New York’s Surprise Wins Delivered the House

A ll the undecided House races have been called, and Republicans will hold 222 seats. That’s just above the 218 they need for a majority.

For the first time in decades, New York was the deciding factor in a midterm election. Fueled by Lee Zeldin’s 47 percent showing against Governor Kathy Hochul, a dramatic improvement over the 36 percent showing of Republican gubernatorial candidate Marc Molinaro in 2018, New York Republicans this year won four Democratic congressional seats: two on Long Island and two in the Hudson Valley. They also held on to an open Syracuse seat that voted for Joe Biden by nine points. Those five victories are the difference between a GOP takeover of the House and Nancy Pelosi’s continuing on as speaker. New York was truly relevant in this election.

Democrats claim that the GOP gains came only because New York’s highest court — none of whose members were appointed by Republicans — struck down the legislature’s gerrymandered congressional map last April, ordering it replaced with a less partisan one drawn by a special master.

But many experts disagree. “There’s no place else in the country that Democrats lost districts that President Joe Biden won by ten points or 15 points,” Michael Li, senior counsel at the liberal Brennan Center for Justice, told a local news channel. “Something else is going on in New York that is unique to New York, and that has nothing to do with the maps.”

Indeed, Republicans had a potent set of issues — led by a crime wave, inflation, and taxes — to use against Democrats in the Empire State. But few pundits thought that even those would work in New York, where so many of the voters who once supported George Pataki, Rudy Giuliani, or Michael Bloomberg have died or moved to other, redder states. After all, Biden carried New York State by 61 to 38 percent in 2020.

The first sign that this election would be different in New York came around Labor Day, when a Trafalgar Group poll showed Democratic governor Kathy Hochul only four points ahead of Zeldin. Trafalgar’s survey got noticed because, as New York magazine noted, “in 2016 and 2020, Trafalgar Group did what many . . . other pollsters could not: come close to accurately portraying America’s support for Donald Trump” by finding his “hidden” or “shy” voters.

In early October this year, a second Trafalgar poll showed Hochul’s lead dropping to just two points. At the same time, the Siena College poll still showed Hochul with a 54 to 37 percent lead, though it noted that the crime issue was surging.

Zeldin’s fundraising picked up, and Republicans began to think there was a good reason to vote. The problem of rising crime clearly dogged Hochul. In her only debate with Zeldin, on October 25,  Zeldin quipped, “Halfway through the debate, she still hasn’t talked about locking up anyone committing any crimes.” Hochul replied, awkwardly, “I don’t know why that’s so important to you. All I know is that we can do more.”

In the week before the election, a third Trafalgar poll showed Hochul and Zeldin effectively tied: Zeldin at 48.4 percent and Hochul at 47.6 percent, with a margin of error of 2.9 percent. “Our own polls show that that’s wildly wrong,” Hochul told MSNBC, but behind the scenes she convinced national Democrats to spend millions on a huge get-out-the-vote effort. Hochul won 53 percent to 47 percent, the closest race for governor of New York in nearly 30 years. But traditional pollsters have nonetheless trashed Trafalgar for “errant polling.”

Robert Cahaly of Trafalgar attributes the error in his numbers to higher Democratic turnout and the likelihood that late undecideds broke for Hochul. “Major pollsters including the New York Times didn’t poll in the last couple of weeks in New York, so if you don’t stick your neck out, it’s easy to say you weren’t wrong,” he told me. Emerson College, the only significant pollster that also issued a survey, had Hochul winning by eight points, rather than her final margin of six.

Nationwide in the generic ballot, Trafalgar was tied for first in predicting that the GOP would win the House vote by over 3 percent. At the state level, Trafalgar clearly missed some key Senate races, such as those in Pennsylvania and New Hampshire. But its prediction was quite close elsewhere, as in Ohio and North Carolina.

When asked about the fact that his polls drew attention to the evident public discontent in a deep-blue New York, a discontent that ended up swinging the House to the GOP, Cahaly said, “A political outcome isn’t ever my goal, but uncovering voter sentiment that others miss is. On that level, I’m very satisfied.”

Cahaly believes that the public’s populist pique in New York that toppled Speaker Pelosi from power isn’t going away. He will tweak his turnout models and be ready to compare his results next time with other pollsters.

He continues to argue that the polling industry needs to rebuild confidence in itself: “Lots of pollsters got things wrong in 2022, and we’ll see who comes closest to fixing it in 2024.”