The Space-Time Dynamics of Criminal Violence and Population Change

Steven F. Messner, State University of New York at Albany
Glenn D. Deane, State University of New York at Albany
Luc Anselin, University of Illinois at Urbana-Champaign
Jon G. Bernburg, State University of New York at Albany

Despite the longstanding interest in the demographic context of criminal violence, our understanding of the relationship between violent crime and population dynamics is limited in two important respects: most research is based on cross-sectional designs, and models are commonly based on the assumption of unidirectional causality. To address these limitations, we analyze annual data on homicide, robbery, and population for U.S. counties, 1977-1998. This design is particularly well suited for tracing the spatial imprint of criminal violence and disentangling the causal processes linking violence with population change. The geographic and temporal detail provides statistical leverage for identifying reciprocal causal processes using newly developed econometric techniques for Granger causality modeling in pooled cross-sectional time series. By applying this innovative method, our analyses shed new light on one of the core substantive issues in sociological criminology - the space-time dynamics of population change and violent crime.