Understanding the Multi-level Interactions Between Physical Environment and Neighborhood Characteristics in Assessing Vulnerability to Crime in Micro-places

Abstract

Spatial crime risk assessment is based on the identification of the environmental conditions that could facilitate crime. Previous applications of this approach mainly rely on single-level analyses neglecting that different contextual factors are likely to influence crime at different geographical levels and to interact with one another in defining crime risk. This study proposes to innovate this approach by using a multi-level analysis and estimating the interaction terms among environmental features and neighborhood characteristics. An empirical test is conducted in a large urban area. The results prove that this method significantly increases the predictive capacity and favors more complete interpretations of the underlying criminogenic mechanisms, thus orienting more effective consequential actions.