The Dismal Science, Figured Out

THE news cycle, in the northern hemisphere at least, is moving toward its dog-days phase. During this period all sorts of events, or apparent events, loom like mirages before a traveler too long on the road. Alert citizens must keep asking themselves - Does this appear important only because nothing else is going on? But in another sense, there is never ``nothing else going on.'' On even the stillest, deadest days of summer, the statistical mills of the United States government grind on. Employment figures, unemployment figures, trade-deficit figures, industrial-production stats, interest rates, and countless others are churned out regularly.

But what do they show us? And if these economic snapshots don't show us enough, is it our fault, or the fault of the figures?

In his budget address in February, President Bush set improvement of economic statistics as a priority for his administration. At the end of this month, the President's Economic Policy Council is to hear from a working group that has been soliciting ideas from the statistical and policy-making agencies of the government, as well as, informally, from private-sector groups such as the National Association of Business Economists and the American Economics Association.

As the economy has changed, with services, information-based industry, and international trade ever more important, the numbers crunchers have had a hard time finding out how best to track them. The well-known consumer price index, which tracks the cost of a given set of typical consumer goods over time, not too many years ago removed Panama hats from its ``market basket.'' But the producer price index, which tracks some 3,100 commodities, figures in hardly any services at all, and leaves out virtually the entire transportation industry.

One Boston bank economist cites employment and unemployment data as an area where nonspecialists are becoming more knowledgeable. Time was when unemployment figures got most of the attention, because joblessness is something most people can identify with personally; emotion attaches thereto. But employment data - covering the numbers of people who have jobs - actually tell more about where the economy is moving, and this is being more widely recognized.

Audrey Freedman, an economist with the Conference Board in New York, suggests that statistics with social implications are particularly problematic. The ``snapshots'' provided by the consumer price index, for instance, are useful. But employment and unemployment statistics would be of greater use with a more longitudinal approach - to get a sense of long-term vs. short-term joblessness, for instance, or of women's long-term participation in the work force.

Another key point from Ms. Freedman: ``Improving our productivity is the only way to improve our standard of living.'' But productivity has always been hard to measure, and in an information-based economy, ``We don't always know what the product is.''

Lyle E. Gramley, chief economist of the Mortgage Bankers Association and a former governor of the Federal Reserve Board, finds the sophistication with which nonspecialists track statistics to have improved in recent years. Financial markets, particularly bonds, still overreact, however, to blips on the statistical curve well within the range of sampling error. ``And then there's the fact that the stock and bond markets focus not on what figures actually are but on what they were expected to be,'' he adds.

In economics as in politics, ``better than expected'' can be as sweet as outright victory - which is trickier to assess in economics, anyway. Mr. Gramley imagines aloud a world in which there were an Earthquake Forecasting Service that would predict the Richter scale readings of the big temblors. ``Something like the great San Francisco earthquake could hit, with a reading of 8.6, and the markets would go wild - because it was expected to be 10.4!''