Elimination of the preferential tax treatment of capital gains should be part of a comprehensive effort to widen the tax base by reducing tax expenditures.

This suggestion may seem more than a little quaint in this day and age. It was, after all, less than a year ago that politicians of both parties were stumbling over one another to see who could cut taxes on capital most deeply. But before you write me off as a lunatic, please hear my reasons.

First, I think there is little doubt that the special treatment of capital gains has led to more complexity in the tax code and more tax avoidance than any other provision in the law. Tax shelters proliferate. Countless resources are spent devising ways to convert ordinary income into capital gains, concocting schemes to ensure that capital losses are short-term while capital gains are long-term, and in other socially unproductive activities. These exercises in ingenuity undermine confidence in the tax system and create an atmosphere in which "Beat the Tax Man" is the national sport. The resources could be better used elsewhere.

Second, because it is mainly the rich who receive capital gains, the equity of the income tax is seriously compromised by taxing gains at lower rates than earnings.

Third, when some types of income are taxed more lightly than others, patterns of saving and investment are distorted, and may bear little resemblance to underlying economic realities. If the economy is to operate efficiently, new investment must flow to the places where it will be most productive, not to the places where it will be taxed most lightly. So preferential tax treatment of capital gains is bad supply-side economics.

All of these problems would disappear overnight if capital gains were taxed like ordinary income.

With no difference in their tax treatments, there would be no incentive to transform ordinary income into capital gains. Most purveyors of tax shelters would go out of business. Investment would be guided by productivity, not by tax avoidance. And public faith in the integrity and equity of the tax system would be shored up.

But isn't there another side to the story? Aren't there good reasons for giving capital gains favorable tax treatment? Three principal reasons for tax preference have traditionally been given.

The first holds that, because capital gains tend to come in large lumps, they will inadvertently be taxed at higher rates than other forms of income under a progressive tax system. For example, someone earning $100,000 per year in taxable income will pay less tax than someone who earns $90,000 per year and realizes a $100,000 capital gain every 10 years.

This argument has always been a dubious justification for lower tax rates on capital gains because the bunching problem is more naturally handled by income averaging. But whatever merit it once had has been largely destroyed by the 1981 tax cuts.

The tax cuts, by bringing all the top bracket rates down to 50 percent, have ensured that most capital gains will now be received by taxpayers who are almost always in the 50 percent bracket. Under the current tax structure, for example, the hypothetical family that earns $90,000 for nine years and then $190,000 in the tenth year will pay exactly the same tax as the family that earns $100,000 every year--if the capital gain is taxed like ordinary income.

A second argument maintains that we must give tax preference to capital gains in order to encourage risk-taking and capital formation. Even if we accept the notion that investment should be subsidized, the tax code already contains numerous incentives for capital formation, including several that were added last year. Besides, the capital gains loophole is a particularly inefficient way to promote capital formation since it parcels out "tax goodies" indiscriminately to those who speculate in gold, works of art or stocks and bonds as well as those who invest in productive business activity.

A third argument for lower tax rates is to undo the overtaxation of capital gains that inflation would otherwise engender. Again, an example will clarify. Suppose I buy $1,000 worth of stock, hold it for 10 years, and then sell it for $2,000. In the eyes of the law, I have earned a $1,000 capital gain. But if prices have doubled in the interim, I have gained nothing at all in terms of purchasing power. And if prices have more than doubled, I have actually suffered a loss--my $2,000 can buy less than the $1,000 I originally invested.

The problem is genuine, but taxing only 40 percent of capital gains (as we now do) is not the solution. What we should do instead is extend the principle of indexing, which was introduced last year for bracket rates starting in 1985, to the definition of capital gains.

Under a fully indexed tax law, only gains over and above the general rise in the price level would be taxed. Price appreciation that fails to keep pace with inflation would be treated as a capital loss.

But, you may ask, doesn't excluding 60 percent of all gains from taxable income compensate for inflation in a rough way? The answer is no.

Research has shown that the current system is a clumsy and ineffective way to compensate for inflation. Some investments are undercompensated, which means that some activities are harshly overtaxed while others are generously subsidized. Changing the tax code so that it taxes only real capital gains, but taxes them at the same rates as ordinary income, would be both more equitable and more efficient.

So the choice is clear. We can keep the tax treatment of capital gains as it is--full of legal complexities, inequities and economic distortions. Or we can shift to a simple, clean and equitable system in which real capital gains are taxed like any other type of income, but in which illusory capital gains produced by inflation are not taxed. Which choice is really more conducive to capital formation and productivity?