Methodological considerations on the spatial and temporal analysis of violent crime patterns and the identification of social and environmental correlates of crime

Abstract

Violent crime which is referred to as crime against persons has been recognized as the most serious species of crime. Violent crime imposes monetary costs, psychological costs, and social stability disruption. Incarceration strategies used in several societies have been inefficient in deterring, incapacitating, and rehabilitating offenders of violent crime. Additional approaches to preventing and reducing violent crime are needed. This dissertation aims to address three objectives. First, empirically investigate the temporal stability of socio-economic covariates of violent crime. Second, investigate the persistence of violent crime hot spots over time, and the socio-economic factors that correlate with said persistence. Third, develop a Google Street View-based environmental audit approach to quickly and systematically collect environmental data, and explore the role of physical features of built environment in inducing and deterring violent crime. The results of model parameter analysis indicate that it is unnecessary to average crime over multiple years if model parameters are stable across years. The results of hot spots persistence analysis suggest that a combination of social disorganization theory and routine activity theory provides an applicable framework for understanding the temporal dimension of violent crime hot spots. By identifying the factors that contribute to the persistence of hot spots of crime, insights gained from the results can help to inform focused crime prevention practices. The results obtained from Poisson regression with spatial filtering and Google Street View-based environmental audit provide a series of environmental correlates of violent crime, and provide both theoretical and practical implications for several theories of crime and crime prevention efforts.