Misdemeanor Policing, Physical Disorder, and Gun-related Homicide A Spatial Analytic Test of "Broken-Windows" Theory

Background: Homicide contributes substantially to the burden of death in the US and remains a key contributor to the gap in white-black life expectancy. It has been hypothesized that "broken-windows" policing is associated with lower homicide rates and that physical disorder may mediate this association. However, the empiric evidence is limited and conflicting. Methods: We used pooled, cross-sectional time-series data for 74 New York City (NYC) Police Precincts between 1990 and 1999 to test the relation between neighborhood misdemeanor policing (an indicator of physical order) and homicide in NYC in the 1990s. We applied Bayesian hierarchical models, including a random effect of place, to account for serial correlations in homicide across adjacent neighborhoods. Results: An increase of 5000 misdemeanor arrests in a precinct with 100,000 people was associated with a reduction of 3.5 homicides (95% credible interval = -5.00 to -1.00). However, increased misdemeanor arrests were associated with lower physical order (posterior median = -0.015 [-0.025 to -0.01]), and physical order was unrelated to homicide. Conclusions: Our study replicated prior findings suggesting that misdemeanor policing reduces homicide rates, but offered no support for the hypothesis that physical disorder is a mediator of the impact of such policing. Factors responsible for the dramatic decline in US homicides in the last decade remain unclear. (Epidemiology 2009;20: 533-541)