Accounting for Spatial Mobility and Sociospatial Context: Consequences for Ecological Studies of Crime

• In Event: Multidimensional Approaches to Studying Crime and the Intersections of Race, Ethnicity, and Community Characteristics

Abstract

In recent years, crime and place scholars have often justified a focus on ever-smaller geographic units based on the existence of variance across micro units within neighborhood units. The challenge with smaller units is that offenders, targets, and guardians all can (and do) move about the spatial landscape. Further, smaller units often neglect the broader sociospatial context relevant in key theories. As a consequence, we propose that there is no single appropriate (or best) geographic unit in which to aggregate data, but rather spatial behavior and sociospatial context must be considered. Based on existing criminological theory we propose that three broad spatial patterns characterize how we should consider aggregating measures: 1) decay patterns; 2) mixing in an environment; 3) relational patterns. We demonstrate measuring these three different patterns based on data in Southern California from 2009-11. Although we create measures based on theoretical expectations of spatial patterns, the actual spatial extent is empirically determined in this study based on relationship with crime at the street segment level. The results demonstrate robust relationships for theoretically derived measures.

Authors

• John R. Hipp, University of California, Irvine

• Seth A. Williams, Rice University