The sudden fashionability of "industrial policy" only demonstrates how little we have learned from the sobering example of Great Britain. Do we have to repeat all of Britain's mistakes?

Industrial policy, in case you hadn't heard, is Washington's newest economic vogue. No one seems to know quite what it means, but there is much scurrying about to find out. Mostly, this attests to a sense of economic helplessness and the capital's enthusiasm for labels. A clever label excites reporters, stimulates congressional hearings, creates executive task forces and employs speechwriters.

But the only new thing here is the label. Government always has made industrial policy. Not just one policy, but dozens. Government rarely has a single policy on anything. Too many interests want to get their hands into too many pots.

So you want industrial policy? Try these:

When Congress decided to rescue Chrysler Corp. last year, it made industrial policy. In effect, it sanctioned a three-year labor settlemment with a pay increase of about 35 percent for workers at a failing firm because the United Auto Workers was powerful enough to extract that deal.

When the Environmental Protection Agency set emission standards for electric utilities in 1979, it made industrial policy. By requiring "scrubbers" on all plants to remove sulfur from exhaust fumes, the EPA gave utilities little incentive to burn low-sulfur western coal instead of high-sulfur eastern coal. Burning western coal might be more efficient, but eastern politicians and coal interests (among others) were powerful enough to get that deal.

When Congress enacts heavy subsidies for gasohol -- the legislation is nearing approval -- it will make industrial policy. Little matter than gasohol (a mixture of grain alcohol and gasoline) will reduce oil imports only slightly or that investment in the distilleries may be wasteful. Farm interests were powerful enough to get that deal.

No one should imagine that policy will stop being made this way: Political life is too fragmented and the competing interests too powerful. The main mystery about the new industrial policy is whether any group, or groups, will be able to turn this particular label to private advantage.

Whatever happens, we should not expect much good.

Talk to people who are studying industrial policy, and they usually will mention one of two ideas. Either the United States must reindustrialize -- that is, revive basic industries such as steel and autos -- or government must pick tomorrow's industrial winners and shove them in the right direction.

These concepts are essentially contradictory. If we're interested in promoting tomorrow's industries, autos and steel will be at the bottom -- not the top -- of the list. But these sectors (and others like them) boast the largest established constituencies and, therefore, stand the best chances of exploiting any industrial policy.

This is likely to be bad for them as well as the country. Given the vast size of the U.S. market, basic industries aren't about to whither before imports. But they no longer enjoy a product of technological monopoly, and their market share will depend on their competitiveness. Steel and auto wages already are among the highest in industry and are increasing more rapidly than average. If industrial unions insist on continuing this -- as they probably will -- they simply will buy themselves fewer jobs and a lower market share.

Government assistance or import protection is likely to perpetuate this destructive behavior by distracting attention from fundamental problems. Even with government help, products get priced too high, profits get squeezed, and innovation and investment suffer. that's precisely what happened in Britain, where the government liberally assisted industries with the largest constituencies.

The same caveat applies to business executives. Both here and in Britain, basic industries adapted only slowly to the new reality of worldwide markets. With their large domestic market, American companies are in a stronger position, but executives need to spend their time worrying about competitive -- not political -- realities.

Actively promoting tomorrow's industries is equally silly. Government is good at managing massive technological projects for which the goal is clear-cut and no private market exists: making the atomic bomb and landing on the moon. But for the economy at large, the sweep of technological change and the development of markets are too complicated and unpredictable to be managed effectively.

One of the superficial attractions of industrial policy is the belief that it has been employed successfully abroad, particularly in Japan. If we could only emulate the Japanese . . .

The danger here is that we will make a myth of Japan's economic success. Americans see the products of Japan's modern, export-oriented industry. Its cars, electronic products and steel are among the best in the world, if not the best. But Japan also has a huge traditional sector of small shopkeepers and lacks America's efficient agricultural base. Despite enormous advances, Japan's living standards remain well below America's.

Nor should we romanticize Japan's industrial and export success as the result of a wondrous partnership between government and industry. The stereotype of Japan Inc. is a vast oversimplification. Japanese ministries quarrel among themselves, and companies quarrel with government. Successful companies owe most of their success to their own efforts: a hunger for new technology, a diligent labor force, high-quality management and thorough market studies. If we think otherwise, we are deluding ourselves.

Japan has -- and we lack -- a broad consensus that successful enterprise is in the national interest. We need to give people and firms rewards for taking risks; major changes in tax policy are needed to redress inflation's impact on investment and profits. But this is much different from handouts to specific industries. We don't need that. Nor do we need to tie up our risk-takers in the bureaucratic process of trying to decide what an industrial policy should be. It's simply more makework.