Do consumer sentiment surveys predict changes in consumer spending?

I have gotten very tired of reading an endless stream of press accounts that misguidedly imply or outright state that some latest reading in a consumer sentiment survey indicates a causal effect on consumer spending in the coming weeks or months, when studies have shown that no such causal relationship exists. In fact I added this as one of my "predictions" for 2008: "Consumer sentiment surveys will continue to be proven as a poor indicator of future consumer spending." Imagine my surprise when I read an article in The Wall Street Journal by Kelly Evans entitled "Holiday Spending May Defy Gloom Of Consumer Polls" that details the case for my view. Thank you Kelly!

She does not list the specific study from the St. Louis Fed that debunks the myth of a causal relationship between current consumer sentiment and future consumer spending, she does note the contributions of the St. Louis Fed to this effort.

The really good news is that she first reports that holiday spending was reasonably decent and in defiance of the recent consumer sentiment surveys:

Early indications that Christmas sales have been decent -- though not spectacular -- suggest that Americans may be opening their wallets wider than consumer-confidence barometers have been signaling they would.

She then delves into the central issue of "a long-running debate about whether confidence numbers are useful in predicting how freely consumers will spend."

She summarizes some of the academic views:

Jeremy Piger, an assistant professor of economics at the University of Oregon who studied consumer sentiment while working at the Federal Reserve Bank of St. Louis, explained that early academic studies of consumer surveys found that there was a correlation between the level of consumer confidence and future economic activity.

But, he said, later studies "got more sophisticated." They took a close look at other economic data released each month to see whether the confidence surveys, in and of themselves, had any predictive power. "The answer has pretty uniformly been, 'No,'" he said. The consumer numbers reflected other developments, on jobs and prices, for example.

A 2006 study by Dean Croushore, an associate professor of economics at the University of Richmond, concluded, "If you are forecasting consumer spending for the next quarter, you should use data on past consumer spending and stock prices and ignore data on consumer confidence."

She provides additional background and a number of other interesting and useful tidbits on this issue. One quote I thought was particularly direct:

Joshua Shapiro, chief economist at the research firm MFR Inc., parses the surveys for his clients but has a disclaimer at the end of his comments: "In our view, consumer sentiment and confidence indicators have never been more than loosely correlated coincident indicators of economic activity, and this relationship has become even weaker in recent years."

Thank you again Kelly for helping enlighten readers on a point that so many so-called "professionals" on Wall Street are so ignorant about.

I just realized that my 2008 prediction should have been worded as "Most of the media and so-called "professionals" of Wall Street will continue to falsely claim that consumer sentiment surveys are a good indicator of future consumer spending."

-- Jack Krupansky