Turnout
The numbers posted on the Secretary of State show that 2,165,744 people voted in the Republican Primary and 2,311,826 people voted in the Democratic Primary. Yes, more people voted in the Democratic Primary, 146,082 more to be exact. As I have mentioned before, this has happened several times over the past 25 years:
2020: Democrats 2,094,428 / Republicans 2,017,167 (Advantage D +77,261)
2008: Democrats 2,874,986 / Republicans 1,362,322 (Advantage D +1,512,664)
2004: Democrats 839,231 / Republicans 687,615 (Advantage D +151,616)
2002: Democrats 1,00388 / Republicans 622,423 (Advantage D +380,965)Three of those were presidential years, and in each of those elections, Texas was still in play for the Democratic presidential nomination while the Republican nomination had already been locked up. You have to go back 24 years to find the last time turnout in the Democratic Primary surpassed the totals on the Republican side in a midterm election.
A lot of you were probably still in diapers in 2002. That year, there weren’t any major battles for statewide office on the Republican Primary ballot. The only true undecided race was for Land Commissioner (Jerry Patterson vs. Kenn George). Meanwhile, the Democrats had a contested primary for senate between Ron Kirk, Victor Morales, and Ken Bentsen. To a lesser extent, the Democrats also had a race for Governor between Tony Sanchez and Dan Morales. (No relation between Dan Morales and Victor Morales).
While there have been contested midterm primaries in the Democratic Primary since 2002, none seem to have had the same driving power that we saw this year.
Early vs. Election Day
I am in complete shock by how few people voted on Election Day compared to how many people voted early. I first thought that might be the case when I showed up at my local polling place at 5pm on Election Day (in Travis County), expecting to see long lines, and there was no line. I was in and out within five minutes. The final numbers and data only confirms my personal experience. Outside of a few pockets, election day voters made up a significantly smaller share of the vote than we have seen in past primary elections. For comparison:
Republican Primary
2020: 46% of votes were cast on Election Day / 54% were cast early or by mail
2022: 46% / 54%
2024: 47% / 53%
2026: 36% / 64%Democratic Primary
2020: 52% of votes were cast on Election Day / 48% were cast early or by mail
2022: 41% / 59%
2024: 41% / 59%
2026: 32% / 68%The splits we saw in both primaries seem more in line with what we would typically see in a November General Election and not a primary. For comparison, in the 2022 General Election, the split was 68% early/mail and 32% election day.
I’ve had a few people ask why I thought Election Day numbers weren’t higher and my theory is this. For months now, Republicans and Democrats have been inundated with text messages, phone calls, mail, ads on social media, ads on TV/radio, ads on streaming services, you name it. By the time early voting rolled around, each likely voter had probably been contacted several times from various different campaigns. I feel like previous primary voters voters: (a) had already decided if they were going to vote or not and (b) if they had decided to vote, they had likely already made up their mind as to who they would vote for. In other words, there was likely very little information that came out between Friday 7:01pm and Tuesday 7:00pm that was going to convince people to vote (if they hadn’t already decided to vote).
There is some possible evidence of this being the reason. During early voting in the Republican Primary, non-primary voters (those who had either only voted in General Elections or had no history at all) made up 14% of votes cast. However, amongst Election Day voters only, this group accounted for 27% of votes cast. On the Democratic Primary side, the group’s share during early voting was 31% while the group made up 51% of the votes cast on Election Day. The previous primary voters had already made their decisions, leaving those who hadn’t been frequently contacted to account for a larger share of the vote.
[…]
Crossover Voters and New Primary Voters
Just over five percent (5.3%) of the votes cast in the Democratic Primary were from people who had most recently voted in a Republican Primary prior to 2026 and of the votes cast in the Republican Primary, only 2.2% were cast by voters who had most recently voted in a Democratic Primary. Democrats voting in the Republican Primary was about half of what we typically see. Once again, this goes back to the fact that the Democrats had a competitive primary which likely kept Democrats voting in the Democratic Primary as opposed to switching to vote in the Republican Primary.
In addition to primary voters crossing over to vote in the other party’s primary, there is always a group of voters who have no previous primary history who show up to vote. Over one-third (37%) of votes cast in the Democratic Primary came from people who had not voted in either party’s primary between 2018 and 2024. On the Republican side, the share was only 19%.
Primary Data and Predictions for November
I have said multiple times, the total number of people who voted in each party’s primary doesn’t necessarily mean anything in terms of predicting what could happen in November. It’s a nice talking point, but as noted above, Democrats have seen higher primary numbers than Republicans four times since 2000 and in each of those years, Republican statewide candidates still managed to win in November.
The total number of people who voted in the primary elections also is only about half of the equation for November. Combined, both parties accounted for 4.5 million votes cast. We will likely see double that number voting in November.
Having said that, while you can’t look at the top line number and point to possible projections for November, there are some takeaways from this month’s election that do help shed some light on November. One of those being how many people voted in each party’s primary who did NOT vote in the 2022 November General Election (i.e., new midterm voters). There were about 546k in the Democratic Primary and 262k in the Republican Primary who fit this criteria. That’s a difference of 284k votes in favor of the Democrats. In 2022, Governor Greg Abbott defeated Congressman Beto O’Rourke by 883k votes. In 2018, Senator Ted Cruz defeated O’Rourke by only 215k votes. So 284k votes could either determine a winner or it could simply just make for a narrower margin of victory.
This is not quite final because there’s still some data he was unable to get as of this report. He got about 97% of Dem data and about 95% of GOP data; as he notes, it’s a little easier to get Dem data because the large majority of it comes from the 20 most populous counties, while there’s a lot of Republican data in a bunch of small counties that aren’t as quick about providing the voter files. There may be another report when he’s at 100% for each, but it likely won’t be all that different.
You can find a full breakdown of the Dem data here and of the GOP data here. This includes more on primary history as well as the ages of the voters. About 46% of Dem voters were under the age of 50, while only 22% of GOP voters were that young. Make of that what you will. The Chron, which also summarized this report, has more.