UC Santa Barbara

Crime is an important issue that any society has to deal with. Spatial analysis of crime is a valuable tool for tackling this problem, allowing patterns to be revealed, deeper understanding on the crime phenomenon to be attained and a more efficient allocation of policing and other public policies to be possible. Mapping and explaining spatial patterns of crime, however, involves many challenges. On this premise, this PhD dissertation has two main objectives: (1) presenting new methodologies to improve the spatial analysis of crime (2) applying these new methodologies to a case study — residential burglaries in Belo Horizonte — to evaluate what best explains the observed geography of crime. The new methodologies presented in this dissertation include a method for deciding on an adequate areal unit to count and map crime; a new approach for standardizing crime rates that provides more robust results than existing methods, and a model for estimating the effects of income and income inequality on the spatial distribution of crime within cities. Finally, some policy suggestions are discussed based on the results obtained by applying the methodology to the case of residential burglaries in Belo Horizonte.