Sometimes the most sensible solution to about three-fourths of Washington's problems seems to be this: Ship half of the city's lawyers to some Caribbean island where they would be publicly supported -- in the splendor to which they are accustomed -- in exchange for not practicing law. Although expensive, this would be a lot cheaper than keeping them in Washington, where they continually generate mischievous make-work.

We are inspired to this satisfying fantasy by a recent excursion into the wondrous world of antitrust law, as practiced by the Federal Trade Commission. The FTC is attempting to stretch the antitrust laws considerably beyond the conventional notions of price-fixing, monopoly and oligopoly (a market dominated by a few sellers).

The case at hand involves Exxon Corp. and its announcement last year of the development of a new device that conceivably could save up to the equivalent of one million barrels of oil (about 2.5 percent of current energy use) by 1990. The claim may be exaggerated, but you might have expected government to exhort Exxon onward: Show us your stuff, Exxon; sell those energy savers.

More fantasy, of course. To promote commercialization of its energy saver, Exxon proposed buying Reliance Electric Co. for $1.2 billion. Any time a company as big as Exxon (1979 revenues: $84.9 billion) wants to buy anything, the FTC's lawyers get indigestion. Naturally, the agency opposed the sale in court.

All this illustrates how the legal process has become a large cost of doing business. Lawyers, like bureaucrats, create work for each other; there are depositions to be taken, motions to be filed, briefs to be answered. Frivolous suits mean senseless costs. In this instance, the FTC's involvement is especially ironic: The government says it wants to promote innovation and competition, but in the name of promoting them the commission is actually doing the opposite.

Exxon's interest in Reliance, a major manufacturer of motors, stemmed from the nature of its energy saver, which is known as an "inverter." When attached to an electric motor, the inverter allows the motor to vary its operating speed and energy consumption. All this may sound dull, but the more than one billion electric motors estimated to be in U.S. factories and buildings now account for about half of electricity consumption. And a lot of that energy is being wasted.

The reason is simple. Although inverters existed in the past, they were expensive. Cheap electricity meant that power reductions usually were achieved mechanically. For example, many electric motors operate pumps that push liquids through chemical plants, oil refineries and factories of all sorts. If less liquid were needed, it was cheaper to tighten a valve than slow the motor and pump.

Exxon now says the use of semiconductors has reduced the cost of inverters dramatically. With power costs rising, Exxon figures that the inverter market will jump from less than $100 million to $1 billion or more by 1990. The two models it tested on refinery pumps cut energy use by 27 percent and 38 percent.

Even the FTC agrees with Exxon that the purchase of Reliance initially would promote more rapid commercialization of the inverters. Reliance already has an experienced marketing and service organization, which knows customers and has their confidence. The FTC's reaction to this is essentially: So what? Denied the opportunity to buy Reliance, Exxon (says the FTC) would enter the market itself. Ultimately, there would be more competitors.

This is kindergarten economics. No one can predict the shape of changing future markets on the basis of present knowledge. Old competitors fight to stay alive, new ones struggle to capture sales. Nor can anyone really prophesy the value of new technology. Some electric companies apparently are skeptical of Exxon's claims. Let them gamble on that hunch. The "public interest" -- which the FTC is supposed to guard -- lies in getting the combat started as rapidly as possible, not adhering to some silly arithmetical standard of competition.

To think that Exxon and Reliance suddenly will dominate the new market is to disregard the whole history of semiconductor technology, which is that lower costs expand sales and attract competitors like flies. Potential competition is abundant. Major motor manufacturers such as General Electric Co. and Westinghouse Corp. easily can stand toe-to-toe with Exxon. Borg Warner Corp. not only just has announced new inverters that sell at one-third of the price of previous models, but also is incorporating inverters in air conditioners. Can the Japanese be far behind?

All this brings us back to law and the lawyers. One of the traditional functions of civil law has been to settle disputes between parties, and lawyers were supposed to assure that each side got adequate representation. But more and more, lawyers abandon this largely passive role. Motivated by financial self-interest, their own sense of self-assurance or a thirst for political power, they create demand for their own services by stretching the interpretation of existing laws or promoting new ones.

The FTC's suit against Exxon involves an expansive interpretation of the Clayton Act. Although the FTC couldn't convince a federal judge to block the merger, he did order that Reliance segregate one of its principal activities pending completion of an administrative antitrust proceeding at the FTC. This could take years. Whether it hinders Reliance's commercialization of Exxon's inverters is unclear. But the time, cost and uncertainty punish a company for pursuing the type of industrial innovation that the federal government says it favors.

There is a larger lesson. The legal process, which attempts to determine right and wrong at a fixed moment in time, is a poor instrument for coping with continuing economic change. The increase desire to control some side effects of those changes -- on the evironment, for example. But if we subject too much of the economy to review, we only will make lawyers wealthy and more powerful without improving either the economy or public respect for the law.