Two Theories for Two Units of Analysis in Place-based Crime Research

• In Event: Communities and Place: Crime Concentrations

Abstract

In their thorough and path-breaking book, Weisburd, Groff, and Yang (2012) and in earlier studies (Weisburd et al., 2004), the authors make a strong case for the street segment being an important unit of analysis. They suggest that it is superior to the neighborhood because crime concentration varies too much within neighborhoods: there are a few segments within neighborhoods with much crime and many segments with no or very little crime. This raises an important question about street segments: How is crime distributed within them? If segments, particularly those with a great deal of crime, have crime scattered their lengths, then the segment is a useful unit, and segment level explanations for crime are necessary. However, if crime is highly concentrated at a few numbers of addresses within many segments, one could make the same argument against segments as a unit of analysis as Weisburd and colleagues make against neighborhoods. Shannon index allows us to examine these internal distributions efficiently, and thereby test the assumption that crime within segments is reasonably uniformly spread.  This test shows how many street segments follow either the social disorganization theory or place management of the routine activities theory.  We discuss suggestions for future research.

Authors

• YongJei Lee, University of Colorado Colorado Springs

• John E. Eck, University of Cincinnati

• SooHyun O, University of Cincinnati