Everything’s bigger in Texas. Including the swing.

For a decade, Texas Democrats kept circling the same date. In 2018, Beto O’Rourke filled arenas, became a national phenomenon and came within 2.6 points of Ted Cruz, the closest Senate race in Texas since 1978. Two years later the state was called a battleground, and Joe Biden lost it by 5.6. O’Rourke tried again in 2022, against Greg Abbott, and lost by 10.9. By 2024, when Donald Trump carried Texas by 13.7 points and won Hidalgo County on the border, “Blue Texas” had become a punchline in both parties.

On Tuesday, James Talarico won the open Senate seat by 12.1 points over Attorney General Ken Paxton. Gina Hinojosa unseated Abbott by 7.1, making him the first sitting Texas governor to lose since Ann Richards in 1994. Democrats won all 15 statewide offices on the ballot, their first statewide victories of any kind since 1994, and 83 of the 150 seats in the Texas House, a majority.

Three things happened at once, and they explain almost everything below. A national environment that favored Democrats (our final forecast had them 9.6 points ahead on the generic ballot, with the president’s approval at 37 percent) reached a state that had shrugged off every previous wave. The Hispanic voters who moved sharply toward Trump in 2024 moved back, further than they had moved away. And the polls missed by about nine points, in a direction almost nobody had prepared for. A congressional map drawn last year to lock in Republican gains ran into all three.

These are unofficial returns. Nine of the counties below were still counting on Wednesday, and the state canvass comes later this month. Close races may change; we will update this piece after the canvass. All the data is downloadable at the end.

Thirty years of red, gone in a night

Before Tuesday, Republicans had won every statewide race in Texas since 1996. In its final ratings, Sabato’s Crystal Ball noted that Texas had the longest streak of any state of electing Republicans to every statewide office. The grid below shows every Senate, executive and high-court election since 1998.

The sweep reached the bench. Democrats won all four Supreme Court races on the ballot, including chief justice, and all three for the Court of Criminal Appeals, the state’s highest criminal court. Both courts had been entirely Republican. In January, Democrats will hold four of the nine seats on the Supreme Court and three of the nine on the criminal court.

Those court races also say something about why the wall fell. In races voters know little about, the result measures the party more than the person. Few Texans could name a candidate for the Court of Criminal Appeals, and the seven statewide judicial races went Democratic by an average of 4.2 points. The executive offices ran a little higher, from D+4.5 to D+7.1. That is roughly the party’s baseline this year: what a Democrat with no particular profile got from the 2026 electorate.

Talarico ran well above it, 7.9 points ahead of the judges and 5.0 ahead of Hinojosa. Some of that is the candidate and some is his opponent. Sabato’s Crystal Ball had called Paxton’s an “astoundingly weak” candidacy, and every other Republican on the statewide ballot drew more votes than he did. About 275,000 Texans voted for Abbott but not for Paxton.

The wave reached the courthouses, too. Democrats won 159 of the 301 district and appeals court races decided so far. In Harris County they won 40 of 41 partisan county offices; in Tarrant, long the most Republican of Texas’s big urban counties, Republicans still won 18 of 32.

A gerrymander built for 2024

In 2025, at President Trump’s urging, the Legislature redrew Texas’s 38 congressional districts in the middle of the decade. The goal was five more Republican seats. On the 2024 vote, the old lines had produced 13 Democratic seats; the new ones, by our estimate, would have produced eight. The map did it by turning five Democratic-held districts (TX-9, TX-28, TX-32, TX-34, TX-35) into districts Trump had carried, most of them Hispanic-majority seats in Houston and South Texas where he had made his biggest gains.

That bet assumed the 2024 swing would stay put. It didn’t. Democrats held four of the five targeted seats; only TX-32 fell. They also took TX-15 and TX-23, which Trump had carried by double digits. The delegation went from 25–13 Republican to 24–14: Democrats gained a seat on a map drawn to cost them five.

The rest of the map held, and that is the part Republicans will point to. The suburban seats were drawn with Trump margins of 15 to 25 points. They moved about as much as the state did, around 16 points, and every one stayed Republican, eight of them by less than six points. Democrats won the House popular vote by 5.7 points and 37% of the seats.

So how much did the redraw buy? Answering that means knowing what the old lines would have produced on Tuesday, and the swing was too uneven to shift every old district by the same amount: districts with Hispanic majorities moved about 1.9 times as far as the rest. So we measured how far each new district moved from the 2020 presidential vote to Tuesday’s House vote, related that movement to who lives there (Hispanic, Black and Asian shares, college degrees, density), and applied the same relationship to the old districts. The fit explains 45% of the variation, and refit with each district held out it calls 37 of 38 winners correctly. It gives Democrats 15 seats on the old map and 15 on the new one, where they actually won 14. On 2024’s vote the redraw was worth five Republican seats; on 2026’s, about one.

The new map is now much less safe than it looked a year ago. Under a simple uniform swing, its lines keep a Republican majority of the delegation until Democrats win the state by about 8.5 points, and then give way fast. Move the slider to see where each map tips.

A Democratic House in Austin

All 150 Texas House seats were on the ballot. Democrats won 83, a gain of 21 seats from a chamber Republicans had held 88–62, counting vacant seats for the party that last won them. Democrats now have seven more seats than the 76 they need, their first majority in the House since Republicans won it in 2002.

The count holds up. In every one of the 88 districts where counties have finished counting, their returns name the same winner as the Secretary of State.

The gains came from everywhere: seven in Dallas–Fort Worth, six around Houston, three in South Texas, two in San Antonio, one in Austin and two elsewhere. Many were close: 10 of the 21 were won by less than three points; the closest, House District 126, by 0.4. The canvass could still move a seat or two, though Republicans would need to win back nine to retake the chamber.

The Senate is another story. Only 16 of its 31 seats were up, mostly safe ones, and none changed hands. Counting the seats not on the ballot, Republicans keep it 19–12. Texas will have divided government for the first time in more than two decades: a Democratic governor and House, a Republican Senate, and a Democratic lieutenant governor, Vikki Goodwin, presiding over it. How much power she has there depends on rules the senators themselves adopt in January.

Where the vote moved

The House districts make the best map of the shift: together they cover the whole state, and each pairs a 2026 race with its own 2024 presidential vote. Grouped by metro, they show where Democrats started and how far each place moved. South Texas and the border went from D+9.5 for Biden in 2020 to R+6.7 for Trump in 2024, and to D+23.1 for Democrats this year. Dallas–Fort Worth went from R+2.5 to D+11.7. Rural and small-city Texas moved too, but from much further back: R+46.5 to R+32.1.

County returns, available so far for the 24 counties that publish precinct results, tell the same story at a finer grain. The biggest move was in Hidalgo County, from R+2.9 for Trump to D+46.5 for Talarico, a 49-point swing; El Paso County moved 33. The big suburban counties that decided O’Rourke’s race in 2018, Tarrant, Williamson, Denton, Fort Bend, each moved 22 points or more and voted for Talarico. The smallest moves came in rural East Texas: Wood County moved 11 points and stayed heavily Republican.

The whipsaw

The clearest pattern in these results is a reversal. In 2024 the most Hispanic parts of Texas moved sharply toward Trump; in 2026 they moved back, and then some. Split the contested House districts into fifths by Hispanic share, and the most Hispanic fifth went from D+17.7 for Biden to D+3.0 for Harris, then to D+33.3 for this year’s Democrats. The least Hispanic fifth moved far less: R+26.7, R+31.4, R+21.5.

Across all 133 contested districts, the Hispanic share of residents explains more of the 2024-to-2026 shift than anything else we measured. Education points the other way: districts with fewer college graduates moved more, which in Texas largely picks up the same Hispanic working-class districts. A regression that includes both, along with Black and Asian shares, keeps the Hispanic effect strong: holding the rest constant, a district 10 points more Hispanic moved 3.4 points more toward Democrats.

The districts that moved most toward Trump between 2020 and 2024 moved back the most: across districts the two shifts correlate at -0.70, and on average each point a district gave Trump in 2024 came back 1.2 points this year. Several South Texas districts that the new congressional map treated as Republican on the strength of 2024 are now more Democratic than they were in 2020.

These comparisons describe places, not people. A district that is 80 percent Hispanic and moved 30 points does not prove that Hispanic voters moved 30 points: turnout changed, and non-Hispanic voters in those districts count too. Voter-file and survey studies, when they come, will say more about individuals.

Who showed up, and how

About 9.9 million Texans voted for governor, 22 percent more than in 2022 and 87% of the 2024 presidential vote, with ballots still being counted. Among counties that had finished, turnout tended to grow more where Democrats were already stronger (r = 0.47). The fastest growth came in Kaufman (+46%) and Hidalgo (+44%); the slowest in Wood and Lamar, in rural East Texas.

Most Texans voted early. In the 23 counties whose returns split the vote by method, 77% voted early in person, 20% on Election Day and 3% by mail. Election Day was the most Republican vote in most counties, but only by a few points: early voters favored Talarico by 28.8, Election Day voters by 19.1. The returns counted late on election night were not much more Republican than the first ones.

Everybody missed

Almost nobody who forecasts elections for a living saw this coming. The consensus going into Tuesday was that Talarico was a modest favorite in a close race and that Abbott would survive. Sabato’s Crystal Ball moved the Senate race to Leans Democratic the day before the election; the Cook Political Report had it as a Toss-up. Prediction markets priced Talarico at about 74 cents. Every major rater had the governor’s race as Lean Republican. Our own final forecast gave Talarico a 73% chance and Hinojosa 34%.

A few saw the opening early. In March, G. Elliott Morris laid out “six data-driven reasons Texas could actually go blue”, several of them about Paxton’s electability, and by August Nate Silver’s model rated the Senate race Lean Democratic. None of them, and none of us, projected a margin anywhere near twelve points.

The forecasts were built on the polls, and the polls were wrong in one direction. Take each pollster’s final likely-voter poll of the two races, as published: every one of the 19 firms underestimated the Democrats, by 5.3 points at best (New York Times/Siena College) and 12.8 at worst (Quantus Insights), and the average poll missed by 9.3.

Texas polls have missed before, usually the other way. In 2020 they overstated Biden by 4.4 points; in 2024 the final average had Trump up about 7.5 in a state he won by 13.7. The last time they underestimated Democrats, in O’Rourke’s 2018 race, the miss was 2.5 points. This year’s is the largest miss in either direction in a competitive Texas race in our records, which go back to 1998. The biggest before it was 6.9 points, in the 2002 Senate race.

A shared polling miss like this is exactly what a forecast is supposed to allow for, and ours did, just not enough. Across the 183 contested congressional and legislative races, results ran 8.0 points more Democratic than our forecast on average, and 90% of them beat it. Because the model simulates errors that move every race in a state together, it still gave Democrats a 15% chance at the Texas House. But only 3.0% of its 40,000 simulations had them winning 83 or more seats. Pulse, the polls-only version, was more optimistic, at 33%. We will publish a full accounting of the 2026 forecast, in Texas and nationally, once results are certified.

What comes next

For a decade, the case for a blue Texas rested on demographics: a younger, more urban, more Hispanic state, and a party that only had to wait. That is not what happened on Tuesday. Texas did not turn because its electorate changed shape; it turned because the voters who had swung hardest to Trump two years ago swung back harder, in a year that ran against the party in the White House. That is a shakier foundation than either party would like. Republicans spent a redistricting cycle locking in a realignment that lasted one election. Democrats now govern on the strength of voters who have shown, twice in four years, that they belong to no one.

The canvass will settle the closest House seats this month. Then a Democratic governor and House will try to govern alongside a Republican Senate, for the first time in a generation. And in 2028 the same border and suburban voters will be asked again, on the same congressional map, with eight Republican seats sitting within six points.