The South Texas Story Is Four Years Older Than You Think

Starr, Maverick, Duval, Webb, Willacy and Hidalgo all voted Democratic in 2020 and Republican in 2024 — six of the ten Texas counties that changed hands from Democratic to Republican that cycle. That is the version of the story that got covered, and every part of it is true. It is also four years late. The counties that flipped in 2024 had already done most of their moving in 2020, in a cycle the rest of the country spent moving the other way.

Here is the fact the coverage skipped. In 2020, the fourteen largest Republican swings in the country relative to the nation were all Texas counties. Not eleven, not twelve. All fourteen. The fifteenth was Miami-Dade. The file holds 18,671 county-cycles across 3,112 counties, and that year's list has one state at the top.

The fourteen

Lean is a county's two-party margin minus the nation's, so it strips out what the country did that year. Relative swing is the change in lean between cycles, negative toward Republicans. In 2020 the national two-party margin, as summed from the county file, moved 2.31 points toward Democrats — D+2.21 to D+4.52. Summed as one bloc, these fourteen counties moved 31.01 points the other way — from a lean of D+42.74 in 2016 to D+11.73 in 2020, and on to R+3.34 in 2024. Starr County's 2020 swing of 58.63 points toward Republicans is the largest single-cycle relative swing in the file, in either direction. The largest toward Democrats is Salt Lake County's 34.84 in 2016.

For all fourteen, 2020 is also their own record: every one posted its biggest single-cycle Republican move that year, not in 2016 and not in 2024.

Two-thirds of it arrived in one cycle

Of the bloc's movement from 2016 to 2024, 67.3 percent landed in 2020. County by county the concentration is higher still: the median of the fourteen put 80.53 percent of its eight-year move into that one cycle, and twelve of the fourteen put at least two-thirds there. Jim Hogg, which never flipped, did 92 percent of its moving in 2020 and has drifted 3.61 points since.

Widen the frame and it gets stranger, because before 2020 these counties were trending the other way. The bloc's lean runs D+25.32 in 2000, D+17.08 in 2004, D+34.76 in 2008, D+43.12 in 2012, D+42.74 in 2016. It peaked in 2012, more Democratic relative to the country than it had been at any point this century. The net across the full twenty-four years is 28.66 points toward Republicans, which badly undersells what happened: from the 2012 peak the move is 46.46 points, and 31.01 of them are one cycle.

2020 and 2024 are not the same event

The dating matters because the two cycles were made of different things. From 2016 to 2020, turnout across the fourteen rose 24.83 percent, Republican votes rose by 71,023, and Democratic votes rose by 4,726. Republicans did not take voters off the other side. They found voters who had not been voting. Starr is the extreme case: turnout up 49.28 percent, Republican votes from 2,224 to 8,247, Democratic votes down 166. In Webb the Republican vote exactly doubled, 12,947 to 25,898, while the Democratic vote fell 487.

2024 was the opposite shape and a smaller one. Turnout across the fourteen fell 2.34 percent. Democratic votes fell by 40,612; Republican votes rose by 32,806. That is close to a straight trade, on a shrinking electorate. Both cycles show up in the swing column as red numbers. They are not the same phenomenon, and averaging them into one arc called "South Texas shifted right" hides the interesting part.

Where it stopped being a Texas story

Run the same list for 2024 and the state monopoly is gone. Eight of that year's fourteen largest Republican relative swings are Texas counties. The other six are Imperial County, California at 19.75; the Bronx at 16.50; Queens at 14.95; Passaic County, New Jersey at 13.69; Miami-Dade at 12.88; and Hudson County, New Jersey at 11.93. In 2020 the list held one state. In 2024 it held five.

Our reading is that South Texas was an early instance of something that later showed up in a lot of places, not a regional exception with regional causes. That is an inference from the composition of a ranked list, and it is the weakest claim in this piece. The list identifies no cause and we have tested none.

What we are not claiming

Edwards County is the exception on our own table: it moved 0.92 points back toward Democrats in 2024, and it is the only one of the fourteen Republicans carried in every cycle since 2000. Starr holds the single-cycle record but not the era record — its 68.49-point total shift since 2000 ranks fiftieth of the 3,111 counties with a complete series, behind Jackson County, Tennessee at 101.83. Everything above uses the two-party margin, as everywhere else on this site; third-party votes are excluded by design.

Where the argument stops

These are also the most elastic counties in the file. Elasticity here is the regression slope of a county's two-party margin on the national one, where 1.0 means moving in lockstep with the nation. Five of the seven highest belong to our fourteen: Maverick at 5.106 is first, Webb's 4.277 is second, Starr's 3.859 fourth, then Hidalgo and Willacy. Historically, a one-point move in the national margin has shown up in Maverick as something closer to five.

That is amplitude, not direction, and it is where we stop. Amplification cuts both ways, and the file has these counties running hard in both directions inside twelve years. We are not going to tell you what South Texas does in November. We will say that if it moves it will move by more than the country does, and that the number to read afterward is not which party carried Starr but how many people voted there.