Crime Places in Context: An Illustration of the Multilevel Nature of Hot Spot Development

Abstract

Objectives: The present study provides an illustration of a statistical test of the Brantinghams’ theory about the formation of hotspots and the effects that nodes, paths, and environmental backcloth have on their development. Methods: We used multilevel Poisson regression analysis to explain variation in the count of incidents at each address. Place-level proximity to nodes and paths was measured by using the Euclidian distance from each location to the closest carry-out liquor store, on-premises drinking establishment, and bus route. The broader environmental backcloth was represented by various census block-group characteristics, including density of commercial land use. A three-way place-level interaction as well as a cross-level interaction involving all four key independent variables were used to estimate the Brantinghams’ concept of the overlay of nodes, paths, and backcloth. Results: The three-way interaction involving the distance to the closest on-premises liquor establishment, the distance to closest carry-out liquor facility, and the distance to the closest bus route was significantly and negatively related to place-level crime incidents. This three-way interaction had effects which varied across neighborhood contexts, with stronger negative effects on crime occurring in neighborhoods characterized by high levels of commercial density. Conclusion: This study supported the notion of a multilevel theory of crime places and has implications for more effectively addressing crime. In particular, those places with multiple nodes and paths in their proximal environments and dense commercial land within their broader environments likely need additional crime prevention measures to get the same benefit relative to places with multiple nodes and paths in the proximal environments yet little commercial density within their broader environment.