Abbott leads O'Rourke by 11 points among likely voters in latest UT/Texas Politics Project Poll
Good morning to all,
We just published the results of the October 2022 University of Texas/Texas Politics Project Poll on the Texas Politics Project website. Along with documentation for the poll, you’ll also find the usual array of thousands of free downloadable graphics of poll results, and an overview post that focuses mainly (though not exclusively) on election-related results.
The poll finds Gov. Greg Abbott leading Beto O’Rourke by 11 percentage points among likely voters, 54%-43%. Green Party Candidate Delilah Barrios and the Libertarian Party’s Mark Tippets each earned 1% support while 2% preferred an unspecified “someone else.”
The overall poll surveyed 1,200 self-declared registered voters from October 7-17, and has a margin of error of +/- 2.83 for the full sample. Likely voters were defined as those respondents who indicated that they have voted in every election in the past 2-3 years, or who rated their likelihood to vote in the November elections on a 10-point scale as a 9 or a 10. This likely voter screen yielded a pool of 883 likely voters, with a margin of error of +/- 3.3% for the full likely voter sample. The likely voter screen was applied in all trial ballots and in items directly related to the election, as indicated in the poll summary (available at the post on the poll).
We also included trial ballots for five other statewide races, three for the first time this cycle (Comptroller, Agriculture Commissioner, and Land Commissioner).
Results from the other statewide races:
Lt. Governor. Dan Patrick 51%, Mike Collier 36%, Shanna Steele 5%, someone else 8%.
Attorney General. Ken Paxton 51%, Rochelle Garza 37%.
Comptroller of Public Accounts. Glenn Hegar 47%, Janet Dudding 35%.
Agriculture Commissioner. Sid Miller 51%, Susan Hayes 39%.
Land Commissioner. Dawn Buckingham 47%, Jay Kelberg 36%.
As in the trial ballot for governor, undecided voters were asked whom they would choose if forced to make a decision. All results for the trial ballots report the results of the initial question combined with this “forced” response. As a result of this procedure (which we typically use in the last poll prior to the commencement of voting), the "someone else" results are the total of those who said someone else in the initial item, plus those who in the force still wouldn't commit to a named candidate. (Respondents' choices in the follow-up are the candidates named on the ballot and “someone else." The poll summary reports the share of voters who expressed no preference in the initial question in each race.)
With more than two weeks to go before election day, this snapshot from last week shouldn’t be taken as a prediction of what the final vote totals will look like. But given that the margins are somewhat wider that what we’ve seen in recent polls of likely voters (which have had Abbott leading by 7 or 8 points), it’s fair to ask how durable the apparent GOP advantages in these results might be, and whether they make sense in context.
The political context expressed in the poll results and the characteristics of the broader election environment are consistent with increasing GOP leads in Texas, even if those advantages don’t remain as large as they are in this poll through the final two weeks of campaigning. Barring major political upheavals in Texas between now and November 8, there’s not much chance of a fundamental reversal of the current dynamic favoring Texas Republicans.
There was a sense during the summer that the convergence of public attention to the overturning of Roe v. Wade, public outcry over mass shooting catalyzed by the Uvalde murders, some Democratic/Biden legislative wins in Congress, and falling gas prices might move sentiment and the public agenda in a direction favorable to Democrats. We saw some evidence of such shifts in national polling over the summer. In Texas, this was accompanied by the continuation of criticism over handling of the Uvalde mass shooting and single-digit GOP statewide leads in polls of registered voters and adults.
In recent weeks, based on national polling and macroeconomic developments, there has been a lot of evidence that politics has reverted closer to the mid-term pattern in which the party out-of-power in a president's first mid-term experiences significant gains as the election becomes a referendum on the president's party. As evidence mounts suggesting this is happening at the national level, the resulting Republican advantages are very likely amplifying the baseline advantages Texas Republicans held at the outset of this cycle. In no particular order, here are a few indications of this dynamic we see in the latest poll.
The issue environment – that is, the alignment of the issues that are receiving the most media attention with the patterns of public opinion on those issues – favors Republicans. Our latest poll finds that the three issues most important to Texas voters are immigration and border security (32%), the state economy (14%), and abortion (13%). The first two play to Republicans’ advantage – Biden’s approval ratings on both issues are deeply negative, while Abbott’s are net-positive. The third, abortion, while topping the list of issue important to Democrats, does not appear to have unified nor mobilized Texas Democrats to the extent many had hoped. While 26% of Democrats say abortion is the issue most important to their vote, it competes with gun violence (16%), the environment/climate change (13%), and healthcare (10%) as Democratic priorities. Among Republicans, immigration and border security continue to unify Republicans like no other issue: 60% say this issue area is most important to their vote – an embrace that encompasses a share of Republicans larger than the total share of Democrats selecting the top three Democratic issues. This makes it much easier to grab Republican voters’ attention and hang onto it through the course of the campaign.
Negative partisanship remains a powerful force in Texas politics, which plays to the advantage of Republicans by reinforcing their numerical advantage in the electorate. Statewide Democrats, including (especially) Beto O’Rourke need to find additional voters beyond established Democratic voters. One possible source of these votes is Republican voters who might be persuaded to cross partisan lines. But the October poll shows that the likely voter pool (even more than the registered voter sample, where this has been in evidence for months if not years) contains very few Republicans who appear open to voting for Democrats, including O’Rourke. Both gubernatorial candidates earned broadly positive reviews from their own partisans, but much more intensely negative views from voters aligned with the opposing party. This hurts the underdog O’Rourke more than it hurts Abbott. O’Rourke needs non-Democratic votes to overcome a baseline deficit in Democratic votes relative to Republican votes. Likely Republican voters’ views of O’Rourke are intensely negative: 85% hold unfavorable views, with a lopsided 81% holding very unfavorable views. In short, there aren’t very many persuadable GOP voters for O’Rourke to attract. The trial ballots in the latest poll found only 3% of Republicans choosing O’Rourke over Abbott – and 4% of Democrats choosing Abbott over O’Rourke. Both numbers are well within the margin of error of the partisan subsamples, itself an illustration of how much negative partisanship reduces the potential for cross-party voting.
True independents in Texas still tilt in a conservative direction, and hence toward Republicans. While there aren’t very many of these voters in the electorate (the true independents that made it though the likely voter screen made up about 8% of the LV pool), they are another critical piece of the puzzle for O’Rourke. But they are not favorably disposed to him, and most are not likely to be persuaded to vote for him. I've made this point in the mailer before, and it still holds, as the graphics below illustrate.
Like most midterm elections, this one is nationalized around the party of the president, and Joe Biden is weighing down Texas Democrats. The poll illustrates Texans’ negative views of Biden, of the economy, and their assignment of blame for economic ills to him. Beto O’Rourke and other Democrats have kept their distance from the President, but the Republican candidates (most notably Abbott and Dan Patrick) have spared no expense in associating Democratic candidates with a president that Republicans and independents view deeply negatively, and using Biden to define the Democratic brand – another payoff of negative partisanship that amplifies the GOP advantage.
Republican voters remain more enthusiastic about the election than Democratic voters. In the overall registered voter sample, the share of Republicans who say they are “extremely enthusiastic” about voting in the 2022 election is 14 percentage points higher than the share of Democrats, 57% compared to 43%. The familiar Democratic response to such data points and to the clear pattern of Republican advantages in turnout is to point toward predictions of higher turnout, and to suggest that increased turnout will accrue to Democrats' advantage. The enthusiasm gap evident in polling, as well as the Republican responses to Democratic efforts in 2018 and, especially, in 2020, all suggest that higher turnout does not automatically provide a net benefit in election wins to Democrats. Recent history underlines the fact that in a state with such a comparatively low level of turnout, and hence millions of eligible but non-voting Texans up for grabs, a non-trivial number of potential Republicans exist among the party of non-voters. Given the conditions above, especially the poor state of the Democratic brand in Texas, potential increases in turnout don’t automatically mean big net gains for Democrats (or, for that matter, any net gains at all).
There is much more to be gleaned from the latest poll, including some very interesting results from a battery assessing which problems Texans expect to be most serious in the upcoming election, Texans still-grim outlook on the economy and the path of the country and state, and much more. We’ve worked fast to get this data in front of the public before early voting start, and haven’t processed it all. So I’ll be back with more next week. That said, there is a lot in the overview of the poll results we posted today, so be sure to have a look at that post - there’s more there that I haven’t touched on here.
Have a good weekend, and keep an eye out next week for more on the poll results, including a Second Reading podcast with Daron Shaw, Josh Blank, and me comparing notes on the poll and the election environment.
Executive Director, The Texas Politics Project
College of Liberal Arts / Department of Government
The University of Texas at Austin