The Effect of Housing Demolitions on Crime in Buffalo, New York

From 2010 through 2015, the city of Buffalo demolished over 2,000 residences. This study examines whether those demolitions resulted in crime reductions. Analysis was conducted at microplaces matching demolished parcels to comparable control parcels with similar levels of crime. In addition, spatial panel regression models were estimated at the census tract and quarterly level, taking into account demographic characteristics of neighborhoods. We find that at the microplace level, demolitions cause a steep drop in reported crime at the exact parcel and result in additional crime decreases at buffers of up to 1,000 feet away. At the census tract level, results indicated that demolitions reduced part 1 crimes, but the effect was not statistically significant across different models. While concerns over crime and disorder are common for vacant houses, the evidence that housing demolitions are an effective crime reduction solution is only partially supported by the analyses here. Future research should compare demolitions in reference to other neighborhood revitalization processes