Researching Crime After a Disaster: What We Can Learn From a Large Survey in New Orleans After Hurricane Katrina

Abstract

In this article, I describe a large (N = 7,000) survey we conducted in greater New Orleans after Hurricane Katrina. We encountered many challenges in conducting the survey and used a number of creative means of collecting the data. Our survey, which included detailed questions about collective resources and can be aggregated to the census tract level, has great utility in investigating not only crime but also such questions as repopulation, blight reduction, resident stress, heart health, Airbnbs or short-term rentals, and foreclosures. The main difficulty was the time it took to conduct the survey because many interviews had to be done door-to-door and face-to-face to produce a representative sample. While we clearly outlined these limitations in papers we wrote, the survey duration raised questions of causal direction, and we had to conduct detailed tests of endogeneity to provide convincing evidence that our analyses were sound. I also briefly describe some of the other data we utilized, other surveys we conducted, and ethnographic and organizational work we did that not only assisted disaster recovery but also gave us insights into the social processes we investigated with our quantitative data.