General Election Projections and Predictions

Note: We are making this article free today for all our website’s visitors and Twitter followers. If you’d like more in-depth analysis of Texas politics and races, please consider subscribing to Texas Election Source. Most mornings, you’ll receive an email with links to our latest reports, and you’ll always have access to all our easily searchable past content. Subscribing is easy and costs less than $20 a month. Click the Subscribe link to the upper left to get started.

Our models project one statewide, six state House and one Congressional incumbent will fall in this general election. All will be Republicans except for the lone Democratic statewide official and lone independent representative. Republicans hold most of the competitive seats coming into the presidential election cycle, including several wrested from traditional Democratic control in 2014 or 2015. We project all open seats will remain with the party whose incumbent is leaving or has left office and all other incumbents will prevail for federal and state offices.

President. We project a 10- to 10.5-point victory for Donald Trump, 53.5% to 43.0%, over Hillary Clinton, which would put her between Barack Obama’s 2008 and 2012 performances. We expect the minor party candidates to receive support in line with their historical performances, but we expect Libertarian Gary Johnson to set a record for his party’s nominee in the state.

Statewide. Republican candidates will sweep statewide races as they have since 1994, winning 125 straight races, including presidential contests. We project the streak will stretch to 133. The lone Democratic statewide elected official, Court of Criminal Appels Judge Lawrence Meyers, will be turned out of office. Meyers has won several statewide contests as a Republican since 1992, but his switch to the Democratic Party in 2013 meant his winning streak was likely over.

Minor Parties. We would not be surprised to see Libertarian RRC candidate Mark Miller exceed 5%, the threshold necessary to guarantee the party access to the 2018 ballot. He received 3% two years ago and has been endorsed by several of the state’s largest newspapers. Because the Democratic Party fielded candidates for all statewide races for the first time since 2002, we expect the Green Party will lose its automatic ballot access. The Green Party lost automatic access in 2002, the last year all statewide offices were contested by both major parties. It did not return until 2010.

Senate. No changes in partisan control of seats are expected. None of the seats are competitive.

House. We are projecting six incumbents will be defeated in the general election, this highest total for a presidential general election in at least 28 years. The number is bolstered by three traditionally Democratic seats won by Republicans in the 2014 general election or a 2015 special election and one traditionally Democratic seat won by an independent in 2016.

Regardless of turnout assumptions or impacts of the presidential races, we expect these four seats will revert to their typical presidential year outcomes:

• HD117 – We project former Rep. Philip Cortez (D-San Antonio) will reclaim the seat from Rep. Rick Galindo (R-San Antonio) by 10 points.
• HD118 – We project challenger Tomas Uresti will reverse his special runoff election loss to Rep. John Lujan (R-San Antonio) by 18 points.
• HD120 – We project challenger Barbara Gervin-Hawkins will easily defeat Rep. Laura Thompson (I-San Antonio) by a margin typical of a major-party candidate over an independent.
• HD144 – We project former Rep. Mary Ann Perez (D-Houston) will reclaim the seat from Rep. Gilbert Peña (R-Pasadena) by 8 points, but very low turnout may make it close from a vote-margin standpoint.

If the presidential race follows our estimate and turnout tracks the way we expect, then our models project Reps. Rodney Anderson (R-Grand Prairie) and Kenneth Sheets (R-Dallas) will also fall to Terry Meza and Victoria Neave in HD105 and HD107, respectively. We expect both of those races to be within a couple of points. There are no minor party challengers, which helps both incumbents. If Clinton underperforms our estimates in these districts, then we estimate Anderson and Sheets would eke out narrow wins.

If Clinton wildly overperforms our estimates, then the seats of Reps. J.M. Lozano (HD43), Jason Villalba (HD114), Cindy Burkett (HD113), Sarah Davis (HD134) and Matt Rinaldi (HD115) could conceivably come into play.

If Clinton badly underperforms our estimates, then it’s possible but unlikely that Rep. Bobby Guerra’s (HD41) seat could become competitive, only if his constituents overwhelmingly vote down a proposition creating the Hidalgo Co. Healthcare District and authorizing an 8-cent per $100 valuation “initial” property tax rate.

State Board of Education. No changes in partisan control of seats are expected. We project ED5, the third race between board member Ken Mercer (R-San Antonio) and Rebecca Bell-Metereau to be the closest at around 6 points.

Congress. We project former Rep. Pete Gallego (D-Alpine) will reclaim the seat from U.S. Rep. Will Hurd (R-San Antonio), unless Clinton finishes below 40%. The district leans about 6% more Democratic than the state as a whole, Democratic congressional candidates outperform other Democrats in the district by another 2% and Democrats should have a slight edge in straight-ticket voting. We also expect Trump’s anti-immigrant rhetoric to hurt Hurd through increased Hispanic/Latino voter participation.

We project no other changes in partisan control of seats. None of the rest of them are competitive.

All models start with assumptions. Here are ours.

Assumption 1: Trump 53.5, Clinton 43.0

Two recent polls show a double-digit lead for Donald Trump, and another shows a 9-point lead with Gary Johnson polling nearly 5 times the historical best for a Libertarian candidate. We estimate a slightly larger than 10-point gap between them, largely because Clinton remains more unpopular than Trump in nearly every poll. Thus, we are projecting Clinton to receive a lower share of the vote than Barack Obama in 2008 but higher than in 2012.

Johnson is likely to set a record for a Libertarian candidate, but it will not look like his polling numbers. The Green Party candidate and 11 write-in candidates will account for around 0.5% combined.

Assumption 2: Straight-ticket voting accounts for 66% of all votes cast

This would be a record-setting level, at least in the modern era of Texas politics. A lack of partisan competition across much of the state and the highly polarized views of the electorate suggest fewer people voting a full ballot.

Assumption 3: The 15 counties with the most registered voters account for 67% of all early voters.

This was the actual figure for 2012 and 2008. In that case, the total number of early voters in Texas is around 6.6M.

Assumption 3A: Election Day turnout of around 2.4M.

The last time we saw record-shattering early voting turnout, Election Day turnout tumbled by 1M voters. The data from the 15 counties with the most registered voters suggest that people are voting earlier than usual, but there are enough new voters that a drop in turnout similar to the one experienced in 2008 is not likely. We are estimating a drop of just over 500K Election Day voters from 2012.

Assumption 3B: Thus, overall turnout will be around 9M, a record by around 1M voters.

We originally mused that turnout might drop for the second consecutive presidential election for the first time in state history. Barring a catastrophic drop in Election Day turnout, it instead appears that a record – at least in terms of number of votes cast – will be set this year.

Assumption 4: Hispanic/Latino voters will set a record for participation.

Evidence from early voting suggests stronger than normal turnout among voters with Hispanic/Latino surnames, and such surnames make up a slightly greater proportion of new voter registrations than the overall pool of previously registered voters. The extent to which this is a reaction to Trump or a reflection of demographic change remains to be seen.

Assumption 5: At the state level, the gains in Hispanic/Latino Democratic voting will be more than covered by increases in Anglo suburban and exurban voters.

We covered the meat of this here. Essentially, the growth in Republican areas, given their greater propensity to register and vote, should wash out any gains in Hispanic/Latino voters at the macro level. However, at local levels, increased Hispanic/Latino turnout will be more noticeable, and it could be enough to swing some close seats toward the Democrats.

Assumption 6: In general, Republicans will fare better in early voting than on Election Day, and Democrats will fare better on Election Day.

This is consistent with recent election cycles. In key races, Democratic candidates historically fare better, sometimes 10 or more points better, on Election Day than in early voting. If Election Day turnout is robust enough, it can be enough to flip results.

Assumption 7: Expect the unexpected.

Polls, even perfectly designed and sampled ones, are at best a snapshot of what voters, and some non-voters, claim they’re going to do. Past trends do not guarantee future results. It might rain. Most voters are angry at Washington and don’t like either major party presidential candidate. How motivated are younger voters, undecided and uncommitted voters, Hispanic/Latino voters and first-time voters to show up on a Tuesday after passing on 12 days of chances to vote? How many adults who favor Clinton and the Democratic Party but who aren’t planning to vote decide their voice needs to be heard? Will women who might otherwise vote Republican break rank and vote against Trump because of his crude comments. Will confusion over identification requirements squelch turnout among some ethnic groups? While there aren’t as many opportunities for unexpected upsets thanks to largely noncompetitive districts and high levels of straight-ticket voting, the game still must be played.

Our live coverage begins at 7 p.m. Central Standard Time, when polls close in all but El Paso and Hudspeth Cos. Polls close there an hour later.

To receive our free live updates, email us at [email protected] or follow us on Twitter at @TXElects.