Exploring the conditional effects of bus stops on crime

Public transportation is a major element of social life in most cities, and the most common mode of public transportation is busing. This study examines whether bus stops are a robust predictor of crime, net of other factors, and whether the effect of bus stops on crime is conditioned by socioeconomic and land use factors. We use geocoded Indianapolis crime and bus stop data for 2010 to predict crime counts in 500-feet × 500-feet square grid cells, using negative binomial models. Net of other factors, bus stops are associated with variation in counts of Uniform Crime Reports reported rape, robbery, aggravated assault, burglary and larceny in a cell. In addition, the effect of bus stops on crime was conditioned by land use characteristics. In particular, the effect of bus stops on crime was greater in commercial and industrial areas, but dampened in areas with high-density residential housing.