Guest blog from Economics Editor Peter Coy
Breaking a tidbit of economic news here: I just got off the phone with Yale’s Robert Shiller, who has some interesting responses to the posting this morning on voxeu.org by Charles Calomiris and others that questions the existence of a “wealth effect” from changes in home prices.
The wealth effect theory, which says that rising home prices stimulated Americans to spend more during the boom years, has been cited frequently by Federal Reserve Chairman Ben Bernanke, and it’s built into the Fed’s main macroeconomic forecasting model. The theory seemed to be corroborated in a 2005 paper by Karl Case of Wellesley and Shiller—the namesakes of the Case-Shiller home price indices, among other badges of honor—and John Quigley of Berkeley.
Today, on voxeu.org, there’s an important posting that denies the existence of the effect. It’s called “The (mythical?) housing wealth effect” and it’s by Charles Calomiris of Columbia and Stanley Longhofer and William Miles of Wichita State. It summarizes a June National Bureau of Economic Research working paper by the authors with the same title.
Calomiris et al. argue their case on both theoretical and factual grounds. In theory, they say, the only people who should be expected to experience a positive wealth effect from rising prices are those who expect to sell their houses soon to cash in. Renters who had hoped to buy should actually experience a negative wealth effect, as they realize they’ll need more money than ever to buy. And people in the middle—most of us—should be neutral. Furthermore, they say, it’s possible that even if consumption does rise when home prices go up, it doesn’t have to be a cause-and-effect relationship. It could be, for example, that both consumption and housing are rising in response to an increase in expectations for future incomes.
Digging into the data, Calomiris et al. reanalyze the Case-Quigley-Shiller data and conclude that there is in most cases no statistically significant effect on consumption from housing wealth once you control for other possible contributions to consumption changes. (The working paper explains how they did this using instrumental variables—read it for yourself if you’re interested.)
I was able to reach Shiller at his office at the height of summer vacation season and he had several top-of-the-head responses, which I will now relay pretty much unfiltered:
“I’m the most behavioral of the three authors on our paper, so this is me speaking: The effects of housing wealth operate through animal spirits rather than cold calculation. … It seems to me that part of the effect is through people’s general sense of the world.”
“This is the dismal law of economics, that it’s always difficult to disentangle the ultimate causes.”
“The whole view of the world is changing all the time. … I’m sorry to be pushing my book [“Animal Spirits”] but [Berkeley’s George] Akerlof and I say that the economy is driven by stories, and one of the prominent stories [during the boom] is the triumph of capitalism.”
In other words, it’s hard to disentangle the income-expectation effect from the housing-wealth effect.
The problem with running regressions, Shiller said, is that it “inherently takes the world as unchanging for long periods of time. … I can’t resist talking about what’s happening now.”
Shiller’s response is obviously a bit loosey-goosey, but he said he didn’t mind being quoted. I think he and Case and Quigley are likely to have a more formal response in the weeks or months ahead.