Texas Prognosis

I posted this as a very long comment elsewhere and figured it would make a more appropriate diary entry. Anyway, here is my prognosis of the Texas primary race, based on polling, turnout and demographics. I'd love it if people more versed in Texas politics can respond here.

I think the race has achieved a sort of equilibrium by now. There's been so much early voting that poll movement at this point (as in Zogby) is increasingly irrelevant (like late movement in California that didn't pick up early voters).  Either way, I'm interested to see what the polling looks like over the weekend in Texas.

The key swing vote, it seems, is white Democrats. White Indies and Republicans are heavily for Obama - and together they will be about 18% of the vote (no other way to explain the massive turnout in GOP suburbs). Latinos will go heavily for Clinton and blacks will go even more heavily for Obama. But white Democrats are split largely on lines of gender and age. They will be about 30% of the total vote. Those in Austin will go heavy for Obama, but elsewhere (especially rural parts of Texas) will go strong for Hillary. The recent Zogby poll, strangely enough, has Clinton only up 1 among women and Obama only up 5 among men. This may be a factor of heavy black and Latino voting, which doesn't reflect gender divides very much. But Hillary will win white women, the question being by how much.

I see the turnout breakdown this way:

18% Indy and GOP crossover (mostly white and some Latino; this number seems high but the 7:3 edge among primary voting for the Democrats means there will be a LOT of crossover voting)
30% white Democrat
30% Latino Democrat
22% African American Democrat

Obama wins 88% of blacks. Hillary wins 64% of Latino Democrats. Hillary wins 56% of white Democrats. Obama wins 65% of Indy and GOP voters.

Add these up and you get a final vote of: Obama 55, Clinton 45.

I think a lot of the polling is undercounting Indy/GOP crossover voting. I separate these out from Democrats because the patterns will look different than for Democrats. Some polls have Latinos as high as 37%, for example. But I bet 7 of those 37 will be Indy and GOP crossover. And if California is any indication, these non-Democratic Latinos will go heavily for Obama (they aren't exactly the Delores Huerta crowd. Many are wealthy 10th generation Tejanos and others are evangelicals with no ties to the Democratic Party anymore). In California, Latino Democrats went for Obama 64-36 and Latino Independents went for Obama by about the same margin in the other direction.  That will probably apply in Texas too.

Hillary runs up her white Democratic numbers through women and older voters. But Austin keeps it real, meaning Hillary's white Democratic totals will be somewhat limited. But by how much?

Crossover votes have gone strong for Obama in other states despite Limbaugh shenanigans in Wisconsin. Those Republicans voting in the Democratic primary in Denton and Collin Counties are not voting for Hillary Clinton - I'm quite confident of that. There may be some voting for Hillary to keep her in the race. There also may be lots of Texas Republicans voting for Obama just to ensure that Hillary Clinton will never be President. But I suspect most of these crossovers are voters genuinely drawn to Obama (and perhaps a small segment genuinely drawn to Clinton). Overall, Indy/GOP turnout will look like Virginia's: 65-35 for Obama.

What might change this? Black turnout will be no lower than 22% but might be higher. Latino Democrats may go for Clinton at a higher clip than 64%, though that seems unlikely. White Democrats might go more heavily for Clinton - especially with older women.  And there may be more crossovers going for Clinton than I calculated. But I think this breakdown makes the most sense at this point. What do y'all think?