Can Texas GOP steal the midterms in advance? Yes, they can

Texas Republicans have an unusual mid-decade redistricting on the agenda for a special legislative session that begins this week. It’s entirely possible that the GOP will claim anywhere between an additional two to five congressional districts for themselves, potentially ending any realistic Democratic dreams of reclaiming the House of Representatives some 15 months before a vote is cast.

There’s not much that can stop them.

John Roberts and the U.S. Supreme Court have ruled that federal courts aren’t interested in partisan gerrymandering claims. State courts certainly exist in Texas, but serve to cement egregious GOP power grabs, not prevent them. Despite California Gov. Gavin Newsom’s midnight-drunkard ramblings that his state might magically undo its gold-standard citizen redistricting commission, there aren’t many gains left for Democrats that they didn’t maximize in 2021.

In other words, it looks like the Democrats are screwed here. Being Democrats, they’ve got one last, desperate ploy remaining: Maybe they can somehow or other scare Republicans into not doing this. Here’s the theory: A new GOP-drawn map might spread Republican voters too thin, backfire and end up electing Democrats.

Um, no. Believing this might happen requires closing one’s eyes and pretending it’s still the 1990s, which, to be fair, has a certain appeal. It also requires closing one’s eyes to all the congressional elections in all of the GOP-gerrymandered states since 2012, where not a single one of the extreme gerrymander maneuvers has backfired, or even come close to backfiring, even in election cycles when Democrats win tens of thousands more votes.

This bizarro-world-backward ploy sounds like a bonkers scheme that Joey from “Friends” might concoct to convince soap opera producers that he’s also a dancer, but we can count on the savvy reporters at Politico to swallow hard and take such nonsense seriously. This week, an article appeared there under the gullible headline “Republicans run a risky strategy for holding the House that rests on redrawn maps,” which for some reason required three bylines. “In Texas,” the article observed, “Republicans are in danger of creating a so-called dummymander, whereby an attempt to draw more seats for one party accidentally benefits the other.”

Reader, they are in no such danger.

Let’s give full marks for “dummymander.” It’s such a great term, attention-getting and alive. It allows reporters to take a simple story with a fairly clear villain — Texas Republicans determined to make their own state’s congressional delegation less representative to strengthen the national GOP’s midterm prospects — and turn it into something counterintuitive, clicky and packed with horserace game theory. The term was coined by Bernie Grofman and Thomas Brunell, two redistricting scholars and expert witnesses, in a 2005 chapter of “Redistricting in the New Millennium.”

But a dummymander has a specific meaning: It’s a gerrymandered map drawn so stupidly and greedily that it backfires. Think Wile E. Coyote and the burning wick attached to a bundle of Acme dynamite that he’s certain will trap Road Runner but leaves him in ashes instead. There are three classic examples. Grofman and Brunell named something real, but when journalists and partisans misuse the term now, they’re generally missing something important.

Technology has changed since 2005. Republicans have gotten really, really good at drawing egregious gerrymanders that don’t backfire. It is not 2005 any longer. These days, an actual GOP dummymander is as rare as someone opening the trunk to change the CDs. That’s not how anyone plays music and, likewise, Texas Republicans are not about to dummymander the state blue. That the media, some pundits and far too many Democrats seem to think that something similar might happen now both confuses wishful thinking with an actual strategy, and fails to understand how technological leaps and GOP mastery fundamentally changed gerrymandering over the last two cycles.

* * *

It is slightly crazy-making that Republicans reinvented redistricting during the 2010 cycle as a sleek, modern way to hold unearned majorities at the ballot box, but that so many people pretend these gerrymanders are like the ones politicians drew in earlier decades, more or less blindfolded and by feel. First, data journalists who should know better perpetuated a narrative that the pseudo-Gladwellian “Big Sort” theory was to blame — that Democrats in Pennsylvania, North Carolina and Wisconsin, for example, had clustered themselves into districts that looked like Donald Duck kicking Goofy. Study after study and election after election, in state after state, has proven this incorrect. (We laid out all the studies and court findings here.)

Now the hot word is dummymander. This is just as silly.

As Grofman and Brunell detailed with insight and wit, a dummymander is what Democrats did in Georgia in 1991, when they needed to draw three new majority-minority seats, but stretched themselves so thin, in an effort to protect incumbents, that the Peach State’s delegation flipped from 9-1 Democratic in 1990 to 8-3 Republican by 1994, and stayed that way the rest of the decade.

That same year in North Carolina, lawmakers drew two majority-minority districts but white Democratic incumbents foolishly tried to save their own seats and stretched too few voters too thinly. Again, under a Democratic map, a 7-4 Democratic delegation in 1990 reversed into a 7-5 GOP edge in 1994.

There is one more recent example of a dummymander — Arkansas in 2010 — but that too was a map drawn by Democrats, whose leadership, let’s face it, might still have CD changers in their trunk. Democrats, who actually controlled a trifecta in Little Rock that year, sought to replicate the 3-1 edge in the Congressional delegation they’d held for much of the 2000s. Alas, the map could not turn back time, and Democrats haven’t won an Arkansas seat since. (If there is the risk of a modern dummymander, look not at Republicans but at the 14-3 Democratic gerrymander of Illinois.)

This is not about to happen in Texas. Texas Republicans already control the state delegation 25-12, with one vacancy in a blue district. Of those 25 GOP incumbents, only one – Rep. Monica De La Cruz in the 15th district – won in 2024 with less than 60 percent of the vote. Even that race wasn’t particularly close. De La Cruz got 57 percent of the vote, beating her Democratic challenger by nearly 32,000 votes. Thirteen of Republicans either won with more than 65 percent of the vote or had no challenger at all.

Quite a few of those red districts have votes to spare. In each of the last two decades, 95 percent of Texas’s population gains have come from people of color. No problem for the GOP, whose  mapmakers “cracked and packed” them and retained 70 percent of the seats regardless.

So, no: A dummymander under these circumstances just isn’t going to happen. The two still-blue seats in South Texas could be crafted red with relative ease. If Republicans wanted to be more aggressive, they could dismantle urban districts and attach slices of city dwellers to larger rural districts. (This was the “pizza slice” strategy that divided blue Austin into multiple red districts last decade, then copied by Utah in 2021, which created four red seats by dropping a small corner of Salt Lake City into each of the state’s districts.)

Donald Trump has said he wants Texas to push five more seats into the GOP column. That’s potentially too many, and while it wouldn’t make the entire map backfire, it might cut some margins too close for incumbents’ comfort if just the right wave crashes just the right way. Trump is likely providing a smokescreen, so that when lawmakers tweak around the edges and push just two or three districts into the red column — say the 28th and 34th, the two most competitive blue-leaning seats — it might look like a compromise.

If they dismantle one largely blue city and attach it via hooks and tentacles to uncontested rural seats, they might get away with more than that without going all the way. Even so, they will double the number of House seats Democrats will need to win to reclaim the chamber, while shrinking the number of competitive targets. Next they’ll turn to a mid-decade redistricting in Ohio to claim two more Democratic seats, and look to the Roberts court to further eviscerate the Voting Rights Act, placing majority-minority seats across the South in the crosshairs.

* * *

At this point, the journalists and partisans who throw around the term “dummymander” would probably say something like: We didn’t mean that the Republicans’ map would backfire completely, according to the actual definition. We just meant that they’d better be careful that they don’t make a few districts more competitive, and maybe lose a couple. After all, they might add, Democrats took back the House with the “blue wave” of 2018, even after all the gerrymanders. Maybe those were dummymanders!

They were not. A simple look at districts won by Democrats in 2018 shows that they reclaimed the House by winning seats drawn by commissions and courts, and by forcing fairer maps via litigation in Virginia, Pennsylvania, Florida and, later, North Carolina. They also won seats in Kansas and Oklahoma that have since been gerrymandered off the board. They took a total of two seats in Texas and two in Michigan on gerrymandered maps, the only such flips on GOP-drawn maps of the entire decade. These maps were pretty much the opposite of dummymanders, yielded no gains at all for Democrats, over an entire decade, in Wisconsin, North Carolina and Ohio. A redrawn map that only gives back a couple of seats, in a wave year for the opposition, is an extraordinarily strong map.

Indeed, this is where congressional elections and state legislative battles since 2012 ought to be instructive when we think about Texas now and the prospect that any GOP-gerrymandered map might backfire. You could look at North Carolina. Or look at Michigan from 2012 to 2020, until citizens won an independent commission to draw lines: More votes for Democrats in every cycle still resulted in GOP state house control. Look at Pennsylvania, before the state courts intervened and ordered more representative maps ahead of the 2018 elections.

In Ohio, just to take one cycle during the last decade, voters had no choice at all in a quarter of state legislative races, no congressional race was within 19 points and the average margin of victory was 36 points. Ohio and Texas vote similarly in presidential elections. The Ohio map could not be dummymandered — as the emails of those who drafted the maps show, they spent late nights refining maps to gain every possible fraction of a percent.

Wisconsin, which is perhaps the purplest state in the nation, can barely be called a democracy after 15 years of gerrymandered GOP legislative majorities. Republicans have held either a 5-3 or 6-2 congressional delegation since 2012, even as presidential elections in the state are decided by a few thousand votes every four years. Under the gerrymandered state Assembly maps in place between 2012 and 2022, Republicans never held fewer than 60 of the 99 seats, even in years when Democrats won many thousands more votes. The map could not be dummymandered.

In the 2018 “blue wave” election, Democrats won the U.S. Senate race in Wisconsin and swept every statewide race. They won some 54 percent of the statewide assembly vote — but Republicans ended up with a 63-36 edge. The Milwaukee Journal-Sentinel did the math. Democrat Tony Evers defeated incumbent Scott Walker in the governor’s race by about 1 percent, but Walker carried 63 of the 99 legislative districts. Democrats were packed into a handful of seats they won overwhelmingly, while Republicans had lots of comfortably safe seats where their candidates got numbers in the high 50s and low 60s. As the Journal-Sentinel noted, not only did “Republicans increase the number of GOP-leaning seats, they increased their partisan advantage in those individual seats.”

Want more sharp takes on politics? Sign up for our free newsletter, Standing Room Only by Amanda Marcotte, also a weekly show on YouTube or wherever you get your podcasts.

In other words, even when hit by a blue wave in a purple state, these GOP maps were impervious to the dummymander.

Who did Texas hire as a six-figure operative to work on redistricting in 2021? Adam Foltz, one of the masterminds of those very same Wisconsin maps.

We don’t know whether Foltz is involved in the current project, but he has become a go-to name. Republicans have cultivated an entire bench of them. To believe that Texas might manage to dummymander its maps ahead of 2026 is to look at the magnificently complete Wisconsin gerrymander and believe that its architect might somehow screw the pooch, even with more sophisticated technology and better voter data.

* * *

In 2016, I published a book called “Ratf**ked: Why Your Vote Doesn’t Count,” laying out the GOP redistricting strategy of that decade, essentially for the first time, as well as the behind-the-scenes machinations in state after state. I spent time in subsequent years at conferences as the only journalist in rooms full of political scientists. This was not usually much fun. The political scientists arrived with slide shows that proved, to them, that redistricting did not matter, and gerrymandering over time had little to no effect at all. They had graphs dating back to the end of World War II that showed how gerrymanders eroded over a decade, how parties very rarely won any enduring edge through redrawing the maps.

I tried to argue that the 2010 cycle was different, that new technology plus extreme polarization had shifted the paradigm, and that even tenured political scientists might want to consider the idea that these modern gerrymanders had nothing in common with those from 1950, 1980, maybe even from 2000. I suggested they talk to the actual mapmakers, as I did, who could explain the magnitude of the changes in the technological power and data they had in 2001 versus 2011. This did not go over well. What about their chart?

Time and multiple cycles have now demonstrated that today’s technologically-driven gerrymanders are a new thing. It is impossible to look at Wisconsin, North Carolina and elsewhere and argue otherwise. The current maps have not lost their power or impact over the course of a decade. They have been, thus far, impervious to waves. They have maintained their strength despite population and demographic change. It’s time to look at this not as a game, or as game theory, but as the mounting threat to meaningful elections and fair representation that it is.

Could demographic trends in Texas accelerate enough to overwhelm a map? There is a nonzero possibility that could happen. We have all experienced surprising elections. But what we haven’t experienced since 2010 is any GOP gerrymander backfiring on itself. Not in a wave year. Not in any year. Not anywhere. It’s one thing to be acknowledge a remote possibility, but banking on one — or making it the focus of coverage the same week this anti-democratic scheme is announced — is to bury one’s head in the sand and play make-believe.

Politico quoted Texas state Rep. Julie Johnson, a Democrat, saying, “They are playing a little bit of roulette with these maps. … In a wave election like what we have a potential opportunity for in ‘26 I think it makes these Republicans very vulnerable.”

It might have been useful to ask when and where a Republican map has been vulnerable during any wave election. Johnson would not have been able to supply an example. Or one could look at the draft map circling in Texas political circles this weekend. It takes Johnson’s district in Dallas-Fort Worth and reorients it as a rural East Texas seat with a long arm slicing into a tiny sliver of the DFW metro region.

And let’s face it: Even if such an imaginary wave were actually to dummymander Texas blue, the GOP state legislature would have a good follow-up of their own. In the “heads we win, tails you lose” world largely created by the gerrymander and the extremists it has elected, entrenched and empowered, they’d just redraw the lines all over again.