Geographic environment and street crime: a spatial stochastic point processes approach

• In Event: Spatial and Temporal Analysis

Abstract

Criminologists have long been interested in studying the correlations between crime and physical environments. Environmental criminologists posit that crime does not distribute randomly in space, but clusters in a few areas. The clustering pattern of crime is affected by the geographic environments and the land use patterns (Brantingham and Brantingham 1995). Because crime is highly concentrated, the locations of crime occurrence are one of the focal interests for researchers.

The dominant methods in analyzing the relationship between physical environment and crime are using spatially aggregated data as unit of analysis, such as census tracts or block groups, and employing some regression approaches to study the correlations between physical environment and crime. These methods have a number of limitations.

This paper presents the first spatially disaggregate analysis of the influences of physical environments on street crime. This article will start by discussing the limitations of using aggregate spatial data. Next will use empirical results to explain the limitations. A new method using disaggregate spatial data will be introduced and discussed. In conclusion I summarize the findings and thoughts on future research directions.

Author

• Jie Xu, Rutgers University