Q&A

A bipartisan pair of pollsters breaks down the state of the 2024 presidential campaign.

Vice President Kamala Harris speaks during a campaign rally in Flint, Michigan, on Oct. 4, 2024. New polling suggests most swing state voters now expect Harris to win. | Carolyn Kaster/AP

By Eugene Daniels

10/05/2024 10:00 AM EDT

Link Copied

Eugene Daniels is a Playbook co-author and White House correspondent for POLITICO.

Nobody trusts the polls these days — not even some pollsters.

“There are a lot of shitty polls out there,” said John Anzalone. And, added Greg Strimple, many hard-core Donald Trump voters aren’t responding to online surveys, which have become increasingly common.

They are two of America’s top political pollsters: Anzalone has done polling work for Joe Biden, Hillary Clinton and Barack Obama. Strimple has polled for John McCain, Chris Christie and Rick Perry.

In a conversation with the Playbook Deep Dive podcast, the bipartisan pair of pollsters laid out their views on the state of the race and who exactly are the undecided voters that both Trump and Kamala Harris are spending millions of dollars to reach.

They also both predicted at least a modest resurgence in ticket-splitting, with some voters backing a Democrat for president and Republicans for Congress, or vice versa.

In fact, new polling suggests most swing state voters now expect Harris to win — a perception that Strimple said could benefit Republicans in House and Senate races: “People are saying, ‘We’re going to put Kamala Harris in the White House. We’re going to have a check and balance on her.’”

This conversation has been edited for length and clarity by Deep Dive Producer Kara Tabor and Senior Producer Alex Keeney. You can listen to the full Playbook Deep Dive podcast interview here:

Listen to this episode of Playbook Deep Dive on Apple, Spotify, or wherever you get your podcasts.

There’s been a lot of chatter about the accuracy of polls — if they’re able to capture Trump’s support, Black voters’ opinions and more. Why should people feel that the polls are more trustworthy now than they have been in some of the recent cycles?

John Anzalone: Well, they shouldn’t. Because there’s so many polls. Like a third of the [polling] companies — I don’t even know who they are. It doesn’t mean they aren’t good polls. But the fact is, there are a lot of shitty polls out there, and what Greg and I do for our candidates and our clients and our corporations is so much different than what media polls do. We spend a lot of money, a lot of effort and a lot of labor using multimodal methodologies so that we’re getting hard-to-reach voters. A lot of times what the media concerns are using aren’t the best methodology.

Greg Strimple: The big thing is that undecided voters vote, right? And so when people see tight races like this and you’re not paying attention to who’s undecided and how they could break, then all of a sudden you have a problem because you’re saying, “Oh, he was ahead 47, 46, he should win.” Well, no. So that’s one thing.

The second piece is there’s been a lot of progress — I speak to this as a Republican — by the use of cell phones and online surveys to address issues with Hispanic people and Black people in our samples. One of the challenges that I discovered was that if you do online surveys of Republicans, you’re going to get more of a country club Republican than a hardcore Trump conservative Republican. So many of these surveys are online, and there’s some bleed among Republican voters who are more centrist, against Trump. So if you have a whole bunch more of those types of voters in a survey sample, it’s going to suggest that Trump isn’t as strong as he is. His folks are not hard to get to on the phone but hard to get to online.

What are the ways in which you think that people could get this election completely wrong?

John Anzalone: I actually think it starts with bad polling. Now there’s a whole industry trying to influence the aggregators like FiveThirtyEight and RCP with shitty, biased polls. What I wish FiveThirtyEight and RCP would do is have a subset of their aggregators, which is just five or six really credible polls. I mean, The Wall Street Journal has a multimodal methodology. I helped co-found it. If you’re going to do one online poll, I would do Pew Research because they have their own panel. And then you can take, if you want, NBC, CBS and maybe Washington Post? CNN — I hate their methodology — so, I wouldn’t include it. Pick five really good polls, say you’re only going to put it in the aggregator if they use likely voters and use the voter file and use multimodal methodology so you have a real take of what’s going on.

Greg Strimple: One of the funniest things I think of is that Fox News uses someone who clearly doesn’t know how to survey and it always makes the Democrat look better.

John Anzalone: It’s true.

Greg Strimple: I think that’s hysterical. The other thing I would say is the far left and the far right dominate the conversation in American politics. The middle of the electorate is very different from them, and they’re going to see it through a different lens. And so much of what’s being talked about in news and who you bring on TV or you’re interviewing in a newspaper........

© Politico