Space matters: An analysis of poverty, poverty clustering, and violent crime

While numerous criminological theories emphasize the theoretical importance of the spatial distribution of poverty, few studies specifically examine the empirical relationship between the spatial clustering of high poverty areas and violent crime rates. In this analysis we examine the association between poverty clustering and violent crime rates across 236 cities. For each city we compute a poverty cluster score that measures the proportion of contiguous high poverty census tracts. We find little support for a direct relationship between the spatial clustering of high poverty tracts and murder, rape, robbery, and assault. However, variables that measure city disadvantage (e.g., poverty) interact with poverty clustering scores in the case of homicide rates. Specifically, disadvantage has a much stronger relationship to homicide in cities with high levels of poverty clustering. Such an interaction effect is strongly supported by the literature.