Natural Disasters in Black and White: How Racial Cues Influenced Public Response to Hurricane Katrina

Shanto Iyengar and Richard Morin
Date Published: 
June 8, 2006

Attached files

table1.pdf150.38 KB
table2.pdf281.67 KB
fig1.pdf38.23 KB
fig2.pdf27.64 KB
fig3.pdf37.66 KB
Natural Disasters in Black and White
How Racial Cues Influenced Public Response to Hurricane Katrina

By Shanto Iyengar and Richard Morin

The Washington Post
Thursday, June 8, 2006; 6:05 PM

(Note: All tables and figures are attached for download)

Natural disasters are typically occasions for political unity rather than controversy. In the aftermath of large-scale death and destruction, Americans reach for their wallets rather than engage in rancorous debates over fixing responsibility and blame.

Hurricane Katrina proved an exception. In the first place, it was quickly apparent that government officials at all levels were utterly unprepared for the scope and severity of the disaster. Thousands of people were left cooped up in the Superdome for days in the most primitive of conditions. The feeble relief efforts provoked a firestorm of criticism leading eventually to the resignation of FEMA Director Michael Brown.

Not only did Katrina raise questions about the government's ability to deal with large-scale flooding, it also rekindled longstanding issues concerning the standing of African-Americans. The people who remained left behind in New Orleans to suffer the brunt of the hurricane's consequences were disproportionately black. Post-hurricane publicity, although sympathetic to victims, was criticized as seeming to be racially biased at times. The media publicized instances of looting by blacks while characterizing similar activity on the part of whites as "looking for food." Other reports alleged that gangs of armed blacks had attempted to shoot down rescue helicopters. Quite unexpectedly, Katrina became a metaphor for the state of race relations in America.

We designed this experiment to investigate how racial cues conveyed in news coverage conditioned Americans' response to Katrina. We wanted to explore whether public outrage over the governmental response was mitigated by frank coverage of the demographics of the victims, their perceived inability to help themselves and in some cases their lack of compliance with rescue efforts.

Racial cues can be conveyed within two distinct genres of news coverage. "Thematic" news reports cover events in general terms providing information about background and context. In contrast, "episodic" news personalizes events by focusing on the experiences of specific individuals. We presented participants with two different thematic frames for Katrina; one, completely lacking in race-related references, focused on the scope of the flooding and destruction in different areas of New Orleans. The second added implicit racial cues to the coverage by focusing on the breakdown of law and order in the immediate aftermath of Katrina. We anticipated that people who read the former news report would favor more energetic government relief efforts than those exposed to the report on lawlessness. In fact, this is exactly what we found.

Our manipulation of race was more explicit in the case of episodic news coverage. We presented study participants (those who were not assigned to either of the two thematic reports) with a typical story about a displaced Katrina victim. By varying the victim's ethnicity we could observe whether the audience responded differently to efforts to help the entire class of Katrina victims when they were presented with a specific case of an African-American, Hispanic, Asian, or white victim. Here our results suggest that the race-ethnicity of the person showcased by the news report was relevant: participants responded more generously (in the sense of recommending higher levels of government assistance to hurricane victims) when they encountered a victim who was white.

The study design was as follows. All participants first read one of the three news reports. Some participants were assigned to the two thematic conditions, but the majority encountered an episodic report featuring a particular individual left homeless by the hurricane. We embedded several manipulations of the victim's personal attributes into this episodic report. The victim's name was either Terry Miller or Terry Medina. Terry was either a mother or father of two children, married or single, and said to be either a school custodian, factory worker, or real estate agent. We also inserted a small headshot photo of Terry into the report; depending on the condition, the photograph showed a white, African-American, Hispanic, or Asian person (see Table 1 for an example of the episodic report). We selected a total of 18 different photographs (9 men and 9 women) from a national database all showing people from the shoulders up with a neutral (non-smiling) expression. We then edited each photograph so as to alter the subject's skin complexion. In effect, for each of the 18 selected faces, we created dark and light-skinned versions of our fictional Terry. (Examples of the skin color manipulation are provided in Table 2.) We then had Stanford undergraduates view all 36 photographs and identify the ethnicity of the person. (They were asked to classify each face as white, African-American, Hispanic, Asian, or ambiguous.) A majority of the undergraduates were able to identify each face as either white, African-American, Hispanic, or Asian.

Approximately 2,300 people completed the experiment. As in our past studies, the sample was skewed heavily in the direction of Democrats and liberals -- only 12 percent of the participants identified as Republican. Eighty-six percent were critical of President Bush's handling of Katrina. The sample was also highly educated -- 84% had completed at least a bachelor's degree. These features of the sample are especially important in light of the results we describe below.

Our principal interest was to trace the connections between different forms of news coverage and participants' willingness to support government assistance to hurricane victims. We asked participants to indicate how much money hurricane victims should be awarded in the form of assistance for housing and general living expenses. For each type of assistance, they could check a box that ranged from $200 per month to $1200 per month. Participants also indicated for how long (from a minimum of three to a maximum of eighteen months) victims should receive government assistance. Based on these responses, we created separate measures of the total amount of recommended assistance and the average length of time for which victims could receive this assistance. (The average total amount of assistance was nearly $1,500 and the average length of assistance was twelve months.) These measures reflect some mix of beliefs about the moral obligation of the government to assist victims of natural disasters on the one hand, and beliefs about how deserving were Katrina's victims on the other.

We began by examining the effects of the different genres of Katrina news on the amount and duration of recommended financial assistance for hurricane victims. Our analysis includes participants of all ethnicities although the vast majority (86 percent) were white. We expected that beliefs about the appropriate level of assistance would vary with the presence or absence of racial cues in the news. As shown in Figure 1, the looting news frame had significant effects. Participants were least generous in their recommendations after reading the report on looting. Episodic framing of the disaster -- presenting readers with an actual flesh and blood victim attempting to restart his or her life -- and impersonal descriptions of the scope of destruction both elicited higher levels of recommended assistance. The data does not permit us to assess whether the significant reduction in the amount and length of financial assistance in the looting condition is attributable to racial cues per se, but many previous studies have documented the existence of a close connection between references to violent crime and implicit racial stereotypes. We suspect that exposure to the news story on looting "primed" people to associate hurricane victims with crime, thus making them scale back on what they considered the appropriate level of assistance.

We can test for the effects of racial cues more directly within the various episodic coverage conditions where participants either encountered a white, African-American, Hispanic, or Asian family uprooted by Katrina. The appropriate comparisons (see Figure 2) demonstrate that beliefs about the appropriate amount of assistance did not vary substantially by the race of the person depicted in the news. However, participants recommended different periods of assistance depending on the ethnicity of the victim they encountered. Those who saw the African-American version of Terry Miller (Medina) awarded a significantly reduced period of assistance. (On average, the difference between the African-American condition and the remaining episodic conditions was nearly one month.) Conversely, participants awarded a significantly longer period of assistance after reading about the same Terry Miller, but who now appeared to be white.

We do not mean to suggest that participants were sensitive only to the race of the person featured in the new story. In fact, they were also affected -- and significantly so -- by gender and occupation. Participants recommended considerably higher levels of assistance after reading about Terry Miller the mother and Terry Miller the real estate agent. Occupation is clearly a proxy for earnings potential, and we suspect that people saw fit to award more generous levels of assistance when they encountered a case of a victim with significant lost earnings. Interestingly, neither marital status nor surname made any difference at all to the level of recommended support.

Finally, we turn to the question of skin color. For each of the episodic news conditions we created a lighter and darker complexion image of the person in question. We anticipated that the impact of skin color would be especially influential when the person in question was non-white. That is, we expected that darker skin color would prompt people to consider race only when they believed the person in question to be non-white. In fact, the impact of the skin color manipulation on the level of recommended financial assistance was striking. (We have plotted the difference in the level and length of disaster relief between the dark and light conditions in Figure 3.) When the hurricane victim in the news was a dark-complexion white, the amount of assistance for hurricane victims actually increased. Perhaps well tanned whites are perceived as vigorous, fit and attractive, thus putting our respondents in a more favorable state of mind concerning hurricane victims in general. But for every other ethnic group -- blacks, Hispanics and Asians -- the effect of skin color ran in the opposite direction. When people saw a dark-skinned black, Hispanic, or Asian, they recommended lower levels of financial assistance. This divergence in the effects of skin color for whites and non-whites was statistically significant. A similar, but weaker pattern emerged for duration of assistance. Here the effects of darkened skin color were to increase the duration of assistance in the white and Asian conditions, but to decrease it in the case of the African-American and Hispanic conditions.

These results suggest that news media coverage of natural disasters can shape the audience's response. Framing the disaster in ways that evoke racial stereotypes can make people less supportive of large-scale relief efforts. News reports about flooding evoke one set of apparently positive images in the reader's mind; reports about lawlessness evoke quite another.

The effects of the racial identity of individual hurricane victims on the prescribed level of government assistance for all victims are suggestive of what psychologists call the "automaticity" of stereotyping. People cannot help stereotyping on the basis of ethnicity despite their best efforts to act unbiased and egalitarian. As we noted at the outset, this particular sample of participants consisted of highly educated individuals who located themselves toward the liberal end of the political spectrum. Many of them live in and around the nation's capital, one of the more racially diverse and cosmopolitan areas of America. We suspect that this group would score at or very near the top of most measures of support for civil rights and racial equality. Yet their responses to Katrina were influenced by the mere inclusion of racial cues in news media coverage. The fact that this group awarded lower levels of hurricane assistance after reading about looting or after encountering an African-American family displaced by the hurricane is testimony to the persistent and primordial power of racial imagery in American life.

Shanto Iyengar is Professor of Communication and director of the Political Communication Lab at Stanford University. Richard Morin is director of Washington Post polling and a staff writer.