THE FIGHT over general revenue sharing began in earnest last winter when the administration and Congress got serious about balancing the budget. For each of eight years, the revenue-sharing program had, with little controversy, transferred several billion dollars of federal revenue to state and local governments for their own use. Now congressional leaders, stung by the demands of over 30 states that they balance the federal budget, have begun to question the wisdom and necessity of turning over federal taxes to state governments, many of which had been cutting taxes and running substantial surpluses.

While most state budget surpluses have disappeared with the onset of the recession, the basic revenue-sharing question has not. Why should the federal government take the heat from the citizens of the various states for raising their taxes -- and then turn the money back to the governments of those states to spend as they please? It's been a long time since we heard about the "fiscal dividend" -- that ever-receding surplus of federal revenues that economic growth and the end of the Vietnam War were supposed to produce, and that, it was argued, could be well used in supplementing the more limited taxing powers of states and localities. And while federal resources have come under increasing pressure, state resources have grown rapidly. All but six states now have income taxes, and all but five have sales taxes.

There are, of course, other arguments for sharing federal revenues. One is that there is a national interest in ensuring that at least the basic needs of all citizens are met, even if those citizens live in states that are unwilling or unable to help them when they are in trouble. But this argues for earmarking the money for particular purposes and establishing standards, as in the so-called "categorical" federal grant programs. By contrast, while states and localities may designate the money for worthy causes, the premise (and, for governors and mayors, the chief charm) of general revenue sharing is that it really has no strings attached.

A case certainly can be made for consolidation of the more than 500 special-purpose federal grant programs that create inefficiency and distort spending priorities at all government levels, but thatS another issue worth a separate discussion.

A second possible argument would be that revenue sharing produced some general equalization of needs and resources among states. This assumes a rough comparability between the distribution of need and the distribution of shared revenues. Judged by its durability, the current formula, an ingenious compromise engineered nine years ago by Wilbur Mills, has certainly been a political success. But a conparison of the percentage of revenue-sharing funds received by each state with the percentage of federal revenues paid by the citizens of that state casts doubt on the program's fairness. Judged by this measure, most of the hard-pressed older industrial states, such as Michigan, Illinois and Ohio, have in recent years been net losers in the revenue-sharing game, while such resource-rich or growing states as California, Arizona, Montana and Alaska have been net gainers.

Both the House and the Senate have now passed a bill extending revenue sharing but suspending authorization for the state, but not the local, portion for next year. Thereafter, money for states would have to be voted each year, not effectively guaranteed as it is for localities. This is probably the best compromise that can be achieved for the moment until Congress can face up to what it really ought to do. That is to abolish general revenue sharing, at least for states, as a program that is both wrong in principle and unfair in practice.