Here Today, Gone Tomorrow? Evaluating Random and Nonrandom Processes in Crime Hot Spots

• In Event: Poster Session

Abstract

Prior research has demonstrated the crime-reduction effects of targeting law enforcement activity in crime hot spots. Hot spot research, however, is limited by insufficient attention to the random component in the spatial distribution of crime and by an operational definition of crime hot spots that incorporates crime-free places. We present formal methods for partitioning the random and nonrandom variance in the spatial distribution of crime that includes and excludes crime-free locales. We then illustrate the utility of these methods with violent crime data from a large high-crime city. Results show that the random component in the data varies considerably by crime type and depending on whether crime-free places are included. We recommend that future hot spots studies examine the functional form of the spatial distribution of crime prior to conducting explanatory analysis and consider limiting the scope of analysis to the most crime-prone places within hot spots.

Authors

• Aaron S. Levin, University of Missouri - St. Louis

• Richard Rosenfeld, University of Missouri - St. Louis