How did it come to this?

Photographer: Justin Sullivan/Getty Images

We're Still Not Sure What Causes Big Recessions

Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.
Read More.
a | A

There is an important, but quiet debate in the economics profession about what leads to big recessions: wealth or debt.

Almost everyone agrees, at this point, that the Great Recession of 2007-09 was caused by the financial system. But that leaves the question of what, exactly, happens in a financial system that leads an economy to crash.  Formal economic models of financial shocks are not very realistic. They usually assume the harm comes from disruption to the banking system, which acts like a supply bottleneck that chokes off economic activity. But the Great Recession and similar episodes look very much like demand shocks, with low inflation and lots of spare capacity.

So economists are asking what kind of financial disasters have the biggest impact on demand. Roughly, the two answers are wealth effects and debt overhangs. The wealth-effects school holds that when asset bubbles pop, people suddenly feel poorer. This causes them to cut spending, which sends demand crashing. The debt-overhang school believes that people have sudden shifts in their willingness to take on debt -- when they go into balance-sheet repair mode, they stop spending.

The argument between these two schools fortunately hasn’t been very politicized (yet). But it does have important policy implications. If wealth effects are the big culprit, then taming asset bubbles becomes the central task for recession prevention. If it’s debt that does it, the key is to stop households from borrowing so much.

The main evidence for the importance of debt overhangs comes from the observation that bubbles that involve lots of borrowing seem to cause more damage when they burst. This has been confirmed -- as much as historical patterns can really be confirmed -- by economists Oscar Jorda, Moritz Schularick and Alan Taylor. That would explain why the recession following the 2000 stock-market crash was so much milder than the carnage after 2008 -- though both involved similar-sized losses of paper wealth, the latter included much more borrowing. Paul Krugman, for example, embraces this reading of history. It’s also the idea of the balance-sheet recession, popularized by Nomura economist Richard Koo.

Others beg to differ. Center for Economic and Policy Research founder Dean Baker says that the collapse in wealth was much more important than the run-up in debt. He attributes the difference between 2000 and 2008 to the fact that the latter crash was a hit to middle-class wealth, while the former affected mostly the fortunes of the rich (who are less likely to cut back spending after suffering losses).

What does the academic literature have to say on the question of wealth versus debt? One of the most famous papers, by economists Atif Mian, Kamalesh Rao and Amir Sufi, found in 2013 that the fall in housing wealth had a big negative impact on consumption. But it also found that the effects of falling wealth were stronger for more indebted households.

The paper’s authors are therefore on the fence in the debate. They say that debt is dangerous -- it enables greater speculation in boom times, which pushes prices up too much, only to fall later. They also say that debt does a poor job of insuring against risk, and can lead to costly and inefficient legal processes like foreclosure. But they stop short of endorsing the idea of balance-sheet recessions.

Other economists who have followed up on Mian et al., however, have come down more firmly on Baker’s side. For example, Greg Kaplan, Kurt Mitman and Giovanni Violante re-did Mian et al.’s analysis using publicly available data. They focused on areas where house prices declined more and checked to see if consumption fell more in those places. This team looked mainly at changes from 2006 -- before the epic housing collapse -- to 2009, the trough of the recession.

Kaplan et al. found big wealth effects on consumption. Basically, when housing prices fall, people spend a lot less. But they didn’t find that housing leverage in 2006 had a measurable effect on consumption. And unlike Mian et al., they didn’t find that it exacerbated the wealth effect. Essentially, Kaplan et al. come down solidly on the side of the wealth-effects hypothesis, and against the idea of the balance-sheet recession.

So how can we reconcile the historical finding that debt exacerbates bubbles with the microeconomic evidence that the wealth effect was the biggest factor? One possibility is that Baker is right, and that housing bubbles -- which generally involve more debt -- simply hit the middle class harder than stock bubbles. If that’s the case, it means that much of our recent focus on household leverage might be misplaced. The real task for policy makers might be not to restrain people from borrowing, but to keep their wealth from dropping precipitously. Preventing asset markets from bubbles and busts might therefore be our top policy priority to prevent a repeat of 2008.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Noah Smith at nsmith150@bloomberg.net

To contact the editor responsible for this story:
James Greiff at jgreiff@bloomberg.net