Those Static Statistics

Major Government and business forecasts of the nation's year-to-year growth have been wrong by 30% to 40% since 1947. Reason: shaky projections based on shoddy statistics. This disconcerting report came last week from the privately financed National Bureau of Economic Research, and it further darkened the shadow of doubt that hovers over many of the indicators used by businessmen to reach their decisions to spend, lend or cut back.

The Government spends $100 million a year to find and refine the modern statistics that measure where U.S. business stands and is likely to move next, and industry and private economic groups spend millions more. But the U.S. economy has changed so quickly in recent years that many statistical standbys have become insensitive, inaccurate or downright misleading.

Some long-revered statistics that U.S. businessmen would do well to treat with caution:

-DEPARTMENT STORE SALES: Last week businessmen were cheered by reports of rising department stores sales for June in the telltale New York City area. This statistic, like many commonly used ones, ignores some recent and significant changes in the economy, since it fails to measure the growth of discount houses and other fast-expanding stores that have no index of their own. Says Roye L. Lowry, executive secretary of the Federal Statistics Users' Conference: "Considering that drugstores now sell pots and pans and that supermarkets sell clothes and fertilizer, the figures on department store sales really don't tell very much." Much more meaningful as an indicator is the overall figure for retail sales, which has been showing gains that are steady--but less spectacular than those of department store sales.

-FREIGHT CARLOADiNGS: Bernard Baruch is reputed to have said long ago that the surest way to gauge the whole economy is to "watch freight carload-ings." That was long before trucks and planes captured such a large share of the changing cargo market, and also before freight cars were built bigger to carry more cargo. Result: freight loadings often go down--as they have for four of the past ten weeks--at the same time that total cargo tonnage goes up. For such reasons, the Pennsylvania Railroad, the nation's largest, last week announced that it will no longer issue carloading figures to the public.

-UNEMPLOYMENT: The problem is serious--the unemployment rate in June rose from 5.1% to 5.3%--but it is not as bad as the figures indicate. In its overly broad definition of "unemployment," the Government counts among the jobless such people as fulltime students looking for part-time work, and the job-seeking wives and children of laid-off workers. What if other countries did the same? Sweden, celebrated land of low unemployment, not long ago adopted the U.S. system; to their chagrin, the Swedes soon found that their national unemployment rate was four times higher than under their old, looser standards.

-WHOLESALE PRICES: The U.S. whole sale price index, table-flat for more than six years, is weighted heavily with food and farm goods, which have been falling because of the agricultural glut. But the much broader index kept by the National Bureau of Economic Research, which measures more manufactured goods, has been showing what worried officials call "fairly dramatic rises" for the past six months. The danger of inflation is thus greater than the Government's price index shows.

-HOUSING: Statistics on housing starts often rise or fall 10% a month, much more sharply than the overall construction market or the economy as a whole. They are the product of a tangled series of estimates, hunches and guesses. To get a housing figure, Government analysts 1) count the number of building permits, 2) guess at the number of houses that might be planned in areas that require no permits, 3) hunch the number of "permit houses" and "non-permit houses" that are actually started. "Then," sighs Chase Manhattan Bank Vice President William Butler, "you take the time of the year into consideration, and you seasonally adjust the whole mess."

Disturbed by such unreliable statistics, both the National Bureau and the U.S. Census Bureau are combing through 150 of the most widely used statistics, will recommend by year's end which of them should be revised or dropped entirely. Meanwhile, Washington's wisest economists ignore such statistics in favor of about five reliable indicators used to assess the here and now economy, and another five to predict what is likely to happen next. For current performance, they look to the Federal Reserve Board's industrial production index, the number of Americans at work, personal income, total retail sales and new consumer credit.

To get a feel for the future, they study the average number of hours in the working week, the prices of industrial materials, businessmen's plans for capital spending, inventory movements and new orders for durable goods. Fortunately for the U.S. economy, nine of these ten reliable indicators--all except industrial materials prices--are now moving in the right direction.

This file is automatically generated by a robot program, so reader's discretion is required.