The mortgage foreclosure crisis has uprooted millions of households and destabilized communities around the country.  News stories have reported growing concerns about the effects these foreclosed homes are having on their surrounding communities and on crime in particular.  In New York City elevated numbers of mortgage foreclosures in micro-neighborhoods increase crime, and effects are larger for properties that linger in the foreclosure process leading up to auction (Ellen, Lacoe, and Sharygin, 2012).  Still, there is scant work exploring the relationship between foreclosures and crime across cities that are regulated by different foreclosure laws and policies. We investigate variation in the impact foreclosures have on micro-neighborhood crime in Atlanta, Chicago, Miami, Philadelphia, and New York City. Using longitudinal, point-specific foreclosure and crime data in all five of these cities, we employ statistical techniques that allow us to estimate the causal impact of foreclosures on different types of neighborhood crime. Our analysis also sheds light on possible mechanisms and explains the variation in impacts across cities by differences in market conditions, the length of the foreclosure process, and state foreclosure policy.

The experience of these cities during the housing crash varied greatly in nature and intensity, and our preliminary results suggest that impacts are greater in the cities that experienced a greater number of foreclosures. Impacts appear relatively modest in New York City which saw a milder run-up and decline in housing prices than in other cities, and moderate foreclosure rates (7.6% of all mortgages were in the foreclosure process in 2011).  Similarly, the Philadelphia metropolitan area experienced the most modest decline in house prices of the five cities and the lowest rates of mortgage distress (5.6%). The price declines in both Chicago and Atlanta were more dramatic, with housing prices declining by 39% between 2006 and 2011 in both metropolitan areas. By 2011, these metropolitan areas faced foreclosure rates of 8.7% and 5.3%, respectively. Miami experienced the most dramatic boom and bust, and the metropolitan area had the highest foreclosure rate in the country (18.9%) in 2011.   

Our main geographic unit of analysis is the blockface, an individual street segment including properties on both sides of the street.  We compare crime levels on blockfaces before and after homes on the blockface enter foreclosure to changes in crime on other blockfaces in the same neighborhood that did not experience a change in foreclosure activity. Given that crime trends are likely to be the same on other blockfaces in the same neighborhood, such a difference-in-differences model can identify if foreclosures lead to higher crime.

To shed light on mechanisms, we consider the impacts of properties at different stages of the foreclosure process and explore whether particular types of crime are more sensitive to foreclosures.  Analyzing data from five cities subject to different foreclosure laws and policies gives us the opportunity to explore mechanisms that may explain how foreclosure effect crime across these different contexts.