How severe was the recent recession? By standard counts of total unemployment, the answer is: very severe. Total monthly unemployment, as measured by the Bureau of Labor Statistics, neared 11 percent, and the bureau estimates that in 1982 one out of three workers suffered some sort of employment problem. But by another measure of unemployment, the insured unemployment rate, things weren't all that bad. This discrepancy is of keen interest, because it is the lower of the two measures that currently determines in substantial part how much additional help the government provides to the jobless.

Many unemployed people are not eligible for unemployment benefits because they have not worked long enough to qualify, have used up all their benefits or have quit or been fired. As a result, the insured unemployment rate is always lower than the total rate. But a recent study by Brookings Institution economist Gary Burtless points out that the discrepancy has increased sharply since 1980, for reasons that are not all readily explainable.

Although the 1982 recession hit a larger portion of experienced industrial workers--who would normally be expected to qualify most readily for benefits--the jobless received much less government help than in the last severe recession. Mr. Burtless calculates that, after adjusting for inflation, the average jobless worker received 40 percent less in unemployment benefits in 1982 than were paid out in 1976. Partly that's because the Reagan administration persuaded Congress to cut back sharply on the rules for providing extended benefits under state programs and was much less generous in providing special federal benefits. And longer stretches of unemployment meant that more of the unemployed ran through their benefits before they found jobs.

These and other policy and administrative changes cannot fully account for the fact that a far lower portion of those reporting unemployment in the BLS's monthly surveys appear to have qualified for regular unemployment benefits. Mr. Burtless concludes, sensibly, that the insured unemployment rate is no longer a reliable indicator of the need for extra help for the jobless. More reliable information can come only from improving the BLS's monthly surveys of total unemployment. Continuity in data collection methods is useful, but when important questions affecting social policy remain unanswered, improvements are in order.