"Latino Power" trumpets the cover of Newsweek. It features the newly-elected Los Angeles mayor, leftwing Democrat Antonio Villaraigosa, flashing his charismatic smile while cavorting on the beach.

Newsweek's macho subhead: "A new mayor in L.A. A decisive showing in '04. Latinos are making their mark on politics as never before. Get used to it."

But last Thursday the Census Bureau released its massive survey of who actually voted in the November, 2004 General Election. It painted a very different picture than Newsweek's cliché-ridden celebration of the cresting tsunami of Hispanic electoral clout.

The Newsweek cover features the triumphant face of Villaraigosa, an unrepentant former chairman of the UCLA chapter of the separatist MEChA organization and a past ACLU apparatchik. But the story on the inside, "A Latin Power Surge," (By Arian Campo-Flores and Howard Fineman ) is all about how the rise of Latino Power has been good for…the Republicans!

"'If the GOP maintains its current share of the Latino vote, says Simon Rosenberg of the New Democrat Network, 'then the Democrats will never be the majority party again in our lifetimes.' How did things become so dire for the Democrats?"

The subhead on the cover story ("A decisive showing in 2004") even implies that Latinos won the election for George W. Bush.

Does Latino Power benefit Republicans or Democrats? Whom should you believe: Newsweek's words or its cover picture of Villaraigosa?

Los Angeles, a city that as recently as 1993 and 1997 elected a Republican mayor (albeit a liberal one), has just given a landslide to Villaraigosa in a Tweedledee vs. Tweedledum struggle with another liberal Democrat, incumbent James Hahn.

So how can things be "so dire for the Democrats?"

My theory: Newsweek had its "Latino Power" articles prepackaged and has been waiting around for some Latino candidate to win something big enough to (almost) justify it.

And what about the disagreement between Newsweek's words and its own numbers about what happened last November.

"President George W. Bush captured roughly 40 percent (the exact figure remains in dispute) of the Hispanic vote …"

It's good to see that VDARE.com's investigative reporting on the errors in the original exit poll claim that Bush had won 44 percent of the Hispanic vote has made Newsweek wary of using that now discredited figure—even though their case would sound more plausible if they had used 44 percent.

(Probably the essay was written on the assumption the old 44 percent figure was true, and an internal fact-checker made them lower it.)

But how does winning 3-2 among a group celebrated for its growth keep the Democrats from ever being "the majority party again in our lifetimes"?

And, even better for the Democrats, Latino political talent is even more skewed toward the left than are Latino voters. In California seven-eighths of the Latino legislators are Democrats. Villaraigosa, who is a genuinely skilled political operator, has made his career tacking back and forth between the hard and soft left.

Villaraigosa won easily over the incumbent in large measure because Hahn was seen as a dull cipher, the second coming of recalled former Governor Gray Davis. Californians seem to believe these days that they are decadent and distracted and need energetic personalities like Villaraigosa and Gov. Arnold Schwarzenegger just to keep their attention engaged on public affairs.

In 1997, Peter Brimelow and Edwin S. Rubenstein's article "Electing a New People" first laid out the math of how importing Democratic-leaning immigrants works against the Republican Party in the long term. The inexorable conclusion: it is in the GOP's self-interest to cut immigration.

Pro-mass immigration enthusiasts on the right, however, inverted this logic to argue that Hispanics were already such an irresistible force that the only salvation for the Republicans was more of the hair of the dog that bit them. The GOP must win over the Latino vote by opening up the borders even farther.

This quickly became conventional wisdom in the news media.

My contribution from 2000 onward has been to make two criticisms of this conventional wisdom.

• First, I noted that opening the borders wider was not the royal road to the hearts of Hispanic voters. Because Latino voters bear so much of the brunt of the immigration wave in lower wages and overwhelmed schools, they are far more ambivalent about immigration than their self-appointed ethnic "leaders" claim. The Latino leadership wants more warm bodies from south of the border to make themselves look more important. But Hispanic voters want better lives for themselves and their children. This was validated last November when the successful anti-illegal immigration initiative Prop. 200 won 47% of the Latino vote in Arizona.

• Second, I pointed out that, even if Hispanic citizens were indeed desperate for more immigration, the much-heralded future of Latino political dominance hasn't quite gone through the formality of taking place. Hispanic voting clout is more limited and growing more slowly than the media assumes. There is still time to limit immigration.

For example, in 2001 I was the first to show that while the press universally claimed that Hispanics comprised 7 percent of the electorate in 2000, the Census Bureau's 50,000 household telephone survey of voters, which is the gold standard for understanding who votes, reported they made up only 5.4 percent.

Not that facts matter much these days.

Two years later, Michael Barone claimed:

"… Hispanic immigrants are the fastest-growing and politically most fluid segment of the electorate. They were 7 percent of voters in 2000 and could be 9 percent in 2004, most of them in big states." 

Barone truly is one of America's leading experts on voting behavior. His biennial Almanac of American Politics is an awe-inspiring 1,800 page trove of data for political junkies.

But Barone's factually-challenged cheerleading for immigration is unworthy of him. Which is why I've criticized him frequently over the years. It's easy to beat up on amateurs, but for me to score so many points off the top pro means I've had to be right about the impact of immigration on voting.

And the only way I've been able to be correct so much more than a master like Barone, who has fifty times my experience and contacts, is if Barone is opening the door by kidding himself about what the numbers say.

So, in May of 2004, I wrote in VDARE.com:

"I hereby declare that, in the tradition of the famous bet between Julian Simon and Paul Ehrlich, I will wager $1,000 that the Hispanic share of the 2004 Presidential vote—according to the November 2004 Census Bureau survey—will be closer to my prediction of 6.1 percent than to Barone's prediction of 8.5%."

Barone didn't take me up on the bet, which is too bad because I could definitely use the money.

Last week, the Census Bureau revealed its results: the 2004 Hispanic vote totaled only 6.0 percent—even less than my forecast of 6.1 percent and a long way from the 9 percent Barone speculated about.

Last November should have been a good election for Hispanic turnout, since the excitement level was the highest in years. Total turnout was up 16 percent over 2000, which should have particularly boosted the Hispanic share.

As a contrasting example, note that during the sleepy midterm election of 2002, Latinos made up only 5.3 percent of voters, down from their 5.4 percent in the more stimulating Presidential election year of 2000.

On the other hand, what Barone claimed was a particular strength of Hispanic voters in the Presidential election—their concentration in large states—turned out to be a decided weakness in 2004. Two states that were completely out of play in the Electoral College, Texas and California, were home to 47.5 percent of all Hispanic voters and a large majority of the Mexican-American voters to whom Bush's Open Borders temporary worker plan is supposed to appeal.

Florida was definitely in play, but its sizable Hispanic electorate is led by Cubans and, increasingly, Puerto Ricans, neither of whom care about immigration policy. (New arrivals from Cuba are treated by current law as refugees, not immigrants, and Puerto Ricans are born U.S. citizens.)

In contrast, in Ohio, the most contested state in the 2004 election, Hispanics cast only 1.6 percent of all votes.

So Hispanics were even less important in the 2004 Presidential election than their 6.0 percent would suggest. (Of course, due to the "rotten borough" syndrome of counting illegal aliens and other noncitizens in drawing up legislative districts, Latinos may be more important in House and state legislature elections than their numbers of voters would warrant.)

But what about the troubled Edison-Mitofsky National Exit Poll, which initially claimed Hispanics accounted for 9 percent of the vote in 2004, before reducing that assertion to 8 percent? Well, that was the weighted result: they raise or lower the raw participation rates of different demographic groups based on various preconceptions.

I downloaded the raw data from Edison-Mitofsky and found that the unweighted Hispanic turnout in their exit poll was only 5.9 percent, almost identical to the Census Bureau's telephone poll.

It was only after it was weighted according to who knows what principles that it inflated to 7.5 percent, which got rounded up to 8 percent in the press.

Many commentators have attributed Bush's better showing in 2004 compared to 2000 to Hispanics. Dick Morris, a campaign consultant for Vicente Fox and Bill Clinton, wrote in the New York Post:

"George W. Bush was re-elected on Tuesday because the Hispanic vote, long a Democratic Party preserve, shifted toward the president's side."

There's no question Bush did well among Hispanics in 2004. He appears to have gone up from winning 35 percent of Hispanics in 2000 to 40 percent in 2004.

But he did better among almost all regions and demographic groups. For example, the exit poll showed him increasing his popularity among non-Hispanic whites from 54 percent to 58 percent.

That means that despite his solid performance among Hispanics, the margin between Bush's share of the white vote and his share of the Hispanic vote merely narrowed from 19 points in 2000 (54-35=19) to 18 points in 2004 (58-40=18).

What about turnout? Karl Rove's political machine seems to have executed well what VDARE.COM insists on calling the Sailer Strategy: getting whites out to the polls. According to the Census Bureau's press release of last Thursday:

"In 2004, turnout rates for citizens were 67 percent for non-Hispanic whites, 60 percent for blacks, 44 percent for Asians and 47 percent for Hispanics (of any race). These rates were higher than the previous presidential election by 5 percentage points for non-Hispanic whites and 3 points for blacks. By contrast, the voting rates for Asian and Hispanic citizens did not change."

Bush pulled 11.6 million more votes in 2004 than in 2000, the majority of that growth due to higher overall turnout. By my calculations, over 80 percent of those 11.6 million additional votes, or 9.5 million, came from non-Hispanic whites.

Whites provided almost ten times as many incremental Bush votes as the next most important ethnic contributor to his growth, Hispanics, at 0.97 million extra votes.

As I've said for years, there's a distinct possibility that Karl Rove knows that his minority outreach talk is mostly a smokescreen to distract the media from his Strategy That Dares Not Speak Its Name: majority inreach.

I'm sure he's right that his disingenuousness with the media is prudent.

But that doesn't mean Republicans need to believe it too.

Or, least of all, the supposedly objective Mainstream Media. (email Newsweek).

[Steve Sailer [email him] is founder of the Human Biodiversity Institute and movie critic for The American Conservative. His website www.iSteve.blogspot.com features his daily blog.]