Of the economic forecasts made for this year, the Reagan administration's is among the most pessimistic. Martin Feldstein, chairman of the Council of Economic Advisers, acknowledging the serious weakening of the economy at the end of last year, persuaded the administration to estimate real economic growth at a tiny 1.4 percent in 1983 over 1982.

For the fourth quarter of 1983, over the fourth quarter of 1982--the kind of comparison most experts consider a better indicator of the economy's progress--Feldstein's forecast implies only a modest 3 percent rate of growth, far below the typical surge in a recovery period.

Obviously, the slower the real growth rate, the longer both unemployment and budget deficits stay embarrassingly high.

Supply-siders have already accused Feldstein and OMB Director David Stockman of "cooking" the numbers so as to push a reluctant Ronald Reagan into supporting tax increases down the road that will ruin supply- side economics.

Now comes Paul Samuelson, the liberal Democratic Nobel laureate from MIT, to suggest that administration economists deliberately underestimated the extent of probable recovery this year, "perhaps to stampede (Reagan) into action on the deficit . . . and to soften up the Congress for cuts that will be necessary in entitlement (spending) programs."

On Metromedia's "Panorama" TV show last week, Samuelson said he expects a growth rate of at least 4 percent in the first year of recovery, less than the 7 percent typical of most of the postwar recoveries, but well over what the administration is now predicting.

"I think they're pitching us a low ball now," Samuelson said, suggesting that the administration is playing "a Machiavellian game" to enable it to say, later, "'See, it's working out even better than we said; you were wrong to be so impatient on Reaganomics.'"

Albert T. Sommers, longtime chief economist of the business-oriented Conference Board in New York, agrees that something funny is going on with the administration's basic economic assumptions. In light of business conditions, which he says will yield close to 5 percent real growth in the fourth quarter of this year, Sommers labels Feldstein's forecast "impossible. . . . 1983 will be a substantially better year."

As a rough rule of thumb, the Congressional Budget Office last September estimated that each additional 1 percent drop in economic growth this year would add $28 billion to the deficit in fiscal 1984. At that time, the CBO projected a 3.6 percent growth rate in calendar 1983 and a 3.7 percent rate for calendar 1984, raising the deficit for fiscal 1984 to $152 billion. Thus, a further two-point drop in the growth rate to Feldstein's 1.4 percent could--by this admittedly imprecise rule of thumb--run the deficit to the $200 billion mark in fiscal 1984.

Sommers says he doesn't know whether the administration projections "are a forecast, a target or a political and economic ploy." But he argues that Reagan, by putting forward a gloomy economic projection, risks having it fulfilled. That is, the Fed may be forced to key its policies to expectations of a wider deficit, and boost interest rates again.

But Feldstein has his defenders. Joseph Pechman, director of research at the Brookings Institution, says the consensus there is that the Feldstein forecast is "quite realistic." Pechman sees a dismal picture--high unemployment, weak business investment and export sectors, state and local governments pressed to the wall.

Yet, there are quite a few substantially more optimistic forecasts out there, in addition to Sommers' and Samuelson's. One comes from Charles B. Reeder, chief economist for the Dupont Company. He won an award last year for being closest, in his forecast, to the disappointing performance of the economy in 1982.

"Today, it is possible that forecasters are underestimating the positive forces in the economy," Reeder says.

With the administration's excessive optimism of last year on his mind, Feldstein may be less Machiavellian than Samuelson and Sommers suggest. Perhaps he just wanted to play this one safe. But as Sommers warns: "If the government forecast is right, there will be no question about where the responsibility for such poor performance lies: it will lie with government policy."