Now comes a study that rejects or qualifies much of this received wisdom [about the housing bubble and collapse]. Conducted by economists Manuel Adelino of Duke University, Antoinette Schoar of the Massachusetts Institute of Technology and Felipe Severino of Dartmouth College, the study — recently published by the National Bureau of Economic Research — reached three central conclusions.Link to free abstract and gated NBER paper, "Changes in Buyer Composition and the Expansion of Credit During the Boom" by Manuel Adelino, Antoinette Schoar and Felipe Severino, NBER Working Paper No. 20848, Issued in January 2015.
First, mortgage lending wasn’t aimed mainly at the poor. Earlier research studied lending by Zip codes and found sharp growth in poorer neighborhoods. Borrowers were assumed to reflect the average characteristics of residents in these neighborhoods. But the new study examined the actual borrowers and found this wasn’t true. They were much richer than average residents. In 2002, home buyers in these poor neighborhoods had average incomes of $63,000, double the neighborhoods’ average of $31,000.
Second, borrowers were not saddled with progressively larger mortgage debt burdens. One way of measuring this is the debt-to-income ratio: Someone with a $100,000 mortgage and $50,000 of income has a debt-to-income ratio of 2. In 2002, the mortgage-debt-to-income ratio of the poorest borrowers was 2; in 2006, it was still 2. Ratios for wealthier borrowers also remained stable during the housing boom. The essence of the boom was not that typical debt burdens shot through the roof; it was that more and more people were borrowing.
Third, the bulk of mortgage lending and losses — measured by dollar volume — occurred among middle-class and high-income borrowers. In 2006, the wealthiest 40 percent of borrowers represented 55 percent of new loans and nearly 60 percent of delinquencies (defined as payments at least 90 days overdue) in the next three years.
Link to free download of pre-published version of the research paper.