Off the Kuff Rotating Header Image

Precinct analysis: The “not Obama” vote, such as it was

I now have my hands on a draft canvass for the Democratic primary in Harris County, and you know what that means: Precinct data! I’ll be doing various analyses over the week so we can get a better idea of just what happened in this election. Let’s start with the Presidential race, such as it was. We’ve established that Texas is not Appalachia, and as you might expect President Obama did very well in the urban areas like Harris, where he collected over 95% of the ballots that were cast. There’s always some variation in the numbers, so let’s look and see how he did in the 24 State Rep districts in Harris County:

Dist NO 1 NO 2 NO 3 Obama Obama % ========================================== 126 18 9 12 1333 97.16% 127 31 16 32 1424 94.74% 128 150 61 75 1396 83.00% 129 70 16 49 2174 94.15% 130 18 5 18 991 96.03% 131 59 28 37 6441 98.11% 132 22 6 14 1085 96.27% 133 31 12 24 1873 96.55% 134 74 39 43 4498 96.65% 135 23 4 18 1252 96.53% 137 48 16 30 1614 94.50% 138 19 17 12 1256 96.32% 139 46 18 40 5423 98.12% 140 52 28 64 1473 91.09% 141 29 16 18 4256 98.54% 142 76 21 54 5055 97.10% 143 321 106 233 3179 82.81% 144 133 49 118 1692 84.94% 145 113 45 81 1867 88.65% 146 58 26 47 7743 98.34% 147 74 36 55 7751 97.92% 148 106 47 73 2538 91.82% 149 21 9 17 1521 97.00% 150 29 6 19 1145 95.50%

“NO” = “Not Obama”. It doesn’t really matter who the other three candidates are, what matters is how many people voted for someone other than the President. If you had asked me beforehand which districts would be the laggards, I would have predicted HDs 128 and 144, as these two “Bubba districts” were where Obama performed the worst relative to other Democrats in November of 2008. It’s a similar story in the Latino districts, with HD143 having a decent dose of Bubba in it as well. Note that in only those three districts did Obama do worse than he did overall, at 88.19%. No real surprises to me in these numbers.

Statewide, there was a broader range to the numbers, but still nothing that raised my eyebrows. There were 23 counties in which Obama did not get a majority of the vote:

County Obama Total Obama % ================================== King 1 7 14.29% Borden 14 51 27.45% Loving 5 16 31.25% Sherman 1 3 33.33% Newton 633 1,758 36.01% Stephens 101 263 38.40% Lynn 2 5 40.00% Kent 27 66 40.91% Knox 21 49 42.86% McMullen 15 34 44.12% Terrell 74 167 44.31% San Augustine 648 1,455 44.54% Hardeman 70 157 44.59% Dallam 16 35 45.71% Crockett 339 741 45.75% Hall 53 113 46.90% Swisher 175 373 46.92% Fisher 156 332 46.99% Upton 191 406 47.04% Foard 68 138 49.28% Cass 701 1,421 49.33% Childress 10 20 50.00% Stonewall 27 54 50.00%

I’ve uploaded a spreadsheet of the results, taken from the SOS Election Night Returns county-by-county page. Only in King, Borden, Loving, and Sherman did Obama not get a plurality. I confess, I don’t think I know where most of those counties are. A total of 7.664 votes were cast in these counties, which is to say a bit more than one percent of the overall total of 587,465. By contrast, the 42 counties in which Obama received at least 90% accounted for 308,888 votes.

There were also a couple of counties where Obama’s percentage was sort of middling where I suspect there may have been some crossover votes. For instance, in Zapata County, where Obama received 78.09% of the 2,519 votes cast, there was a grand total of 21 GOP primary ballots. Frio County, which went 74.21% for Obama, had 1,842 Democratic votes and 185 GOP. Reeves County, in which Obama got 67.68% of 1,828 votes, apparently did not have a GOP primary. A few others I’d attribute to the Bubba effect – San Jacinto (52.29%), Jasper (52.34%), Orange (74.95%). Still no real surprises.

No real surprises on the GOP side, either. In only three counties did Mitt Romney fail to get a majority. Oddly, there were exactly 20 votes cast in each of them – Hudspeth (Romney received 7), Culberson (9 for Mitt), and Starr (10). Only 17 counties fell outside the 60 to 80% range for Romney, whose overall percentage was 68.98%, and one of them was Ron Paul’s back yard of Brazoria, where Romney got 59.92%. (Paul got 25.10%, his best showing.) I was going to say that Romney’s performance struck me as the less impressive, but John Kerry got 67.11% of the vote in the same situation in 2004 as the Mittster, so I guess that’s just how we roll here. I’ll have more precinct results over the next week or so.

Related Posts:

6 Comments

  1. Bill Shirley says:

    Does it matter where King County is? There were 8 people that voted. Actually, it would be an interesting news story to interview who the one person who voted for Obama – i suspect they’re not secretive about it.

  2. OK, you made me look:

    http://en.wikipedia.org/wiki/King_County,_Texas

    “King County is a county located in the U.S. state of Texas. As of the 2010 census, its population was 286.[1] Its county seat is Guthrie.[2] King County has the third smallest population of any county in the United States, ranking behind only Loving County, Texas, and Kalawao County, Hawaii. The county was named for William Philip King, who died at the Battle of the Alamo.”

    It’s between Wichita Falls and Lubbock on US82, north of Abilene. So now we know.

  3. Taylor says:

    I noticed that you didn’t take undervotes into account – how would undervotes change this analysis? Could undervotes indicate a lack of enthusiam on the part of the Democratic electorate (not that it makes a huge amount of difference in presidential politics come November)?

    Tangentially related: How on earth do overvotes happen? I could have sworn that the e-slate machines won’t allow you to vote twice in the same race….

  4. Greg Wythe says:

    I think a bit more research is needed on where the “Not Obama” vote originates. You’re suggesting that the biggest culprit is “Bubba” voters. Yet, when I plug in the 10 most homogenous Hispanic precincts, I see only 84.6% voting for Obama. And following Taylor’s suggestion, I see 75.1% Obama when factoring in the undervote. The countywide comparisons are 95% Obama among votes cast and 90% Obama when accounting for the undervote.

    I think this paints a few more colors to the complexity of the Hispanic vote that hasn’t been a part of the conversation in politically active circles. I’d argue that its not entirely dissimilar to what we saw with support for Mayor Parker in the recent COH elections.

  5. You’re right, I did not count undervotes. The problem is that we don’t know if the undervote was because the person specifically skipped that race, or because s/he showed up for the purpose of voting in only one or two races – I believe Greg is familiar with this phenomenon. Some people may have not voted in the Presidential race because it didn’t occur to them that they needed to vote in it. It’s not like anyone is actively campaigning as a D against the President. Given that, I thought the case where people did bother to vote and chose someone other than Obama was the one where I could try to draw some conclusions. Your mileage may vary.

    Taylor, overvotes can occur on absentee ballots, which are done on paper. The eSlate will not let you overvote, but there’s nothing to stop you from checking more than one box on an absentee ballot. Those are thrown out.

  6. […] to speculate why he did better in some places than in others. Romney’s 16 point range is as wide as Obama’s was on the Democratic side, and a close mirror of the bulk of his countywide numbers. In other words, […]