Better Government Statistics Are Needed

To most Americans, Government statistics are meaningless, dust-dry columns of astronomical figures printed in small type in dull pamphlets. Few citizens actually read the figures, but not one is unaffected by them.

The consumer price index put out monthly by the Bureau of Labor Statistics directly controls the wages of well over 1,000,000 workers; a fractional miscalculation could cost U.S. employers millions of dollars in paychecks. The farm support program is based on the Agriculture Department's calculations of how much farmers must pay for such essentials as tractors, seed and fertilizer. Since these calculations in turn affect the prices of farm products, the U.S. housewife meets Government statistics every time she stops at the check-out counter of the supermarket. Department stores use Federal Reserve Board retail-sales figures to plan buying and to control inventories. Railroads use Government figures for rate bases; makers of building materials and appliances base production on Commerce Department housing reports; shippers set up their schedules after looking over the import-export figures.

But despite the wide use of official facts and figures--and the critical need for even more accurate measurements--Congress treats Government statistics like an unwelcome stepchild. The most pressing statistical needs now are new censuses of business, manufactures and mineral industries. All are old; the most recent--business--was last made in 1948. Since that time radical changes have occurred; whole new industries have sprung up to make the six-year-old figures obsolete. Although the Budget Bureau authorized $8,340,000 for such censuses, the House hooted the appropriation down as an elegant, high-domed boondoggle.

Some criticism of Government statistics is justified, because they are far from perfect. Says Arthur F. Burns, head of the President's Council of Economic Advisers: "Although the construction industry is of strategic importance in our economy, statistical information about the industry is very weak . . . Prices of consumer durable goods entering into the Consumer Price Index are official or quoted prices; they are not 'actual' prices taking full account of variations in markdowns, discounts, trade-in allowances, or premiums and other forms of surcharges."

The trouble with some Government figures was dramatically demonstrated in January when the Census Bureau sampled 68 communities and produced an estimate of 2,359,000 unemployed, then sampled 230 communities and jacked its estimate to 3,087,000. The difference of 728,000 cast doubt on the whole method of collecting unemployment statistics (TIME, March 15). Rivalry among Government bureaus leads to more confusion. The Commerce Department's June census reports showed agricultural employment at 7,600,000: the Department of Agriculture put the figure at 9,600,000. Why the 2,000,000-man discrepancy? The Commerce Department does not count part-time farmers, or working children under 14; the Agriculture Department does. Commerce counts some agricultural processing workers; Agriculture does not.

Isador Lubin, former Federal Commissioner of Labor Statistics, says wage statistics are inaccurate now because they omit fringe benefits (which may equal as much as 20% of cash wages). Businessmen complain about having to go to the expense and trouble of keeping two or more sets of statistics for the Government, one for the Bureau of Internal Revenue, others for the ICC, FTC or some other Government agency. Most experts believe that a central statistical bureau in Washington would be impractical. But they hope that some Government official, perhaps a presidential assistant, will be appointed to end the conflicts in the Government figures.

The cost of all Government statistics is surprisingly low, about $50 million a year. Private statisticians believe that the Government could easily improve its statistical services for perhaps $10 million more. The Administration is already speeding up reports of personal income and corporation profits, planning more Federal Reserve Board surveys of consumer spending, bringing out more frequent reports on new business, short-term interest rates and public works. The President's economic advisers propose more figures on labor turnover, a handy catalogue listing major Government statistics, and what they mean.

The biggest need now is for "benchmark statistics," the actual nose-counts of population and industry that become the basic figures for sampling and periodical reports. The benchmark censuses that have been turned down by the House may still be saved in the Senate. They would provide the basis for more accurate reports making for sound planning in business, farming and government.

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