Built environment and violent crime: An environmental audit approach using Google Street View

Abstract Recent studies empirically support the role of the built environment in inducing or hindering violent crime. Particularly, studies of the broken window theory have provided evidence that physical disorder is an environmental correlate of crime. This includes broken windows, vacant/abandoned housings, abandoned cars on street, graffiti, and decayed street lighting, among other things. Current studies are limited by the difficulty involved in collecting fine-scale quantitative environmental data. The conventional environmental audit approach, which aims to assess environmental features, is costly, time-consuming, and burdensome. In this study, we use Google Street View to study the relationship between violent crime and physical features of urban residential environment. More concretely, a Poisson regression model with spatial filtering is used to identify socio-economic correlates of violent crime. Parting from the hypothesis that omission of built environmental factors results in systematic residual pattern, we proceed to analyze the spatial filter to select sites for virtual environmental audits. A series of physical environmental factors are identified using contingency table analysis. The results provide both theoretical and practical implications for several theories of crime and crime prevention efforts.