Introduction

In the days following hurricane Katrina, thousands of Americans were stranded in the Gulf Coast with little access to food, water, shelter, and the outside world. The images of human suffering that emerged following this disaster led Americans to wonder how one of the richest nations in the world was unable to help entire communities that were devastated by the storm. The fact that those most devastated by the storm were predominately Black led many to wonder whether racism was responsible for the poorly conceived and executed disaster response. Such racialized explanations for the disaster response were prominent in national magazines and newspapers, and on news and radio broadcasts.Footnote 1 , Footnote 2 This sentiment was captured by musician Kanye West’s controversial remark: “George Bush doesn’t care about Black people.” This paper draws upon theoretical perspectives on system justification and social justice to examine whether the racialized explanations that were prominent in the post-hurricane Katrina media coverage affected Black and White Americans’ racial attitudes.

An understanding of the effects of exposure to racialized explanations can be informed by a growing area of social psychological research on claiming discrimination. This research demonstrates that members of low-status groups (e.g., African Americans) incur negative interpersonal costs, such as being targeted by retaliation, when they claim to be the target of discrimination (see Kaiser, 2006; Stangor et al., 2003 for reviews). Recently, scholars have drawn upon theoretical perspectives on justice and legitimacy to understand the social costs of claiming discrimination (Blasi & Jost, 2006; Jost & Burgess, 2000; Kaiser, 2006; Kaiser, Dyrenforth, & Hagiwara, 2006). For instance, Kaiser and her colleagues argued that when groups at the bottom rungs of the status hierarchy claim to be victims of discrimination, these claims are at odds with the system justification motivation—an ideological belief system aimed at defending, rationalizing, and legitimizing aspects of the status quo, such as the status structure in a given society (Jost & Banaji, 1994; Jost, Banaji, & Nosek, 2004; Jost & Hunyady, 2005). Specifically, because discrimination claims convey that the US status system is illegitimate and that some groups face unfair barriers that block them from reaching the upper rungs of the status hierarchy, these claims challenge perceptions of system legitimacy. Indeed, Kaiser et al. (2006) found that White Americans who strongly endorsed system-justifying beliefs (e.g., Social Dominance Orientation, Just World Beliefs, Protestant Work Ethic beliefs) were particularly threatened by an African American who blamed a failing grade on a racist course instructor and they responded particularly harshly toward that individual. White Americans who strongly endorsed system-justifying beliefs did not experience increased threat or behave harshly toward an African American who blamed a failing grade on a source of unfairness that was not relevant to system legitimacy beliefs (a course instructor who was a “jerk” to all students).

The discrimination claims stemming from the tragically botched disaster response following hurricane Katrina are particularly ripe for studying through the theoretical lenses of system justification and social justice (see also Adams, O’Brien, & Nelson, 2006; Belle, 2006; Napier, Mandisodza, Andersen, & Jost, 2006). For example, these racialized explanations posed a strong challenge to the belief that all individuals, regardless of group membership, have equal opportunities in America. Because post-hurricane Katrina racialized explanations suggested that racism continues to deny Blacks an equal chance in life (indeed, in New Orleans, racialized claims implied that one’s race could determine whether one lived or died following the storm), these explanations can be conceptualized as a threat to the legitimacy of the social system, and particularly to the legitimacy of the status hierarchy. Because White Americans are especially invested in maintaining the legitimacy of the status structure in which their group resides at the top, they should be motivated to engage in system-defensive behavior following exposure to system threat (Jost & Banaji, 1994; Jost et al., 2004; Sidanius & Pratto, 1999).

Individuals have at their disposal a number of strategies that they can employ in the service of defending and justifying their system beliefs (see Jost et al., 2004 for a review). This paper focuses on ingroup favoritism as one type of system defense. In a cumulative review of research on system justification, Jost et al. (2004) argued that threats to system legitimacy motivate ingroup favoritism among groups at the top of the status hierarchy. By emphasizing the worthiness of their group and seeing it in a particularly positive light, high-status groups are able to rationalize the system and the status hierarchy. Consistent with this idea, Whites who are chronically preoccupied with system justification (e.g., those who possess a conservative political ideology) show more racial ingroup favoritism relative to individuals who are less concerned with system-justifying needs (e.g., those who endorse a liberal ideology) (Jost et al., 2004; Sidanius, Levin, Frederico, & Pratto, 2001). Accordingly, a system justification perspective predicts that when White Americans encounter threats to the legitimacy of the status hierarchy, such as those posed by racialized explanations, they may be particularly likely to respond by showing greater allegiance to their racial group and sticking firm to the belief that their group is worthy of its high-status.

The present research tests this logic by examining whether White Americans express more ingroup favoritism following exposure to media clips arguing that racism was responsible for the poorly conceived and executed hurricane Katrina disaster response. Furthermore, these reactions are compared to reactions following exposure to media clips contending that the government’s incompetence was the cause of the ineffective disaster response. Although governmental incompetence is still troubling, involves a judgment of blame, and can even be conceptualized as a general system threat (e.g., Napier et al., 2006), this type of explanation does not directly challenge beliefs that legitimize status differences in the US. Thus, increased ingroup favoritism is unlikely to directly resolve this type of general system threat.

Additionally, this research examined how these two types of explanations for the ineffective disaster response would affect Black Americans. Because Black Americans have more experience with injustice and disadvantage compared to White Americans, the former are more likely to reject system-justifying beliefs (Jost & Thompson, 2000; O’Brien & Major, 2005; Schmader, Major, & Gramzow, 2001; Sidanius & Pratto, 1999). Therefore, Black Americans may have less motivation to defend the legitimacy of the status system after exposure to racialized explanations. Indeed, in the weeks following hurricane Katrina, Black Americans reported lowered endorsement of Protestant Work Ethic beliefs than they did both prior to the storm and several months after the storm (Levy, Freitas, Mendoza-Denton, & Kugelmass, 2006). White Americans did not show a dip in Protestant Work Ethic beliefs in the weeks following hurricane Katrina. This suggests that Whites were actively rationalizing the system, whereas Blacks, at least in the short term, were willing to forsake this type of system-justifying belief. If Blacks did feel a need to defend the system, this would be manifested in outgroup favoritism (more positive attitudes toward Whites) following racialized explanations (Jost et al., 2004). As this research employs explicit, rather than implicit, measures of system defense, it is unlikely that Blacks would show pro-White bias following racialized explanations (Jost et al., 2004; Jost, Pelham, & Carvallo, 2002). Nonetheless, this research explored whether racialized explanations would affect Blacks’ intergroup attitudes.Footnote 3

According to a system justification theoretical perspective, White Americans who are exposed to racialized explanations should express increased levels of ingroup attachment and ingroup favoritism compared to White Americans exposed to non-racialized explanations. As racialized explanations pose less of a threat to Blacks’ beliefs about the legitimacy of the status system, these explanations should not cause defensive intergroup bias among Black Americans. Ingroup favoritism is conceptualized as especially favorable attitudes toward the ingroup, namely, ingroup love (Brewer, 1999). The expression of ingroup love is more easily justifiable and hence more socially acceptable to express than its counterpart, outgroup hate (Brewer, 1999; Dovidio & Gaertner, 2004), making it a sensitive and appropriate way to measure intergroup attitudes.

Method

Participants

Participants were 93 White (59.1% women; M age = 18.8 years, SD = 1.1 years) and 60 Black (66.7% women; M age = 19.1 years, SD = 1.6 years) undergraduates who participated in exchange for $10. Participation occurred between mid-November 2005 and mid-January 2006.

Procedure

Small groups of participants arrived at the lab and were met by a non-Black experimenter who explained that participants would watch computerized digital videos describing hurricane Katrina and would then provide their reactions toward the videos. The experimenter then escorted participants to small rooms with privately enclosed computers equipped with MediaLab software (Jarvis, 2004), 17-in. CRT monitors, and headphones.

In both experimental conditions, the video began with 5 min of clips showing the devastation caused by the storm and the growing chaos developing in the days following the storm (e.g., deteriorating conditions at the Superdome and New Orleans Convention Center, people stranded on rooftops waiting for helicopter rescues). This footage came from National Geographic’s, “Inside Hurricane Katrina.” Although these videos communicated that the disaster response was clearly inadequate, they did not convey any judgments of blame for this inadequacy.

At this point, the video content diverged for the two experimental conditions. Participants in the “Race Blame” condition watched a 6-min series of video segments in which Katrina victims, public figures, and journalists claimed that the government responded slowly to the disaster because the majority of the hurricane victims were Black. They claimed, for example, that patients at Gulf Coast hospitals serving primarily minority populations were left to die while patients at hospitals serving primarily White populations were rescued, that the government would have responded more rapidly if those primarily affected by the hurricane were White, that the media mischaracterized Black hurricane survivors as violent looters rather than as victims trying to survive, and that the government more generally abandoned its Black citizens. Thus, the race blame condition conveyed that the inadequate government response, as well as depictions and perceptions of hurricane Katrina victims, stemmed from racism.

Participants assigned to the “Government Incompetence” condition watched a 6-min series of video clips in which Katrina victims, public figures, and journalists attributed the ineffective disaster response to the government’s incompetence. For instance, these individuals claimed that the state and national government failed to adequately communicate with each other and were unresponsive to requests for help from the local government, that the government took a callous “wait and see” attitude rather than preparing for the worst prior to the storm, that Federal Emergency Management Agency leaders were inadequately trained, and that the United States government was severely lacking in leadership during this emergency. Thus, these videos communicated that the government’s incompetence was to blame for the poorly conceived and executed hurricane Katrina disaster response.Footnote 4 , Footnote 5

When the video presentation was complete, participants watched a 3-min computerized slide show containing 96 photographs of the physical damage in the Gulf Coast and the human suffering following hurricane Katrina. These photos were obtained from an Associated Press archive. This slide show provided participants with a brief distraction from the blame manipulation. Research on belief threat suggests that defensive processing is enhanced when participants receive a small delay after exposure to threatening information (Greenberg, Solomon, & Pyszczynski, 1997).

After the slide show, participants completed measures assessing ingroup attachment, attitudes toward Black and Whites, and an open-ended manipulation check asking participants to describe the major reason conveyed in the video for the poor hurricane Katrina disaster response. Finally, participants provided their race, gender, and age and were then carefully debriefed.

Measures

Ingroup Attachment

Ingroup attachment was assessed with 6 items derived from theoretical and empirical research on this construct (e.g., Jackson, 2002). Items were: “I feel strong ties to my racial group,” “I feel a common bond or connection with other members of my racial group,” “If I could drop out of my racial group, I would” (reverse-scored), “I feel held back by my racial group” (reverse-scored), “I feel separate or independent from other members of my racial group” (reverse-scored), and “I don’t care what happens to my racial group as a whole (reverse-scored).” Scale endpoints were 0 (strongly disagree) and 6 (strongly agree) (α = .74).

Attitudes Toward Blacks and Whites

Participants completed feeling thermometers assessing their attitudes toward Blacks and Whites (these were embedded among filler items, assessing attitudes toward other groups). The feeling thermometers ranged from 0 (extremely cool) to 100 (extremely warm), and participants could respond in 10° increments (i.e., there were 11 response options per thermometer). Higher ratings represented more favorable attitudes. By supplying ratings about the ingroup and outgroup, participants provided information about both ingroup love and outgroup derogation.

Manipulation Check

The effectiveness of the attribution manipulation was assessed by asking participants to provide an open-ended response describing what explanation for the ineffective hurricane Katrina disaster response they remembered being made most often in the video.

Results

Manipulation Check

A majority of participants (92.2% of Whites, 86.7% of Blacks) in the race blame condition wrote that the videos primarily conveyed that racism was responsible for the ineffective disaster response. A majority of participants in the government incompetence condition (81.0% of Whites, 80.0% of Blacks) wrote that the videos primarily conveyed that some type of government ineffectiveness was responsible for the disaster response. Thus, participants correctly recalled the blame attributions conveyed in the videos.Footnote 6

Ingroup Attachment

A 2 (Participant Race: Black or White) × 2 (Video: Race Blame or Government Incompetence) ANOVA on ingroup attachment revealed a main effect of participant race, F (1, 149) = 93.83, p < .01, which was qualified by a Race × Video interaction, F (1, 149) = 4.20, p < .05. Consistent with hypotheses, Whites in the race blame condition (M = 3.98, SD = .96) reported greater ingroup attachment than Whites in the government incompetence condition (M = 3.56, SD = 1.01), F (1, 149) = 5.29, p < .05. Blacks reported strong levels of ingroup attachment regardless of whether they watched the race blame video (M = 5.11, SD = .72) or government incompetence video (M = 5.26, SD = .68), F (1, 149) = .61, p = .44.

Attitudes Toward Whites

The 2 × 2 ANOVA on attitudes toward Whites revealed a main effect of participant race, F (1, 149) = 25.73, p < .01, which was qualified by a Race × Video interaction, F (1, 149) = 5.18, p < .05. Consistent with hypotheses, Whites in the race blame condition reported more favorable attitudes toward Whites than did Whites in the government incompetence condition, F (1, 149) = 3.99, p < .05 (see Table 1 for descriptive statistics). Blacks tended to endorse less favorable attitudes toward Whites after watching the race blame video than the government incompetence video, but this difference was not significant, F (1, 149) = 1.73, p = .19.

Table 1 Means and standard deviations on the racial attitudes measures
Full size table

Attitudes Toward Blacks

The 2 × 2 ANOVA on attitudes toward Blacks revealed a main effect of participant race, F (1, 149) = 44.64, p < .01. As can been seen in Table 1, Blacks felt warmer toward Blacks than Whites did. This main effect was not qualified by a Race × Video interaction, F (1, 149) = 1.10, p = .30. It is worth noting that Blacks’ explicit ingroup attitudes were quite favorable; they were nearly at the top of the scale.

Discussion

The findings in this study are broadly consistent with recent research suggesting that racialized explanations pose a threat to Whites’ beliefs about the legitimacy of the status hierarchy (Blasi & Jost, 2006; Kaiser, 2006; Kaiser et al., 2006; Napier et al., 2006). Whereas past research on racialized explanations has focused on documenting negative interpersonal reactions toward individual discrimination claimants, the present research expands on this literature by showing that exposure to discrimination claimants also affects the intergroup attitudes of groups at the top of the status hierarchy. Specifically, Whites reported greater attachment to their racial ingroup and greater ingroup love when exposed to videos suggesting that racism, rather than government incompetence, was responsible for the poorly executed hurricane Katrina disaster response. Thus, this research is consistent with system justification theory and advances empirical research on claiming discrimination.

The findings differ in one important respect from prior research on claiming discrimination. Whereas past research demonstrates that Whites express outright negativity toward Black discrimination claimants (e.g., Kaiser & Miller, 2001; 2003), the present research reveals that Whites exposed to racism claims express especially positive attitudes toward Whites, rather than negative attitudes toward Blacks. This difference may have emerged because in the present research, Whites did not evaluate the individuals who aired the discrimination claims; instead they evaluated Blacks and Whites at the group level. This methodological difference, in combination with social pressure to avoid expressing prejudice toward Blacks as a group (Dovidio & Gaertner, 2004), likely contributed to this discrepancy with past research.

In contrast to the findings observed among Whites, Blacks exhibited high levels of ingroup attachment and ingroup positivity, regardless of experimental condition. Because Blacks endorse most system-justifying beliefs to a lesser degree than Whites (Jost & Thompson, 2000; O’Brien & Major, 2005; Schmader et al., 2001), the racialized claims presented in this study may not have posed a strong challenge to their beliefs about the status hierarchy; in fact, simply thinking about hurricane Katrina more generally may have caused Blacks to see the system as illegitimate (Levy et al., 2006). Additionally, because Blacks in both experimental conditions watched videos depicting ingroup members suffering, exposure to these images may have led to enhanced feelings of solidarity with the group, which could result in the especially positive (in fact, near ceiling level) ingroup attitudes observed in this study (Branscombe et al., 1999; Napier et al., 2006). Alternatively, because this research used explicit attitude measures, it is possible that Blacks simply were unwilling to profess anything but extremely positive ingroup attitudes—especially in the face of such blatant images of ingroup suffering. Because members of low-status groups are more likely to engage in system-justifying responses when these responses are measured implicitly (Jost et al., 2002; 2004), different effects may have been observed if ingroup attachment and attitudes had been measured implicitly.

The goal in this paper was to provide a test of the hypotheses in the context of an externally valid, naturally occurring racialized explanation. Thus, relying on the gripping images that actually aired in the days following the disaster was important in achieving this goal. Indeed participants reported that these images were especially engaging; several participants reported that the images of human suffering brought them to tears. Thus, the experimental context closely mirrored the experiences one might have while watching the media coverage from the comfort of one’s own living room. This focus on external validity was critical for addressing these questions. One could argue, however, that this decision detracted from traditional, tight experimental control. This research attempted to circumvent this problem by ensuring that both the race blame and government incompetence conditions contained clips showing a variety of speakers. By doing so, this research highlighted general messages rather than the idiosyncrasies of any one speaker.

One could also question whether the effects observed among Whites occurred because of racialized explanations (as we suggest) or because of increased salience of race in the race blame condition. The latter explanation seems unlikely given that participants in both conditions watched 11 min of videos and 3 min of photographs featuring hurricane Katrina survivors, most of whom were Black. Race was very salient in both experimental conditions.

Additionally, although the video manipulation caused differences in Whites’ reports of racial attachment and ingroup love, the lack of a control condition in which participants were not exposed to the videotapes or photographic slide show makes it impossible to conclude whether racialized explanations caused increases on these measures. In future studies, such a control condition would help to clarify the direction of these differences.

Although these data were explored from a system justification perspective, there may be alternative theoretical lenses for understanding these data. For example, at first glance the findings seem consistent with those offered by theories of group justification, such as Social Identity Theory (e.g., Tajfel & Turner, 1979). That is, Whites’ expression of ingroup regard in the race blame condition could have been a defensive effort to justify the integrity of their racial group. However, it is difficult to reconcile how expressing extra attachment toward a group that was just portrayed in an extremely negative light (i.e., as racists) would be group justifying for Whites. Indeed, several recent investigations demonstrate that high-status group members who are most invested in their group experience shame and guilt when they are made to consider their group’s racist past, and this shame and guilt leads to efforts to distance themselves from the group. For example, the more Americans identified with their national identity, the more shame they experienced when they recalled instances when their group engaged in blatant discrimination, and the more they desired to distance themselves from Americans as a group (Johns, Schmader, & Lickel, 2005; see also Doosje, Branscombe, Spears, & Manstead, 1998; Iyer, Leach, & Crosby, 2003; Swim & Miller, 1999). For this reason, a system threat analysis provides a more plausible explanation for the effects observed in this study. Nonetheless, this assumption should be examined in future studies. For example, one could assess whether the effects reported here are intensified or attenuated for individuals who vary in system justification beliefs and White identity.

One could also argue that White participants resisted the argument that racism was responsible for the botched Katrina disaster response. If this were true, then expressing positive attitudes toward Whites could still serve to justify the integrity of Whites as a racial group as well as justify the system. However, as discrimination claims were so prevalent in the media, social discourse, and most importantly, in the images and information conveyed in the videos in these studies, it is difficult to argue that a majority of White college students completely disregarded the possibility that race and social class played some role in the Katrina disaster response. Nonetheless, this alternative explanation is still plausible. It would be useful to assess participants’ pre-existing beliefs about the role of racism in the Katrina response prior to exposing them to the video conditions. If attachment to Whites was moderated by participants’ beliefs about whether race played a role in the disaster response, then this would provide important insight into how system- and group-justifying motivations contributed to the findings reported here.

Finally, this research focused on racial attitudes as a way of defending the system from threats posed by post-hurricane Katrina discrimination claims, but this is not the only way individuals could have engaged in system justification. Black and White participants could have justified the system by expressing increased admiration for the United States, for example. This type of justification might occur in both the race blame and government incompetence conditions (as both present different types of system threat), but it might be stronger in the government incompetence condition because expressing allegiance to the attacked system could be an effort to reaffirm its legitimacy. Alternatively, if given the opportunity, Whites and perhaps even Blacks in the race blame condition might have preferred this latter type of system justification over racial ingroup pride, as it would serve to assuage system threat without raising concerns about possessing racially divisive attitudes.

In sum, this research suggests that the system threat resulting from the post-hurricane Katrina media coverage focusing on racialized explanations likely affected White Americans’ racial attitudes. Furthermore, these shifts in racial attitudes could have potentially important consequences for intergroup relations in the United States. For example, insofar as White identification is positively associated with self-reports of racism (Jost & Thompson, 2000; Knowles & Peng, 2005; Major et al., 2002) and opposition to policies that harm Whites (Lowery, Unzueta, Knowles, & Goff, 2006), racialized explanations may lead to more intergroup tension. It will be important for scholars who study system threat to begin identifying strategies that make individuals less vulnerable to behaving defensively after exposure to system threat. Finally, this research highlights how theoretical perspectives from social psychology can be applied to an understanding of important societal issues and social problems.