How We Analyzed the Foreclosure Data
This story was also published in The Seattle Times on Sunday.
ProPublica and Seattle Times reporters and researchers collected a random sample of some 1,200 foreclosure filings to conduct our analysis. We pulled about 400 filings from each of three central counties, which include the Seattle, Phoenix and Baltimore areas, and collected nearly 60 data points on each foreclosure.
We chose a time period of 2005 to 2008 to straddle the housing crash and provide enough time to see the outcomes of later foreclosure filings.
For each loan, we gathered deeds of trust and promissory notes where available. We also looked for prior bankruptcies for each owner. Some of that information was available only by visiting a courthouse or recorder's office.
Additionally, certain information was not available in every county. In King and Maricopa counties promissory notes are seldom on file, for instance. That document is usually the only source for the interest rate on a fixed-rate loan. Only in Baltimore were we able to track the interest rate on nearly every loan. In the other two counties, we collected interest rates on adjustable-rate loans. Analysis of interest rate information for the three counties in total was based on the adjustable-rate data only.
Reporters also sent a letter to each borrower in the sample. But some had lost or sold their homes and moved, leaving no forwarding address. Reporters then contacted dozens of borrowers by phone. Many of those reached chose not to be interviewed.
We defined predatory loans as those that had balloon payments, prepayment penalties or interest rates that were at least three points above the prime rate at the time the loan was issued. One caveat: In most cases, prepayment penalties are against the law in Maryland.
In a few cases, borrowers in our database who did not lose their house in a trustee sale faced later foreclosure filings.
Request the Data
Sign up, and we'll send you the data from our foreclosure analysis.