MEXICAN President Carlos Salinas de Gortari made a big splash when he promoted a U.S.-Mexico Free Trade Area (FTA) during his visit here a few weeks ago, predicting that the agreement would "create jobs in Mexico, and in the United States as well," and help to stanch the flow of illegal Mexican immigrants to this country. Now the Bush administration has gone him one better, proposing an FTA that would blanket the entire Western Hemisphere and its population of nearly 700 million people, twice as numerous as -- and some have suggested, a counterpoise to -- the European Community.

There are likely to be some benefits for everyone in the free-trade proposal, particularly Mexico. But behind the big population numbers are some hard realities -- and some problems.

Take Mexico as an example. Proponents of a bilateral FTA stress that Mexico, with rapid recent increases in imports from the United States, is our third most important trading partner, after Canada and Japan. But there is a sharp drop from No. 2 to No. 3. Our two-way trade with Canada and Japan last year totaled, respectively, $167 billion and $138 billion compared to only $52 billion with Mexico. Indeed our trade with Canada, a country of 25 million people, exceeds by 1.5 times our trade with all of Latin America, with a population of about 400 million. Trade with the European Community is still larger.

With respect to trade, we are vastly more important to Mexico than Mexico is to us. They account for 6 percent of our total trade. We account for more than 60 percent of theirs. Much has been made of the fact that Mexico's population is about 90 million, larger by far than any European Community country, including a unified Germany. But in purchasing power, it takes about 10 Mexicans to match one West German. That's partly because West Germany's per capita income is so much higher but also because income is distributed so inequitably in Mexico. World Bank data show Mexico's top 10 percent of households taking 40.6 percent of total income, and the bottom 20 percent of households accounting for less than 3 percent.

Nor is a free-trade area likely to do much to discourage illegal Mexican immigration into the United States, now estimated to add as many as a quarter of a million to the U.S. resident total each year. Mexico's going wage rate is a scanty $6 a day in the north and only $3 a day in the south. Estimates of unemployment and underemployment run as high as 40 percent, and one million new job seekers join Mexico's labor force each year. Even should the FTA generate the immense number of jobs necessary to put all Mexicans to work, jobs in the United States that pay 10 or 20 times as much and offer better benefits and social services will still be a powerful magnet. Moreover, some experts believe that improved job opportunities in Mexico will accelerate the flow of illegal immigrants to the United States by making it easier for poor Mexicans to accumulate a stake to pay for their move.

The economic realities thus suggest that the benefits of a U.S.-Mexican FTA have been oversold, at least in terms of what it can do for the United States. But political realities also counsel caution.

Inward-looking, statist policies have impeded Latin America's development for decades. During the 1980s, the de la Madrid government made a good start on moving Mexico away from such policies. Salinas has advanced that process and has also taken steps toward ending the one-party dictatorship that is all Mexico has known since its revolution early in this century. But there is still a way to go on both scores.

As an example of the opening up of the economic system, the Mexican Government has trumpeted its program of privatization of state-owned industries. But the reality is less impressive than the rhetoric. A case in point: The sale of the highly inefficient state telephone monopoly TELMEX has been "suspended" while the technicians study its "true market value." This kind of backsliding doubtlessly reflects a hot debate within Salinas's Revolutionary Institutional Party (PRI). But there is little debate in the opposition Democratic Revolutionary Party (PRD), led by Cuauhtemoc Cardenas, who, absent election fraud, might have been the real winner of the 1988 elections. Events in Eastern Europe and Nicaragua notwithstanding, the populist, leftist PRD still tends to see Mexico's problems as the result of Yankee imperialism and the solutions in economic nationalism and socialism. The PRI used to pursue the same ideology. One consequence: a "Third World" foreign policy that was driven by an emotional need to be on the opposite side of the fence from the United States.

If Salinas does the right thing and follows through on his commitment to open up the political system, it is quite possible that he will be defeated by Cardenas in 1994. Some of Cardenas's advisers have already made it clear that they do not support an FTA. And even should the PRI win again, it is far from clear that the next president will be able to consolidate the political transformation to democratic capitalism from authoritarian quasi-socialism.

Thus, aside from the question of FTA benefits for the United States, questions about shared values and institutions need to be pondered.

Latin America's record with respect to post-World War II economic integration programs has been poor. That failure, I believe, is due in part to some of the same cultural obstacles that have impeded the progress of individual Latin American countries toward political pluralism, economic dynamism, and social equality.

The Latin American Free Trade Association (LAFTA), of which all the larger countries including Mexico were signatories when it went into effect in 1961, is moribund. The Central American Common Market, established soon thereafter, is also a shambles, and not just because of the upheavals of the past decade. It was in trouble in the relatively peaceful 1970s. And the Andean Group, created in 1966 within LAFTA as a counterpoise to the Big Three -- Brazil, Mexico and Argentina -- has also gone nowhere.

Certainly a U.S.-Mexico FTA, or a broader hemispheric FTA, could in the long run make sense for all parties, notwithstanding a probable disproportionate benefit for Mexico. But the question marks that surround these new initiatives remind us of a far more urgent priority of U.S. trade policy: movement away from regional trading blocs and toward a worldwide trading community.

Lawrence Harrison, a visiting scholar at Harvard's Center for International Affairs and author of "Underdevelopment Is a State of Mind -- the Latin American Case," has just completed a second book on culture and progress.