We have reached a nadir of dishonesty in the budget debate when some Republicans justify their opposition to cuts in Medicare benefits by defining such cuts as a "tax increase on the elderly." We also have reached a nadir of illogic when Democrats propose to exchange higher top-bracket tax rates for a cut in the capital gains tax.

Although the rich should certainly pay more taxes, there is nothing wonderful about higher tax rates as such. Higher tax rates do discourage economic activity. There are plenty of ways to sock the rich without raising marginal rates -- by closing loopholes instead. Sen. Lloyd Bentsen's idea of raising the "alternative minimum tax" -- the tax you pay if you take too many deductions -- is a good technique for doing this through the back door.

To raise marginal rates and then give back the money through a capital gains break is doubly crazy. Fairness aside, the problem with a capital gains break is that it violates the principle of tax neutrality: letting the market, rather than tax considerations, dictate economic decisions. Neutrality is a good capitalist notion: it holds that the free market knows better than the government how resources can be used most efficiently.

Raising the ordinary tax rate while cutting the capital gains rate makes the distortion even worse. The bigger the gap, the bigger the opportunity to create tax shelters. Many of the Democratic Party's contributors and fundraisers are in exactly the lines of business -- real estate, oil and gas, entertainment -- that would benefit most from the combination of a higher top tax rate and a new capital gains break. I say no more.

The New York Times advocates "a tax cut that is limited to new investment in venture capital and other risky start-up companies." Many others are tempted by the idea of sorting out the "new" or "risky" investments and only giving a tax break to these. But the problems of definition -- and the certainty that clever lawyers will bend them out of shape -- merely illustrate the conceptual fallacy. Why give a break to a new company but not to an established firm that wishes to branch out into a new product line? And so on.

The tax code already gives a big break to capital gains compared with other forms of investment income. Capital gains accrue tax-free until you choose to liquidate the investment. By contrast, you pay tax on interest income every year. For an investment yielding 10 percent annually and held for 20 years, the after-tax profit on a capital gain is almost 50 percent larger. And capital gains on investments held when you die escape tax completely -- a $5 billion annual loophole.

To index capital gains for inflation -- another frequent suggestion -- would increase the bias against other forms of investment. Like a capital gain, interest income is partly compensation for the erosion of your principal. And if you start indexing the tax on profits, you also should index the deduction for borrowing costs. Part of the interest paid on debts also merely compensates for the erosion of principal. If people can borrow money and deduct all the interest, then invest it but pay tax only on the profit above inflation, they are going to screw up the economy and the federal treasury but good.

Some people -- most notably the editors of The Wall Street Journal -- argue that the limit on deducting capital losses makes the tax on capital gains unfair. Here we reach the farthest shores of capital gains dementia. Watch closely. Capital losses are fully deductible each year against capital gains of the same year. Beyond that, you can deduct up to $3,000 a year of capital losses against your ordinary income. And you can "carry forward" any losses beyond $3,000 for use in future years.

So ask yourself: Who is hurt by the limit on capital losses? Two groups come to mind. First there are those who, year after year, lose more than $3,000 on their investments. Such people are surely to be pitied. But they are hardly the key to a thriving economy. And the capital loss obsessives seem remarkably well-dressed to pass for such losers.

In truth the capital loss limitation affects a larger and more prosperous group: those who can arrange their affairs to generate capital losses while actually doing quite well, thank you. The key point about capital gains and losses is that you can choose when to take them. With no limit on capital losses, it would be easy to create transactions that generate huge gains and losses in roughly equal amounts. You take the loss (thereby reducing the tax on your other income) and postpone the gain -- perhaps indefinitely. Et voila!

In one of its recent screeds on this subject, The Journal gave this game away. "If you allowed full deduction of losses, the capital gains tax probably wouldn't raise any revenue and might as well be abolished," it sneered. But surely The Journal does not envision an economy with no net capital profits. That would be an ironic result of taking The Wall Street Journal's advice. What The Journal actually envisions is an economy with lots of profits, but no tax on them. There's nothing wrong with profits, but there's nothing wrong with taxing them either.