The Paxton Variable
Democrats have not won a statewide race in Texas since 1994. G. Elliott Morris, a data journalist and former elections analyst at The Economist, knows this history well -- he once called Texas the party's "white whale." But in a March 2026 analysis for his Substack publication Strength In Numbers, Morris lays out six data-driven reasons why 2026 might genuinely be different.
The linchpin of his argument is the Republican primary. If Ken Paxton wins the nomination over incumbent John Cornyn, Morris believes Democrats could have a real path to flipping a United States Senate seat in the Lone Star State.
Paxton has been dogged by scandal after scandal for over a decade. He was indicted on felony securities fraud charges in 2015 for persuading investors to buy stock in a tech company without disclosing he was being paid for it.
Morris catalogues Paxton's legal history with prosecutorial precision: the 2015 securities fraud indictment, the 2023 impeachment on 20 articles by the Texas House, a federal corruption investigation, and a messy public divorce. Polling data backs up the damage. A Ragnar Research survey from November 2025 showed Democrat James Talarico running even with Paxton while trailing Cornyn by six points. That three-point gap between the two Republican candidates could be decisive in a state where Beto O'Rourke lost by just 2.6 points in 2018.
Talarico's Coalition Advantage
On the Democratic side, Morris argues that primary voters made the right call selecting state House Representative James Talarico over United States Representative Jasmine Crockett. The evidence is granular.
Talarico led among white voters 71-29 and Hispanic voters 60-39. He also won among both 2024 Kamala and Trump voters, with Crockett carrying non-voters.
Morris validates this crosstab analysis by pointing out that the final Emerson College poll nearly matched the actual primary result -- 52-47 predicted versus 53-46 actual. That accuracy lends credibility to the demographic breakdowns. Talarico's adjusted favorability rating hit 65% in the University of Houston's Hobby School poll, compared to Crockett's 56%.
Critics might note that favorability ratings this early are heavily shaped by name recognition and that Talarico remains relatively unknown to most Texas voters. High favorability among people who have actually heard of you is not the same as broad appeal.
A Turnout Earthquake
The enthusiasm numbers Morris presents are striking. Democratic early vote totals in the March 2026 primary ran 274% above the 2022 midterm primary. Republican early votes increased too, but only by 106%.
In Dallas County, nearly 188,000 Democrats voted early compared to 64,000 Republicans -- a 3:1 ratio. That's the highest Democratic primary turnout since 2008, when Obama was on the ballot.
The county-level data is even more telling. In Tarrant County, long a Republican stronghold where Trump won by double digits in 2024, Democrats outpaced Republicans during early voting. In El Paso, voters under 30 nearly tripled compared to 2022.
Perhaps the most revealing statistic comes from election analyst Derek Ryan's finding about voter history:
28% of Democratic early voters had previously only voted in November general elections. These are people who don't normally bother with primaries at all, but evidently decided this year was different. On the Republican side, that figure was just 13%.
Democrats are activating general-election voters at twice the rate Republicans are. That matters because primary-only voters are already locked in. The persuadable middle is showing up on the Democratic side.
A counterargument is that Morris himself acknowledges the limits of primary turnout as a predictor. In 2008, Democrats cast two-thirds of Texas primary ballots and then lost the state to John McCain by 12 points. Primary enthusiasm does not automatically convert to general election victory, especially in a state with Texas's structural Republican advantage.
The Latino Swing and a Diversifying Coalition
Morris points to a January 2026 special election in Texas's Ninth Senate District as a leading indicator. Democrat Taylor Rehmet beat Republican Leigh Wambsganss by 14 points in a district Donald Trump had carried by 17 points just 14 months earlier.
Majority-Hispanic precincts swung an average of 34 points toward Rehmet compared to the 2022 baseline. VoteHub estimated Rehmet won roughly 79% of the Hispanic vote in the district, compared to the ~54% Democrats captured among Texas Hispanics in 2024.
Morris supplements this with primary data from Hidalgo County, where 52,000 voters chose Democratic ballots versus 15,000 Republican -- in a county Trump won in 2024. He also tracks the broader shift in the Democratic primary electorate, which is becoming more demographically representative of the state. The white share of Democratic primary voters in Dallas County climbed from 34% to 39%, while Black voters' share declined from 42% to 36% -- not because Black turnout dropped, but because other groups surged in.
A more representative primary electorate means you get candidates who are better candidates for the general election.
The logic is sound. A primary electorate that mirrors the general electorate produces nominees who can compete statewide rather than candidates who appeal mainly to a narrow base.
Trump as an Anchor
Morris's sixth factor is Trump's declining approval in Texas. University of Texas tracking polls show the president's approval sliding from 52% at inauguration to 45% approve versus 49% disapprove by February 2026. Among independents, the collapse is dramatic: net favorability went from negative 22 to negative 48 over the course of a year.
Even on his best issue, border security, Trump's approval rating among all Texas voters is just +12. On the economy, which voters consistently name as their top issue, he's underwater at 41-48.
Morris argues this creates an asymmetric problem for the two potential Republican nominees. Cornyn has an independent brand built on decades of constituent service and could plausibly outrun Trump's numbers. Paxton, who has explicitly branded himself as Trump's candidate, is tethered to the president's performance.
When the president is at 45% approval in your state with independents at net -47, you have to do some major work to dig yourself out of his hole.
Bottom Line
Morris builds a cumulative case that is stronger than any single data point within it. The Paxton electability drag, Talarico's coalition breadth, record Democratic turnout, a diversifying primary electorate, Latino recoil from the Republican Party (often abbreviated as the GOP, or Grand Old Party), and Trump's sinking approval all point in the same direction. His final assessment -- a 45-50% Democratic win probability if Paxton is the nominee -- is carefully hedged and defensible.
The vulnerabilities in the argument are real but acknowledged. Primary turnout has been a false signal before in Texas. Special elections with 15% turnout are imperfect analogues for a general election. And the entire thesis depends heavily on the Paxton variable; if Cornyn wins the May 26th runoff, Morris's own polling data shows the race reverting to a more conventional Republican lean.
What makes this analysis valuable is its discipline. Morris does not claim Texas is likely to go blue. He claims the conditions exist for it to happen, and he backs each condition with specific polling data, county-level turnout figures, and historical baselines. After 32 years of "this is the year" false starts, that restraint is itself a signal worth noting.