Measuring the Problem: Foreclosures

On Credit Slips we have posted about the difficulties in obtaining timely and reliable bankruptcy data. (See here and here for some examples). Each of us is involved in the 2007 Consumer Bankruptcy Project, a new iteration of a decades-long effort to gather detailed data about a sample of bankruptcy cases to enrich (and sometimes critique) the very limited publicly available data. Apparently, similiar attention to the quality of released data and scholarly efforts to improve data are needed for foreclosures. Tara Twomey pointed me to this article in today's Los Angeles Times, Getting a Fix on Foreclosure Data, that explains growing concern with the divergent foreclosure numbers being hawked by different companies. The article itself is worth a read, offering some juicy quotations from testy company analysts asserting that their numbers are accurate.

Foreclosure data typically come from public real estate records. Since all companies look same place, how do they get different numbers? The answer may partly rest with quality issues with public data, but mostly comes from how one defines a "foreclosure." The LA Times writer, David Streitfeld, nicely explains: "Foreclosure is popularly understood as an event. . . Yet, foreclosure is really a process, one that can stretch over a year and vary from state to state." Does putting a "1" in the foreclosure tally require that a family must actually lose their house? Or merely that a foreclosure action has been filed? How do we count judicial and non-judicial foreclosures, which proceed in very different ways? Does a foreclosure occur if a lender files a notice of default but the homeowner then sells the house?

Neither federal, nor state, nor local governments keep accurate counts of foreclosures, or even apparently agree on what they should be counting. This creates serious problems for policymaking designed to aid distressed homeowners. Empirical researchers could help by conducting even micro-research on particular jurisdictions, bringing their knowledge of the legal process to bear on the development of a shared and consistent definition of foreclosure. Scholars could devise a different term for filed but abandoned foreclosures and consider the usefulness of alternate ways to measure the variety of outcomes from default. Housing economists are doing work in this field, but they typically use private lender-based data sets and adopt whatever definitions the lender's database incorporates. A shared public norm about what constitutes a "failed" homeownership experience and hard numbers to document these outcomes are crucial to thoughtful and useful policy-making to advance homeownership success.