A history of early voting turnout in the primaries

Early voting is over. We’ve already discussed the final EV totals, but there was something of note that I didn’t mention yesterday that I want to comment on today:


Year          Mail    Early     Total
=====================================
2018 Dem    22,695   70,152    92,847
2018 GOP    24,500   61,425    85,925
  
2022 Dem    17,810   82,631   100,441
2022 GOP    11,064   97,941   109,005

2026 Dem     8,894  211,390   220,284
2026 GOP     4,891  121,622   126,513

The decline in mail ballots from the past years to this year is kind of shocking. We know that the Republicans have made it harder to vote by mail, and we saw the effect that had in the first elections following that change. I don’t know if the people affected by this have adjusted, with a greater number of them deciding it’s better and easier to just vote in person if possible, or if there are now a significant number of people who don’t vote anymore. Note that the biggest dropoff was on the Republican side. Someone would need to dig into the voter rosters from the past years to see what has happened with the mail voters from then, but if there is a preventative effect, I have to think the Republicans have gotten the worse of it.

(Note: There was a small correction to the number of Democratic mail-in ballots, which was reflected in a later report that was sent out on Saturday.)

The big remaining question is, what might this early voting turnout mean for Tuesday? In previous early vote updates, I’ve noted Derek Ryan’s observation that Republicans now have more of a preference for Election Day voting, thanks to Donald Trump and his delusions (not how Derek put it, but this is my blog, so). If that is true, if Republicans are more likely to wait for Tuesday to cast their vote, then we can expect the Democratic to shrink by some amount in the end. Similarly, if some number of Democrats have shifted their behavior towards early voting, because of enthusiasm or whatever other reason, we might expect to see a smaller turnout from them on Tuesday. But first we have to know what we’re dealing with. So I went back and looked at final totals for all primaries since 2010, to see if there were any obvious patterns. Here’s what I got:


Dem
Year    Early    E-Day    Total   Early%
========================================
2010   40,963   60,300  101,263   40.45%
2012   38,911   37,575   76,486   50.87%
2014   31,688   22,100   53,788   58.91%
2016   87,605  139,675  227,280   38.48%
2018   92,847   75,135  167,982   55.27%
2020  145,158  183,338  328,496   44.19%
2022  100,441   67,687  168,128   59.74%
2024  105,913   72,195  178,108   59.46%  

GOP
Year    Early    E-Day    Total   Early%
========================================
2010   64,141   95,680  159,821   40.13%
2012   79,507   84,473  163,980   48.49%
2014   77,768   61,935  139,703   55.67%
2016  134,827  194,941  329,768   40.89%
2018   85,925   70,462  156,387   54.94%
2020  107,587   88,136  195,723   54.97%
2022  109,005   80,181  189,186   57.68%
2024  109,921   93,693  203,614   53.98%

Other than the surge in E-Day voting for both parties in 2016, and in 2020 for Dems, the early voting share has been pretty consistent over this span. I can explain 2016 and 2020 by noting they were both competitive Presidential years for the former, and for Dems for the latter, as in those races hadn’t been decided by the time Texas voted. You could maybe argue that the Republican share of early voting dipped from 2022 to 2024, but it’s basically the same as the previous Presidential year of 2020, so I wouldn’t take that too seriously.

That doesn’t mean that this year won’t be different. It wouldn’t surprise me if Republicans vote more, if only as just a share of the total vote, on Tuesday than Dems do. It wouldn’t surprise me if a greater share of Dems voted early this year than one would project from past performance; indeed, if something closer to 70% of Dems have already voted, it wouldn’t shock me. But this is what we have seen in the past. It’s more consistency than a trend. And if anything, having a competitive race where the outcome is not clear, as is the case in both Senate primaries, makes it more likely that Tuesday will be busier than expected, not less busy. There’s only one way to find out.

So that’s where we stand today. If we assume about 60% of the Democratic vote has been cast, which would be in line with what we saw in 2024, 2022, and 2014, then Dems should end up in the range of 330K total votes, which would just outdo 2020 for second place overall. If it’s more like 65% of the vote is already in, then Dems might fall a bit short of 300K. If Dems turn out on Tuesday as they did in 2020 – and note that 2020 had the highest early vote participation before this year – then Dems will approach 400K total, just short of the 2008 record of 410K. Voting in 2020 was happening at the very start of the pandemic, so I wouldn’t consider that to be in any way representative, though note that Republicans had the usual pattern for themselves that year. Bottom line, I think a wide range of possible outcomes is in play.

A couple of links to wrap this up. Trump was in Corpus Christi on Friday attempting to fire up Republican voters, who would seem to need the nudge. He’s also been endorsing candidates not endorsed by Greg Abbott, so that’s a bit of drama to look for after Tuesday. Rep. Jasmine Crockett threw a reporter from The Atlantic out of a rally in Lubbock recently, apparently because she objected to a profile the reporter had written of her in 2025. And finally, here’s a presentation from the Harris County Clerk called “Prepping for the Primary”, which mostly contains entry-level info that most of us here will know well, but there are some interesting bits about accommodations for voters who need assistance and about working elections, for which one can get paid $17/hour. Read through it, it’s all good to know.

I will have a look at voting in other counties tomorrow. Have you voted yet?

Related Posts:

This entry was posted in Election 2026 and tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

2 Responses to A history of early voting turnout in the primaries

  1. TexMike says:

    I vote by mail and find it hard to remember each January to download and send in the filled in forms. The number should increase during the year as people realize they have to do it each year.

  2. ken roberts says:

    Anecdotally, there seemed to be quite a few very elderly and disabled voters this year across the 10 voting centers where I worked as a clerk for this election. I talked to a few of the ones who were really struggling to walk into the polling center and several either didn’t trust the USPS or voting by mail. Others just liked voting in person. Curbside voting is available at every location, but that process can be very slow.

    As a caveat, I had only worked one of these 10 polling places in previous elections, so I don’t know if their demographics skewed elderly more than in prior years. I am well aware the voting population skews old, generally, especially in primaries.

    There seemed to be a fairly normal number of people who had a ballot mailed to them, but then decided to vote in person, anyway. This number is negligible enough to be ignored.

Comments are closed.