Study Sees Flaws in Consumer Sentiment Data

Consumer confidence indexes help move stock markets, influence corporate decisions and alter governments’ economic outlooks.

But a study says they’re essentially useless for forecasting Americans’ spending patterns. The two closely tracked indexes -- from the University of Michigan and the Conference Board -- may reveal the mood of Americans when they are surveyed, but they don’t predict whether people will spend more or less, said Dean Croushore, an economics professor at the University of Richmond.

“People may say they’re dissatisfied with the economy, but then they go out and buy a car,” Croushore said.

Croushore said that when he began his study, he believed the confidence readings had value and that his analysis might show they were more insightful than perceived.

“But the more investigation I did, trying all different types of models to make them work better, they kept getting worse and worse,” he said. “I think the reason, a lot of times, is that consumer confidence just reflects the past. You lose your job, your confidence falls. There’s not really anything new there. What we really want are indicators that look forward.”

Croushore, a former vice president at the Federal Reserve Bank of Philadelphia, said he spotted a fundamental flaw in older studies that found the indexes somewhat useful in predicting consumption. Those researchers had been comparing the consumer confidence indexes to widely used economic data that had been revised several times -- sometimes drastically.

It makes more sense, he said, to compare the indexes with “real-time” data -- information available to consumers and researchers at the time the surveys were taken.

When he crunched the numbers using the unrevised data, he found that in most cases, the indexes fell after recessions and rose after economic expansions -- but not before those phases in the economic cycle. And the indexes provided little guidance on consumers’ spending, Croushore said.

For instance, Croushore said the two main indexes -- the University of Michigan’s index of consumer sentiment and the Conference Board’s consumer confidence index -- fell after the stock market crashed in 1987. But neither suggested that consumer spending would come back as strong as it did, he said.

Croushore said the Michigan index performed worse in his unpublished study. In the late 1990s, for example, consumer spending increased, but the Michigan readings had showed confidence at a rather flat level while the Conference Board index had risen.

Michigan’s consumer surveys were created after World War II as a way of measuring consumers’ optimism or pessimism. Each survey now includes about 50 questions that deal with topics as varied as refrigerator purchases and inflation. Consumers are asked a number of follow-up questions, with a goal of understanding why they make certain financial decisions.

The sentiment index is based on the answers to five questions in the survey, including consumers’ views of their financial situations currently and in the future, their expectations about employment, and whether it’s a good time to buy a major household appliance.

A minimum of 500 telephone interviews are conducted each month, and a score is calculated based on the responses.

The Conference Board’s index, which dates to the 1960s, works in similar fashion. Its monthly survey is mailed to 5,000 U.S. households, with a response rate of 70%.