Don't Listen to the Consumer
By Rob Norton

(FORTUNE Magazine) – Stocks slumped yesterday after a crucial gauge of consumer confidence declined, stoking fears that Americans would close their wallets and deprive the sluggish economy of one of its main props. --Reuters

Sound familiar? It should. The big question hanging over the economy like a thundercloud is whether consumer spending will falter and drag down the economy. It's been the big question for a long, long time. That particular quote, in fact, is from a story published Aug. 29, 2001.

The robustness of consumer spending was the big surprise of last year's recession, helping ensure that the downturn was mild and setting the stage for the recovery that began last winter, when GDP surged 5% in the first quarter. But now that we know that the recovery fizzled in the second quarter, when GDP growth slowed to a puny 1.1% rate, the question is back and more important than ever.

The consumer confidence indexes would seem a logical place to look for answers. They purport to reveal how consumers feel about the economy and how they're likely to behave. And like virtually all the financial and economic news this summer, they've been unrelievedly grim. Confidence plunged in July to its lowest levels since last winter.

Unfortunately for economic soothsayers, the consumer confidence indexes are poor forecasters. They've been particularly squirrelly lately, often out of step with what's happening and what's about to happen. After the Sept. 11 attacks, consumer confidence sank to levels that hadn't been seen since the last recession, and although the indexes recovered a bit in December, nothing about the way they behaved hinted at the first-quarter rebound. Confidence then came surging back in March and stayed high in April and May--at the same time, it turns out, that the economy was slowing sharply.

In reality, the consumer confidence indexes have never been very reliable forecasting tools. Research indicates that they're actually more reactive. A formal investigation of their power was published in the Federal Reserve Bank of New York's June 1998 Economic Policy Review. The Fed economists found that of the two leading measures, data from the Conference Board's consumer confidence index made a standard economic forecast only modestly more accurate. Data from the other one, the University of Michigan index of consumer sentiment, actually made the forecast worse. (One reason the Conference Board's index performs better, the economists conclude, is that its questions seem better designed to elicit information about the state of the economy. Another is that it's based on a bigger sample than the Michigan index--3,500, compared with 500.)

A new academic study by Johns Hopkins University's Christopher Carroll suggests what may really be going on with the consumer confidence numbers. Carroll looks back at the University of Michigan's survey of consumer expectations about inflation and unemployment for the past two decades. He assumes that most people form their opinions about the economy from the news media, which in turn report the views of professional economic forecasters. He tested the data to see whether the behavior of expectations can be explained by a statistical model in which consumers are merely reflecting the opinions of the forecasters filtered through the news media. Carroll found that this model does a "remarkably good job of capturing much of the predictable behavior of the Michigan inflation expectation index," and that in most respects it performs even better in explaining the unemployment expectations data.

So what does it all mean for consumer confidence, circa summer 2002? A good bet is that the cratering of confidence in July had more to do with what consumers were reading in the papers and seeing on TV than with how they actually felt about the economy. The news in July was about as depressing as it gets--dominated by the corporate accounting scandals, the volatile and speedy decline of stock prices, and stories about the disappearance of people's retirement savings.

What it also means is that the slump in confidence doesn't answer the big question facing the economy--whether consumers will stop spending and cause a double-dip recession. (For more, see "Is This Where the Economy Is Headed?") For that, we'll have to wait and see what consumers do over the next several months rather than listen to what they're saying now. The most recent hard data on spending, from June, were pretty positive: Personal consumption was up a strong 0.5%, and the growth of personal income was even stronger--its best increase since 2000. Even if those trends continue, as long as the news remains downbeat, consumer confidence will too.

ROB NORTON, a former FORTUNE executive editor, is a freelance writer, editor, and consultant in New York City. He can be reached at rob@robnorton.com.