Here is where things stand a month before the midterms. President Trump’s approval rating hit 31 percent in last week’s Associated Press poll. The Boston Globe’s James Pindell notes that this is the lowest for any president heading into a midterm in 36 years. For comparison, Trump went into 2018, which was a Democratic wave year, at 44 percent.
The AP-NORC survey was conducted Sept. 24-28 among 2,140 adults. It also found 69 percent disapproval and the lowest number of either of his terms, below a previous low of 32 percent from late 2017. Political Wire reported the same headline figure. Pindell writes that about two weeks ago Republicans started publicly distancing themselves from Trump in a way they hadn’t since he became president, and that it is now “every candidate for themselves.”
His cleanest piece of evidence is the money. Polls can be spun. Ad reservations are harder to spin, because nobody spends real money defending a seat they think is safe or chasing one they think is lost. On that measure, Republicans have stopped spending on the open Senate seat in North Carolina and on their challenge to Senator Jon Ossoff in Georgia. Both parties’ Senate money is going instead to Texas, which Trump won by 14 points in 2024 and where the Senate and governor’s races are now tied, according to the Globe. It is also going to Kansas, where Democrats have not won a Senate seat since 1932, and to Alaska, which has not voted for a Democratic presidential candidate since 1964. Pindell adds that Senate races in Nebraska, South Carolina and Florida are still rated easy Republican wins, while polling there says they are anything but.
So the battlefield has moved onto Republican ground. That part is fairly intuitive. The Florida House situation is more interesting.
The gerrymander as a leveraged trade
Ron DeSantis signed a new congressional map on May 4. It was designed to take Florida’s House delegation from 20 Republicans and 8 Democrats to 24 Republicans and 4 Democrats, a net gain of four seats from redistricting alone. Opponents sued, and the Florida Supreme Court turned the challenge away, with its majority saying it lacked jurisdiction. At the time, several Republicans described the plan as the thing that would keep Democrats from winning the House. (Rews covered the wider mid-decade redistricting fight last month.)
Pindell’s argument is that the map is about to backfire, and the reason is mechanical. A state has a fixed number of Republican voters. To draw more Republican districts, you have to take Republican voters out of seats that are safely Republican and spread them into seats that are currently Democratic. Every seat you gain has a smaller cushion than before. You have turned a few very safe seats into many reasonably safe ones.
Reasonably safe holds up fine in an ordinary year. Imagine you have a portfolio of 24 positions, and each one pays off unless the market falls more than a certain amount. Most years that is a nice steady income. In the year the market falls further than that, all 24 positions go wrong together, because they all depended on the same thing. A tight gerrymander works the same way. Its main risk is a correlated wave, and you don’t get to pick the year.
The Globe puts numbers on it. The Florida map works for Republicans until Democrats have about a five-point advantage. In the latest RealClearPolitics average of the generic congressional ballot, Democrats lead by 8.3 points. The Cook Political Report counts 14 newly drawn districts where the Republican advantage is 8 points or less. Add those 14 to the 4 seats drawn for Democrats and you get the Globe’s ceiling of 18 Democratic seats. Florida has 8 Democratic seats now, so 18 would be a gain of 10. The swing compared with the plan is 14 seats, in one state, from a map that was supposed to be the safe part of the Republican strategy.
A caveat. That ceiling is a national generic-ballot number laid uniformly across Florida districts, and real elections are messier than that. On the other hand, the Globe reports that Florida’s own Senate and governor’s races are now tied, which suggests the state is not immune to the shift.
The number that matters
Pindell’s view is that the most important number in the coming flood of polls is how many Republicans say they are excited to vote. In a low-turnout midterm you win by getting your own side to show up, more than by changing anyone’s mind. Emerson College’s latest national poll found 66 percent of Republicans “very excited” to vote, against 78 percent of Democrats. Because that is a poll of likely voters, it leaves out people who have already decided to stay home, so real Republican enthusiasm may be lower than it looks. The White House seems to read it the same way, which is why Trump is spending the final stretch in Texas, Oklahoma and Alabama. A president doesn’t usually need to campaign in Alabama.
The Fetterman buffer
In the Senate, Democrats need to gain four seats, and Pindell writes that they now have several ways to do it, even if Republicans flip a seat in Michigan or New Hampshire. Any of those paths runs through at least two states Trump won by more than 10 points.
Then there is the question of what counts as a majority. Pennsylvania Senator John Fetterman is a Democrat and says he has no interest in switching parties. He also just appeared at the Republican midterm convention, and last weekend he made it fairly clear, in the Globe’s account, that he would enjoy being a thorn in Democrats’ side if he ended up as their 51st vote. If you are Chuck Schumer, a 51-seat majority that depends on Fetterman is not one you can count on. Pindell’s conclusion is that the number Democrats actually need may be 52.
So the Republicans’ plan was built on margins too thin to survive a wave, and Democrats now face the same problem with a majority so thin it may hinge on Fetterman.