Neighborhood Disorder and Crime: An Analysis of "Broken Windows" in New York City.

For more than 20 years, the relationship between disorder and crime has been the focus of a contentious debate in criminology and policing. Wilson and Kelling's "Broken Windows" essay emphasized the importance of policing "the little things" in the prevention of serious crime, motivating many cities to integrate an order-maintenance component into their community policing startegies. However, critics question the suggested causal links between disorder and crime, citing difficulties in the measurement of disorder, and in modeling the suggested "developmental sequence" that the theory asserts.In this paper I study the "Broken Windows" hypothesis, focusing on conditions in New York City throughout the 1990s. Combining data on neighborhood disorder from the New York City Housing and Vacancy Survey (HVS) with police measures of violence, I use a series of ordinary least squares, fixed effects, and change models to assess the link between physical neighborhood disorder and crime in subsequent years. I also perform an Instrumental Variables analysis leveraging a citywide renovation initiative of the late 1980s and 1990s, to determine whether the randomly scheduled renovations lead to a reduction in disorder, and whether this reduction can causally be linked to violent crime. I find that while disorderly neighborhoods tend to also be high in crime (evidenced by strong cross-sectional associations) this relationship can be explained by citywide trends in crime and disorder, and unobserved, time-invariant properties of neighborhoods. This suggests that addressing conditions of disorder, without other efforts to improve neighborhoods, are unlikely to significantly reduce crime. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]