Short-term job losses and other fallout from Katrina, high gas prices and higher heating bills this winter all could give consumers less discretionary income and more reason to lose confidence going into holiday shopping. That conceivably could turn a recovery into recession despite the stimulus that will come from post-Katrina government and insurance-financed rebuilding.
While the problems are real, September's confidence numbers will tell only part of the story. "You don't put all that much weight on one month," said Lynn Franco, director of the Consumer Research Center at the Conference Board, producer of the signature Consumer Confidence Index.
Even if the numbers are bad, optimists still can hold out hope: Short-term shocks to the consumer psyche don't necessarily foretell troubles for consumer spending or the overall economy. "You can expect it to take a little bite out of consumer confidence," said Ms. Franco, one of the optimists. "I don't expect it to curtail consumer spending to the point that it would drive us into recession."
Economists and investors each month pore over confidence numbers-for good reason. Consumer spending accounts for about two-thirds of gross domestic product, and the indexes measure how Americans feel about the economy, jobs, income and spending. "It gives you a quick read on the general tenor of the economy," said Jason Bram, an economist at the Federal Reserve Bank of New York.
The indexes come out days or weeks before official government data, giving economy-watchers some nearly real-time indicators. "Even though they are fuzzy, they are early," said David Wyss, chief economist at Standard & Poor's. "You'd like to get a fuzzier picture of the present rather than a clearer picture of the past."
The Consumer Confidence Index and the University of Michigan's Index of Consumer Sentiment, the two best-known confidence surveys, generally track with GDP, falling as the economy heads into recession and rising as recovery takes hold.
The numbers also move with the job market; a question in the Conference Board survey about job availability closely tracks with the unemployment rate (see chart, this page). That's no coincidence; consumers know when they, family and friends are getting hired and fired.
Unfortunately, the correlation that marketers really would like to see-what confidence today means for consumer spending tomorrow-is imprecise. Mr. Bram, who has studied the relationship between confidence indexes and spending, said confidence is a useful incremental tool-along with data about the stock market, interest rates, income and employment-to help forecast spending. "We found it does help a bit, but it helps by a fairly modest amount." He added: "It's not a silver bullet."
Ms. Franco said confidence is a better gauge of durable-goods purchases such as appliances than it is of retailing overall. Confidence isn't the main driver of spending. "What dominates consumer spending is not consumer attitudes but consumer income," said Mr. Wyss.
Low confidence numbers don't necessarily foretell weak spending, but low numbers do give marketers early warning that they may need to get more aggressive to draw consumers. Case in point: Wal-Mart Stores, hurt because its customers are pinched by fuel prices, this month vowed to be more aggressive on holiday discounting.
Monthly moves in consumer confidence generate headlines, but trends over several months are more relevant. Rajeev Dhawan, director of the Economic Forecasting Center at Georgia State University, said he doesn't worry if fluctuations are less than 5% in a given month. "Look for very big jumps up or down" over several months, he advised. Mr. Bram said a 20-30 point drop (or rise) over a couple months is reason to worry (or be jubilant).
Ms. Franco said a Consumer Confidence Index of 90 or above is indicative of a growing economy. She offers a rule of thumb: Watch out when the index falls below 80 for two to three months, which she said is generally a warning sign of a looming recession.
Going into September, the Conference Board index stood at 105.6, meaning consumer confidence was 5.6% above the baseline year of 1985. The university index stood at 89.1, below its 1966 baseline of 100.
Consumers are good at reading the economy-especially knowing when things are bad. The debilitating 1973-75 recession began at the business-cycle peak in November '73, when the Consumer Confidence Index was a healthy 107.5 (not far from last month's number), and hit bottom in March '75, three months after the index sunk to an all-time low of 43.2.
In the 1980 recession, the Index of Consumer Sentiment slumped to an all-time low (51.7) two months before the business cycle hit its trough. During the 1981-82 downturn, the Confidence Index hit bottom one month before the economy reached its nadir. Both indexes had a dramatic upturn in March 1991, the start of a decade-long expansion.
The indexes at times give false readings: The Consumer Confidence Index skidded in 1992, for example, but there was no recession in the offing.
The last business cycle was more intriguing. Investors didn't know at the time that the Dow Jones Industrial Average had peaked Jan. 14, 2000, or that the bubble was about to burst when Pets.com and a pack of dot-com dogs took over Super Bowl XXXIV later that month. But consumers called it: Both the Consumer Confidence Index and the Index of Consumer Sentiment reached peaks that month.
The numbers drifted downward-and then plummeted in September 2001. After 9/11, the Index of Consumer Sentiment and Consumer Confidence Index dropped to the lowest points since 1993 and 1996, respectively. But even before the attacks, the numbers were falling, reflecting pessimism amid increased unemployment, higher inflation and a spike in fuel costs. Economists later would rule a recession had begun months earlier, in March 2001.
Economists advise caution in making broad assumptions about confidence given that consumers' responses about how they feel don't always translate into how they act, even over the short term.
Consumer spending on cars rocketed 28.5% in October 2001 vs. a year earlier and broke industry sales records, driven by General Motors Corp.'s "Keep America Rolling" zero-percent-interest promotion and similar offers from rivals. Two months after 9/11, the recession hit bottom and recovery began. The university's index surpassed its pre-9/11 level in January 2002, just four months after the attack.
The auto industry's ability to drive demand after 9/11 demonstrates the power of marketing: Even when consumers are most pessimistic, they can be enticed into the store if the deal's too good to pass up. That's assuming consumers have money to spend and fairly good feelings about job security; such was the case in the last recession, a relatively minor downturn.
Confidence counters have been at their game for many decades. Economist George Katona began the University of Michigan's pioneering survey in 1946; Richard Curtin, his successor, took over in 1976. The Conference Board, a nonprofit industry research group, in 1967 created the Consumer Confidence Index, compiled since inception by market researcher National Family Opinion (now TNS).
The Consumer Confidence Index and the Index of Consumer Sentiment each are based on five questions, but differences in the questions help explain why the indexes don't always track. The Conference Board asks about jobs, business conditions and family income, looking out six months. The university asks consumers how they feel about their financial picture, their view about making major purchases and opinions about business conditions, looking out one year.
Both indexes use two of their five questions to create a "current situation" sub-index; they use the other three questions to produce an "expectations" sub-index.
The Consumer Confidence Index and the Index of Consumer Sentiment generate the headlines, but the sub-indexes are important. In fact, the university's Index of Consumer Expectations is one of 10 components in the closely watched Index of Leading Economic Indicators, published by none other than the Conference Board. (Why does the Conference Board use the university's index when the board has its own? Because the Commerce Department created the leading index, incorporating the long-established Michigan survey; the board took over the leading index in the `90s and kept the university's survey in it.)
The confidence indexes have rivals. The ABC News/Washington Post Consumer Comfort Index offers a weekly read based on phone interviews by ICR-International Communications Research. Ipsos Public Affairs fields the monthly RBC CASH (Consumer Attitudes and Spending by Household) Index. The Institute for Business Cycle Analysis, based in Copenhagen, promotes an index of U.S. purchasing intentions produced by TNS-Intersearch, sibling of the TNS unit that does the Conference Board's index.
But the Conference Board and the university are the big guns. Which is the gold standard? Michigan got there first, and it has the prestige of inclusion in the leading index. The Conference Board has a bigger sample size (5,000 surveys mailed and about 3,500 completed, vs. the university's 500 phone interviews), allowing the board to do regional breakouts.
"We think we have a terrific tool and a terrific product," said Jon Harding, president-U.S. client service at TNS, the Conference Board's partner. "Its predictive power has certainly [been] proven out over the years." Mr. Curtin, director of the university's surveys of consumers, stands by his numbers: "If you look at the data trends over time, you will see exceptional performance of our much smaller scientific sample, which is drawn to be representative of the entire nation."
The surveys at times diverge in the short term. The university's index fell in August, which Mr. Curtin blamed on high gas prices; the more job-focused Consumer Confidence Index rose in August, which Ms. Franco attributed to improving job prospects. Over longer periods, the surveys tend to track. Said Ms. Franco: "They pretty much tell the same story." That should give data consumers more confidence in both indexes.