WASHINGTON'S MISLEADING MAPS OF THE ECONOMY
Last autumn, Main Street was shivering in what seemed like the depths of recession. Yet, the government's official estimate of gross national product--its broadest measure of the nation's economic health--suggested things weren't so bad. The economy was contracting, according to the Commerce Dept., at a modest annual rate of 1.7%. So why did the economy feel so much worse?
One reason is that the official figures may be badly out of whack. A decade of budget cuts and benign neglect has left the government's main producers of statistics, the Labor and Commerce Depts., unable to track a rapidly shifting economy. Indeed, according to a new government measure of GNP, the economy was actually declining at a troubling 3.1% pace in the fourth quarter of 1989, nearly twice as fast as initially estimated. The cause for the dramatic shift: Commerce couldn't adjust its figures to reflect dramatic changes in the cost and quality of computers. That meant it was overestimating the true cost of the equipment and thus the real output of computer manufacturers. The new calculations should solve the problem, but they won't be officially published until next November.
The problems with the government's numbers go far beyond a single set of GNP figures, however. From retail sales to the trade gap, the statistics needed by policymakers and businesses are either misleading, late, or just not available. Observes Federal Reserve Board Chairman Alan Greenspan: "The economy has been changing faster than our ability to measure it."
DANGER SIGNS. That's especially troubling when the economy is near a turning point. Looking back, it seems clear that the economy was showing early signs of distress by late 1989. Yet, after looking at relatively upbeat statistics (chart), the Federal Open Market Committee decided not to lower short-term interest rates at a critical Mar. 27, 1990, meeting. Fed officials insist they received no "false signals" about the early signs of a downturn. And no one will ever know if the Fed would have cut rates in the spring of 1990 had it known how serious the slowdown was. But uncertainty over the reliability of data takes a toll. "It's not that the Fed went off 180 degrees," says William C. Melton, chief economist of IDS Financial Services Inc. "But it means policymakers tend to go real slow."
Capitol Hill, too, may have been misled. Congress hasn't moved to counter the recession, in part because lawmakers were convinced the downturn would be relatively mild. "The line has been that the recession would be short and shallow," says Joint Economic Committee Chairman Paul S. Sarbanes (D-Md.). "It would have been harder to sell that with more accurate figures."
Now that the U. S. economy may be heading upward, the lack of good data could once again be getting in the way of sound decisions. If, as many believe, consumer spending will lead the U. S. to recovery, the government's monthly measure of retail sales should be a major clue to how things are going. Yet, for the past two months, the initial report has seriously underestimated sales (chart). For example, in April, Commerce reported that March sales declined by 0.8%. The report was widely interpreted to mean the recession was still in full swing. Then, on May 15, officials revised their estimates, reporting that March sales had actually increased by 0.4%. Government statisticians blame the problems on late reporting by businesses, and on a survey that hasn't kept pace with the rapid growth of new types of retailers, such as video rental outlets.
DRAWING BLANKS. The paucity of accurate and timely information has even more serious consequences for policy debates over such issues as international trade and competitiveness and health care reform. The oft-heated debates are riddled with assertions. But the hard evidence policymakers need to steer them in the right direction just doesn't exist. "Increasingly, the government is run by people who think by the numbers," says Labor Statistics Commissioner Janet L. Norwood. "But the economy is so much more complex, you really need better data than we have ever had."
For example, it has become an article of faith that U. S. service-sector productivity is lagging. But government statisticians say they simply don't know how to count what many of these businesses produce. It's easy to measure the output of a steel mill. But what does a bank produce? "There's a general belief that we've had this long-term decline in productivity," says Patricia Ruggles, an economist at the Urban Institute in Washington. "It may perfectly well be true. But our output measures are so bad, there is no way to know."
That's especially true of health care. It's the fastest-growing sector of the economy, yet government statisticians have no handle on it at all. "We don't know how to define the output of the medical care industry," says Norwood, "and we don't know how to measure it." Not only do these problems distort productivity trends in services, but they also cast doubt on whether the overall growth of the U. S. economy is being accurately measured.
It's the same problem with international competitiveness. Almost everyone agrees research and development is critical, but no one knows how many researchers are working here or what they are doing. And no one knows how much money is going into key technologies, such as high-definition television or advanced ceramics. "We haven't a clue," says Jules J. Duga, senior research scientist at Battelle Memorial Institute.
The basic trade statistics are also flawed. In theory, the U. S. trade deficit should be roughly equal to the amount of foreign capital coming into the country, since the U. S. must borrow from abroad to finance the shortfall. But in 1990, the official current-account deficit was $73 billion more than the government's tally of the flow of money into the country. Says Rudolph G. Penner, a member of the National Association of Business Economists' statistical committee and former head of the Congressional Budget Office: "It's conceivable that if we did everything right, we'd find we don't have a competitiveness problem."
OUT OF SYNC. The Bush Administration has taken some tentative steps to improve government data. Council of Economic Advisers Chairman Michael J. Boskin wants to spend $36 million to upgrade statistical programs--the first significant spending increase in more than a decade. But except for a handful of lawmakers, such as Sarbanes, there is very little interest on Capitol Hill in pumping up the budgets of statistical agencies.
For now, Labor and Commerce are trying to hold their tattered statistical operations together. The government's efforts to better measure the real output of computers are also a big step forward. Both agencies are upgrading their survey methods by refining both questions and targets to better reflect the changes in markets and the labor force. These improvements may help the government track cyclical changes. But the agencies have a long way to go before they can follow broader trends in productivity and international capital flows.
Indeed, without new resources, government statisticians will fall further behind a rapidly changing economy. The demand for instant, accurate information will only grow. But the ability of government agencies to deliver will continue to shrink.Howard Gleckman, with John Carey, in Washington