The polls have been fantastic for Democrats since Labor Day. They’ve led in nearly every survey of the crucial Senate races they need to win to flip the chamber this November.
And yet they didn’t lead in today’s New York Times/Siena survey of Maine, one of three relatively Democratic-leaning states where we’re releasing polls this morning. Here are the results:
This is the first poll to show Susan Collins ahead in Maine since Graham Platner dropped out of the race in July. Since then, Troy Jackson, who replaced Mr. Platner as the Democratic nominee, has led by an average of three to four points in the available polling, including a three-point lead in a CNN/SSRS poll taken before Labor Day.
It’s important to caution that our result is a modest outlier — outside the range of other polls. I’ll say more about that in a moment, but it would be a mistake to dismiss the poll because it looks different from the others.
Without Maine, Democrats’ path to Senate control would get a lot more challenging. They would need to make up for it by winning not just two but three seats in states where President Trump won by double digits in 2024, like Ohio, Iowa, Alaska, Texas or Kansas. That’s a daunting task but, incredibly enough, the polls have been so good for Democrats that this doesn’t seem entirely inconceivable.
The idea that Democrats could win three red states and lose much bluer Maine might seem far-fetched. Yes, Ms. Collins pulled off something similar back in 2008 — when she easily won in an anti-G.O.P. year in which Democrats reached 60 Senate seats — but that was nearly 20 years of polarization ago. In today’s era, it can be hard to imagine how Democrats could seriously be competitive in Kansas if they can’t even win Maine. It sounds impossible.
But if recent elections are any indication, it might not be as outlandish as it seems. It might even wind up making a certain kind of sense.
But first: Should we believe this poll?
There’s nothing more uncomfortable for a pollster than a result that looks different from that of other polls. It naturally gets more attention, but it also raises the question of whether the poll is accurate.
A good rule of thumb is to focus on the average of all of the polls. In the Maine contest, the average would be close to a tie after adding this latest Times/Siena poll to the mix, since there haven’t been many recent polls of the state.
The average is typically more accurate, and it mitigates the risk of being misled by a statistical outlier — a survey result that’s off, simply by the chance inherent in random sampling. This is the textbook explanation for why outliers happen, and with that in mind, I’ve often downplayed outlying Times/Siena poll results over the years.
But looking back, it’s often been a mistake to downplay them. More often than not, outlying Times/Siena results — defined here as different from the prior average by at least four points — have been more accurate than the average of other polls at the time.
And in this particular case, it’s even harder to dismiss this poll. It has a sample that looks reasonably favorable to Democrats. In fact, it looks almost exactly like the last high-quality poll of the race, the CNN/SSRS poll that found Mr. Jackson up by three points.
The similarity with that poll doesn’t prove the Times/Siena poll is right, but it means the result can’t be easily dismissed for having too many Republicans in the sample, or the other things that people usually point to when an outlier is released.
Instead, Ms. Collins may be leading for a simple reason: People like her better and think she would be a better senator.
She has the edge over Mr. Jackson on almost everything tested in the poll. She’s viewed positively by voters, with 51 percent holding a favorable view and 48 percent having an unfavorable one. Mr. Jackson, on the other hand, is viewed quite unfavorably, with 41 percent holding a favorable view compared with 49 percent who view him unfavorably.
At least to me, Mr. Jackson’s low ratings are the most striking result of the poll. He may not be a particularly charismatic figure, but he hasn’t been embroiled in a major scandal, and that’s what it would usually take for 18 percent of Kamala Harris’s supporters to view a Democrat unfavorably. For some perspective, Mr. Jackson’s ratings are even worse than Mr. Platner’s were in June, when he was fending off scandal before an allegation of sexual assault ended his run.
The poll didn’t provide a great explanation for why he’s this unpopular, but the poll isn’t an outlier in this respect, either. Yet again, the numbers are similar in the CNN/SSRS poll that found him ahead: Just 36 percent had a favorable view of Mr. Jackson in that poll, compared with 42 percent with an unfavorable view.
Can Collins really win in this environment?
Our poll isn’t simply an outlier compared with the other polls of Maine. It also just seems inconsistent with Mr. Trump’s extraordinary unpopularity, Maine’s Democratic lean and the last two weeks of exceptional polls for Democrats, including in states much more conservative than Maine, like Kansas or Texas.
Nonetheless, it’s not a scenario that ought to be dismissed. Beyond 2008, Ms. Collins pulled off a similar feat in 2020. And in every election I’ve covered, surprising things happen — things without such obvious precedent. In this polarized era, there’s always some demographic group, some region, some district that swings enormously — far more than anyone would have imagined.
Here’s one way to see it in the numbers: Over the last 20 years, every presidential election has featured at least one congressional district that swung at least 20 points from the last presidential race. At the state level, that’s easily enough to turn Maine “red” or Kansas “blue.”
Big swings are even more common in midterm elections. After all, every race is different. The election is less nationalized without a presidential election, and down-ballot races get a lot more attention. The turnout is lower and more variable.
The polls sometimes show signs of these swings in advance, but they’re easy to dismiss given all of the uncertainty inherent in political polling. Back in 2022, I downplayed a Times/Siena poll of Kansas’ Third District that showed a Democrat up 14 points. I did say that abortion rights was a plausible driving issue, but the poll simply seemed too good for Democrats, as even Democratic pollsters told me (and others).
But the poll proved accurate, and once the actual results arrived a clear pattern emerged on protecting democracy and abortion rights — not just in Kansas but across the country. And even when major shifts aren’t so obvious, they can sometimes be clear years later. In hindsight, the Republican sweep in Florida in 2018 was the beginning of the party’s rise to dominance in what had until then been one of the nation’s pre-eminent swing states.
What will we say after 2026?
The polls certainly offer signs that Democrats will surge among young and Hispanic voters. These voters are plentiful in a state like Texas, but Maine — one of the nation’s oldest and whitest states — is lacking in both.
Or maybe we’ll look back and say this was an election dominated by candidate quality — when voters flocked to trusted candidates with independent streaks at a time when both parties had been discredited by two consecutive unpopular presidencies.
I’m not predicting that Maine will be red and Kansas and Texas will be blue, of course. But if somehow that’s what happens, we might look back and say it made a certain kind of sense.
The post Collins Leads in Maine in Our New Poll. Is That an Outlier or an Omen? appeared first on New York Times.
