An Economy On Steroids?

New Commerce numbers may show that U.S. growth has been even more robust than previously thought

Most revisions of government statistics get the attention they deserve: next to none. But the Commerce Dept.'s new look at the U.S. economy could send a tremor through the markets. These new calculations, due on July 31, may show that America's already healthy growth has been stronger than previously thought--adding almost a full percentage point to growth for the last two years--and help clear up some mysteries about the economy's recent performance.

The key to the increase: an unusually large--and growing--gap between two measures of the economy. In theory, gross domestic product, the government's main yardstick for tracking U.S. production of goods and services, should be equal to U.S. gross domestic income: the sum of wages, salaries, profits, and other income generated by the economy.

But since early in 1993, gross domestic income has apparently grown by $178 billion more than GDP. About 70% of this gap, known as the "statistical discrepancy," has accumulated over the past two years. Measured by income, the economy has posted a sizzling 3.7% growth rate, after inflation, during that period, compared to 2.9% growth in GDP.

"90 MPH." These revisions will try to reconcile the two sets of numbers. If they boost the official GDP closer to the income data, "the record will show that the economy has been driving at 90 mph without any damage to the engine," declares Everett M. Ehrlich, president of economic consultants ESC Co. and until lately the Commerce Dept.'s top economic official. "The economy is capable of much faster growth without any feedback in inflation."

Which figures are more likely to be correct? Government analysts still believe that the income data are less solid than the figures used to calculate GDP. "We continue to feel more confident about the product data," says Robert P. Parker, chief statistician at Commerce's Bureau of Economic Analysis. For example, government accountants don't get detailed income figures from tax returns until three years after returns are filed. That lag can lead to statistical misunderstandings, since capital gains from the stock market are taxed as income--but they're supposed to be counted as increases in wealth, not income, in the national economic accounts. If some capital gains are miscounted in the preliminary tax data, they could inflate the income numbers. Bottom line: When the revisions are done, "GDP might go up, or income might come down," the BEA's Parker says.

But many economists are betting that GDP will be revised up. Treasury Dept. analysts say that capital gains can't account for the recent surge in tax revenues, including the $45 billion-a-year windfall that helped seal Washington's balanced-budget deal. Those revenue gains are so large that "even the gross domestic income numbers may be too low," says Salomon Brothers Inc. economist David Hensley.

GONE AWOL. Upward revision of GDP would help solve several mysteries of the current expansion. For one thing, revisions could capture the productivity growth that has apparently gone AWOL during the expansion. The official productivity statistics, which are based on the GDP figures, show a meager annual nonfarm business productivity growth of only about 1% over the past two years. At that rate, rising wages and other labor costs should be squeezing profits--but Corporate America continues to report strong earnings increases.

Replacing the GDP growth rate with the income figures, though, boosts productivity growth to a respectable 1.7% since early 1995 (chart). Higher productivity can offset labor cost hikes--and still leave plenty of room for profits to grow. "Every measure of productivity suggests that GDP growth has to be revised up," says Gordon Richards, chief economist of the National Association of Manufacturers.

That sits well with Federal Reserve Chairman Alan Greenspan. He has long argued that official measures aren't capturing all the productivity created by corporate restructuring and heavy investment in new technologies. A sharp upward revision in growth would alarm some Fed officials, who fear that faster growth will lead to higher inflation. Greenspan would argue, however, that higher growth matched by higher productivity poses no inflationary risk.

The revisions could also help explain the missing "wealth effect." Economy watchers, including the Federal Reserve, have been puzzled because the stock market boom, which has added $3.5 trillion to household wealth since the end of 1994, has apparently not led to higher levels of consumer spending. That goes against past experience, which typically showed a significant wealth effect from soaring stock values as consumers spend some of their gains.

FRESH RIDDLES. But if there is an upward revision to growth, much of the gain would show up as an increase in consumer spending, suggesting that there has been a wealth effect. By Salomon's calculations, substituting the higher income figures raises consumption by some $110 billion. That's close to what wealth-effect models would predict from the stock market boom, since research indicates that each $100 increase in household net worth should boost consumption spending by $3. The downside: The extra consumption wipes out reported gains in the household savings rate.


But the income figures don't just solve mysteries: They create them, too. If income is higher than output, the economy must be producing something that's not being counted. But what? Some of the extra income is coming from small businesses, suggesting that "high-tech startups and Internet commerce are creating shadow output," according to economist Ehrlich. "No way," however, is there "$90 billion in commerce on the Internet," says a Clinton Administration economist. Another explanation: Financial services, which have been generating enormous profits, are very poorly counted in GDP. And output in the health-care industry, where both wages and employment are growing, is undercounted, says John D. Hancock of Regional Financial Associates Inc.

Even July's revisions won't stamp "Case Closed" on these puzzles. Given the data lags and the need to upgrade government statistics, it may be years before economists fully understand the current business cycle. But this first cut at balancing the books should help make some sense of the numbers--and put an even brighter shine on the long-running expansion of the '90s.

Before it's here, it's on the Bloomberg Terminal.