Clare Malone: This is The New Yorker Radio Hour. I'm Clare Malone. There's just no other way to put it, this presidential election has been shocking with the assassination attempt on Donald Trump coming as the latest in a series of historic and troubling events, and then there's the ongoing movement to convince President Biden to drop out of the race. At a press conference on July 11th, the reporter Haley Bull from Scripps asked him whether Kamala Harris should replace him.

Haley Bull: If your team came back and showed you data that she would fare better against former President Donald Trump, would you reconsider your decision to stay in the race?

President Biden: No, unless they came back and said, "There's no way you can win." Me. No one's saying that. No poll says that.

Clare Malone: Biden won the primaries already. There isn't really a precedent for this, but recent polls in key swing states show his support fading, and that has Democrats, including Congressman Adam Schiff, advocating desperate measures. To find out what we know at this point, the best assessment we can make about what's going to happen in November, I called up Harry Enten. Harry spends most of his time buried in the crosstabs of polls. He works at CNN as a data reporter and we worked together for years at the website FiveThirtyEight. We did reporting and analysis from polling and other data.

I reached him in Milwaukee at the Republican National Convention. Harry, Donald Trump was just the subject of an assassination attempt. This is unprecedented and I want to ask you how you think the shooting will affect the race.

Harry Enten: You say it's unprecedented, but I recall back in my younger life when Teddy Roosevelt was also the subject of an assassination attempt, in fact, got shot and was still able to give a speech afterwards in 1912 when he was running as the Progressive or Bull Moose Party candidate, but of course, that's not the era in which we're in right now, right?

Clare Malone: Yes.

Harry Enten: This is a very different era, a very highly polarized era. There is no real historical precedent for this as long as any of us have actually been alive. The only one really in polling history was when Ronald Reagan was president and obviously got shot. What we saw was his approval ratings did go up at least initially. According to Gallup, rose by a little bit less than 10 points, but two months later, it declined. I think that this is bigger than the numbers.

I think it's a larger reflection of where we are right now, that we are in an incredibly politically divisive time, one in which members of the other party truly dislike each other to a degree that, simply put, has not been the case as long as we've had polling data around.

Clare Malone: Do you think that Trump could potentially get a bump in polling and that it could, because of our partisan era and because of the really specific contours of this race, do you think something like that could be sustainable after this event?

Harry Enten: Because of the timing, it's going to be very difficult to disentangle what might be a bump because of the events that happened in Pennsylvania with a normal Republican National Convention bounce. Maybe if there's no movement in the poll, that might answer it, but if there is any sort of movement, it is going to be very difficult to figure out whether that was a normal RNC convention bounce or was that something had to do with the fortunately failed assassination. Emphasis on the failed, obviously.

I don't think, though, to be honest, in this era, if Reagan's any guide, if the debates have been any guide, if anything along those lines, something may peak at least initially, but then for it to sustain itself, that's something an order of magnitude more difficult.

Clare Malone: An additional factor that I'm thinking of is the Biden campaign pulled a lot of ads off the air in the wake of the shooting in Pennsylvania. They canceled events. The tenor of the media coverage and the Biden campaign's partisan rhetoric is very different now. I'm curious, do you think something like that could affect the contours of the race, the polling of the race, in the long term?

Harry Enten: Unprecedented times make for difficult-to-forecast times. I think that the general belief is it's going to be very difficult to truly change the contours of the matchup between Biden and Trump. Maybe what will end up happening is if the campaign becomes more positive, then maybe both men will be better liked than they currently are. It's difficult for them to be more disliked, at least in the polling, in terms of the unfavorable ratings than each of these guys are.

We do have something different going on in this election right now, but if you just look after the debates and you look at where Biden's numbers were pre-debate and post-debate, the numbers really didn't change despite the fact that most nonpartisan analysts would agree that Biden's performance in the debate was one of the worst in the modern era, if not the worst. Before the debate, a record number of people thought Joe Biden was too old to be an effective president. That number may be slightly higher now. It's not really a big movement pre-debate versus post, it's really a big moment from 2020 to 2024.

Clare Malone: I want to go back to President Biden and his pretty disastrous debate performance that is, at this point, infamous, I think. I saw some national polls that didn't show that much of a difference in the race after the debate, but in swing state polls, Biden was down. Can you explain that?

Harry Enten: Yes, I think that there are a few things going on there. I think nationally and in the swing states, for the most part, there has been a small movement away from Biden post-debate, but we're talking on the magnitude of closer to two points than, let's say, 10 points. In any individual poll, you could see no change. You might even see a slight shift towards Biden, but in the aggregate, you're seeing a small movement towards Trump.

Clare Malone: Who exactly are the voters who are leaving Biden?

Harry Enten: Who are the voters that are leaving him? It depends on which polls that you're looking at, but it did at least appear to me that independents did move by around five points towards Donald Trump compared to where we were pre-debate. Of course, independents are such an important group, a group that Trump won in 2016. Then Joe Biden won them by double digits back in 2020. If Trump is winning them by six, seven points as appears to be true in the national aggregate, then that is a very, very good sign for his campaign and one that will be awfully difficult for Biden to overcome if it does, in fact, hold.

Clare Malone: Do you think the media is blowing out of proportion the polling shifts away from Biden? Because that's been the headline, the shifts after the debate.

Harry Enten: I think what's going on for a lot of folks is that they're realizing that Donald Trump is actually leading in the polling in the average. Even if there hasn't been a massive movement, I think there's finally recognition in some quarters that Donald Trump is ahead, not by a ton but by more than enough, especially given that he seems to be running stronger in the important battleground states than he is nationally.

Clare Malone: Demographically, who are those independents who might be moving towards Trump? Are they young? Are they voters of color? Who are those voters?

Harry Enten: They're both, younger people, voters of color, especially Hispanic voters tend to be the voters who are more likely to identify as independent than perhaps in past years. Because independents, a larger share of them are younger or Hispanic than the populace at large, and so there's no doubt that there is that overlap there in a nice Venn diagram, and so that is a significant part of that movement that we've seen away from Joe Biden versus four years ago.

Clare Malone: Who are the voters that are sticking with Biden even after this really bad debate performance?

Harry Enten: Aggressives, a very liberal voters, older Black voters, older Black women, in particular, most Democrats. That is the situation. It's not like magically these Democrats are going to be like, "You know what? I want Donald Trump to be the nominee or want him to be the president." Donald Trump is still a very unpopular man in this country, at least according to the polling data.

Clare Malone: How do we square the fact that he's unpopular according to polling data and yet he's leading in the polls?

Harry Enten: The way you square it is that you have Joe Biden, incumbent president, whose disapproval rating is in the mid-50s, whose approval rating is 40% or below. There's just no president who's been reelected with those types of numbers ever. Donald Trump, despite himself being unpopular, finds himself in a similar position that he did in 2016 with the Democratic Party, for reasons, pick a nominee that left the door open for somebody who is as unpopular as Donald Trump is. It's really that simple. Donald Trump just has many more paths to victory at this point than Joe Biden has.

Again, in this country, in which elections are won and lost, in which has not had a single major party candidate win the election by double digits since before either you or I were born, a shift of a few points could make all the difference in the world.

[music]

Clare Malone: Harry Enten, senior political data reporter for CNN. We spoke last week while he was in Milwaukee covering the RNC. I'm Clare Malone, and this is The New Yorker Radio Hour. More to come.

[music]

Clare Malone: This is the New Yorker Radio Hour. I'm Clare Malone, a staff writer with The Magazine. In presidential election years, we all want to know what's going to happen, but we're drowning in poll numbers, national polls, swing state polls, favorability ratings, issue polls, you name it, we've got numbers for it, but it's really difficult to separate the wheat from the chaff. There are partisan polls, which may well be biased, junky online polls, and there are high-quality polls that still take the time to call hundreds of people and ask them questions, but even those types of polls can seem suspicious. How can you get an accurate sample when so few of us actually pick up strange numbers calling our phones?

Are pollsters putting their fingers on the scales? I called up Ann Selzer because I wanted to hear from her how polling actually works now, how the polls know what they know. Selzer is one of the best people to explain that. Her company does the polling for The Des Moines Register, and over the years, she's established a track record for accuracy.

Ann Selzer: There's a particular cluster of people here in Des Moines whose nickname for me is harbinger of doom. Now, these are all people who have worked on Democratic campaigns and thought that they were coasting to a win, and then my final poll comes out and it shows their Republican opponent, in fact, leading with a solid lead. It's an example of what happens if you don't have good polling. Would they have done things differently in their campaign had they been more apprised of what was actually happening?

Clare Malone: One of the reasons Selzer has been so effective as a pollster is a method she calls polling forward. She looks at the current demographics of the country instead of who turned out in previous elections.

Ann Selzer: We make no assumptions about what the electorate is going to look like. We don't look at past elections, and there are pollsters that do. They go back and look at the exit polls and say, "This is what the racial makeup was, so we're going to make our data look like that." I call what I do polling forward. The future electorate is going to reveal itself through that methodology.

Clare Malone: That's interesting because my familiarity is with this idea of, hey, this election in 2016 or 2020 looked this way, we're going to weight our data based on that. How do you actually concretely pull forward? What does that look like?

Ann Selzer: The moment that it was clear to me that my method was on to something was in 2008 with the Iowa caucuses, and our final poll was showing a surprise, that Barack Obama was winning easily, but the bigger surprise in our data was that 60% of the people who planned to caucus on the Democratic side, it would be the first time they had ever gone to caucus. That 60% just created a furor and an uproar and there was lots of media talk, lots of campaign memos going about saying, "Pay no attention to this poll because this is a number unheard of."

I had one of Hillary Clinton's state co-chairs, a friend of mine, he's been a client, called to say, "I've knocked on 99 doors, I haven't found this lurking Obama support," and I said, "Tell me about the 99 doors." He said, "Oh, we're focusing on previous caucus attenders and registered Democrats" and I made no assumption.

Clare Malone: What are you seeing now in your polling forward for this election cycle?

Ann Selzer: We're seeing a couple of things that are interesting the way the electorate is aligning. Traditionally, older people have been more on the Republican side, and conventional wisdom is the older you get, the more conservative you get. The age group of 65 and over is Joe Biden's strongest demographic group, AND younger people are aligning more on the Republican side. I think this is a perfect example of if you're not doing polling, you're going to behave the way you behaved in the past and assume that the older people who show up in proportionately larger numbers on election day are going to be voting Republican, and the opposite is true.

Clare Malone: I don't think it's a secret to you that polling has become controversial in American life. Sell me on it, what are the benefits of good polling? Give me your best pitch.

Ann Selzer: I think it depends on how high you think the stakes are. If you want to have a good idea of the way the electorate is moving, because it does move, you need to have polling that will reveal that. I think it's very important to stay on top of how the mood of the nation is changing, how the mood in states is changing. If you don't know what it is, you're just shooting in the dark.

Clare Malone: Talk me through how polling actually works. What I mean by that is who pays for it, who writes the questions, who makes all the actual phone calls?

Ann Selzer: I can talk with most accuracy about how we conduct our polls. I use The Des Moines Register as my example, but we've pulled in other states, we've pulled nationally. With the Iowa Poll, we put together a poll committee. We all come to the table with ideas. From that conversation, I began crafting the questionnaire and I'm keeping a lot of things in my mind as I do that. I need the interviewer to easily pronounce the words and that the syntax is easy for them to follow. I need the respondent to know what the heck we're talking about. At the end of the day, we want to look at the opinions of a few people and project it to the opinions of the entire electorate.

That's the beauty of random sampling. You can end up with a good cross section. We get a sample of real phone numbers, we drop the last two digits, and we put on random digits. Whether you're a listed number or not, whether you're a cell phone or not, we have the ability to get all of those phone numbers, and then draw a sample from it.

Clare Malone: I want to ask about that because it's so easy to screen unknown numbers these days and that obviously didn't use to be the case. What do you do to deal with that relatively new social behavior that we all have? How does that affect polls?

Ann Selzer: It's not relatively new.

[laughter]

Clare Malone: That's true.

Ann Selzer: That's been going on a long time. As soon as people got caller ID, that's decades. People say, "I just don't answer the phone if it says unknown caller," and I said, "Fortunately for every person like you, there's a doppelganger who's willing." I think the important point I want to make is that our polls are known for their accuracy. If there was a skew in who would answer and who wouldn't, we could not be accurate, and so it makes me worry less. I worry.

Clare Malone: Sure.

Ann Selzer: I worry, but I worry less. When we call you, we want to find out if you're a likely voter. In our world, that means if you don't say definitely you're not a likely voter to us, then we're going to terminate the call, but we will have captured, before we hang up on you, your age, your sex, and what area of the country you live in. Now that's the general population, and from the census, we know what that should look like. We wait everybody we talk to, whether you're a likely voter or not

Clare Malone: What do we mean by waiting?

Ann Selzer: Waiting is, to novices, it sounds very anti-democratic. Some people count more and some people count less. It's not one person, one vote. If we have too many older people, let's say, we'll count their vote a little less, we're going to weight them down, so give some other groups a little bit more weight overall. It sounds complicated, it's simple. Everybody does it.

Clare Malone: Let me ask this, in an ideal world, would you have any rules that you'd want reporters to follow when they talk and write about polls for their audience?

Ann Selzer: About a million.

[laughter]

Clare Malone: What about three?

Ann Selzer: I'll give you one that's controversial.

Clare Malone: Sure.

Ann Selzer: There are people who say if it's a two-point race, you can't really call that a two-point race, you have to call that a statistical tie, but it could also be true that it's a four-point race, that the gap is actually larger. If you want to say it's a tie, go ahead, but you should also say that it's equally likely that the race is a wider margin than reported, and to do one and not the other is exposing bias.

Clare Malone: Have you ever gotten a call for a poll yourself?

Ann Selzer: Like last year?

Clare Malone: Really?

Ann Selzer: The Iowa Poll called me up.

Clare Malone: What'd you do?

Ann Selzer: First time ever. I said, "I'm so sorry, I'm not able to participate in this poll because of my job, but thank you."

Clare Malone: You didn't say I run this poll.

Ann Selzer: I did not.

Clare Malone: [laughs] Ann Selzer's company runs the Iowa Poll.

[music]

Harry Enten: What's the point of the horse race? People are interested in the horse race.

Clare Malone: Journalists like me are constantly getting scolded for covering the horse race. In other words, talking about the polling results all the time. Who's up and who's down. The argument is that the horse race distracts from substantive reporting on issues or from the real stakes of an election, so I ran that argument by Harry Enten, who we heard from earlier. He's from CNN and he was my former colleague at FiveThirtyEight.

Harry Enten: I hate to blow some people's bubbles here, but I very rarely get asked by my friends, "Harry, in this latest poll, what was the top issue?" Their number one question, and oftentimes, their only question is, who is ahead and who is behind? If legit news organizations were not going to do the polls, then that poll would be filled by other folks who might not spend as much on the polls, might do shadier polls, and then all of a sudden, we have a very different view of where the race is heading into election day than we might otherwise should and could have. I think Donald Trump has actually provided a very good reason for why we should care about the horse race.

We are at an all-time low, at least from some sections of our country who believe elections and the vote counting is legitimate, which for everything that we see, it is. Donald Trump's case to say this election was stolen is significantly higher if, both going into the election, we didn't have a good baseline level to understand where the race actually stood.

Clare Malone: You're saying it gives us data prior to election day that in fact legitimizes the results after election day.

Harry Enten: Correct. In my opinion, absolutely. Absolutely. Absolutely does that, and it does that on both sides of the aisle, right?

Clare Malone: Harry, you're obviously steeped in polls, you know how to differentiate different polls, but what about people who are just lost in all the swirling numbers? What should they be doing with all of this data that we have?

Harry Enten: Look, we've spoken about this a thousand times in our lives; average, average, average. Yes, you take a look at the average.

Clare Malone: The average of the polls.

Harry Enten: The average of the polls, that'll give you the best indication. That's always what I try to do. You can find something that you might not have ever found before, but as a casual lay consumer, you average the polls and read the polls from the people that you trust.

Clare Malone: Yes, and there are places that do the average of the polls for you somewhere like RealClearPolitics.

Harry Enten: The New York Times, CNN has a Poll of Polls.

Clare Malone: There's places for people to go. They don't have to sit there with their calculator.

Harry Enten: No. Although granted with our phones these days, folks, you learned math in school, why don't you put it to some good usage?

Clare Malone: There you go. A public service announcement from Harry Enten.

Harry Enten: Math is fun.

Clare Malone: I want to ask about how people should be consuming polls. I think there's more polls than ever now. Should people be paying attention to these big national polls or should they really, if they want to be a smart consumer and know what's going on, should they only really care about who's up and who's down in swing states?

Harry Enten: Look, this election is going to be determined in the electoral college. It's just a matter of what it is that you're looking for. If you're looking for what are the issues that are driving this election, then the national polling is going to give you a pretty good idea, but if we're looking at a race that's tied, then yes, you want to be paying attention to those swing state polls because at the end of the day, we have seen numerous times in recent history where the swing states go one way or the electoral college goes one way, and the national popular vote goes another way.

Every single election since 1988 in this country on the presidential level has been decided by single digits nationally. It is the longest such stretch.

Clare Malone: I think the trends that have stood out to a lot of people who watch this stuff closely are the changing sentiments of Black voters and Latino voters. Black voters in particular have historically been a pretty reliable Democratic voting block. The top line in this election is that we're not seeing as much support for Biden as you might think. I wanted to get you to break it down a little bit.

Harry Enten: One of the trends that we do know is that Black voters, traditionally speaking, Black voters who are more religious actually tend to, if anything, be slightly more democratic depending when you control for some variables than Black voters who aren't so religious, which is the complete opposite pattern that you see among whites and Hispanics. Now all of a sudden, churchgoing is heading downward. The church helps keep Black voters into the Democratic fold in a way that I think is very unfamiliar with a lot of white audiences. Now all of a sudden, you're seeing dropping religiosity. Same is true with younger Black voters.

Now all of a sudden, that thing that is keeping them in the fold may not be keeping them in the fold any longer. All of a sudden, you have all of these different things coming together, which is lower religiosity, you have movement away from Joe Biden among younger Americans generally speaking, and you have fading memories of the civil rights movement. I'm not saying there aren't any other factors, but those three I think are a starting point to help understand what it is that we're seeing.

Clare Malone: Very interesting. Most people in the US don't have a bachelor's degree. What's so interesting to me is that Joe Biden's appeal was always Scranton Joe, down to earth, working class, but now his support is trending much more elite, right? How is he doing? Let's maybe talk broadly with voters who don't have a bachelor's degree, and how has that shifted since 2020?

Harry Enten: What we're seeing is across the board voters without a college degree shifting more and more towards the Republican column. You are seeing that with white voters without a college degree. I wouldn't be shocked if a margin if Joe Biden lost them by more than 30 points this time around.

Clare Malone: White voters with a college degree used to be like a lean Republican demographic, correct?

Harry Enten: Absolutely.

Clare Malone: They flipped a little bit.

Harry Enten: A little bit.

Clare Malone: They have, they flipped.

Harry Enten: A little bit, no.

Clare Malone: I want to go back to young voters. Are young voters moving away from the Democrats to the Republicans? Are they just saying, "I'm unaffiliated." Do they not like Biden, but maybe they'd come back to the Democrats once there's a different candidate? What's going on there?

Harry Enten: There's certainly a larger share that say that they're independent just generally speaking.

Clare Malone: 51% according to the latest Gallup data.

Harry Enten: There you go. Now, if you were to look at the polling data, what it would tell you, especially in a lot of these Senate races, is that if you go to, let's say, Nevada, Michigan, Pennsylvania, Wisconsin, Joe Biden is doing worse than the Democratic candidate for Senate in all of them. Who he's doing especially poor amongst on average tends to be those core parts of the Democratic coalition that we've spoken about here, voters of color and younger voters as well. If we have a different democratic candidate, could things differ? Absolutely.

It's oftentimes difficult to know if we're hitting a mere bump in the road for democratic support among younger voters or whether this is something larger and unforeseen. If I could take a step back in 2000, young voters split evenly between Al Gore and George W Bush. By 2008, young voters--

Clare Malone: The Obama coalition.

Harry Enten: The Obama coalition, those under the age of 25 went for Barack Obama by, I think, the mid-'30s.

Clare Malone: Have we for the past decade or so been living in the shadow or the conventional wisdom of the Obama coalition and now things are just maybe reverting to a historical norm?

Harry Enten: Yes, or a new history. These coalitions last until they don't. If you go to 1976, the Democratic coalition for Jimmy Carter that beat Gerald Ford was the southeast plus the industrial north. The west was a wasteland for for Jimmy Carter. Gerald Ford, I think, basically won every state west of Texas except for Hawaii. That obviously is no longer the case. The question you have to ask yourself now is whether or not we are in a new coalition or an emerging coalition, and it's certainly plausible that we are, and in five years, we'll go, "Should have seen it coming."

Clare Malone: I have to say, it feels that to me, that 2024 could potentially be one of those elections where you look back and say, "That's where the trends really solidified themselves. That's when it was an inflection point."

Harry Enten: It is possible that we are. Now, watch, of course, we'll get to election day, and then the exits and everything else will look the same and we'll say, "Oh, it was just a false start." At this particular point, yes, I think you're onto something, Clare.

Clare Malone: President Biden and the Biden campaign, they keep saying they don't believe the polls, the polls are broken, they're not accurate. What do you make of that view? Are they deluding themselves?

Harry Enten: The polls are the same polls that were taken four years ago that had Joe Biden's favorable rating about 15 points higher.

Clare Malone: When he says, "Oh, the polls are broken now," you're saying they're not more broken than they were four years ago when they were favorable to him?

Harry Enten: Correct. He loved the polls in the same way that Donald Trump loved the polls. Remember Donald Trump would quote the polls every freaking time in 2015 on his rise to the Republican nomination. Then all of a sudden, the polls went in direction he didn't like. It turns out that at least when it comes to polling and reading polls and complaining about polls, that Joe Biden and Donald Trump have very similar viewpoints. When the polls are good for them, they like the polls. When the polls are bad for them, all of a sudden, the polls are not so great and they complain about them.

Clare Malone: This is the thing that I know people who deal in polls or who are polling analysts who are pollsters themselves get really riled up about, which is when people say, "The polls missed this result, the polls were inaccurate." What's your response to that? Give me two examples. One where the polls were really accurate and another when the polls were really off.

Harry Enten: The polls in 2016, the national polls, were pretty much right on.

Clare Malone: People were surprised by the results of the 2016 election.

Harry Enten: Yes, the polling in those battleground states did show that Hillary Clinton was ahead, but if you know how much a poll can miss by or how much an average can miss by in a state, you know that it can miss by, depending, three, four, five, six percentage point is not something that is crazy. If you recall at FiveThirtyEight, which we were a part of, if I remember correctly, Hillary Clinton had a 71% chance of winning on the eve of the election, and Donald Trump had a 29% chance.

Clare Malone: You and I were in the same newsroom and I still remember feeling surprised when I realized Trump was going to win. Is that an emotional response?

Harry Enten: Yes.

Clare Malone: Because I was consuming all of these polls. I still was consuming the media narrative that, I guess, helped create that Hillary Clinton was probably going to win.

Harry Enten: Yes, that's part of it, but I think also this was something that was so unique, and when you combine that with the fact that if you lived in New York City at the time, you lived among people who had a college degree, which many in the media did, then, yes, I think that it becomes very, very easy to get lulled into a sense of, what am I actually seeing here on these pieces of paper? What are these polls actually telling me?

Clare Malone: People should be more aware of these margins of error. I want to hit you with one last question. When do the polls actually count? When are voters really most likely to be locking into the positions that they'll have on election day?

Harry Enten: Here's the deal. The deal is that you really should wait until after the conventions before really understanding where this race actually stands. Labor Day is another way to look at it, but don't let the polls run your life.

Clare Malone: Touch grass, kids.

Harry Enten: Touch grass.

Clare Malone: Buckle up. Harry Enten, thank you so much for coming. It's so great to see you.

Harry Enten: The pleasure was all mine. Thank you.

[music]

Clare Malone: Harry Enten is the senior political data reporter at CNN, and he hosts their podcast, Margins of Error.

[music]

 

Copyright © 2024 New York Public Radio. All rights reserved. Visit our website terms of use at www.wnyc.org for further information.

New York Public Radio transcripts are created on a rush deadline, often by contractors. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of New York Public Radio’s programming is the audio record.