The Washington PostDemocracy Dies in Darkness

Did underwater mortgages kill the economy?

The housing market, finally, is recovering. Home prices are up 8 percent over the past year. And that is providing relief to a particular set of homeowners – the underwater borrowers who owe more than their properties are worth. Over past year alone, 1.7 million homeowners have climbed out from being underwater.

These positive developments have boosted the overall economy, even as it faces other headwinds. But they also raise a difficult question: How much has the phenomenon of people being underwater on their mortgages – rather than simply the decline in home prices – held back growth?

A recently revised paper by Atif Mian of Princeton, Amir Sufi of the University of Chicago Booth School of Business and Kamalesh Rao of MasterCard Advisers suggests that underwater mortgages have played a significant role in holding back the recovery. The paper, “Household Balance Sheets, Consumption and the Economic Slump,” was first released several years ago, but it has been revised to include an important new calculation.

It is widely recognized that the fall in housing prices had a “wealth effect” that led homeowners across the country to cut back on spending. In the updated paper, Mian, Sufi and Rao measured how much more underwater borrowers probably  cut back on spending compared to borrowers without an overhang of mortgage debt. (More precisely, they measured how much homeowners cut back on auto spending for each dollar loss of housing wealth. But that’s important; the decline in auto sales was a significant part of the economic contraction.)

The authors found that being underwater makes a big difference. As the chart below shows, Zip codes with fewer than 15 percent of homeowners only cut back only a little – spending only half a cent less for every dollar their home fell in value. But in Zip codes where more than 50 percent of homeowners were underwater, borrowers cut back five times as much – spending 2.5 cents less on car purchases for each dollar of reduced housing wealth.

In an e-mail interview, Sufi says he and his co-authors believe the paper is the first to show that borrowers with very high leverage – which would include people who are underwater – are likely to cut back significantly on spending in a housing decline.

“This shows that a decline in household wealth will have larger consequences for housing spending if losses are concentrated on high leverage, low net worth borrowers,” he says. “The distribution of losses matters, not just the level.”

Sufi says the findings underscore the notion that principal reductions – reducing the overhang of mortgage debt left by the financial crisis – should have been more widely embraced as part of the policy response to the housing bust.

“Facilitating mortgage debt write­-downs would have softened the blow to household spending. A dollar lost by a creditor has less of a negative effect on consumption than the positive effect on consumption from a dollar gained by an underwater homeowner,” he says. “But some caution is warranted from a policy perspective: there may have been other costs of mandatory write­-downs that our estimates don't capture. For example, faith in contracts may have been undermined.”

Some critics argue that the overall decline in housing wealth easily explains the economy’s subpar performance from 2007 through much of 2012, and that you don't need effects from underwater mortgages to account for the drop in consumer spending. And while debt may have had an impact, the real problem was that housing values declined dramatically across the country. Writing down mortgage debt, this argument goes, would not have had a significant impact on the overall economy as long as home values were depressed.

“There was a big negative housing wealth effect,” said Dean Baker, co-director of the Center for Economic and Policy Research. “It might be somewhat stronger for people who are underwater and for lower-income people compared to higher-income people, but the story is that we lost $8 trillion in housing bubble wealth and that was going to have a huge impact on consumption even if no one was underwater.”

But Sufi maintains that critics too often assume that the decline in housing wealth was uniform across all borrowers. The reduction in consumption, he points out, was much greater for the most indebted people.

“For the same dollar decline in housing value, high-leverage, low-net-worth households cut spending the most,” he says. “We are fixing the dollar decline to be the same. It is not just about the decline in house prices.”