 1 in which we did not manipulate focus.

Participants and Design
One hundred and seventy Utrecht University students (71 men and 99 women) participated in our study. 1 Their ages ranged from 18 to 34 years (M = 21.33, SD = 2.02). The study had a 3 (Focus: self vs. other vs. control) 9 2 (BJW threat: high vs. low) between-subjects design. Between 25 and 31 participants took part in each of the six conditions.

Procedure and Materials
The researchers approached students at different restaurants at the Uithof campus of Utrecht University and were asked whether they wanted to fill out a short questionnaire. When students were willing to participate, a questionnaire was handed to them and the researchers continued to approach other potential participants. After about 15 min the researchers came back to collect the questionnaire and participants were thanked and debriefed. The study started with our focus manipulation. Subsequently, participants read a newspaper article about an accident and negative reactions toward the victim were measured. The experiment ended with some demographic questions.

Focus Manipulation
In the other-focused condition, participants were asked to think of a situation in which they were focused on another person and to describe that situation in detail. This was followed by three open-ended questions: (1) ''Please describe, as specifically as possible, how that person acted in the situation,'' (2) ''Please describe what feelings you think this other person had in that situation,'' 1 A total number of 180 participants took part in Study 1. Nine participants were excluded from analyses because of missing values and 1 because he had a Cook's (1977) distance score more than 3 SDs above the mean in the main analyses (Cohen et al., 2003). This resulted in an effective sample of 170 participants.

BJW Threat
In an ostensibly unrelated experiment, participants read a newspaper article about Thomas Weijers (a typical Dutch name). Thomas had spent his evening with friends, but decided to go home earlier than his friends at 2 a.m. and left by himself. When he was cycling home, he got hit by a car on a dangerous crossing. Thomas loses consciousness and wakes up in the hospital. He suffered a severe head injury and became paralyzed from the waist down because of this accident. The driver of the car did not stop after hitting Thomas and was therefore not caught. To induce a high or low BJW threat, proximity was manipulated. In the highthreat condition, Thomas was described as a fellow Utrecht University student and the accident happened in Utrecht. In the low-threat condition, Thomas was described as a Groningen University student and the accident took place in Groningen (a different Dutch city far away from Utrecht).

Indirect Negative Reactions
After an explicit blaming and derogation questionnaire (Bal & Van den Bos, 2010) and some filler questions about the article, participants were asked with an open-ended question how many glasses of alcohol they thought the victim had drunk before he was involved in the accident. Whether or not the victim had drunk alcohol was not explicitly mentioned in the newspaper article. Alcohol intoxication indicates that the victim was (partially) to blame for the accident according to the participants (e.g., Girasek, Gielen & Smith, 2002), so we took the amount of glasses of alcohol participants' estimated Thomas had consumed 2 We conducted a separate manipulation check on the computers in the lab to see whether out focus manipulation worked as expected. In this study, 105 students (44 men and 61 women) participated. Their ages ranged from 18 to 38 years (M = 20.76, SD = 2.91). The study started with our focus manipulation, creating three conditions in which between 33 and 35 participants took part. Subsequently, we measured whether our manipulation worked as intended by measuring participants' focus. Specifically, we asked participants to what degree they were (1) focused on themselves, (2) focused on others, (3) thinking about themselves, (4) thinking about other people, and whether they were more self or other -focused and thinking about themselves or others more at this moment. In the self and other-focus condition the order of the questions fit the manipulation (i.e., self-focus: 1-2-3-4; other-focus: 2-1-4-3). In the control condition, the order of the questions was counterbalanced to check for order effects. No order effects were found in this condition, F(1, 33) = 0.30, p [ .59, g p 2 = .01. An ANOVA on the manipulation check with focus as independent variable revealed a main effect of our focus manipulation, F(2, 102) = 6.46, p \ .01, g p 2 = .11. Participants in the other-focus condition (M = 3.72, SD = 1.25) reported being more other-focused than participants in both the self-focus condition (M = 2.81, SD = 0.77), p \ .01, and the control condition (M = 3.19, SD = 1.08), p \ .04. The self-focused condition did not differ significantly from the control condition, even though the pattern of results was in the expected direction, p [ .14. This may indicate that people generally adopt a self-focused mindset.
as an indirect measure of participants' negative evaluation of the victim. We indicated to participants that they could fill in ''0'' if they thought the victim did not drink any alcohol that night.

Results
A 3 (Focus) 9 2 (BJW threat) ANOVA 3 on glasses of alcohol revealed a significant main effect of BJW threat, F(1, 164) = 5.09, p \ .03, g p 2 = .03. This effect was qualified by a two-way interaction between focus and BJW threat, F(2, 164) = 3.57, p \ .03, g p 2 = .04. As can be seen in Fig. 1, statistically significant effects of the focus manipulation were found in the high BJW threat condition, F(2, 164) = 3.70, p \ .03, g p 2 = .04, and not in the low BJW threat condition, F(2, 164) = 0.90, p [ .41, g p 2 = .01. To further interpret these effects, we performed a least significant difference test for multiple comparisons between means (p \ .05), with the six cells of our design serving as the independent variable. Figure 1 shows the results of this test. As hypothesized, under a high BJW threat self-focused participants evaluated the victim more negatively (M = 4.59, SD = 2.89) than participants with an otherfocus (M = 2.84, SD = 2.35). Furthermore, participants in the control condition (M = 4.17, SD = 2.37) also evaluated the victim more negatively than otherfocused participants. The control condition did not differ from the self-focus condition.
Additionally, the least significant difference test revealed that, when focused on themselves, participants thought that the proximal victim had drunk more than the distal victim (M = 2.92, SD = 3.38). In the control condition, participants also indicated that the proximal victim drank more glasses of alcohol than the distal victim (M = 2.50, SD = 2.33). When focused on another person, this effect attenuated and was not statistically significant (other-focus & distal victim: M = 3.44, SD = 2.36). 4

Discussion
The results of Study 1 revealed that self-focused people blamed the proximal victim more for what happened to him than other-focused people, indicated by the 3 In Studies 1 and 2, including gender in our analyses had no significant effects on the variables analyzed. We therefore dropped gender from the analyses reported. 4 In Study 1, we also included the negative victim evaluation questionnaire (a = .92; cf. Bal & Van den Bos, 2010. A 3 (Focus) 9 2 (BJW threat) 9 2 (Gender) ANOVA on negative evaluation revealed a main effect of BJW threat, F(1, 158) = 6.41 p \ .02, g p 2 = .04, a main effect of gender, F(1, 158) = 9.34, p \ .01, g p 2 = .06, and a significant three-way interaction between focus, BJW threat, and gender, F(2, 158) = 2.16, p \ .04, g p 2 = .04. Conducting separate analyses for men and women (Bal & Van den Bos, 2010), analyses revealed that men only show a significant main effect of BJW threat, such that they evaluated the proximal victim (M = 2.64, SD = 0.88) more negatively than the distal victim (M = 2.29, SD = 0.67), F(1, 65) = 4.20, p \ .04, g p 2 = .07. For women, a significant two-way interaction of BJW threat and focus was found, F(2, 93) = 3.57, p \ .04, g p 2 = .07. Only in the self-focus condition did women evaluate a proximal victim (M = 2.52, SE = 0.19) more negatively than a distal victim (M = 1.76, SE = 0.20), F(1, 93) = 7.79, p \ .01, g p 2 = .08. These gender differences were absent, in our more subtle measure of victim blaming that asked about alcohol intake. amount of alcohol they estimated the victim to have consumed before the accident. Moreover, this effect was only there following a high BJW threat as opposed to a low BJW threat. In the control condition, people also indicated that the victim drank more alcohol following a high BJW threat as opposed to a low BJW threat. This effect attenuated in the other-focused condition. As such, reactions observed in the control condition were more similar to responses obtained in the self-focused condition than to those observed in the other-focused condition. Perhaps this indicates that a self-focus might be the default when confronted with an innocent victim. Future studies can be conducted to substantiate this idea.
We would like to emphasize that the newspaper article was presented to our research participants as an unrelated study. As such, participants could not take the victim's perspective in the focus manipulation. Moreover, many situations participants described in the other-focus condition had nothing to do with taking an other's perspective (e.g., listening to a lecturer). In our view, an other-focused construal of the world around you could encompass, but does not exclusively entail taking another person's perspective. Furthermore, taking another's perspective, can lead to more, but also to less, empathic reactions. That is, earlier research showed that two forms of perspective taking exist (Lamm et al., 2007). People can either imagine themselves in the other person's shoes or imagine how the other must feel. Lamm et al. (2007) showed that while the former enhanced feelings of distress, only the latter also enhanced feelings of empathic concern. Hence, we believe that focus is different from perspective taking in several ways and propose that an other-focus specifically will lead to more benign reactions toward innocent victims. In Study 1, we found that derogatory reactions toward victims were reduced in the other-focus condition as opposed to the self-focus condition. To see how more benign ways of dealing with a BJW threat are affected by focus, in Study 2, we examine victim helping as a reaction toward innocent suffering. Glasses of alcohol High threat Low threat Fig. 1 Participants' estimates of glasses of alcohol drank by the victim as a function of focus and BJW threat (Study 1). Note Means with no letters (a, b) in common differ significantly (p \ .05), as indicated by a least significant difference test for multiple comparisons between means (Kirk, 1982) Study 2 In Study 2 we focused on positive reactions toward innocent victims and examined whether an other-focus as opposed to a self-focus would enhance helping of innocent victims. The experimental setup was largely similar to Study 1 in that we used the same focus manipulation and victim scenario. However, in Study 2 we measured positive reactions following the confrontation with an innocent victim. That is, participants could help the victim by raising money for his rehabilitation. Thus, we measured actual helping and not merely helping intentions. We expected that an other-focus would lead to more help for the victim (i.e., spending more time raising money) than a self-focus. This effect was expected to occur especially under a high BJW threat, so when participants were confronted with a proximal as opposed to a distal victim. As in Study 1, we also included a neutral control condition, in which we did not manipulate focus, to obtain additional evidence about how people normally react to innocent victims from a self-focused or an other-focused perspective.

Participants and Design
One hundred and thirty-seven Utrecht University students (70 men and 67 women) participated. 5 Their ages ranged from 17 to 43 years (M = 21.33, SD = 3.00). The study had a 3 (Focus: self vs. other vs. control) 9 2 (BJW threat: high vs. low) between-subjects design. Between 20 and 27 participants took part in each of the six conditions.

Procedure and Materials
Participants were invited to take part in this study with flyers posted at different locations on the Uithof campus of Utrecht University. Upon arrival in the laboratory, participants were seated behind a computer in individual cubicles. The study started with our focus manipulation. Subsequently, participants read the newspaper article used in Study 1. Following this, participants were asked to volunteer some time to help the victim by solving math problems to raise money. The experiment was programmed on the computer from start to finish.

Focus Manipulation
The focus manipulation was the same as in Study 1.
BJW Threat BJW threat was manipulated through using the same newspaper article as in Study 1. 5 A total number of 164 participants took part in Study 2. Eighteen participants were excluded from analyses because they had participated in Study 1; 5 because they indicated that they had not complied with the instructions; and 4 because they had a Cook's (1977) distance score more than 3 SDs above the mean in the main analyses (Cohen et al., 2003). This resulted in an effective sample of 137 participants. Soc Just Res (2015) 28:274-292 283 Helping the Victim After reading the newspaper article, participants learned that the victim suffered severe neurological damage, because of the accident. Without help from a personal study coach, he would therefore not be able to finish his study. Participants could voluntarily help the victim and raise money for a personal study coach by solving math problems. This procedure was based on Freerice.com, a website hosted by the United Nations World Food Programme that provides rice to hungry people. On the website you can answer trivia questions in several categories and earn rice grains for each correct answer. The rice that you earned is donated to the World Food Programme. We programmed a similar paradigm for our experiment, which was focused on helping Thomas by raising money for a personal study coach. Participants were asked to solve math problems of increasing difficulty ranging from simple additions and subtractions (e.g., 5 ? 8) to multiple divisions and multiplications (e.g., 9 9 7 9 3). For each set of two correct answers, participants raised € 0.05 for Thomas. Participants could spend as much time as they wanted on solving the math problems with a maximum number of 200 problems. They could opt to stop throughout the task by clicking a button that remained on the screen during the task. Money raised in euros and time spent on the math problems in milliseconds constituted our main dependent variables.

Time Spent on the Sums
As scores on this variable were skewed to the right, we performed a logtransformation on the scores before analyzing the data (Fazio, 1990). For ease of interpretation, untransformed means and standard deviations are presented in Fig. 2 and below. A 3 (Focus) 9 2 (BJW threat) ANOVA on time spent on the sums revealed a marginally significant main effect of BJW threat, F(1, 131) = 3.20, p \ .08, g p 2 = .02. This effect was qualified by a significant two-way interaction between focus and BJW threat, F(2, 131) = 5.30, p \ .01, g p 2 = .08. As can be seen Time spent on the sums High threat Low threat Fig. 2 Time spent on the sums as a function of focus and BJW threat (Study 2). Note Means with no letters (a, b) in common differ significantly (p \ .05), as indicated by a least significant difference test for multiple comparisons between means (Kirk, 1982) in Fig. 2, statistically significant effects of the focus were found in the high BJW threat condition, F(2, 131) = 6.63, p \ .01, g p 2 = .09, and not in the low BJW threat condition, F(2, 131) = 0.76, p [ .47, g p 2 = .01. To further interpret these effects, we performed a least significant difference test for multiple comparisons between means (p \ .05), with the six cells of our design serving as the independent variable. Figure 2 shows the results of this test. As hypothesized, under a high BJW threat self-focused participants spent less time on

Money Raised
For the amount of money raised for the victim, the effects were comparable to time spent on helping the innocent victim, although they were less strong. A 3 (Focus) 9 2 (BJW threat) ANOVA on money raised revealed a marginally significant main effect of BJW threat, F(1, 131) = 3.73, p \ .06, g p 2 = .03. This effect was qualified by a marginally significant two-way interaction of focus and BJW threat, F(2, 131) = 2.62, p \ .08, g p 2 = .04. As can be seen in Fig. 3, statistically significant effects of the focus manipulation were found in the high BJW threat condition, F(2, 131) = 3.46, p \ .04, g p 2 = .05, and not in the low BJW threat condition, F(2, 131) = 0.64, p [ .52, g p 2 = .01. To better understand these effects, we performed a least significant difference test for multiple comparisons between means (p \ .05), with the six cells of our design serving as the independent variable. Figure 3 shows the results of this test. As hypothesized, under a high BJW threat self-focused participants raised less money for the victim (M = 0.80,

Money raised
High threat Low threat Fig. 3 Money raised for the victim as a function of focus and BJW threat (Study 2). Note Means with no letters (a, b) in common differ significantly (p \ .05), as indicated by a least significant difference test for multiple comparisons between means (Kirk, 1982) Soc Just Res (2015) 28:274-292 285 SD = 0.81) than participants with an other-focus (M = 1.38, SD = 0.67). There was no significant difference between the control condition (M = 1.10, SD = 0.77) and the self-focus condition or the other-focus condition. Additionally, the least significant difference test revealed that, when focused on themselves, participants raised less money for the proximal victim as opposed to the distal victim (M = 1.35, SD = 0.71). The effects of BJW threat in the other two focus conditions (distal victim and other-focus: M = 1.22, SD = 0.89; distal victim and control: M = 1.48, SD = 0.74) were not significant, ps [ .11.

Discussion
The results of Study 2 indicate that other-focused people help a proximal victim more than self-focused people. Furthermore, self-focused people help a high-threat victim less than a low-threat victim. These findings fit with and extend the findings from Study 1 and earlier findings on negative reactions following a BJW threat. That is, self-focused people seem to help an innocent victim less, when they pose a high BJW threat as opposed to a low BJW threat. This effect attenuated in the otherfocused condition. In the other-focused condition people helped a victim more, regardless of BJW threat.
In the control condition of Study 2, participants seemed to help an innocent victim less under a high BJW threat as opposed to a low BJW threat, although this effect was definitely not statistically significant. Together with the results of Study 1, this could indicate that people generally adopt a self-focus after being confronted with innocent victims of injustice. Under a high BJW threat though, the control condition of Study 2 seemed to fall in between the self-focused and the otherfocused conditions. Further studies into supportive reactions following a confrontation with an innocent victim are necessary to further disentangle these effects. Nevertheless, from both studies it is clear that self-focused people helped an innocent victim less and blamed her more than other-focused people, especially under a high BJW threat.

General Discussion
Coming back to the example which we began with in our introduction: What would you do if you would read about the car accident victim? Would you spend time raising money for his rehabilitation? Or would you think, why was he so careless; everybody knows how dangerous that crossing is? Our results suggest that your focus-whether you view the situation from a self-focused or an other-focused perspective-could influence whether you would be more inclined to help or blame the victim. That is, in two studies we found that a self-focus enhanced derogatory reactions (on measures of victim blaming and derogation and alcohol intake) and suppressed helping of the victim (on measures based on the Freerice game), whereas an other-focus enhanced helping and decreased derogatory reactions toward the victim, especially when the victim involved posed a high BJW threat.
Taken together, our studies complement the existing body of literature of justworld theory by studying both derogatory and benevolent reactions toward innocent victims of injustice (Lerner, 1977(Lerner, , 1980. When just-world theory was introduced, both victim blaming and derogation and helping were proposed as strategies to restore the belief in a just world after a confrontation with innocent victims (Lerner, 1980). Yet, benevolent reactions have only received minor attention in research on just-world theory. Moreover, the studies that have been conducted focused largely on individual differences in just-world beliefs to explain differences in helping (e.g., Bègue et al., 2008;DePalma et al., 1999). Nevertheless, more recent studies have focused on a broader array of possible reactions following a BJW threat and on explaining when people choose to react in a certain way (see, e.g., Hafer & Rubel, 2015;Harvey, Callan & Matthews, 2015). Our current studies fit within this line of research and aim to further our understanding of the mechanisms underlying diverging reactions toward innocent victims. Importantly, we have added to the existing body of literature by introducing an indirect measure of victim blaming and by studying actual helping behavior (instead of helping intentions).
In our studies, we manipulated BJW threat to gain insight into the processes that take place when people are confronted with an innocent victim and that make blaming or helping more likely reactions to resolve the BJW threat. According to just-world theory, people will first try to help or compensate the victim and only resort to the more negative derogatory strategies, when benevolent strategies are deemed futile, too costly, or simply impossible (Lerner & Goldberg, 1999;Lerner et al., 1976). However, our studies suggest that people's spontaneous reaction could be self-focused, resulting in victim blaming and derogation. When people are otherfocused, they will more likely try to help the victims involved. In the other-focused condition, differences between the high and low BJW threat scenarios attenuated. Hence, other-focused reactions may not be aimed at resolving a BJW threat (primarily) and as such may not be related (solely) to the BJW.
In both Studies 1 and 2, we found some support for the idea that people may become more self-focused when confronted with an innocent victim, although these effects are less pronounced with regard to benevolent reactions. Self-focused people indirectly blamed the victim more when she posed a high BJW threat as opposed to a low BJW threat in Study 1 and they helped the high BJW threat victim less as opposed to the low BJW threat victim in Study 2. These effects of more blaming (and less helping) of after a high BJW threat as opposed to a low BJW threat are in line with earlier findings on victim blaming and derogation (e.g., Bal & Van den Bos, 2010Correia et al., 2001;Hafer, 2000a, b;Van Prooijen & Van den Bos, 2009). Moreover, we found a similar effect in the control condition of Study 1 and similar pattern in the control condition of Study 2. These findings could suggest that people might spontaneously adhere to a self-focus when confronted with innocent victims of injustice. Of course, individual and cultural differences (e.g., Cohen, Hoshino-Browne & Leung, 2007) could impact on these effects and future studies can now look into these differences.
In both studies we used, a BJW threat manipulation of crime proximity. We used this manipulation of BJW threat, as we believe it to be a clear manipulation of BJW threat (see also Bal & Van den Bos, 2010). Moreover, although we did not include an explicit manipulation check in our current studies, earlier studies did show that social proximity heightens a BJW threat (e.g., Bal & Van den Bos, 2010;Correaia et al., 2007). Moreover, our findings in the control condition of Study 1, showing that people blamed a proximal victim more than a distal victim, are in line with many earlier studies that included various other successful BJW threat manipulation (e.g., Hafer, 2000b, Loseman & Van den Bos, 2012Van den Bos & Maas, 2009;Van Prooijen & Van den Bos, 2009). Hence, we believe crime proximity to be a valid manipulation of BJW threat.
That being said, one may argue that our crime proximity manipulation could be associated with participants' group membership and as such might be related to our focus manipulation. While we acknowledge this point, we believe that if our effects of crime proximity were driven by group membership instead of BJW threat, an opposite pattern of effects would have been more likely. That is, based on ingroup favoritism and outgroup derogation (Taifel & Turner, 1979), one would expect the proximal victim to be blamed less and helped more than the distal victim, because of heightened identification with and enhanced feelings of empathy for the ingroup victim. Our findings do not support this line of reasoning, but instead follow a BJW threat interpretation of the effects (showing more blaming and less helping of proximal as opposed to distal victims in crucial conditions in our experiments). Hence, we believe that our findings will hold across different manipulations of BJW threat. That being said, we hope that the current studies will inspire future research to examine the role of different BJW threat manipulations in reactions toward innocent victims of injustice.
Furthermore, we proposed that underlying these reactions of blaming versus helping, a self-focus would elicit feelings of threat following a confrontation with an innocent victim, whereas an other-focus would elicit feelings of concern for the victim (cf. Lamm et al., 2007;Lerner et al., 1976). These different interpretations of the event would make either victim blaming or helping a more likely reaction, respectively. In the current research we focused on the subsequent reactions that follow these appraisals. It would be worthwhile for future research to further look into and measure these different appraisals of threat and empathy that are possible after a confrontation with an innocent victim. We note this because helping may not constitute a BJW threat-related reaction solely and other psychological factors are important to understand the social psychology of helping as well.
Moreover, we assessed blaming and helping with two different measures. We also would applaud future research that attempts to measure helping and blaming in one study with one measure that simultaneously reflects either low levels of blaming and high levels of helping (or vice versa). This way a greater tendency to either blame or help a victim could be tested and it could be examined when people choose what type of reaction instead of studying (the strength of) one type of reaction. For now, we conclude that our results fit with our line of reasoning, as we found in two empirical studies that a self-focus made negative reactions more likely, while an other-focused evoked more helping of the victims involved.
Our manipulation of self-and other-focus shows some resemblance to the distinction between independent versus interdependent self-construals (Singelis, 1994). However, it is important to distinguish between the construal of the situation (self-focused vs. other-focused) and self-construal (independent vs. interdependent). Self-construal refers to whether people describe themselves in terms of group membership or in terms of individual qualities. A self-focused versus and otherfocused construal of the situation refers to whether people construe a certain situation by focusing on personal consequences (i.e., a self-focus) or by taking into account other people's feelings, consequences and outcomes (i.e., an other-focus). Earlier studies, focusing on self-construal and empathic and derogatory reactions, have shown mixed results. On the one hand, studies have shown that interdependence is associated with greater empathy, perspective taking and recognition of others' emotions (e.g., Cohen, Hoshino-Browne & Leung, 2007). This could indicate that an interdependent self-construal is comparable to an other-focus. On the other hand, research that focused specifically on reactions following BJW threats (Van Prooijen & Van den Bos, 2009) showed that people with an interdependent self-construal tend to blame innocent victims more than people with an independent self-construal. This indicates that an interdependent self-construal is more comparable to a self-focus instead.
In line with the explanation of Van Prooijen and Van den Bos (2009), we propose that these different findings may be explained by the fact that an interdependent self-construal leads to an assimilation with others. In general this may lead to greater empathy, but when people assimilate with a victim specifically, this may heighten the fear of a similar ill fate bestowing on them. Our findings fit with this idea, as a self-focus also led to more victim blaming and less helping than an otherfocus. Hence, an assimilation with a victim may be caused by a focus on the self. Thus, while similar in some regards, independent and interdependent self-construals and self-focused and other-focused construals of the situation, respectively, can lead to very different reactions toward innocent victims who pose a BJW threat.
More broadly, our findings add to the seminal theories that make a distinction between people as rational economic decision makers and social-oriented individuals (e.g., Baumeister & Leary, 1995;Deci & Ryan, 2000;Fiske et al., 2007;Van Lange et al., 1997). In the current research, we show differentiated reactions toward victim of injustice following a self-focus versus an other-focus. Whereas the homo economicus more likely resolves unjust situations by blaming or derogating the victim, the homo socialis more likely tries to alleviate the victims' suffering by helping or supporting them.

Coda
Benevolent reactions following a confrontation with injustice are often visible in the real world. People can go to great lengths to help innocent victims of natural disasters, charity raisers can be a great success and people take to the streets to protest against grave injustices all over the world. We would like to suggest that it now is the time to also move research on social justice and the reactions people have toward innocent victims of injustice beyond the derogatory reactions and integrate help and support in the scope of reactions that can occur after a confrontation with injustice. In our research, we hope to have taken a step in this direction by showing that a self-focus makes victim blaming a viable way of dealing with injustice, while an other-focus increases the likelihood that people will help the victims involved. As such, we have tried to put both the ''I' ' and ''us'' in ''justice.''