Robert Samuelson on the Wealth Effect and Consumption

While Mark Thoma thinks Robert Samuelson makes some good points, I’m not so sure:

By estimating all three sources, Greenspan and Kennedy reached annual grand totals, shown on the table below. It provides the figures both in billions of dollars and as a percentage of people’s ordinary disposable (after-tax) personal income. The housing money is extra, on top of personal income.

2000 | $204 bil. | 2.8%
2001 | 262 bil. | 3.5%
2002 | 398 bil. | 5.1%
2003 | 439 bil. | 5.4%
2004 | 599 bil. | 6.9%

Whoa! Consumers had a lot more to spend than ordinary income, almost $600 billion more in 2004. How much of that was actually spent (as opposed to being put into bank deposits, stocks or mutual funds) is unclear … The economy has depended heavily on all this extra cash. And, before the housing bonanza, there was the stock boom. From year-end 1985 to year-end 1999—just before the market peaked—the value of households’ stocks and mutual funds grew from $1.4 trillion to $12.8 trillion. Economists figure that consumers spend between 2 percent and 3 percent of their extra stock-market wealth. That’s also a lot of purchasing power. No one has fully explained what caused these immense wealth gains. My own oft-stated belief is that lower inflation is the main cause, because it gradually reduced interest rates.

OK, let’s dispense with this last sentence before addressing my reservations about Samuelson’s main theme. If nominal rates fell by the same amount that inflation fell, the Fisher effect notes no change in real interest rates and hence no wealth effect. The point, however, is that nominal rates fell by more than the drop in inflation rates. Now to the main event and let me put on some Ando-Modigliani life-cycle model glasses for this.

The wealth effect that Samuelson refers to represents a one-time addition to disposable income and not a permanent increase in it – assuming that massive capital gains won’t go on forever. Let’s hope Samuelson realizes that one cannot count the same increase twice. But he does suggest that the marginal propensity to spend out of newly created wealth is modest, which is consistent with the Ando-Modigliani life-cycle model.

What bothers me is his representation of the allegedly immense increase in wealth. Since I have blasted the folks over the National Review for touting the over-time increase in nominal aggregate wealth as opposed to the more modest increase in real per capita wealth, I thought I’d provide a couple of simple graphs for the 1985 to 2004 period. One graph shows consumption relative to GDP, which has grown from 64.5% in 1985 to around 70% lately. The other graph is household net worth relative to GDP, which grew substantially from 3.5 in 1994 to 4.55 by the end of 1999. From 1999 to 2002, this ratio fell even as housing wealth rose because of the erosion of stock market value.

The rise in housing values did, however, partially offset the fall in stock wealth. In a similar vein, the collapse of business investment, which caused a fall in real interest rates, was partially offset by increases in residential investment. Let’s say Mr. Samuelson is correct that housing wealth is about to fall. Note wealth stock wealth is rising so the net effect on wealth could be positive. Hopefully, business investment and export demand will also continue to rise, which would support what James Hamilton said – But you said that more saving was a good thing.