US House races 2026

Ah yes. That’s the hard one. People who’ve invested too much emotionally to walk away, like gamblers in over their heads?

This is, of course, false, almost sentence-by-sentence. But a happy update: the Republican attempt to sabotage the Democratic primary has failed.

Sex therapist Maureen Galindo, who had promised “prison for American Zionists”, received powerful support from Republican donors to the tune of $1 million, outspending her Democratic opponent. Last night must have been terribly disappointing for them, as Johnny Garcia trounced her 63.8% to 36.2% with 95% of the vote in.

While Republicans were clearly infatuated with the antisemitic sex therapist, Democratic progressives denounced her: U.S. Rep. Alexandria Ocasio-Cortez said, “This is absolutely disgusting… This bigoted garbage and antisemitism should be nowhere near our politics.” House Minority Leader Hakeem Jeffries called Galindo’s comments, “Disqualifying”.

To be absolutely fair, I’ll quote from Galindo’s website, via the Guardian. She attacks the media and makes some powerful denials:

“Maureen NEVER said that she wants anyone in internment camps!! She never said ‘internment camps’; that was a very unethical journalist desperate to take her down through a screenshottable article title.

Ok then. The website continues:

“She does want Karnes ICE facility to be turned into a prison for the billionaire zionists who have profited off genocidal prison state materials and trafficking. Prosecution has nothing to do with religion - they could be Evangelical, Catholic, Mormon, Jewish, etc.

She also want pedophiles to be castrated at the Karnes ICE facility. Very normal, at least for the RW Qanon crowd. I nominate Maureen Galindo for Sock of the Year.

That’s just mean. What have socks ever done to you?

Hopefully a Blue Wave in November will overcome even this.

This is mostly about the House.

Politico: Bernie Sanders’ candidates just keep winning

So far, so good; but I’ll withhold my applause till his candidates win their generals.

The Hill: Democrats hold 10-point lead over GOP on congressional ballot: Poll

https://www.nytimes.com/2026/06/23/nyregion/ny17-primary-cait-conley.html?unlocked_article_code=1.slA.7icd.aPZ_7f1XGnAl&smid=url-share

Cait Conley (gift link) will be the Democrat facing Mike Lawler in the Hudson Valley district of New York State, one of the few truly competitive districts.

Mamdani allies have a big night vs. Establishment Dems in the New York primary.

CNN Poll: Voters favor Democrats by an 8-point margin heading into the midterms

Take away 4 points for gerrymandering (just my own educated guess), and that leaves a current 4-point margin. Not bad. Expect this to narrow as GOP smear campaigns go into full swing, unless this is balanced by prolonged war and gas prices.

Oopsi-

Republican senator Bernie Moreno on Sunday publicly called for Representative Max Miller, his ex son-in-law, to step down from Congress amid allegations of domestic and child abuse.

As recently as Friday, the Ohio senator had avoided giving his opinion directly on accusations made by his daughter, Emily Moreno, against the congressman. “If there are any basic standards of character required to hold elected office, Max Miller fails them. He should not serve in the House of Representatives,” the Ohio senator wrote in a post on X on Sunday. “I believe Max Miller needs to seek professional help to end the clear pattern of abuse he has left in his wake. I believe he should not be free to continue endangering others until he does.”

Moreno also said he had hoped the situation would remain private but added that Miller’s “increasingly erratic and dangerous behavior has made that impossible”.

Generally Dem candidates do the right thing, even if they deny the allegations.

Except that that’s not the way gerrymandering works. Gerrymandering can erase a small swing, but it turns large swings into landslides.

I’ve mentioned it before but the way gerrymandering works is your side has, say, five districts with a 1-point edge and two where you dumped all of “those” people. I’m hoping enough voters are POed at Republican policies and shenanigans that those 1-point edges turn into 6-point deficits.

Except they don’t create +1 districts with gerrymanders. A +1 district is a swing district, could go either way, which is not what is desired. They usually want to create +6 to +8 districts, which is a more comfortable but not excessive majority.

That said, I’m fairly sure some of those gerrymanders (especially in Texas) are going to be dummymanders, where the Ds pick up seats rather than lose them.

In any single race, you’re right. But in the aggregate across dozens of races, when one party has more gerrymandered districts than the other, this can “translate” into a roughly calculable advantage that erases some of the other party’s averaged, overall polling advantage – if the goal is, say, to generate probabilities of a certain number of seats changing party. The folks at FiveThirtyEight did this pretty routinely. (I’d offer a cite, but sadly most of their articles have been removed from the intertubes).

The thing that made 538 stand out from the other forecasting sites is that they accounted for correlations in different events. And that’s exactly what prevents gerrymandering from being a simple linear shift.

Yes, I was going to mention that. As you said, FiveThirtyEight was always careful to note that there is some correlation among races (in terms of direction of polling error and certain other figures), and they built this into their predictive models.

But some districts are more gerrymandered than others. If the average shift nationwide ends up being, say, three percent, this will be enough to overcome the built-in gerrymander advantage in some districts, but not in others. This might seem wrong because it’s is using a general number to make predictions about specific races, but if your goal is to get a general sense of how the group of races are likely to shake out in the aggregate, it’s a useful “first pass.”

“If Democrats win 50% of the popular vote, we estimate them to win around 210 out of 435 seats. It is not until the Democratic party wins 51.1% of the vote that they would receive half of the seats in the House.” (This happens to be from 2023. Also, some of that effect is from geographic clustering – cities leaning Dem).

My 4% penalty estimate was probably pessimistic…but pessimism may be a healthy way to go (not just regarding gerrymandering effects, but in general), given recent elections.

True, if the nationwide shift is truly massive toward the party that had been “gerrymandered” in more states – currently the Dems – then you can reach that tipping point, unique to the gerrymandering process (i.e., distributing voters to create many “quite safe but not unsurmountably safe” districts), that favors the gerrymandered party in a bunch of races. If this 8% average polling advantage holds steady until November, and survives the polling errors that have tended to underestimate GOP votes, then this gerrymandering “backfiring” might actually happen. That would be great, but it’s historically rare.

ETA: Some call this “dummymandering.” It backfired against the GOP in the 2018 Texas state senate races (cited in linked article). Has it ever in US congressional races?

https://ohiocapitaljournal.com/2025/08/14/three-reasons-republicans-redistricting-power-grab-might-backfire/

(Also, I erred – FiveThirtyEight didn’t do predictive models, but rather forecast models. Subtle but important distinction.)