The impact of Hurricane Harvey on crime in Houston, Texas: A partial test of routine activity theory at the neighborhood level

Abstract

To enhance the understanding of potential theoretical applications, this research examines the impact of Hurricane Harvey on crime in Houston, Texas. Utilizing crime data obtained from the Houston Police Department (HPD) and employing Difference-in-Difference (DID) for panel-data analysis, the study explores the influence of varying levels of housing damage caused by Hurricane Harvey on four categories of crime: theft, residential burglary, robbery, and aggravated assault across Houston's neighborhoods. The findings reveal distinct crime patterns associated with varying levels of housing damage caused by flooding. After Hurricane Harvey, the average rate of residential burglaries remained largely unchanged over two months, except in neighborhoods with major housing damage, where the average count increased by 0.92 incidents (a 49 % increase) per census tract. In contrast, other types of crime generally declined. For example, the average count of robberies per census tract decreased by 0.57 incidents (a 42 % decrease) in areas severely affected by flooding. These results highlight the applicability of Routine Activity Theory in studying the relationship between natural disasters and crime. The study also explores policy implications and acknowledges its limitations.