Four out of every nine early votes for Tuesday’s CD34 (Lean R) special election have been cast by Democratic primary voters, according to our analysis of voter rosters for all 11 counties in the district. Not quite a third of special election voters participated in the March 1 Republican primary, and just under a quarter of special election early voters did not participate in either party’s most recent primary election. Early voting by personal appearance ended Friday.

We previously analyzed early voting rosters in Cameron Co., which at the time accounted for about three quarters of all votes cast. That analysis found there were about 1,200 more Democratic primary voters than Republican primary voters in the county. The Democratic margin has grown to just over 2K voters in the county, which provides essentially the entire Democratic margin across the district.

Across all 11 counties in the district, Democratic primary voters outnumber Republican primary voters by just over 2K:

  • 7,098 Democratic primary voters (44.5%) – 5,323 in Cameron Co. and 1,775 elsewhere
  • 5,048 Republican primary voters (31.7%) – 3,255 in Cameron Co. and 1,793 elsewhere; and
  • 3,789 non-primary voters (23.8%) – 2,878 in Cameron Co. and 911 elsewhere.

Assuming primary voters remain with candidates of their chosen party, the Republicans would need 77% of non-primary voters to cast ballots for them to draw even with the Democrats in early voting. Adding Election Day voters reduces this percentage. The more who vote on Election Day, the smaller the percentage needed to overcome any early voting deficit, assuming Election Day voters lean more Republican than Democrat (We expect they will.).

County by County

Partisan splits for the 11 counties, listed in order of highest percentage of early votes cast to least, are as follows:

  • Cameron (72% of early votes cast) – 46% Democrats, 28% Republicans, 25% non-primary voters
  • Hidalgo (9%) – 52% Democrats, 25% non-primary voters, 23% Republicans
  • Kleberg (5%) – 39% Republicans, 37% Democrats, 23% non-primary voters
  • Bee (4%) – 57% Republicans, 24% Democrats, 19% non-primary voters
  • Jim Wells (3%) – 51% Democrats, 38% Republicans, 11% non-primary voters
  • Dewitt (2%) – 64% Republicans, 23% non-primary voters, 13% Democrats
  • Willacy (1.5%) – 51% Democrats, 27% Republicans, 22% non-primary voters
  • San Patricio (1.5%) – 44% Democrats, 38% Republicans, 19% non-primary voters
  • Goliad (1%) – 78% Republicans, 16% Democrats, 7% non-primary voters
  • Gonzales (<1%) – 76% Republicans, 15% Democrats, 9% non-primary voters; and
  • Kenedy (0.1%) – 89% Democrats, 5% Republicans, 5% non-primary voters.

Narrowing the gap in Cameron Co. on Election Day is crucial to Republican efforts to flip the seat, at least for now. Maintaining that gap is crucial to Democratic efforts to keep it going into the general election, which we rate as Likely Democratic.

About 21% of people who voted in the Republican primary and 18% of Democratic primary voters have already cast ballots in the CD34 special election. For the primary, Democratic voters outnumbered Republicans, 40K to 24K. Democrats have the numerical advantage among the not-yet-voted partisan voter pool, 33K to 19K. There are of course many, many more registered voters who did not vote in either primary and have not voted yet in this election, but the percentage of those non-primary participants who vote in the special election will be much, much smaller than the participating percentage of primary voters.

Absentee and In-person Voting

Just over 15% of early votes have been cast by mail, 59% of which were cast by Democratic primary voters and 14% by Republican primary voters, a difference of 1,104 voters. Non-primary voters cast the remaining 670 absentee votes, nearly double the number of Republican absentee voters but less than half the number of Democratic absentee voters.

Not quite two-thirds of voters who cast absentee ballots in the Democratic primary have already voted in the CD34 special election, leaving around 700 Democratic primary absentee voters’ ballots to be returned. Just over 60% of voters who cast absentee ballots in the Republican primary have voted so far in the special election, leaving around 180 Republican ballots to be returned. These figures do not include voters who cast ballots in person for the primary but are voting absentee for this election (or the other way around).

In-person voting is much closer between the parties. Democratic primary voters have cast 42% of in-person early votes, and Republican primary voters have cast 35% of them, a difference of 946 voters. Non-primary voters account for 23% of in-person votes, a slightly smaller percentage than their share of absentee votes.

About one in six Democrats who voted in the primary in person (including Election Day) have already voted in the special election, as have about 18% of Republicans who voted in the primary in person (including Election Day). Democrats outnumber Republicans, 18K to 11K, among those who voted early in person in their respective primaries but have not voted early so far in the special election. The margin is less than 1K among those who voted in the primary on Election Day but have not yet voted in the special election.


As we noted in the Cameron Co. analysis, these numbers suggest a runoff is likely between the top Republican, almost certainly Mayra Flores, and the top Democrat, likely former Cameron Co. Comm. Dan Sanchez. It is unlikely, though not impossible, for either to capture a majority, but each candidate’s ceiling is weighed down by another candidate on the ballot with their respective partisan label: Republican Juana Cantu-Cabrera received 7% of the vote in the four-way Republican primary won outright by Flores (60%), and Democrat Rene Coronado, who did not run in the primary election. Neither candidate met the fundraising or spending threshold to file a campaign finance report prior to this special election.

Assuming a 9-to-1 advantage for Flores among Republican voters and a 6-to-1 advantage for Sanchez among Democratic voters, that would put approximate early voting results as follows:

  • Flores would receive 4,500 votes to Cantu-Cabrera’s 500 among Republican primary voters; and
  • Sanchez would receive 6,000 voters to Coronado’s 1,000 among Democratic primary voters.

Flores would thus need a 1,500-net vote advantage among non-primary early voters and all Election Day voters to move ahead of Sanchez. On top of that, she would need to net above Sanchez at least one-half the total received by Cantu-Cabrera and Coronado to win outright. Sanchez would have a smaller hurdle under these assumptions, but probably still too great of one to win outright.


To conduct this analysis, we compared the counties’ voter rosters for the special election early voting period to the rosters for the March 1 primary elections, both early voting and Election Day. For Gonzales, Hidalgo and San Patricio Cos., we confined the primary voting rosters to those precincts which are involved in the special election. At the time we ran the analysis, the in-person voter rosters for June 10, the last day of the early voting period, from Jim Wells and Willacy Cos. were not available. Out of convenience, we did not go back any farther than the 2022 primary to determine primary participation. For purposes of this analysis, we assume partisan fidelity among primary voters, and any variations are assumed to be negligible and/or offsetting.

©2022 Texas Election Source LLC