In the face of disaster, race and class still count.

Race, class, and Hurricane Katrina: Social differences in human responses to disaster

Hurricane Katrina pounded the Gulf South at the end of August 2005, devastating lives and raising questions about how race and class influence human, as well as institutional, responses to disaster. This study uses survey data collected from over 1200 Hurricane Katrina survivors to examine these influences on a wide array of responses, ranging from evacuation timing and emotional support to housing and employment situations and plans to return to pre-storm communities. Results reveal strong racial and class differences, indicating that neither of these dimensions can be reduced to the other when seeking to understand responses by survivors themselves. This intersection renders low-income black home owners from New Orleans those most in need of targeted assistance as residents work to put themselves and the region back together.

Like other disasters before it, Hurricane Katrina offers a unique laboratory in which to study the social infrastructure of its affected region. In this case, it is a region cross-cut by deep and complex divisions of race and class that have hardened over time without direct, excessive interference from outsiders. As media images streamed from the region to the nation and the world following the storm, a public debate emerged over the relative importance of class and race for shaping individual, as well as institutional, responses to the disaster. In the absence of hard data, this debate came to sound much like a skit famously performed by Richard Pryor. In this skit, the comedian tells the story of his wife coming home to find him in bed with another woman. Incredulous, Pryor leaps from the mattress, denies the scene, and blurts, “Come on, baby, who’re ya gonna believe, me or your own lying eyes?” As television screens filled hour after hour, day after day, with images of black Americans desperate for assistance after the storm, many viewers could not help but “see” race and racism at work. Others explained that their eyes were lying to them, and that what looked like race was in fact class in disguise. As commonly occurs in such situations, however, the vigor and volume of this debate soon dwarfed the quantity and quality of information available to assess it.

In the present study, we have sought to fill some of this empirical gap by using data from the largest, most comprehensive survey of Hurricane Katrina survivors currently available. This source cannot determine why the levees failed in New Orleans or why government officials took so long to respond, but it can begin to answer questions about how residents themselves responded to the nation’s costliest natural disaster during and shortly after it occurred. Overall, results indicate that both race and class played important roles in shaping these responses and that neither can be readily reduced to the other. Thus the real issue is not either/or, but where and to what degree.

With respect to race, there are two broad areas where racial differences seem to have mattered. The first involves timing of evacuation and sources of emotional support, that is, behavior more or less under the control of individuals themselves. Our findings indicate that blacks across the region were less inclined than whites to evacuate before the storm, mostly because they did not believe that the hurricane would be as devastating as it eventually was. Previous experience and public assurances suggest that this personal risk assessment may not have been as irrational as it now appears. Reports indicate that had the levees been built and inspected with the integrity typically expected of the Army Corps of Engineers, they would have likely survived the storm, sparing the city from the massive flooding that eventually covered 80% of its area.

As for emotional support, our findings indicate that blacks and whites differed, at least over the short term. Specifically, blacks were more likely to report “leaning on the lord” while whites were more likely to report relying on friends and family. We have suggested that this difference might be more a matter of interpretation and world view than actual differences in network support. Another possibility is that blacks’ friends and family were more likely to be adversely affected by the storm and even more widely dispersed than whites’, making them more a source of concern than support. Both scenarios could easily have worked together to produce the strong racial differences observed throughout the region.

The second and more troubling set of racial difference involves something largely outside survivors’ control, namely job security. Our findings indicate that black workers from New Orleans were four times more likely than white counterparts to lose their jobs after the storm, all else equal. But of course, all else is not equal. When we factor income differences and their effects into the equation, results indicate that the “average” black worker in New Orleans is actually closer to seven times more likely to have lost his or her job than the “average” white worker. This disparity will certainly have a strong effect on who is able to return to the city as it rebuilds and who is not.

This issue of return is also where class standing, specifically home ownership, exhibits its strongest and most consistent effect. We suspect that this effect cuts two ways. On the one hand, home ownership provides survivors power over when and to what extent personal return and rebuilding will occur; on the other hand, it can also create a financial weight in the form of mortgage obligations that limit resettlement options elsewhere. This interpretation is supported by the consistent effects of home ownership across the region and by aggregate analyses which indicate that less affluent home owners are more likely to say they will return than more affluent homeowners. This pattern is also consistent with findings from Hurricane Andrew in the Miami area during 1992. In reflecting on the post-storm plight of many low-income home owners, Morrow explained that, “While they may have acquired some of the trappings associated with economic success, they may lack the ‘defense in depth’—the economic security, political and social influence, and personal power of the professional classes which can be especially crucial in times of crisis.”

Overall, these findings refute the apparent randomness of natural disasters as social events as well as the notion that racial differences are somehow reducible to more “fundamental” class divisions when considering human responses to such disasters. Both “axes of variation”—race and class—appear to have mattered in response to Hurricane Katrina, and while the entire region will continue to require the nation’s ongoing support for years to come, results here indicate that it is low-income homeowners in particular who will need the most assistance in putting their lives and the region back together again. This will be especially true for black residents of New Orleans, who are the most likely to need new jobs as the city recovers, revives, and rebuilds. Failing this targeted assistance, mortgage foreclosures and precarious employment opportunities threaten not only working-class residents from the region but also the futures of children and grandchildren for whom they still care.

In addition to these and related assistance programs, more general efforts to improve policy and planning initiatives for future disasters may benefit from considering the following possibilities. First, with respect to evacuation, our results affirm that poor inner-city residents are often the least likely to heed formal evacuation warnings, some because they lack transportation and others because they fail to take such warnings seriously. Our findings regarding the centrality of religious faith for racial minorities, women, and the elderly coupled with the negative association of this centrality for early evacuation, suggest that emergency planning initiatives can be improved by assisting local civic and faith-based organizations in developing a coordinated, grass-roots system of hazards education and warning dissemination. The basic idea would be to buttress top-down warnings with ongoing planning and preparedness orchestrated through trusted local associations, similar to how school teachers help to educate and evacuate their own groups of students when an ominous but distant fire bell sounds and the entire school must evacuate. At a regional level, such efforts would require a great deal of organizational creativity, money, and time, but if communities are serious about disaster mitigation and saving lives, such investments seem well worth the expense, effectively reinforcing official proclamations in times of emergency with bottom-up planning and social organization.

Second, after the evacuation is over and residents begin returning to the damaged region, housing and jobs are critical to individual and community recovery, almost by definition. While such adjustments produce stress, secondary results from our analyses also indicate that the mere act of evacuation can create high levels of anxiety about the future, regardless of one’s job and housing situation. In the case of New Orleans, this heightened anxiety has taken many anecdotal forms, from (even heavier) drinking and swearing to (even greater) gun purchases and racial paranoia. While it is difficult to pinpoint how or when to intervene to ameliorate such stress and its myriad manifestations, one possibility might be to rethink how military personnel are deployed and organized in post-disaster settings. Borrowing concepts and practices from community policing, one could imagine a proliferation of local substations or mobile command units that integrate themselves into respective neighborhoods over the first six to twelve months following initial search and rescue operations. Through the cultivation of sustained and highly localized relations with residents in the region, these substations could help to serve as well as protect, while simultaneously minimizing the distressful sense of living in an occupied territory, which the passing of anonymous military vehicles and the hovering of distant aircraft helped to produce after Hurricane Katrina.

Surely these will not be the last or only policy lessons proffered in the wake of the Hurricane Katrina, but one thing does appear certain: How the nation responds to this current and ongoing problem will help to define it not only as a society but as a civilization capable of communal expression of awakened conscience.