We construct a database of non-prime hybrid adjustable and fixed rate mortgages from New York City that augments a rich set of loan and borrower risk characteristics with a variety of census tract level neighborhood characteristics. We find that these neighborhood characteristics are important for default behavior, even after an extensive set of controls. First, default rates increase with the rate of foreclosure notices and the number of lender-owned properties (REOs) in the tract. Second, default rates for home purchase mortgages are higher in predominantly black tracts, regardless of the borrower’s own race. We explore possible explanations for our findings.
Last May, the Furman Center, with support from the Ford Foundation, convened leading housing researchers, policymakers, lenders, and nonprofit housing organizations to discuss how best to leverage public and private resources to reuse foreclosed properties in a manner that helps stabilize neighborhoods. The Furman Center has produced a White Paper, Transforming Foreclosed Properties into Community Assets, that documents that roundtable conversation, summarizes much of the discussion’s substance, and includes links to resources—ranging from existing research papers on related topics to listings of REO properties—that we hope will be useful to practitioners, researchers and policymakers involved in neighborhood stabilization projects.