Back to the future: if Mexicans in America fail to assimilate, we could see the return of something ugly.

MEXICAN immigrants like it in America. They are much more likely than other immigrants to rate life in the United States as superior to life in the country where they were born, according to survey research by Public Agenda in 2003. They come to work: They are much less likely than other immigrants to cite political freedom as the reason for their migration. They show near zero interest in radical politics: Aside from a few loony college professors, the mystic cause of reuniting the southwestern United States to Mexico seems to excite almost nobody. They crossed the desert to escape Mexico, not to rejoin it. Whatever is going on, it isn't a Reconquista.

So ... good news, right? Assimilation is working? Not exactly. Mexican immigrants may like America, but they are having serious trouble joining it. Well into the second and third generations after arrival, they remain much poorer than other Americans--with unsettling long-term political and economic consequences for the United States.

Only 7 percent of Mexican immigrants arrive in the United States able to speak English. Few possess much formal education. These deficiencies shunt them into low-wage sectors of the economy. The economist George Borjas calculates that Mexican Americans earn almost 40 percent less than American-born workers with American-born parents. Low wages hold Mexican Americans in poverty--and the evidence suggests that their families will remain poor into the second and third generations.

Lacking English and formal education themselves, Mexican-American immigrants do not seem to attach much importance to their children's acquiring either. While 82 percent of immigrants from Europe feel that all immigrants should be expected to learn English, and even 61 percent of non-Mexican Latinos agree, only a bare majority of Mexican immigrants, 54 percent, think English essential, again according to the Public Agenda survey. While 67 percent of non-Mexican immigrants think that public-school classes should be taught exclusively in English, only 51 percent of Mexicans think so.

Perhaps unsurprisingly, the children of these migrants do not flourish in school. A Manhattan Institute study of the high-school class of 1998 found that only 54 percent of Hispanic students graduated with their class, as compared with nearly 80 percent of their white counterparts. These dropouts may later return to school or earn a high-school equivalency certificate. But they will have difficulty catching up to their classmates who finished on time. And of course the children of high-school dropouts are more likely to drop out in their turn.

Borjas points out that the differences in earnings between U.S. immigrant groups and the native population persist from generation to generation. Experience would suggest that if Mexican immigrants earn 40 percent less than natives, their children will earn 20 percent less, their grandchildren 10 percent less, and so on. That's not a reassuring estimate, but the reality may well be even worse. The "experience" Borjas cites is that of immigrant families settling in the United States in the 1930s, '40s, and '50s: the golden age of blue-collar America. Even if they did not speak much English, even if they did not possess much formal education, they could find work in a factory or on a dock that paid better relative to other sectors of the economy, and offered higher inflation-adjusted wages, than the same jobs today. Post-1970 America has become a much tougher environment for those without higher education. In 1970, a 30-year-old man with a college degree earned a little more than twice as much as his counterpart with a high-school diploma. By 1990, he earned three times as much. Today, the disparity gapes wider still.

The Center for Immigration Studies has found considerable evidence that today's immigrants do considerably worse than previous immigrants, decades after arrival. In 1970, only 25.7 percent of immigrants who had lived in the U.S. for ten to twenty years were poor, compared with 35.1 percent of natives. By 2000, 41.4 percent of long-settled immigrants were poor, as compared with 28.8 percent of natives.

Behind all these numbers is an emerging social reality. The immigration policies of the past two decades have imported into the United States a large population that will remain ill educated, incompletely fluent in English, and significantly poorer than the rest of the country well into the 2030s and 2040s. And President Bush's proposal to settle this population permanently --and to increase the flow of new migrants--will make the problems bigger, deeper, and more intractable.

It's troubling to turn on the television and see thousands of illegal immigrants marching behind the Mexican flag, chanting slogans that denounce America's right to control its borders. But I'd worry a lot less about the Republic of Aztlan than about a future in which the American economy rests on a linguistically distinct subclass of ill-educated low-wage workers.

Imagine America in 2031. That's the not-so-distant future: It's a date as far ahead of us as Ronald Reagan's first inauguration is behind us. Suppose the Bush immigration plan of 2006 or something like it has been enacted. Illegal aliens have been legalized; family-reunification and guest-worker programs continue to bring millions of Mexicans and Central Americans north.

In 2031, daily life for you and your children (if I may make a demographic assumption about you, reader) probably proceeds more or less as it does today. The roads are more crowded, real estate and gasoline probably more expensive, taxes higher, and government services worse. On the other hand, many onetime luxuries have steadily declined in price. Domestic-cleaning services like Merry Maids have proliferated, as have in-home childcare and eldercare, gardening services, prepared-meal services, and car detailing. Like Alicia Silverstone in Clueless, you have little need to know how to parallel park-everywhere you want to go, they have valet. In short: The southern-California lifestyle has spread throughout the country.

But what of the migrants whose labor sustains your pleasantly seigniorial life?

How do they feel about their relative poverty--a poverty cushioned but not significantly ameliorated by the food stamps, Earned Income Tax Credits, housing subsidies, and Medicaid benefits for which they are now eligible? At first, perhaps, they didn't complain much. Life in the United States represented a huge improvement over life in Mexico or Central America. First-generation migrants worked too hard, felt too insecure, and trusted government too little to agitate for a better deal. But what of their American-born but ill-educated and low-wage children?

They have the vote. They have expectations of a better life. Will they not find politicians ready to mobilize them for a new era of populist redistributionism--a redistributionism made more powerful and more exciting by the ethnic and linguistic differences between haves and have-nots?

Certainly that is the future left-wing Democrats expect. Immigration reformers often express wonder that the political Left has welcomed an immigration whose main effect is to lower the wages of less-skilled workers, especially black men. But maybe the Left is playing a longer game here--where the short-term depression of living standards for working people becomes a necessary price to pay to reignite the radical economic movements that inflamed the United States from 1870 to 1940.

Americans have become so used to political and economic stability that they have forgotten that their country was once disgraced by the most violent and bloody labor conflicts on earth. These conflicts were often intensified by the ethnic differences between strikers and the American population at large--differences that exerted a follow-on effect in electoral politics. As Michael Barone observes in Our Country, his magisterial history of 20th-century American politics, outside the South the division between support for and opposition to Franklin Delano Roosevelt's New Deal was grounded on ethnicity as much as social class, with old-stock Americans against and immigrants and the children of immigrants in favor.

Politicians like Los Angeles mayor Antonio Villaraigosa plainly hope that the trick can be repeated in this new century. As Harold Meyerson wrote in The American Prospect, the "Latinoization [of California] has also transformed California's fiscal politics." The state that led the tax-rebellion movement in the late 1970s and fed the national conservative political resurgence passed "a massive $9.2-billion school bond measure and a $2-billion initiative for parks and open-space preservation" in 1998, and reduced the power of anti-tax voters to veto local-government bond measures. "With the uptick in Latino voting ... the gap between the voting public and the people who need public services began to narrow."

Many conservative immigration advocates insist that Villaraigosa and Meyerson have it wrong--that the migrants and their children and grandchildren will be socially conservative and politically quiescent. It's worth noting, though, that Republican political professionals are ceasing to believe this: That's why so many of them, from Karl Rove on down, now emphasize guest-worker programs as a way to benefit from immigrant labor without having to face the consequences of immigrant votes.

But as the French and Germans have discovered, there is nothing less temporary than a temporary worker. The migrants will settle, will take up their political rights--and will use those rights to advance their interests: interests that may sharply differ from those of more-established inhabitants.

For more than half a century, American society has been divided along lines of race. The most polarizing issues--busing, affirmative action, welfare, crime--have explicitly or implicitly involved race. Race has exerted so strong an influence that when new grievances came along--the rights of women, of the disabled, of gays--the grievance-bearers unselfconsciously squeezed and shoved their demands into the forms left behind by the civil-rights battles of 30, 40, and 50 years before.

Race and race analogies so mesmerize us that we unthinkingly assume that the challenges presented by immigration must fit into this ancient envelope. It's almost impossible for us to imagine anything else--least of all something so antiquated, so remote as conflicts along the lines of class.

But everything old becomes new again. And it may well be that the greatest threat from today's immigration is not that the United States will be racially balkanized, but that it will relive a past in which classes invoked ethnic solidarity in a struggle over wealth and power. The danger is not that immigrants won't "Americanize." They will. The danger is that they will reintroduce America to an authentically American history that once seemed long and well forgotten.

Mr. Frum, a contributing editor of NATIONAL REVIEW, is the author of The Right Man: The Surprise Presidency of George W. Bush and (with Richard Perle) An End to Evil: How to Win the War on Terror.