The Contadora Central American peace initiative has run its course. Mexico, its chief sponsor, has made public what observers of its foreign policy already knew: it is pulling back from Central America to devote more time and energy to domestic matters. Colombia, another key country, has been battered by natural and political disasters. Panama and Venezuela were never very excited about the mediation effort to begin with.

The reasons for the demise of Contadora are many but easily stated. The United States clearly wants no agreement that would tie its hands in its efforts to get rid of the Sandinistas. The Sandinistas, for their part, will not -- quite rightly -- sign any treaty that does not guarantee an end to U.S. support of the contras. And the Contadora countries are too weak and too divided to impose a solution on either party.

The impasse at the negotiating table derives from a stalemated situation on the ground -- and growing Central American paralysis in Washington. In El Salvador the FMLN insurgents are perhaps weaker in geographical terms, but with greater unity among their armed factions than ever. They are continuing to inflict heavy casualties on the Salvadoran army, probably heavier than before. They are not winning the war, but they are as far from losing it as at any time since 1981.

Likewise in Nicaragua. The contras are not making any significant headway against the Managua regime despite renewed congressional aid. The Sandinistas are putting Soviet helicopters and Honduran acquiescence to their "hot pursuit" incursions into Honduras to good use. Widening -- and understandable -- discontent within Nicaragua is not translating into greater political support from an armed opposition that in Nicaragua -- and most of Latin America -- will always bear the stigma of Anastasio Somoza's National Guard. The contras' war of attrition against the Sandinistas may bleed them but will not unseat them. Nor will it force them to negotiate.

Indeed U.S.-sponsored or directly applied economic, political and military pressure is either too small or too great: not enough to force the Sandinistas from power or into negotiations, too much to allow them to negotiate from strength with their opposition. For practical purposes, U.S. policy today is back to square one: unsuccessfully trying to roll back revolution in Nicaragua, stubbornly containing revolution but not defeating it in El Salvador. The difference, in addition to the 30,000 lives taken and the several billion dollars spent, is that President Reagan no longer has two four-year terms ahead of him. In fact, U.S. Central American policy has acquired a lame-duck, inertial tint to it. Neither of the two ways -- direct U.S.-Sandinista negotiations or a direct U.S. invasion of Nicaragua -- out of the present stalemate seem to be in the cards, and there is every reason to believe that it will be 1988 before any significant change comes about. But change then means formulating new ideas now, and looking for a lasting, realistic modus vivendi in Central America that will not damage U.S.-Latin American relations for years to come.

A new approach probably implies a unilateral and precise statement of the limits of what the United States can live with in Central America before its national security becomes endangered. It would seem to require that the United States count on itself -- and not on regional mediators or local oppositions -- to ensure that those limits are not overstepped. For the statement of interests and the decision to enforce them to be credible and effective, such an approach would probably have to be truly consensus-backed, and not marred by liberal idealism or conservative arrogance. It would also have to be truly geo-political, and not ideological: Latin America, and other U.S. allies, can understand and respect U.S. national security, but not its ideological preferences for one sort of political regime or another. Nor can they accept ironclad linkage, where U.S. national security interests in such tiny nations are ensured only by making the domestic politics in these countries to American liking.

Finally, the United States and Central America should begin to accept that there are other countries with valid and significant national interests in the area. The obvious case is Mexico. Whatever its present government -- under preshink, many Mexicans believe that their country has important national interests in Central America and the Caribbean; that certain regional governments and societies further those interests and others do not; that it should have an assertive, nationalistic foreign policy that supports those who do and tries to influence those who don't; and lastly, that those national interests are not, nor will ever be, identical to U.S. national interests, although they cannot be diametrically opposed to them either.