The Myth of Consumer Confidence

At the end of a very illuminating column on how U.S. companies are not creating the number of jobs they used to, NYT's David Leonhardt leaves us with this note:

In related news, a survey of consumer confidence released yesterday showed the sort of decline that, according to Merrill Lynch, "is rarely seen outside of an impending recession or a natural disaster."

Since consumers account for over two-thirds of the American economy, knowing the future path of spending is key to forecasting recessions. Unfortunately, consumer sentiment/confidence surveys are one of the worst predictors of spending habits. This from Dean Croushore from the University of Richmond:

The indexes of consumer confidence are not of significant value in forecasting consumer spending. In fact, in some cases, they make the forecasts significantly worse.

And this from Indian economist Roy Batchelor, who does find some predictive power in confidence surveys, but only on the way up out of a downturn:

Rising consumer confidence significantly increases the probability of recovery. But these relationships do not yield a reliable method for anticipating business cycle turning points.

The markets, of course, have figured out the puny forecasting powers of the Reuters/University of Michigan and Conference Board consumer surveys. When I was a reporter at Reuters, part of my job was to call up traders/strategists/economists on market moves to find out what was going on. (A futile practice, especially for the stock market, but you'd be surprised how predictably Treasuries, the market I covered, moved on economic news.) On days when a jobs, inflation, or GDP report was released, that would be the first thing out of their mouths, but I can't recall one time that any of them attributed a market move to consumer moods.

I'm not arguing that we're not headed for a recession, or are not already in one, but that the popularity of consumer surveys don't match their actual usefulness.