SHOULD WE POLICE DISORDER? A LOCAL LEVEL EXAMINATION OF THE SPATIAL AND TEMPORAL ASPECTS OF THE BROKEN WINDOWS THEORY

In 1982, Wilson and Kelling introduced the Broken Windows theory (BWT) arguing that policing neighborhood disorder would reduce serious crime while enhancing the quality of life in neighborhoods. However, the questions and critiques of this theory continue today. On the one hand, empirical research testing on this theory has produced mixed or inconsistent results, while on the other, policing disorder activities created more tasks for the police. In order to place these critiques in their proper context, this study utilizes large-scale operational data—two years (2010-2011) of call for police service data from the Houston Police Department (HPD). A Geographical Weighted Regression (GWR) model is employed to examine the spatial and temporal relationships between itemized disorder issues, and reported violent and property crimes. The results of global models show that the residents’ reported disorder issues in the neighborhood are significantly related to violent and property crimes with limited temporal effects. However, the GWR model indicates significant spatial effects on reported minor offenses and the crime links, and those effects vary for individual minor offenses. Since the relationships between minor offenses and neighborhood crime vary, three models should be applied to categorize social disorder policing: pro-active, supervision, and liaison.