Neighborhood Analysis of Theft Using a Geodemographic Classification

Abstract

This study examined the variations in crime rates of three different types of theft among social area types and identifies high-risk neighborhoods. Using a geodemographic classification based on multidimensional socioeconomic attributes of residents and crime open data for 12 cities in the Tokyo metropolitan area, our analysis showed that the geodemographic system effectively distinguishes crime patterns across regions for all types of theft. In addition, the result of the multilevel Poisson regression models revealed that crime rates vary significantly among these social area types, especially for burglaries, even if the neighborhood built environment and city-level differences in crime patterns are controlled for. These findings suggest the usefulness of geodemographics for policy interventions targeting neighborhood crime prevention.