Portal de Periódicos da CAPES

What Really Happened in New Orleans? Estimating the Threat of Violence During the Hurricane Katrina Disaster∗

2008; Taylor & Francis; Volume: 25; Issue: 4 Linguagem: Inglês

10.1080/07418820802290504

1745-9109

Timothy Brezina, Joanne M. Kaufman,

Criminal Justice and Corrections Analysis

Abstract Although research indicates that initial media reports greatly exaggerated the severity of violent crime in post‐Katrina New Orleans, it is not yet clear whether violence in general was a problem for storm victims. In this study, data from a unique survey of storm victims are used to estimate the overall threat of violence faced by those who were hardest hit by the storm and who lived through the disaster in New Orleans. The findings suggest that, while the threat of violent victimization was very real for some respondents, it did not dominate the experience of these storm victims. These findings are discussed in terms of their implications for the social construction of law and order, the persistence of disaster myths, negative stereotyping of the urban poor, and crime control issues in the context of disaster management and planning. Keywords: disasterHurricane KatrinaNew Orleansviolence Notes 1. For instance, the initial description of the New Orleans Convention Center as “a nightly scene of murders, rapes, and regular stampedes” was not supported by a subsequent investigation (Thevenot, Citation2006, p. 36). Despite a rumor that “30 to 40 bodies” were stuffed in a freezer at the Convention Center, only four bodies were recovered at the site and only one of these deaths “could be called a suspected homicide” (Thevenot, Citation2006, p. 30, 36). By December of 2005, four Katrina‐related homicides were being investigated by the Orleans Parish District Attorney, including one occurring at the Convention Center, one at the Superdome, and two on the street (Mabin, Citation2005). As observed by the state medical examiner, four homicides would be typical for a week in New Orleans (Cataldie, Citation2006). The reports of rape at the Convention Center could not be confirmed. At the Superdome, members of the police sex crimes unit established a permanent presence and reportedly “ran down every rumor of rape and atrocity,” making two arrests for attempted sexual assault (Dynes & Rodriguez, Citation2007, p. 30). The head of the sex crimes unit concluded that other rumored attacks never happened (Dynes & Rodriguez, Citation2007). 2. In 2004, the year before Hurricane Katrina, New Orleans recorded a homicide rate of 56 per 100,000 residents—a rate nearly 10 times the national average. In 2005, the Katrina disaster disrupted the crime statistics reporting system in New Orleans for the remainder of the year (Mitchell, Citation2007), so it is only possible to estimate what the homicide rate might have been had the disaster not occurred. Just prior to Hurricane Katrina, New Orleans recorded 202 homicides. Researchers estimate that, at this pace, the city would have recorded an additional 107 homicides by the end of 2005, for a rate of 68 per 100,000. This rate “would have made New Orleans the murder capital of the United States” (Mitchell, Citation2007, p. 27). 3. The authors wish to thank the Kaiser Family Foundation for providing access to the survey data. 4. An additional limitation of the survey is that it does not cover those who, tragically, perished during the storm. If violence was responsible for a substantial number of such deaths, then their exclusion from the survey would lead us to underestimate the extent to which violence, or the threat of violence, was a problem in the city. However, available evidence indicates that only a small number of known murders occurred (Mabin, Citation2005) and that virtually all of the deceased in New Orleans were drowning victims (Dewan & Roberts, Citation2005). 5. To help address this issue, we performed a supplemental set of analyses in which we controlled for the perceived quality of the respondent’s evacuation shelter. In particular, we ran the analyses with a select group of respondents; namely, those individuals in the survey who described the conditions at their shelter as being “good” or “excellent” (86 percent of the sample). Presumably, within this group, those who report being “threatened by violence” are unlikely to be referring to experiences occurring at their evacuation shelter, and more likely to be referring to threats experienced during the disaster itself. In these supplemental analyses (not shown), an identical pattern of findings emerged (full set of results available from the first author on request). 6. Our findings are consistent with those obtained in another Houston‐area survey that was based on a small convenience sample of Katrina evacuees. Although violence was not a focus of their study, Coker and colleagues (Citation2006) report that a minority of their survey respondents witnessed violence or experienced a direct threat of violence (p. 91). However, among the participants who responded to the violence items in their survey, it is not clear how many were actually in New Orleans during the immediate aftermath of the storm. 7. It is possible that some prestorm evacuees in survey (those who fled to nearby areas) returned to New Orleans immediately after the storm passed, perhaps in search of friends or family members or to assist other storm victims. Indeed, many individuals (including some celebrities) came from outside to assist in the immediate aftermath of the storm well before the government‐assisted evacuation was underway. Thus, even though they evacuated before the storm, some individuals may have spent time in the flooded city nonetheless. Although it is not possible to determine how many of our respondents fall into this category, it will be possible to control for time spent in certain locations in New Orleans during the disaster. 8. The percentage of respondents who report threats of violence is relatively high among married men (23 percent), but it is even higher among the married women in the sample (34 percent). Based on this observation, we tested for a sex‐by‐marriage interaction effect in multivariate analyses, but this effect was not statistically significant (complete set of results available from the first author on request). 9. In Table 5, the sample size varies slightly across the equations due to the listwise deletion of missing values. We also ran these analyses after excluding the total cases with missing values from both equations (i.e., a sample size of 471 for both equations) and obtained an identical pattern of results. 10. Likewise, Cromwell and colleagues (Citation1995) observed that citizen patrols and other informal social control efforts “formed almost immediately” in the aftermath of Hurricane Andrew (p. 63). Perhaps as a result of these efforts, the hurricane‐affected areas included in their study experienced a poststorm reduction in crime. 11. Some threats of violence reportedly came from local authorities. When a large group of storm victims tried to leave New Orleans by walking across a bridge, police officers from the neighboring city of Gretna blocked their path and pointed their weapons at individuals who did not retreat immediately. The officers explained that there “would be no Superdomes in their city” (Brinkley, Citation2006, p. 469). Later, a police helicopter hovered low over the hungry and dehydrated evacuees, kicking up dust and pelting them with debris, until they all retreated (Brinkley, Citation2006).