Confession & Confusion

The Bureau of Labor Statistics took a deep breath last week, squared its shoulders and owned up to a gigantic mistake: its employment index has been cockeyed for at least two years. In its last estimate, made in June, the bureau had put U.S. employment (nonagricultural) at 43,733,000, which, it now turns out, was nearly 1,000,000 too high. In 1948, BLS had been wrong by 930,000, in 1947 by 599,000.

BLS's troubles can be blamed on World War II. The bureau keeps a file of 120,000 companies which it uses as a sample for measuring overall employment. During, and after, the war BLS failed to adjust its totals for new industries or industries that had declined in importance. Future estimates, hopes BLS, will be good and true.

But after listening to one bureau confess a mistake, economists and businessmen raised their eyebrows at the Bureau of the Census, whose optimistic employment estimates for August (51,400,000) had set off a hallelujah chorus of hope for a big upturn. The Bureau of the Census coldly replied that it was not in error, pointed out that it uses a different computing method, and that it includes several types of employment not covered by BLS.

But one of the Government agencies was still haywire. Though they have always differed on their employment figures, there was still the whopping difference of more than 7,000,000 in their estimates of nonagricultural employment.

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