We’re getting to the point in the election cycle where we’re starting to get a lot of polls — enough that it will occasionally start to feel as if the polls are swinging from left to right and back, every day.
This is one of those days.
This morning, the latest New York Times/Siena College polls of Georgia, Arizona and North Carolina find Donald J. Trump leading Kamala Harris in all three states, with a lead of three points in North Carolina, four points in Georgia and five points in Arizona.
In one sense, the finding is very clear: It’s a good set of numbers for Mr. Trump. It’s not necessarily an exceptional one, as Ms. Harris’s easiest path to victory lies in the Northern battlegrounds. But Mr. Trump’s path to victory begins in the Sun Belt; if he wins these three states, he wouldn’t need many more.
In another sense, this result might give you a bit of polling whiplash. As I wrote in my weekly polling summary, the polls have been relatively stable over the last week — or even a hair better for Vice President Harris than polls taken before the debate. These latest Times/Siena results don’t fit that story at all: On average, the results are five points better for Mr. Trump than they were when we last polled these states, in mid-August. And in each state, these results are the best results for Mr. Trump from a high-quality pollster in weeks.
In circumstances like this, our rule of thumb is to toss the results into a polling average — like our Times polling average.
Of the three states, Georgia is the one where the balance of polling is most clearly favorable to Mr. Trump. Ms. Harris hasn’t led a poll of Georgia taken in September, and Mr. Trump has led in several high-quality polls, like a three-point lead in a University of Georgia poll last week.
North Carolina, on the other hand, is arguably the one where the Times/Siena data most starkly departs from the most recent data. Our poll average previously gave Ms. Harris a narrow lead, as she’s had strong results from multiple high-quality polls — including our last Times/Siena poll of the state in August. This one is a very different result even if it’s also a return to normalcy in a way: Mr. Trump won the state narrowly in 2020.
Arizona is the murkiest. There hasn’t been as much polling of the state, period, and there’s never been a time when the polls there have strung together a consistent story. That goes for the last Times/Siena poll, which had Ms. Harris up five in August but now has Mr. Trump up five today. This is our biggest poll-to-poll swing of the year, and it’s hard to believe the race has really moved so much. It’s a helpful reminder that polls of this size (about 700 respondents) are subject to meaningful random sampling error. Nonetheless, Ms. Harris has led only two of the 18 Arizona polls fielded since that last Times/Siena poll; it was hard to make the case for a Harris lead even before this survey.
Now, many of you won’t be satisfied by the suggestion to toss these poll results in the average. And you know what: I don’t find it satisfying, either. Like me, you may want to understand the differences between the various polls. Better still, you might want to know which is “right.” Unfortunately, it’s impossible to know. But here are a few things for consideration:
-
The field period. The Times/Siena polls were fielded Sept. 17-21. This is more recent than a lot of the polling you may have seen lately (the NBC News poll may have been released Sunday, but it was fielded Sept. 13-17) and it comes entirely after the news of the second assassination attempt on Mr. Trump on Sept. 15. Last Monday, I speculated that the assassination attempt could put a dent in Ms. Harris’s debate bounce; perhaps that’s a factor.
-
Other recent Times/Siena polls. Last week, we released a post-debate national poll that was favorable for Mr. Trump (a tie) and a Pennsylvania poll that was quite good for Ms. Harris (a four-point lead). After adding these polls into the mix, the Pennsylvania result starts to look a little like the odd result out. This raises the possibility that, for whatever reason, the Times/Siena poll is producing relatively favorable results for Mr. Trump right now (though the strong results for Ms. Harris in many of these same states back in August throws a wrench in this theory).
-
Response rates. You might think that Mr. Trump’s gains might be driven by big changes in the likelihood of Republicans or Democrats to respond to a poll. But the balance of response rates by party and race stayed exceptionally similar to what they were in August, even as the results were very different. To take Arizona, where our results swung the most: In August, 2.4 percent of white Democrats responded compared with 2.4 percent of white Republicans; in this poll, 2.5 percent of white Democrats responded, compared with 2.5 percent of white Republicans. Whatever’s driving Mr. Trump’s gains, it’s not evident in the response rates by party.
-
MAGA candidates still struggle. While the presidential race swung a lot, the down-ballot races did not. In Arizona, the Democrat Ruben Gallego led Kari Lake, 49 percent to 43 percent, just a three-point swing from his nine-point lead in our last poll — even while the presidential race in Arizona swung 10 points.
In North Carolina, Josh Stein led Mark Robinson, 47 percent to 37 percent, essentially unchanged from our last poll (most interviews were finished before the revelation that Mr. Robinson had said on a porn forum, among other things, that he was a “Black Nazi”).
It’s interesting that Mr. Trump fares so much better than his MAGA allies, whom you might think of as fundamentally similar political figures — and it’s something to monitor to see if it holds through November. For now, the big swing in the presidential race, without a large down-ballot swing, suggests there’s something specific to the presidential race at play.
I’m not sure those explanations will prove especially satisfying. I guess that’s why “toss it in the average” is such a good rule of thumb.
Source: nytimes.com
