Measuring the U.S. unemployment rate

Press, 13 November 1985, Page 20

Measuring the U.S. unemployment rate

By

GENE GRABOWSKI,

When the United States Government wants to know how many people are out of work each month, it doesn’t count every unemployed individual. Instead, it makes what amounts to one of the world’s most educated guesses. Since 1940, the Labour Department has measured unemployment by polling thousands of selected households in big cities, small towns and rural hamlets during one week each month. For the latest batch of figures, it surveyed some 60,000 households nationwide, representing about 120,000 people — or slightly more than one for every 1000 in America’s work force.

“It might not sound it, but it’s a very accurate sample,” said Jack Bregger, chief of unemployment analysis for the Labour Department’s bureau of labour statistics. Mr Bregger said the error in federal unemployment numbers was within 0.2 per cent. Once a household is selected, a census worker visits to find out how many people in it worked the previous week. That household is polled, usually by telephone, for three more months, then it is left out of the survey for eight months. Finally, it is included for another four-month period and dropped from the sample. A job requiring 35 hours or more

of Associated Press, in Washington

a week is considered fulltime.

People are unemployed if they had no job during the survey week, provided they also were available for work during that time and tried to find a job sometime in the previous four .weeks. Other people counted among the unemployed are those not looking for work because they were laid off and are waiting to be recalled, and those expecting to report to a job within 30 days, according to Labour Department regulations. The unemployment rate is calculated by dividing the estimated number of unemployed by the total number of the labour force. The labour force is the sum of the employed and unemployed.