What that Wilson Perkins poll really tells us

There’s something about that poll of Texas last week that I don’t think has gotten enough attention. I only mentioned it in passing myself, so I want to go back and spend a little more time on it.

As I said at the time, I thought the poll’s racial distribution – 60% Anglo, 26% Latino, 14% African-American – was rather favorable to Democrats. The April PPP poll was 69% white, and the full sample of the May UT/TT poll, which was 46-39 Romney before they applied their weird “likely voter” screen, was 62% white. If whites are 70% of the electorate, with blacks and Latinos 15% each, Republicans win more or less without firing a shot. If Republicans get 70% of the white vote, then every statewide Republican starts out with 49% of the vote, and will likely cruise to a 55% win, give or take. If they only get 65% of the white vote, their opening stake is 45.5%. Even with a negligible share of the black vote, getting 30% from Latinos is more than enough to put them over the top; given that Romney was pulling 32%, they’d still be in good shape. They don’t need to worry until whites vote less than about 63% Republican, and I don’t think that’s going to happen in Texas this year.

But if whites make up only 60% of the electorate, with Latinos 25% and African-Americans 15%, the math changes completely. They’re still in good shape if they’re winning 75% of that white vote, as the Wilson Perkins poll claims, since that starts them out with 45%, and from there a mere 25% of the Latino vote is enough to bring them home. It starts to get tight once the white vote drops to 70% for the GOP. That’s a 42% share to start out, and they need 32% of Latinos, which Romney is pulling in that poll, to make it, but now there’s no margin for error. Given that Republicans tend to undervote in the downballot races more than Democrats do in Presidential years, that could be a formula for the GOP winning Texas’ electoral votes, but losing every office except maybe Senator.

Here’s a chart. Let’s assume that African-American participation remains a constant, and that Republicans get a miniscule total from African-American voters, less than a point overall. Let’s also assume that Latinos vote 33% for Republicans consistently. This is how much Republicans need to get from the white vote to reach 50% overall for different levels of the white and Latino vote:

W share H share W needed ========================== 70 15 64.3% 69 16 64.8% 68 17 65.1% 67 18 65.7% 66 19 66.2% 65 20 66.6% 64 21 67.2% 63 22 67.8% 62 23 68.2% 61 24 68.9% 60 25 69.5%

“W share” is the share of the electorate that is white; “H share” is the share of the electorate that is Latino’ “W needed” is the percentage of the white vote that Republicans need, assuming they get 33% of the Latino vote, to get to 50%. Now it’s certainly possible that Republicans will be able to amp up white support sufficiently to hold off the demographic tide. That got me to wondering how Romney is doing among white voters elsewhere. I poked around Real Clear Politics for a bit, and this is what I found. First, national polls, with all percentages given for white voters only:

Fox News, September 12: Romney 52%, Obama 40%

Esquire/Yahoo!, September 13: Romney 50%, Obama 44%

Now, Texas isn’t like the rest of the country, so I went looking through state polls to see what else I could find. Here are results from various swing and not-so-swing states, again courtesy of RCP and again with percentages given for white voters only:

North Carolina, PPP, September 10: Romney 60%, Obama 37%

Florida, PPP, September 12: Romney 54%, Obama 42%

Missouri, PPP, August 30: Romney 59%, Obama 35%

Arizona, PPP, September 11: Romney 56%, Obama 41%

Georgia, Insider Advantage, July 24: Romney 65.6%, Obama 26.7%

There’s nothing in these numbers to make me think that 77% white support for Romney in Texas is more likely to be accurate. It feels like an outlier to me, and if that’s the case, then that Wilson Perkins poll looks even more odd, both for that and for the unexpectedly large non-Anglo sample. More to the point, these results suggest that it won’t be so easy for the GOP to supersize its Anglo support in Texas. If that’s the case, then maybe things aren’t quite as dire here as some might think.

So I disagree with those who say this poll was a non-story. It’s a peek into the future, or at least a possible future. How far off that future may be, that’s the big question. This Kos diarist ran some numbers and projected a 10-15 year time frame. I think it can be a lot shorter than that, but it’ll be much harder to achieve without an investment from the national party. And again, it would be very nice to see more polling data so we can get a better idea of where we are on that chart. I hope this post has given you something to think about.

Related Posts:

This entry was posted in Election 2012 and tagged , , , , , . Bookmark the permalink.

One Response to What that Wilson Perkins poll really tells us

  1. Pingback: TPA Roundup 9/24/2012 | BlueBloggin

Comments are closed.