Crime fluctuations in response to hurricane evacuations

Abstract

Research regarding how natural disasters impact crime is largely mixed. Most studies focus on whether aggregate postdisaster crime levels differ from predisaster ones and pay less attention to how emergency procedures impact the timing of crime fluctuations. A recent study of Hurricane Rita in Houston, Texas, uncovered a surge in burglary prior to the storm, suggesting that the prestorm evacuation increased the opportunities for burglary by reducing guardianship. This suggests that researchers should examine crime fluctuations that may occur before, during, and after natural disasters. Using nonparametric kernel regression models, we examined crime trends surrounding Hurricane Harvey that occurred in Houston 12 years later where no prestorm evacuation was ordered. We observed no crime surge prior to the storm. Instead, we observed substantial increases for some crime types after the hurricane made landfall that coincided with poststorm evacuations. This supports previous findings that evacuations may create certain crime opportunities.