Hasta la Vista, Baby: Bush's Hispanic strategy comes unraveled.

On March 12, two quite separate events combined to undermine the Bush administration's strategy for building a new GOP majority by winning Hispanic votes with such policies as an amnesty for illegal Mexican immigrants. The first event was the result of the Democratic primary in Texas, in which conservative millionaire Tony Sanchez handily defeated former attorney general Dan Morales with a campaign that stressed the rise of Hispanic power. The second was the near defeat in the House of Representatives of Section 245(i)-a measure to allow more than 200,000 illegal immigrants to remain in the U.S. while regularizing their status, rather than requiring them to return home to apply for U.S. entry from there.

The Texas primary strengthened the evidence that the Hispanic vote is drifting firmly into the Democratic camp-irrespective of the GOP's immigration policies. And the House vote signaled that in the aftermath of September 11 most Republicans want to tighten immigration policy rather than liberalize it. Together, they suggest that the Bush administration's Hispanic strategy is falling apart.

In particular, the House decision-in which the Republican leadership averted defeat by a single vote-established that the White House no longer has the Republican votes to push through its larger plans to amnesty 3 million illegal Mexican "guestworkers" as a favor to Mexico's President Fox. Not only did a clear majority of Republicans, including some close to the leadership, rally to the standard raised by Colorado representative Tom Tancredo in opposition to 245(i); but those who voted against it included all the Republicans (and some Democrats) who are considering a run for higher office this year, with the sole exception of New Hampshire representative John Sununu Jr. The measure achieved its narrow victory only with the support of congressmen like Lamar Smith of Texas and Judiciary Committee chairman James Sensenbrenner of Wisconsin, who generally favor tighter immigration controls and would almost certainly oppose the broader amnesty proposal.

The measure now faces an uncertain future in the Senate, where Robert Byrd of West Virginia has announced that he will prevent its passage under the "unanimous consent" provision that was its best hope of an early win. He expressed theatrical astonishment that the House and the White House should be so keen to pass "what amounts to an amnesty for hundreds of thousands of illegal aliens, many of whom have not undergone any background or security check." The politics of an immigration amnesty just got more perilous.

It may have helped the opponents of 245(i) that the previous week President Fox, in between eloquent appeals for a warm American welcome for Mexican immigrants, had handed back to Castro's secret police the handful of Cubans who had sought asylum in his own embassy. But that merely provided them with a nice secondary justification: Their main incentive was changing public opinion. Those Republicans with the most urgent reason for getting public opinion right-their own electoral interests-voted against the White House. One congressman, when taxed by a loyalist, gave his reason simply as "September 11th." Tancredo's immigration-reform caucus, which a year ago had a membership in the low teens, now boasts more than 60 adherents. And Robert Byrd has just reminded the GOP that even if the national Democratic party favors Hispanic immigration even more fervently than the White House does, local Democratic candidates may still flay them for a vote that seems to endorse and encourage illegal immigration.

The lesson for the White House is-or should be-clear: It can only pass the broader immigration amnesty it has been promoting over and against the votes of the majority of Republicans. That course will doubtless be urged upon it by some political analysts and pressure groups, citing the precedent of Clintonian "triangulation." That precedent, however, suffers from an obvious flaw: Clinton's triangulation meant supporting a welfare reform that was overwhelmingly popular with the American public, whereas illegal immigration is highly unpopular. Indeed, pollster John Zogby reports that 83 percent of Americans believe immigration laws are too lax. So the GOP majority would have public opinion on its side in resisting any move to make immigration easier. In which case the White House cannot deliver the goods on which its electoral outreach to Hispanics is based-and it would therefore be well advised to adopt a different strategy.

The good news from the Texas primary is that this may not matter very much, since the old strategy was doomed to fail anyway. It was based on a whole series of assumptions about Hispanic voters, each one of which was either plainly false or highly questionable: for instance, that Hispanic-Americans favor high levels of immigration. In fact, opinion polls clearly show that Hispanics differ only slightly from other Americans on immigration. A clear majority of Hispanics favor either the current or lower levels of immigration. Hispanic voters are swayed much more by the general policy stances of both parties than by immigration.

Another questionable idea is that Hispanic voters are "natural Republicans" because of their conservatism on moral questions such as "gay marriage" or abortion. Sure, in a California referendum on gay marriage, Hispanics voted disproportionately against it. But Hispanics tend to be liberal on economic questions, and when it comes to voting and party identification, in the self-satisfied but accurate words of liberal California analyst Harold Meyerson (now of The American Prospect), "their economic progressivism has consistently trumped their moral conservatism."

Are Hispanics likely to become more Republican the longer they stay in the U.S., and the more they rise up the income scale? No. A study by political scientists James G. Gimpel and Karen Kaufmann showed that Hispanics became more Democratic the longer they stayed in the U.S., and though Republican identification did indeed rise with prosperity, the Democrats retained a 10-point lead even at the highest levels of income.

The Texas primary confirmed these gloomy results for the GOP even before the results were tabulated. Hispanics were 12 percent of the Texas electorate in 1998, and are expected to be 20 percent-the "tipping point" at which their rise will make Texas a Democratic- leaning state-within six years. As GOP pollster Matthew Dowd, a longtime booster of the Hispanic/amnesty strategy, conceded to Dan Balz of the Washington Post: "The question this year is whether the Sanchez campaign advances that [i.e., making Texas a competitive swing state rather than a reliably Republican one], compressing six years into six months." It might do so; Sanchez combined an ethnic appeal to Hispanics-objecting to his opponent's wish to answer questions in English and Spanish rather than solely in Spanish in a televised debate-with an economic appeal to moderate middle-class whites, calling for low taxes.

For that very reason, however, his looks like a transitional candidacy even if he wins in November. For as Hispanic voting strength grows, so it is likely to reflect in Texas the liberal economic voting patterns celebrated by Meyerson in California.

What lies behind this political drift in Texas? Exactly the same force that is pushing once-reliable GOP states like California and Florida into, first, the "undecided" and eventually the "Democratic" column: demographic change driven by immigration. The Hispanic share of the population has risen sharply in these major states in the last 30 years; the Hispanic share of the electorate is now catching up, as immigrants become citizens and register to vote; and their votes heavily favor the Democrats. What has happened in California and now Texas is destined to happen in all the states with large concentrations of Hispanic immigrants. This is not a political prediction; it is a mathematical relationship.

As the study by Gimpel and Kaufmann demonstrated, moreover, this drift will be very hard to reverse. Republican hopes for major gains in the Hispanic electorate are without foundation. Democrats lead the GOP by large margins in every Hispanic group except Cuban-Americans. There is no sign that any significant group of Latino voters is "in play." Because Hispanic voters lean to the Democrats on economic and social grounds, the GOP would have to change almost all its policies (on taxes, welfare, regulation, labor law) to have any hope of attracting Hispanic crossovers in the long term. Above all, insofar as there is a modest drift rightwards among Hispanics as they rise economically, that is more than canceled out by the fact that continuing immigration channels new, poor Hispanic voters into the Democratic ranks.

Of course, there are Hispanics-between one-quarter and one-third of the total Latino electorate-who loyally pull the Republican lever. But they are the very voters who are least likely to favor sectional appeals to a separate Hispanic identity, such as an amnesty for illegals, and most likely to respond to traditional Republican arguments for patriotic assimilation. In the post-9/11 atmosphere, other Hispanics might be won over to their side by a patriotic appeal of that kind. But unless the Bush administration wakes up to the electoral impact of continuing immigration, the most the GOP can hope for is to slow the pace of its decline.