They tell the story of the high school football coach, frustrated on the sidelines as his team was losing badly, who shouted out to the huddle: "Give the ball to Leroy!"

Later, the quarterback explained to the coach: "Leroy, he didn't want the ball."

I get something of the same feeling listening to establishment figures debating the Third World debt problem. In Phoenix the other day at a bankers' association meeting, International Monetary Fund Managing Director Jacques de Larosiere bemoaned the fact that the commercial banks were not making new loans to developing countries.

"No debt strategy can succeed without the broad and active participation of commercial banks," de Larosiere said. "A resumption of net new lending at a reasonable pace . . . ultimately provides the best protection for the value of outstanding loans to those countries."

Similarly, Treasury Assistant Secretary David C. Mulford told the same gathering that, "Traditionally, banks have worked with troubled clients because they have believed it to be in their own self-interest. The present international debt situation is no different."

Despite "coaches" de Larosiere and Mulford, the banks, like Leroy, aren't too interested in taking further punishment. The banks seem to be taking the position that there's little point in throwing good money after bad.

This raises serious doubts about the prospects for the U.S. "debt initiative," which its author, Treasury Secretary James A. Baker III, originally described as a three-legged stool. He meant that the plan depended on three equally important elements: economic policy reforms in the borrowing countries that will enable them to pay back their old loans, an increase of $9 billion in multilateral bank lending and a $20 billion increase over three years in commercial bank loans to 15 key countries.

Baker implores Third World borrowers to shape up -- meaning adoption of market-oriented policies -- to be eligible for new loans. Simultaneously, he assures Congress that he won't ask for new World Bank money until the countries start adopting pro-growth policies. And then tells the commercial banks that they aren't expected to kick in until the countries and the development banks play out their roles.

But, according to de Larosiere, 11 of the 15 countries targeted by the Baker plan already are undertaking the kind of growth-oriented economic reconstruction, with the help of the IMF, that would qualify them for help under the Baker plan. And the World Bank, newly rediscovered by the Reagan administration as a key institution in the process (much as it belatedly rediscovered the IMF four years ago when the debt crisis first broke out), has been pumping loan money out as fast as it can find takers. The bank has new loan programs under way for 13 of the 15 countries, Baker recently told a congressional committee.

But that gets us back to the third leg of Baker's stool -- the commercial banks. Where's their new money?

At the Phoenix meeting, Mulford acknowledged, first, that many smaller banks say that they "want out," and that many of the larger commercial banks are "cautious," waiting to see what loan conditions the IMF and World Bank extract from the debtor governments. He continued:

"I am concerned that not enough has been done by commercial banks to assure that, when the time comes, they will, in fact, be ready to lend. Have the modalities for additional lending within current lending syndicates been agreed? Have the differences between the larger and smaller banks been resolved? I doubt it."

Mulford went on to reiterate that, if the banks are dragging their heels in the hopes of getting either government or World Bank guarantees for new commercial bank lending, they might as well forget it.

Baker remains optimistic. He believes that the process of shifting the borrowers toward the market-oriented reforms he set as a requisite for new money must take place -- for political reasons -- with little public fanfare. But when in place, he expects those reforms to trigger more IMF and World Bank loans, persuading commercial bankers to part with their money.

Skeptics such as Rep. Charles Schumer (D-N.Y.) argue that something more dramatic is needed. Schumer wants a study by the Federal Reserve -- not the Treasury -- to explore the following question: How much of the existing debt can the Third World countries really repay without cutting their living standards much further? How much should be forgiven? He calculates that the maximum repayable amount is somewhere around 70 percent.

Schumer tried, without success, to get his idea into the House trade bill last week. Instead, the House adopted a wider ranging, but more restrained proposal by Rep. Stan Lundine (D-N.Y.) that calls, among other things, for a Treasury study of commercial bank exposure in Third World countries. It also attempts to tie World Bank lending to the liberalization of trade policies by the borrowers.

The Lundine proposal is the first cohesive congressional approach that links the debt crisis to the trade deficit. It recognizes, for example, that the combined U.S. trade deficit with developinng countries "is essentially as large as our much more widely lamented deficit with Japan."

Yet, it does not provide specifically for debt relief. Instead, it calls for Treasury studies of future options, but Baker to date has firmly opposed any write-offs of commercial bank loans. Schumer will continue to press that idea, when the Congress next considers appropriations for the multilateral development banks.

To suggest a write-off of about 30 percent of Third World debt is not as radical as it may seem: The stock market, as Schumer suggests, has been making its own, tough evaluations of commercial banks' real worth, often in direct ratio to the amount of Third World loans they have made.

To prove his critics wrong, Baker will have to produce some real action from the commercial banks, and then demonstrate that the loans offered are enough to generate not only improved standards of living among the debtor nations, but a resumption of imports from the United States.