Counting unemployed: a number for all seasons

President Reagan says ''no,'' but his officials in the Bureau of Labor Statistics (BLS) say ''yes.''

The question to which they respond so differently: Do government unemployment figures, issued monthly, tell us what really is happening out there in the job market?

Last month, according to the BLS, the US jobless rate rose from 9 to 9.4 percent. Certainly a lot more Americans were out of work, weren't they?

Testifying before Congress, Janet L. Norwood, the commissioner of the BLS, said, ''Unemployment, which generally drops substantially each year from March to April, declined much less than usual this year -- by 330,000.''

The number of jobless Americans in April was 330,000 fewer than it had been in March. How, then, could the overall jobless total have risen?

That's the question President Reagan has been asking for two months running. Both in March and April, he declared, the BLS said fewer people were unemployed than the month before.

''I wonder,'' he says, ''if the news media wouldn't serve us better if they would give us more of the statistical information on unemployment provided by the Bureau of Labor Statistics.''

Certainly it would be a lot better for him and for Republicans in this election year if the headlines each month said that unemployment was shrinking.

But it is not. William A. Niskanen Jr., a member of the President's Council of Economic Advisers (CEA), said Monday that unemployment is likely to grow over the next couple of months, ''possibly to slightly over 10 percent.''

Mrs. Norwood referred to what is known as raw data, uncorrected for seasonal factors, when she told Congress that unemployment in April had dropped by 330, 000 persons.

If the press stressed that figure it could be misleading -- indicating that the jobless situation is getting better, when overall it is getting worse.

That, at least, is the dominant view at the BLS, which adjusts the unemployment statistics to account for strictly seasonal factors.

Mrs. Norwood went on to say that ''businesses customarily increase staff as the spring weather sets in, and large numbers of people resume job search activities as opportunities for outdoor work increase.''

Highway crews are enlarged to fix potholes. Managers of parks hire people to groom lawns and shrubs. Construction, stalled by snow and winter weather, picks up as the weather warms.

When BLS officials take these seasonal elements into account, the result doesn't indicate an underlying trend of more robust employment throughout the nation. The construction industry, for example, remains in near-depression state , because of high interest rates.

At other times of the year, seasonal factors produce an opposite effect -- ''raw data'' unemployment is misleadingly high. Each summer hundreds of thousands of young people leave school and descend on the job market. Many get work, but many do not, so the actual number of people looking for work -- i.e., the raw data of unemployment -- rises.

Mr. Reagan might be less eager to have the raw data number stressed in months when it is higher than the seasonally adjusted unemployment figure.

The aim of the BLS is to iron out the wrinkles of seasonal aberrations and to try, based on the experience of past years, to determine what the true state of the job market is.

''The seasonal adjustment factor,'' said a BLS official, ''produces a monthly jobless rate composed of two things.'' These include:

* The actual number of people unemployed, as counted in monthly household and payroll surveys. This furnishes the raw data number.

* Seasonal factors -- Christmas employment, construction layoffs due to cold weather, teen-age summer influx -- and others.

By studying the performance of past years, the BLS then melds these numbers into an average designed to express the underlying trends of the labor market within a given month.

''No one pretends the system is perfect,'' said a BLS official. ''But the accuracy of it can be checked.''

Take the raw data numbers for 12 months, January through December, add them up and divide by 12 to strike an average for a particular year.

Do the same for the seasonally adjusted numbers published by the BLS for that year. The averages should be very close to the same.