LIES, DAMN LIES, AND RUMORS: AN ANALYSIS OF COLLECTIVE EFFICACY, RUMORS, AND FEAR IN THE WAKE OF KATRINA

Abstract

In the days following Hurricane Katrina, many displaced residents from New Orleans evacuated to the Baton Rouge area. As a result, many Baton Rougians became increasingly concerned about crime in their community. This concern, coupled with a lack of official information, led to the widespread dissemination of rumors of criminal victimization. The purpose of this study is to examine the relations between collective efficacy, rumors, and fear during this trying time for Baton Rouge. The results are based on telephone interviews with Baton Rouge residents conducted two months after Katrina. As predicted, collective efficacy fosters the transmission of rumors. These rumors then lead to increased personal and altruistic fear of victimization; therefore, collective efficacy indirectly increases fear through its effect on rumors. The implications of these findings for public and emergency management policy are discussed as are concrete suggestions for future research.

I would like to thank Matthew R. Lee, Johanna M. Thomas, the anonymous reviewers for Sociological Spectrum, and the reviewers for the Mid-South Sociological Association's 2006 Graduate Student Paper of Distinction Competition (which was awarded to an earlier version of this article) for their invaluable comments and suggestions.

Notes

¹This debate may be of little import as the current survey instrument specifically references fear of criminal victimization. Further, this study examines perceived change in fear rather than levels of fear.

²Although the forced option technique was used to minimize fence-sitting, interviewers were instructed to offer the “stayed the same” option if the respondent seemed hesitant to choose one of the other options.

³Questions referencing intervention are standard measures of informal social control (Sampson et al. 1997; Gibson et al. 2002); however, it is somewhat ambiguous as to how one could intervene if someone were merely “hanging around.” While the exact meaning of “intervene” is left to the respondent, possible interventions would include: calling the police, informing neighbors, keeping an eye on the stranger, or confronting the stranger.

⁴Income and education were combined into an SES index based on the results of a principal components analysis.

⁵Because the outcome measures are ordinal level indicators, all models were replicated using ordinal logistic regression techniques. These models are substantively the same as the OLS results reported here and are, thus, omitted.

∗∗p ≤ .01.

∗p ≤ .05 (Two-tail tests).

∗∗p ≤ .01.

∗p ≤ .05.

Unstandardized coefficients with standard errors in parentheses.

∗∗p ≤ .01.

∗p ≤ .05.

Unstandardized coefficients with standard errors in parentheses.

∗∗p ≤ .01.

∗p ≤ .05.

Unstandardized coefficients with standard errors in parentheses.

⁶I would like to thank an anonymous reviewer for his/her insightful comments and suggestions that aided in the development of the policy recommendations described above.