We continue our look at districts potentially in play in November with our first statewide exploration of straight-party voting trends in Republican-held districts. Prior analyses have looked at districts that are bluer than the state as a whole, districts won by Hillary Clinton in 2016 and districts where challengers have raised at least half as much as the Republican nominee. We also previously published our high-level analysis of federal and state and legislative seats we rate “lean Republican” or “toss up.”
There are two types of voters, at least for one more election.
One type of voter casts a straight-party ballot by marking their party of choice by using a single punch, mark, pull or other action. This single-punch, straight-party vote goes to every candidate of that party up and down the ballot. Within this group are straight-party Republican, straight-party Democratic and a much smaller number of straight-party voters supporting the minor parties. Single-punch, straight-party voters cast three out of every five ballots during the last three general elections.
The other type of voter does not cast a straight-party vote. We refer to them as “full-ballot” voters because they mark their choices up and down the ballot, although some may only vote in a handful of races. Full-ballot voters cast two out of every five votes in the last three general elections.
Any deficit a candidate faces with the first group of voters must be made up within the second. In almost every race on any ballot, the second group is smaller, so a candidate needs a higher percentage of that group than the other candidate got from straight-party voters to win. Sometimes it’s a much higher percentage. Often, it’s a bridge too far, especially since full-ballot voters tend to exhibit a similar partisan breakdown as the district’s true partisans.
At the state, legislative and congressional level in 2016, just six candidates prevailed when facing a deficit from straight-party voters:
- Rodney Anderson (R-Grand Prairie) overcame a 3,344-vote deficit to win HD105
- Matt Rinaldi (R-Irving) overcame a 1,835-vote deficit to win HD115
- Linda Koop (R-Dallas) overcame a 1,099-vote deficit to win HD102
- Cindy Burkett (R-Sunnyvale) overcame a 975-vote deficit to win HD113
- J.M. Lozano (R-Kingsville) overcame a deficit of around 60 votes to win HD43; and
- S. Rep. Will Hurd (R-San Antonio) overcame a 15,987-vote deficit to win CD23.
Anderson needed 62% of the full-ballot vote to eke out a 64-vote victory. Hurd received 60% of the full-ballot vote head-to-head against former Rep. and U.S. Rep. Pete Gallego (D-Alpine) to win by just over 3K votes out of 229K cast. Rinaldi received 57% of the full-ballot vote, comfortably over the 54% he needed, to go on to win by just over 1K votes. Burkett and Lozano won 70% of the full-ballot vote in their races.
But, 2016 was a presidential election year. Voting patterns change in a typical gubernatorial election year. For starters, a lot fewer people vote across the board. Compared to the preceding presidential election years, the number of votes cast in the last three gubernatorial elections fell 41% (2006), 38% (2010) and 41% (2014). Significant chunks of both types of voters historically sit out gubernatorial elections, but a greater percentage of straight-party Democratic voters stay home compared to straight-party Republican voters. Compared to 2012, 50% fewer people cast straight-party Democratic votes in 2014. Straight-party Republican voting fell 38%.
Thus, these Republican incumbents and one open-seat candidate (HD113) may not face the same straight-party deficit in 2018. Four years ago, all six districts had a Republican straight-party advantage. However, changing demographics and shifting voting patterns may nonetheless force these, and potentially other, Republican-held seats to be defended against a straight-party deficit or a relatively easily surmountable straight-party advantage.
Between 2012 and 2016, Democrats either cut into the Republican advantage or added to their own in 127 of the state’s 217 legislative and congressional districts, including 66 seats currently held by Republicans, 63 of which are on the 2018 ballot. This includes all of the districts listed above except for HD43, which became redder, at least in terms of straight-party voting.
The chart below ranks the top 25 shifts in straight-party advantage toward the Democrats between 2012 and 2016, measured as a percent of the votes cast in 2016.
Rinaldi’s district experienced the largest shift in straight-party advantage toward the Democrats of any Republican-held seat between the last two presidential elections. In 2012, Republican nominee Bennett Ratliff (R-Coppell) enjoyed a 5,471-vote advantage in straight-party voting out of nearly 53K votes cast. Four years later, Rinaldi overcame a 1,835-vote deficit in straight-party voting out of nearly 59K votes cast. The district’s straight-party advantage shifted blue-ward by 7,306 votes – equivalent to 12.4% of all votes cast in that 2016 race – in just four years.
In 19 Republican-held districts, the swing in straight-party advantage toward Democrats was equal to at least 7.5% of the votes cast in 2016. The second-highest swing was recorded in HD26, a Fort Bend Co. district held by Rep. Rick Miller (R-Sugar Land). In his case, the Republican straight-party advantage fell to 5,482 votes in 2016 from 13,748 votes in 2012, reducing a Democratic challenger’s full-ballot threshold to 66% in 2016, which is within the realm of possible for a candidate to obtain, from 92%, which is practically impossible.
The only way to overcome a straight-party deficit is to win the full-ballot vote by at least one more vote. In 2012, Democratic nominee Mary Clare Fabishak needed 67% of the full-ballot vote to overcome Ratliff’s straight-party advantage (She got 44%.). In 2016, Democratic nominee Dorotha Ocker needed 46% of the full-ballot vote to maintain her straight-party advantage (She got 43%.).
As the swing toward the Democrats increases, the margin needed among full-ballot voters to flip the seat decreases.
So, what does this mean for 2018?
To evaluate seats that may potentially be in play in 2018, we started by taking the straight-party advantage in 2014, measured as a percent of votes cast in that year, and then applied the 2012-2016 shift in straight-party advantage, again measured as a percent of votes cast. We took the resulting hypothetical straight-party advantage and calculated the share of full-ballot vote a Democratic candidate would need to flip the seat under those circumstances and, importantly, holding the ratio of straight-party to full-ballot voters constant.
We found that Republicans would still hold a theoretical advantage in straight-party voting in all but two of the seats they’re defending, under those circumstances.
In this purely hypothetical exercise, we found CD23 as the only district where the Democrat’s 2014 share of the full-ballot vote exceeded the percent share needed to flip the seat, in 2018.
But – and it’s a big but – this is a different political climate. Democrats are the ones running against Washington, switching places from Republicans in 2014. A record number of women are running, including 11 of the Democratic candidates running in the top 15 districts on that second chart. The Democrat at the top of the ballot is running the most competitive statewide campaign since at least 2002.
Instead of focusing on the specific numbers for each district – They are not predictions. – and taking those numbers as the hurdle needed for a Democrat to clear, this analysis should instead be viewed as a relative ranking of districts based on applying recent historical swings in straight-party voting to the 2014 results.
If we ran them off the 2016 results, the numbers would look much more favorable for the Democratic candidates. Somewhere in between is the likely scenario.
We’ll take a different look at straight-party voting trends next week to see if we can hone in more closely on the districts likely in play. The following week, we’ll revisit our analysis of challengers’ fundraising.
©2018 Texas Election Source LLC
