It's official: The economy is in its seventh postwar recession, and the downturn began formally in January.

That's the pronouncement of the National Bureau of Economic Research -- the private, Cambridge, Mass.-based organization that is the generally acknowledged arbiter of business-cycle stages here.

After months of scholarly indecision, the bureau's seven-member Businss Cycle Dating Committee convened in Cambridge yesterday and concluded officially that economic activity in the U.S. "peaked" last January.

"Unless there is an extraordinary sharp and quick reversal," the panel added, January also will mark "the onset of the recession." The bureau has decided the start and finish of each recession back through 1934.

In deciding on the January 1980 date, the committee rejected suggestions from some economists that it mark the recession's start from March 1979. Most key economic indicators have been declining slowly since that time.

However, Martin Feldstein, the Harvard University economist who serves as president of NBER, said in a telephone interview yesterday that the panel had concluded that the March 1979 turnaround was "a false peak."

Feldstein noted that the number of jobs in the economy continued to grow visibly through much of last year and retail sales continued strong."You can't call that a recession of any duration," he asserted.

The NBER president also defended the committee's decision to wait until yesterday -- four months after the downturn began -- to declare the current slide a recession.

"To have done this three months ago really would have put us in the forecasting business," Feldstein said. The bureau traditionally has been cautious in labeling changes in the business cycle.

In a statement yesterday, the panel cited January downturns in three key economic indicators -- industrial production, retail sales (adjusted for inflation) and number of hours worked -- for dating the recession from that month.

It said some other indicators, such as personal income (adjusted for inflation), actually peaked in December, while others, like employment, fell during February. Since February, all major economic indicators have declined.

The decision, ironically, came before the government actually reported any decline in the quarterly gross national product estimate, the broadest measure of economic activity in the United States. Those figures aren't due out until July.

However, the committee said that, in view of the sharp slide in monthly economic statistics, it was "unnecessary to wait" for the publication of the GNP figures. Private estimates show real GNP declining at a 7 percent to 8 percent annual rate.

The committee's decision came several months after the White House formally recognized the downturn as a recession. Indeed, Treasury Secretary G. William Miller declared last year the recession was "half over."

However, Feldstein said yesterday the panel "would have been reluctant" to act any sooner because it didn't believe there was enough evidence to call. "Lots of other people were declaring the recession was half over," he said.

Besides Feldstein, the meeting yesterday included five of the panel's six other members: William Branson, Princeton University; Robert J. Gordon, Northwestern; Robert Hall, Stanford; Victor Zarnowitz, Chicago University; and Geoffrey Moore, former commissioner of labor statistics.

The sixth, Benjamin Fried, a Harvard economist, didn't attend.