The GOP is rightly concerned that its share of the Hispanic vote has dropped to historic lows. What’s interesting about that drop, though, is that it occurred at the same time that the GOP’s share of the white vote increased. And that’s not how it usually works.

In 1996, Bob Dole captured only 21% of the Hispanic vote. But he also only captured 46% of the non-Hispanic white vote. (He beat Clinton among whites, by the way – Republicans winning the white vote and losing the election is nothing new.) In 2000, Bush won 55% of the white vote and 35% of the Hispanic vote. In 2004, he won 58% of the white vote and 41% of the Hispanic vote. In 2008, McCain won 55% of the white vote and 31% of the Hispanic vote. Over these four elections, the gap between the Hispanic and non-Hispanic white vote that went Republican has ranged from 17% to 25% – narrower in better years for the GOP, wider in worse years, but the two blocs otherwise moving roughly in tandem.

This year was different. Romney won 59% of the white vote but only 27% of the Hispanic vote. Not only was the gap historically wide, but the two numbers moved in opposite directions – Romney improved on McCain’s margins with white voters, but did worse with Hispanic voters. That hasn’t happened in recent memory. So why did it happen?

Romney made gains in the white vote across the board – but his gains were smallest in the white swing states. He didn’t move the needle at all in Iowa. He didn’t move the needle much in New Hampshire. And another funny thing – Romney gained the most in states where the white vote share fell the most. He gained 8% of the white vote in Missouri – but the white vote share fell 4%. He gained 5% in Ohio – but the white vote share fell, again, by 4%. He gained 6% in New Jersey – but the white vote share fell by 6%. He gained 5% in Florida – but the white vote share fell by 4%. He gained 7% in California – but the white vote share fell by 8% By contrast, states where the white vote held up rather well – Virginia, New Hampshire, Iowa, Connecticut, Massachusetts – are also states where Obama’s share of the white vote held up well. The major exception to this rule is Wisconsin, where Romney increased his share of the white vote and the white vote held up well.

As for the Hispanic vote, Obama’s gains come overwhelmingly from four states: Colorado and Arizona primarily, and secondarily Florida and Pennsylvania. Obama’s share of the Hispanic vote actually dropped in California. All this is according to the limited exit poll data that we have.

Speaking of California, Leon Hadar has a good piece on the website about the Asian-American vote, noting that Obama is now winning that vote by overwhelming margins. But if you look at the exit polls, what you see is that this is overwhelmingly driven by California. In 2008, Asians were 6% of the California electorate and gave 64% of their vote to Obama. In 2012, Asians were 11% of the California electorate and gave 79% of their vote to Obama. I agree with much of Hadar’s analysis, but we need to know whether that analysis has much to do with the country at large, or whether it’s really a California story. And given how strongly Democratic California is, I wonder whether Republicans will really care about winning back any particular voting bloc there.

So what conclusions do I draw from this mix of data? Well, let me tell you where my instincts go.

The white vote share of the electorate dropped too much to be accounted for merely by demographic change. And the absolute number of voters dropped this year relative to 2008, in spite of the increase in the voting-age population. And that increase looks “browner” than the electorate as a whole. So all that tells me that a significant number of white voters who turned out in 2008 did not turn out at all in 2012. Given what happened to the partisan split in the white vote, I suspect these were predominantly Obama voters.

The “Sailer Strategy” consists of trying to win overwhelming majorities of non-Hispanic white voters to compensate for increasing losses among Hispanic and non-white voters. The Tea Party version of the GOP comes as close to such a strategy as we’re plausibly likely to see – which does not mean that the Tea Party’s policy preferences are the same as those of Steve Sailer. But it does mean that the GOP in its current incarnation was perceived, by a wide swathe of the electorate, as the “white party.” And that strategy has proven inadequate in Presidential years.

But positioning the GOP as the “white party” isn’t the only way to appeal to white voters. And this year’s low turnout suggests, in fact, that it isn’t even optimal. Compare, for example, the GOP of 2004, which positioned itself not so much as the white party as the Christian party. That positioning won it a substantially higher turnout among white voters – and an electoral victory.

I’m not suggesting a return to that positioning – indeed, I’d be pretty turned off by an explicitly theocratic turn in the GOP. I’m just suggesting that “win more white votes” and “act like you care only about white people” aren’t the same thing at all.

It seems to me that the biggest problem the GOP has in its current incarnation is that it doesn’t have an economic message that means anything to these voters that it lost in 2008 and failed to win back in 2012. Mitt Romney ran on jobs and the middle class and opportunity, but there was essentially no policy substance behind those rhetorical appeals. Apart from the basic fundamentals of a modestly growing economy, an opponent with the benefits of incumbency, and a reasonably well-managed foreign policy, that’s why he lost. More to the point, that’s why the lost so many down-ticket races.

The GOP could moderate its stance on social issues, or could switch its position on immigration, or what have you. And these things might help or might hurt – I can see arguments for both sides. But I suspect that, ultimately, those questions are window dressing. the GOP cannot become a governing party again if it doesn’t find an economic message that has some relation to the actual economic situation in the country. I suspect that if it did, the old pattern would reassert itself, and it would win a larger share of both the white vote and the non-white vote. If not, the rump GOP base will continue its slide into petulant fury at the country it “lost.”