SUPPOSE that you are the attorney general, confronted with the steel merger decision. Two large corporations want to fold themselves together - and their steel-making subsidiaries are, respectively, the seventh and eighth largest producers in an industry already heavily concentrated. What do you do?

First, before you get to the legal issues, you have to make up your mind about a question of fact. One company, the Lykes Corp., says it is in immediate danger of going bust. That would be the end of its Youngstown Sheet and Tube subsidiary. Youngstown, you remember, is the company that shut down a number of obsolescent mills in Ohio last September, with a loss of some 4,700 jobs in one-industry towns that may never recover. Collapse of the parent company would mean 6,000 more permanent layoffs in the same region. The first thing that you, as attorney general, need to examine is whether the danger of failure is real. The evidence strongly suggests that it is both real and imminent.

Then you need to consider the recommendation of your own Justice Department's antitrust division, which sharply opposes the merger. That advice is not necessarily controlling. There are cases in which it is the antitrust division's responsibility to make the narrow case against the merger and leave the broader considerations to the attorney general. If, for example, the merger were forbidden and Youngstown folded, its share of the market would very probably go to the bigger survivors - U.S. Steel, Bethlehem, Armco and the rest. Is it not better to have two middle-size producers merge, to stay in the battle with the giants, than to have one of them fall away and leave the giants a little bigger?

Attorney General Griffin Bell followed roughly that logic, we surmise, in arriving at his decision Wednesday to let the Lykes Corp. merge with the LTV Corp., which owns Jones and Laughlin Steel. The merger carries certain dangers to competition, but the alternative carries even greater dangers. Beyond the legalities, there is a political threat. Foreign steel is now the strongest force for price competition in steel, and the only serious check on the domestic industry's built-in tendency toward inflation.

The import's share of the U.S. market has been rising lately, and the American companies bitterly attribute it all to unfair subsidies by foreign governments. There's some of that. But there's another explanation closer to home.The U.S. companies have done a poor job of holding their costs down, particularly their labor costs. Wages in steel and coal have risen faster over the past decade than in any other major American industry, which helps explain the foreign producers' inroads. Steel's labor-management drive for import quotas gets a lot of sympathy in Congress. The collapse of a major company could result in sweeping protectionist legislation doing enormous harm both to American foreign relations and to the struggle against inflation. Mr. Bell is right not to run that risk.

You will notice that, as an industry becomes more heavily concentrated, the public policy options become more unattractive. There's a point at which antitrust enforcement paradoxically begins to look very much like government-imposed market-sharing - the kind of arrangement known as a cartel. To keep dominant competitors from getting more dominant, the government in increasingly forced to prop up weak competitors with special dispensations. Sometimes, as in the present case, the government is left only with a choice as to which kind of concentration is least objectionable.

The steel industry brings to mind the 10 little Indians. Last week there were nine American steel producers with annual sales over a billion dollars each. Two of them merged, and then there were eight.