In Praise of the Wealth Effect

Mark Thoma pointed last week to a typically catchy headline from Slate: "Debunking the 'Wealth Effect'"

For the uninitiated, the wealth effect refers to an increase in spending as a result of a rise in wealth (in this case through things like stock or home price appreciation). In his story, writer Christopher Flavelle says:

...the idea of a wealth effect doesn't stand up to economic data. The stock market boom in the late 1990s helped increase the wealth of Americans, but it didn't produce a significant change in consumption, according to David Backus, a professor of economics and finance at New York University. Before the stock market reversed itself, "you didn't see a big increase in consumption," says Backus. "And when it did reverse itself, you didn't see a big decrease."

However, more Americans own houses than own stocks--shouldn't a change in home equity have a bigger impact on spending than a change in the stock market? Not so, says Backus. "There wasn't much of a wealth effect on the way up [for housing prices], " says Backus, "and probably there won't be much of a wealth effect on the way down, either."

(Flavelle lets Backus do the talking here, and Backus is an excellent economist, but not the best verbal communicator. The other time I heard him speak was at a debate in which he, along with a couple of others, were tasked with defending the position that Markets Are Moral. Unfortunately, Backus took a couple of moments at the beginning of his alloted time to talk about how economists like to get together, drink beer, and talk shop. This left him with much less time to actually argue his case for the morality of markets. It wasn't all his fault, but before the debate, 68 percent of the people in the audience favored the competing view that markets are not moral. After the debate, that contingency grew to 76 percent. This should serve as a note to economists: no more cute anecdotes about how much your profession likes beer, we get it.)

But back to Slate.

It just so happens that there's plenty of evidence that there is a wealth effect, and that the effect from housing is likely greater than the effect from stocks.

For example, Flavelle himself cites a Congressional Budget Office report which, citing seven other studies, says:

Analysts generally agree that an increase in housing wealth due to higher real home prices (that is, resulting from home prices rising faster than inflation) permanently raises real consumer spending by an annual amount that is a fraction of the increase in housing wealth

and also this:

... most estimates of the wealth effect in the United States fall within a range of about 2 cents to 7 cents of extra spending per extra dollar of housing wealth

Flavelle uses this last point as evidence that the wealth effect from housing is small. (Wait, I thought he said the wealth effect "doesn't stand up to economic data"?)

But let's look at the numbers. Between 2001 and 2006, housing wealth grew by $4.8 trillion. If consumer spending increased by 2 to 7 cents-on-the-dollar as a result, that would translate into a $96 billion to $336 billion rise. Consumer spending over that period rose by $2,170 billion, so the wealth effect from housing accounted for 4 percent to 15 percent of that rise. More recent research from John Muellbauer of Oxford suggests that thanks to the more widespread use of home equity lines of credit since the late 90's, the wealth effect from housing is closer to 7 cents-on-the-dollar than 2 cents. A double-digit rise in consumer spending seems like it would be worthy of being labeled significant.

In the same study, Muellbauer -- as well as this San Francisco Fed study that looked at the relationship between wealth and spending in three different countries -- found that the effect from housing was greater than the one from stocks. (Although, it's not an open-and-shut case. This Australian study found the opposite.)

In defense of Backus, his specialties are in "international capital markets, fixed income and currency derivatives, and Asian and Latin American economic history" -- not in household consumption dynamics.

What's more troublesome is that in the Wikipedia entry for the Wealth Effect, the only reference is this Slate article. Wisdom of Crowds, don't fail us now.