Pretty decent result that will hopefully put the “double digit lead” narrative out to pasture.
Sen. Wendy Davis
Republican gubernatorial candidate Greg Abbott leads Democratic opponent Wendy Davis by 9 percentage points, 49 percent to 40 percent, according to the latest Texas Lyceum poll, released Wednesday.
The poll showed Davis, a state senator from Fort Worth, held a “clear lead” among Hispanics, by 36 points, and African Americans, by 80 points, while Abbott, the state attorney general, held a “slight lead” with independent voters, by 6 points, and women, by 2 points.
“Davis is running slightly ahead of other Democrats on the ballot and over- performs compared to Democrats from recent statewide races,” University of Texas Professor Daron Shaw, who conducted the poll, said in a statement. “But the number of candidates who have made up this kind of deficit in the last month, in a state where party ID favors the other side so consistently, is close to zero.”
Immediately following the release of the poll, Abbott sent out a press release noting his campaign had $30.1 million on hand for the “final push” of the race, raising $7.8 million since July. In July, Abbott had $35.6 million on hand, the largest amount ever recorded in the state, while Davis had $8.8 million.
The poll showed that in the lieutenant governor’s race, state Sen. Dan Patrick, R-Houston, was ahead of state Sen. Leticia Van de Putte, D-San Antonio, by 14 percentage points, 47 percent to 33 percent. Van de Putte led with African Americans by 64 points and Hispanics by 24 points, but did not perform as well as Davis with either group.
The Lyceum did actually poll this race before, last September. Most likely the reason you (and I) hadn’t heard of that poll before now is because fully half the voters who responded had no opinion. Abbott led 29-21 among the rest.
The Lyceum’s index page for their 2014 polls is here. Their press release, on which just about all of the coverage I’ve seen has been based, is here, and the Executive Summary is here. Questions and toplines are here, and crosstabs are here. You have to be a little careful in reading their summaries (and therefore the coverage), because they don’t always specify when they are talking about likely voters (LVs) instead of registered voters (RVs). Scroll down to page 111 of the crosstabs to see where the raw numbers are. I’ve summarized below:
Governor RV tot RV pct LV tot LV pct
===========================================
Abbott 362 43.04 265 48.89
Davis 314 37.34 215 39.67
Glass 32 3.80 12 2.21
Parmer 20 2.38 9 1.66
Lt Gov RV tot RV pct LV tot LV pct
===========================================
Patrick 330 39.29 256 47.23
Van de Putte 289 34.40 177 32.66
Butler 32 3.81 18 3.32
Courtney 28 3.33 12 2.21
There are also Senate numbers, which I’ll get to in a minute. The point here is to notice that the RV numbers are much tighter than the LV numbers – indeed, the Lt Gov race is close to being within the margin of error for the RV numbers. This is the clearest illustration of what the effect of turnout may be, and it’s why I’ve been so critical of the polling we’ve seen to date. I don’t think previous public polls have considered the question of turnout and the Battleground Texas effect at all so far. Sure, not all of these registered voters will show up, but you can see the potential here. The contrast in the Lite Guv race, where for reasons I don’t comprehend they’ve decided there are a lot fewer “likely” voters, is especially stark. If you need a reason to believe in what BGTX is doing, this is it, right here.
One other point to note is in the reporting of female voters, where the topline shows a lead for Abbott. The crosstabs tell a slightly different story, however:
Governor Male Male % Female Female %
==============================================
Abbott 198 46.48 164 39.42
Davis 142 33.33 173 41.59
Glass 18 4.23 15 3.61
Parmer 12 2.82 7 1.68
Lt Gov Male Male % Female Female %
==============================================
Patrick 181 42.49 149 35.99
Van de Putte 129 30.28 160 38.65
Butler 21 4.93 11 2.66
Courtney 20 4.69 8 1.93
Yes, I noticed the one-vote discrepancies for Davis, Glass, and Parmer, and no, I don’t know what’s up with that. My point is that the press release is apparently giving LV totals for female voters, but they don’t break out those numbers in the crosstabs. We don’t therefore know what the exact RV/LV gap is there, but we know there is one. Like the single digit/double digit distinction, this would change the narrative if anyone other than me were to notice it. It’s also where the rubber meets the road for BGTX, given the heavy Democratic lean that exists for single women and women of color along with their lesser propensity as a rule to show up. If the final result winds up being more Democratic than what the polls have suggested so far, the first place to look for an explanation in the exit polls will be among female voters.
As for those Senate numbers:
Senate RV tot RV pct LV tot LV pct
===========================================
Cornyn 358 42.62 263 48.43
Alameel 237 28.21 163 30.02
Paddock 33 3.93 20 3.68
Sanchez 63 7.50 22 4.05
Here I would submit that Alameel’s poorer numbers are one part much lower name recognition, and one part “Spicybrown” Sanchez getting a disproportionate share of the total. Third party candidates tend to poll higher than their final totals, and that’s what I expect is happening here. In the end, it is likely that most of Spicybrown’s supporters (who, if you check the crosstabs, are disproportionately Latino) will wind up voting D anyway. It won’t shock me if she does better than the average Green statewide candidate, but her ceiling is maybe five percent, probably lower. PDiddie, Texas Politics, and TRail Blazers have more.