The Federal Reserve, concentrating on the still sluggish U.S. economy and the value of the dollar on foreign exchange markets, has essentially abandoned its money supply growth targets rather than take steps that would increase interest rates.

A majority of Fed policy makers apparently is convinced that, for reasons they don't fully understand, the traditional link between money growth, the economy and inflation has been broken, at least temporarily. As a result, they have not clamped down on money growth even though it has soared far above the Fed's target range.

"This year you just can't say much about M1," declares Fed Gov. J. Charles Partee. "It is spectacular. It is stronger by far than expected . I have no explanation for it and I don't think anyone else has."

This is the second time in less than three years that the relationship between the most closely watched measure of money, M1, and the economy has gone astray. M1 is the measure of currency in circulation and money available in checking accounts -- in other words, the money that can readily be used in transactions.

In the latter part of 1982, Partee recalls, "the policy makers decided . . . to say the economy is so bad, let's not pay any attention to M1."

The Fed could have stuck with its targets out of concern that a rapid money supply increase, even though it would boost the economy, would also mean more inflation. But in 1982, they put those concerns aside, and the gamble worked.

After the Fed temporarily dumped M1 as a guide to policy in 1982, there was a strong recovery and little, if any, acceleration in inflation, despite such rapid money growth that monetarist economists predicted it would set off a new round of sharp price increases. That experience, Partee says, made it "not so hard this time" to abandon the targets.

Now there is another "temporary aberration that we don't understand," he noted. "There probably will be a return to a more stable relationship, but every time you have one of these episodes, M1 becomes a less useful indicator."

The Federal Reserve's vice chairman, Preston Martin, and Henry Wallich, a governor, have also expressed concerns about M1's reliability.

Even if the decision to abandon targets turns out well a second time, the Fed still has a major problem on its hands: It has no very good candidate to replace M1 as a tool for influencing the economy. The money supply is something the Fed can control reasonably well over periods of several months. And as long as what the Fed did to M1 had a predictable effect on the economy, that made it extremely useful to the policy makers there. They could aim at that intermediate target with some confidence about the consequences on the economy and inflation -- the central bank's ultimate targets.

At the moment, Partee and a number of other Fed officials say they are looking primarily at the economy itself rather than any sort of intermediate target, such as money or credit growth, the value of the dollar or interest rates to determine what actions they should take.

In fact, several members of the Fed's policy-making group, the Federal Open Market Committee, indicate they are watching just about everything they can for clues as to what is going on in the economy and what the monetary policy response should be. Partee specifically says this is not the same thing as using changes in current-dollar gross national product as a guide to what the Fed should do, an approach a substantial number of economists have suggested the Fed should take.

Instead, he says, "the intermediate targets are the indicators, the signs in the wind" that show whether the Fed's general projections for economic growth and inflation are on track. "But you get whipsawed there, too," Partee adds. "You never know where you stand . . . and it leaves you analytically more vulnerable. Any move in monetary policy based on these indicators leaves you subject to criticism" because officials cannot say with confidence that the moves will produce the results they expect.

"That may make us, as an institution, move more gingerly, and that could leave us behind," Partee worries. Being behind means making policy responses to events that come too late or are too limited to keep the economy and inflation headed in the right direction, he explained.

This sort of eclectic approach to setting monetary policy is anathema to monetarist economists, who believe that changes in M1 do determine future changes in current-dollar GNP and prices.

Some monetarists, including Council of Economic Advisers Chairman Beryl W. Sprinkel, warn that the links between money and the economy and prices have only been bent, not broken.

Unless money growth is quickly curbed, Sprinkel said last week, there will be a serious resurgence of inflation next year. A group of monetarist economists known as the Shadow Open Market Committee issued the same warning last week after one of its semi-annual meetings in New York.

Money growth must be brought down, the committee argued, even if the process pushes the economy into a mild recession. Once inflation takes off, the cost of stopping it will be a more severe recession, the committee said.

Fed officials admit that they are uneasy about the possibility that this year's rapid money growth could generate more inflation in the future. But in the absence of any sign so far that inflation is about to get worse, they have chosen to keep short-term interest rates stable rather than risk damaging the economy's immediate prospects.

Most financial analysts believe that the FOMC, which meets this week, will make little if any change in the current setting of the monetary policy dials. A number of FOMC members are uneasy about the rapid increases in M1 even though they have disregarded them and at least one member, Robert Black, president of the Richmond Federal Reserve Bank, has not abandoned M1 as a policy guide. Even so, there is only a modest chance that the FOMC would choose to ease policy, the analysts say.

However, given the uncertain signs about whether economic growth is going to rebound from the 2 percent pace it has averaged over the past year, and given the new commitment by the Reagan administration and the Fed to try to reduce the value of the dollar, there is even less likelihood that the central bank officials will decide to tighten policy and let interest rates rise, the analysts believe.

The central bank's normal method of slowing money growth involves making cash less readily available to financial institutions. That usually increases interest rates, reduces the demand for credit and ultimately slows money supply growth.

So far this year, the money supply has expanded so rapidly that 1985 may turn out to have had the fastest growth of M1 in history, according Jerry L. Jordan, a member of the Shadow Committee who is vice president of First Interstate Bancorp of Los Angeles and a former CEA member. That is particularly remarkable -- and particularly dangerous, say the monetarist economists -- in a year in which inflation is running at only a 4 percent rate.

Since the fourth quarter of last year, M1 has increased at a 12 percent annual rate, well above the 4 percent to 7 percent target range the Fed set last winter.

In July, the FOMC decided more or less to ignore the unexpectedly rapid rise in M1 in the first half of the year. It did so by setting a new target range of 3 percent to 8 percent for the second half of this year. Growth was to be calculated not from last year's fourth-quarter average, but this year's second-quarter average.

At the same meeting, according to the policy record released later by the Fed, the FOMC said it expected its day-to-day operations in the money markets to be consistent with an increase in M1 at a 5 percent to 6 percent rate between June and September.

Instead, M1 has shot up at more than a 15 percent rate since June, triple what the Fed expected.

Meanwhile, current-dollar GNP continued to rise more slowly than M1, reversing again a relationship that had remained more or less intact for decades. During the 1950s, 1960s and 1970s, current-dollar GNP generally increased about 3 percent a year faster than M1.

"The link between money and GNP which previously had been short and predictable suddenly shifted in 1982 to being long and variable," recently wrote Michael W. Keran, chief economist of Prudential Economic Research and a former research director of the San Francisco Federal Reserve Bank.

Instead of GNP growing 3 percent a year faster than M1, Keran said, "nominal GNP has grown at about 1 percent slower than M1 since 1982. Thus, in the short run, the same amount of growth in M1 has not put as much pressure on nominal GNP or aggregate demand as before. That means in the long run the rise in the money supply does not put as much pressure on inflation . . . The rate of inflation since 1983 has been lower than that which would have been predicted by the past growth in the money supply." (See chart.)

Keran, like many other economists, believes that the steady 3 percent ratio of the rise in GNP to the increase in the money supply was primarily the result of a general upward trend in inflation and interest rates. Until relatively recently, none of the components of M1 were interest-bearing and most of them -- currency and regular checking accounts at commercial banks -- still are not. Thus, when interest rates rose, the implicit cost of keeping money in a non-interest bearing form also went up and both individuals and businesses were encouraged to economize on their M1 balances.

The upward trend in inflation and interest rates was broken by the 1981-1982 recession, and so was the regularity of the link between money and the economy and prices. Keran thinks that in a world of stable inflation, nominal GNP would not rise more than 1 percent faster than money, with that reduction in the margin due to technological advances in the financial world.

In addition, Keran argues, if the Fed continues to tighten its policies to head off inflation whenever the economy appears to be surging ahead too strongly, changes in M1 will not be a reliable guide to where the economy will be going.

Monetarists such as Sprinkel and the members of the Shadow Committee -- he is a member but on leave -- agree that the old relationship between money and current-dollar GNP is no longer as reliable a guide as it used to be. And they agree that the likely trend in the the ratio of year-to-year increases in the GNP to money is less than 3 percent.

But they argue vigorously that there is no reason to believe that the trend in the ratio has turned negative, that money will continue to rise faster than current-dollar GNP. Almost no matter what one took for the new trend, declared William Poole of Brown University, another former CEA member who is on the Shadow Committee, there would still be no way to justify the sort of money growth the Fed has permitted this year.

Some members of the Fed's policy-making group, on the other hand, found it harder to justify raising interest rates to slow the growth of money at a time when production of goods and services was rising only very weakly, inflation was stuck at or below 4 percent and the nation's currency was at an unsustainably high level on foreign exchange markets.