Broken windows and order-maintenance policing in gentrifying Washington, DC

ABSTRACT The “broken windows” hypothesis has led to millions of citizen/police encounters through aggressive enforcement of minor infractions. Despite extensive research on the relationship between broken windows and crime, there is less research on whether police officers focus on areas with broken windows, and the few studies that exist have mixed findings. Another set of studies finds high levels of policing in gentrifying neighbourhoods. On the one hand, then, studies show there is more policing in neighbourhoods with visible signs of disorder as police departments that adhere to the tenets of broken windows policing would be likely to send officers to patrol areas with broken windows, litter, and unkempt lawns in an effort to prevent crime in these areas. On the other hand, as high-income people move into these neighbourhoods, the signs of disorder should dissipate. This raises the question: Is there more policing in neighborhoods with signs of physical disorder or in neighbourhoods with signs of reinvestment? We measure this using three data sources: 1) an original housing survey using Google Street View that evaluates the level of physical disorder; 2) geocoded data from the Washington, DC Metropolitan Police Department on stop and frisks; and 3) Census and ACS data on population characteristics. These data allow us to answer our research questions: Are residents more likely to be stopped and frisked in neighbourhoods with high levels of physical disorder? Or, conversely, are residents more likely to be stopped and frisked in neighbourhoods experiencing an in-migration of middle-class residents?