Economy’s Strength Depends on What Numbers You Use

We have always known that the views of economic forecasters are best taken with a barrel of salt. Now it turns out that not only don’t economists have a clear view as to where the economy is headed, they can’t even tell with any accuracy where it’s been.

In Britain, analysts were shocked when the Office of National Statistics (ONS) decided that the economy has been growing twice as fast as it previously thought. It seems that someone in the until-now merely useless Department of Trade and Industry had lowered it to downright dangerous by badly misestimating the level of construction activity. The revision raised from impossible to probable the prospect that the Treasury’s prediction that the nation’s finances would come right would prove spot on, rather than miserably off the mark.

In America, economists are trying to decide not only whether the current 5+ percent growth rate is a mere spurt, due to peter out when consumers have finished disposing of the tax rebates they have been getting in the mail, or is the start of another long-term period of above-trend growth. Some, like economists at Goldman Sachs, see danger ahead: “Conditions are ripe for setbacks in homebuilding and consumer purchases of durable goods. . . . Mortgage loan refinancing, one of two key drivers, has already plunged; the tax cut—the other driver—is having its biggest impact . . . in the third quarter. . . .”

Adding to the gloom are declines in the index of the Institute of Supply Management and various indices of consumer confidence, the refusal of China to allow its currency to rise against the dollar so as to increase the competitiveness of American-made exports, and OPEC’s decision to cut back on oil production so as to shore up prices.

Enter the optimists. The Institute of Supply Management’s index, they point out, may have dropped a bit, but is still above 50, the level that indicates that the manufacturing sector is expanding. Consumers may not be in a chipper mood, but they continue to spend: the latest figures indicate a powerful 7.4 percent increase in consumer spending during the early part of the just-ended quarter, a year-on-year rise of 13 percent and 5.8 percent, respectively, in the number of vehicles sold by General Motors and Ford in September, a mini-boom at the nation’s (indeed, the world’s) largest retailer, Wal-Mart, and continued strength in the housing market. Better still, business spending recorded a second quarter gain of 7.3 percent, the biggest rise in three years, and likely to be sustained if reports that America’s corporations are raking in the highest profits in several years prove accurate.

Add these two anecdotes. At dinner last week one of America’s premiere dealmakers told me that there is so much buying power available to finance acquisitions that a new wave of corporate takeovers is about to break upon us, to the joy of the investment banking and legal communities. And a few days ago, in the course of reviewing the state of the economy with a leading investment advisor, I learned that those close to the market, as she is, just don’t believe their own economists’ predictions of slackening growth. “We are in the midst of a worldwide economic recovery,” this normally conservative advisor, responsible for billions of dollars of clients’ money, told me. She has chosen to ignore the readings of her own economists, just as she shunned the advice of her irrationally exuberant colleagues during the booming 1990s.

It was ever thus—the economic tea leaves readable only through a glass, darkly. But now economists are admitting that not only are we uncertain about the future, we are not sure we know much about the recent past.

Consider the most politically sensitive of all U.S. economic statistics: the employment and unemployment rates. Professor Alan Krueger of Princeton University asks his readers to pick one of the following: “The jobless rate dropped in August; it didn’t; we don’t know.” Writing in the New York Times, Krueger points out that the reported decline from 6.18 percent to 6.08 percent might be due to the error inherent in the sampling procedure used by the Bureau of labor Statistics in its sample survey of 60,000 households. So the correct answer to the good professor’s quiz is “we don’t know.”

There are even bigger problems with job market data. Many economists look at the lay-off announcements of major manufacturers and conclude that the job market is somewhere between shrinking and soft. Not so, says Professor Allan Meltzer of Carnegie Mellon University. When a manufacturing firm outsources, say, its cafeteria operation, it reports a decline in employment. The start-up that inherits that function continues to employ the same workers, buts its job total is not included in the government’s so-called Establishment Survey, which covers employers of about one-third of all workers. As Jon Hilsenrath points out in a recent Wall Street Journal comment, “In the past, the survey has underestimated employment in the early stages of upturns because it missed the jobs being created by very young companies.” Indeed, a study by John Kitchen, an economist with the House Budget Committee, found that in the first twenty months of the five previous economic recoveries, the government has had to revise up its early estimates of the total number of jobs.

We are not talking about small differences. Meltzer notes that for the year ending in August, the Establishment Survey shows a loss of 463,000 jobs, while the Household Survey shows that the economy added 313,000 jobs. More important to the White House, the 2.7 million lost jobs reported by the Establishment Survey and used by Democrats to attack the Bush economic record shrinks to 220,000 jobs lost in the Household Survey.

Unfortunately, major data revisions are unavoidable. So when you sup with an economist trumpeting the latest, the very latest information, use a long spoon.

This article appeared in London’s Sunday Times on October 5, 2003.