The oddest item on the Bush administration's now-you-see-it, now-you-don't list of possible new taxes is a securities transfer tax -- a levy of perhaps half a percent on the sale of stocks, bonds, etc. This would raise several billion dollars. The impetus apparently comes from Treasury Secretary Nicholas Brady, who is obsessed with the idea that American finance is plagued by too much wasteful speculation and "short-term time horizons."

The Bush proposal for a capital gains tax cut is also intentionally biased against short-term investments, reserving its greatest blessings for those held for more than three years. Still, it is hard to reconcile the two tax ideas. On the one hand, the administration wants to cut the taxes on your net profits when you sell a stock. On the other hand, it wants to start taxing your gross revenues.

And not just that. One alleged benefit of the capital gains cut is that it will "unlock" old investments, encouraging investors to redirect their money in more efficient ways. Indeed the administration's rosy predictions that a capital gains cut would increase tax revenues in the early years are based entirely on people cashing in old investments, not on the incentive for new ones. Yet the transfer tax is designed to discourage exactly this type of "churning."

What sensible people want from a tax is to raise as much money as possible with the least disruption to the economy. (What good people want from a tax, as well, is to take more from those who can more easily afford it. But that is not our sermon today.) A securities transfer tax, deliberately designed to create friction in the wheels of commerce, is a strange bird to be hatched by a Republican administration. Does the government know better than the free market what "time horizons" the economy ought to have?

Lamentations over "short-termism" often confuse three different issues. One is whether U.S. corporations focus too much on this year's profits, slighting projects with distant payoffs such as research and development. Two is whether financial markets are too tumultuous, with traders preferring the quick in-and-out to a more extended investing relationship. Three is whether we as a society are consuming too much of our resources while saving too little for a more prosperous future.

Number three seems undeniably true to me, but that is a moral judgment, not an economic one. If Americans decide to live more for today and less for tomorrow, the result -- through a chain of causation involving interest rates, money managers, etc. -- will be that General Motors invests less in developing robots or whatever. That is financial markets working correctly, not incorrectly, and "throwing sand in the gears" (a phrase of Nobel economist James Tobin, who favors it) is no solution.

As for too much "churning" in financial markets, economists Lawrence and Victoria Summers of Harvard point out that the volume of financial transactions has exploded. In 1960, 766 million shares traded on the New York Stock Exchange all year. By 1987, the exchange was averaging 600 million shares a week. The Summerses estimate the total cost of operating the securities markets at $75 billion a year, or one-fourth of total corporate profits. And that includes the cost of using up some of America's best minds in the buying and selling of shares. The contribution of financial markets in efficiently allocating capital is worth something -- but is it worth that much? The Summerses compare a lot of the effort invested by financial traders to people standing up to see better at a football game: it leaves them all, on average, no better off.

Lawrence Summers and Andrei Schleifer, an economist at Chicago, have developed a set of fancy theories to explain why financial markets have a built-in bias toward the quick payoff. It has to do with the interplay between the market professionals and the "noise traders" -- meaning idiots like you and me. What they can't explain nearly as well is why this short-termism among the paper shufflers should have any impact on what really matters, which is the time horizons of corporate managers.

After all, even if I only plan to hold a stock for two weeks, I will be selling it to someone else who will be selling it to someone else and so on. How much I think these people will pay affects how much I will pay. The prospective value of a stock 10 years from now ought to have the same effect on its value today, irrespective of how often it changes hands in the interim. Even assuming that corporate managers care only about raising their stock price in the short run, there is no obvious reason why stock market churning should discourage them from projects with long-term payoffs if those projects otherwise make sense. Shleifer has concocted some non-obvious reasons, but he admits they are merely "plausible, but not necessarily true."

Apart from all the arcana, there is the question whether a securities transfer tax would drive the financial industry overseas. Most big industrial countries already have such a tax -- Japan raises about $12 billion a year this way -- but the trend is toward reducing or eliminating them, as finance goes global and markets have to compete for business.

The transaction tax, though, is more interesting for what it says about the Bush administration than for what would happen if it was enacted. It certainly wouldn't do much good, but it probably wouldn't do much harm, either. I would just like to hear an explanation from someone up there about what exactly they think the point of it is, and how it fits in to their general economic world view, if any.