NPR's Mary Louise Kelly speaks with Anna Greenberg, a Democratic pollster, about what polls captured (and underestimated) after Democrats performed stronger than expected in midterm elections. Transcript MARY LOUISE KELLY, HOST: Polls predicted a good showing for Democrats yesterday. If anything, pollsters underanticipated the blue wave that was coming, Democrats flipped at least seven Senate seats, including in red states like South Carolina, Kansas, Texas.

Races predicted to be close were not close. Counties that went for Donald Trump two years ago saw double-digit swings toward Democratic candidates. So what did the polls get right? Where did they fall short? To answer those questions, we have called Democratic pollster Anna Greenberg. Welcome. ANNA GREENBERG: Thank you.

KELLY: What's your top-line assessment? This was a good night last night for Democrats, but a great night for polls?

GREENBERG: (Laughter) We needed one. Certainly, the polls were directionally correct. I think they anticipated a blue wave. I think they probably did not anticipate the tsunami, and I think that's for two reasons. You know, one, I think justifiably, many pollsters are nervous because we know there's a bias in who takes surveys, and Democrats tend to be more likely to take surveys. And so we tend to make our surveys a little more conservative than they actually kind of come in normally. And that probably diminished the share of the Democratic vote in preelection polls. Better to be safe than sorry (laughter).

But the other piece is that a lot of new people voted. And to do a poll, you have to know who to talk to. And if a lot of new people vote, you can't predict who they are very well. And that certainly happened in previous election cycles. It certainly happened in '16 and '20 and '24, where a bunch of new people came out and voted for Donald Trump, and we did not anticipate that they were going to come out and vote. And the same thing happened here. So I think we saw a lot of new people voting in primaries, and so I think we had some hints that that was coming, but the magnitude of it, I think, was unanticipated.

KELLY: That is interesting because I was going to ask you how much this being a midterm election matters, you know, compared to a presidential election where you usually would see higher overall turnout.

GREENBERG: Well, I mean, this midterm looks like it's comparable to 2018 in terms of size. But in terms of impact, it's closer to something like 1976. So I'm not even sure it's a normal midterm. In fact, it's, you know, possibly a realigning election in the sense that some people move permanently into the Democratic camp, and the Republican Party has to figure out what their future is. So I think it's even hard to compare it to any other election. It's such a - you know, the wave was such a repudiation of the status quo that it kind of defied what you normally see happen in midterms. It just happened everywhere, which is more like a presidential.

KELLY: Any major surprises where you looked at the results coming in and thought, wow, that is not what we pollsters had anticipated?

GREENBERG: I mean, the movement of young voters was pretty remarkable. But really the movement of young men. If you think about the posture of the 2024 presidential campaign on the Republican side - of course, which was a very bro-ey, male-podcaster kind of election - young men moved massively towards the Democratic Party. And so I think it's a very interesting dynamic that what they thought worked in 2024 had no impact in 2026 with young men.

KELLY: Go back to what you said at the very beginning, where you said we needed a good night, we pollsters. What is the conversation among people who do what you do today?

GREENBERG: Well, I think that many of us have felt a little aggrieved because the public polling, with some very notable exceptions, has been pretty mediocre and deceptive in a lot of ways. And so a lot of places seemed closer than they really were or were missed during primaries. And everyone said, oh, you know, the polling industry. And then there were the fake polls. You know, the polling industry needs to get itself together. And I think most of us who do campaign polling think, well, that's not us.

A lot of this is cheap online polling that they just push out for news coverage, not to try give strategy or campaign or that sort of thing. And so I think it's helpful for the campaign pollsters to say, look, we actually know what we're doing, and this public polling is actually the problem.

KELLY: Last question, which is, do you think what we saw in these midterms represents a profound political realignment or, you know, low approval ratings for this particular president or something else?

GREENBERG: It's an interesting question. I mean, realigning elections usually have some kind of exogenous event like a Great Depression - right? - that fundamentally changes the coalition, right? And so I don't - not sure that - you know, Trump may be that exogenous shock to the American political system, right? He certainly has had a major impact on our politics for the last 10 years. But I think you need to see something more like kind of permanent realignment in the coalitions.

And I think that remains to be seen. And so I think usually you identify realigning elections, like, way after they happen. And I think we need - you know, I think we need some time to see if some of these coalitions stick.

KELLY: Democratic pollster Anna Greenberg of GQR Insights & Action. Thanks so much. terms of use permissions pages at for further information.

Accuracy and availability of transcripts for NPR programming may vary. Transcript text may also be revised to correct errors or match updates to audio. Audio may be edited after its original broadcast or publication. The authoritative record of NPR’s programming is the audio record.

Please note that NPR platforms and products also include some programming produced by third parties; NPR cannot guarantee transcripts for that content and some transcripts may be produced using automated or other tools.