Washington Wednesday: Taking America’s pulse

MARY REICHARD, HOST: It’s Wednesday the 8th of November, 2023. This is The World and Everything in It. Good morning, and thanks for listening. I’m Mary Reichard.

NICK EICHER, HOST: And I’m Nick Eicher. It’s Washington Wednesday.

Tonight, five Republicans meet in Miami for Republican presidential debate number three. The group has narrowed down to Ron DeSantis, Nikki Haley, Tim Scott, Chris Christie, and Vivek Ramaswamy. North Dakota Governor Doug Burgum and former Arkansas Governor Asa Hutchinson did not qualify…and former Vice President Mike Pence has dropped out.

REICHARD: On Sunday, the New York Times released a set of polls it conducted with Siena College. The results show Donald Trump leading Joe Biden by a range of four to ten points in five key battleground states.

But after missing the mark in the presidential election of 2016, many voters just don’t trust polls to give an accurate picture. So how do polls work, and can they be trusted in 2024?

EICHER: WORLD’s Washington Bureau reporter Carolina Lumetta caught up with a reputable pollster at a political conference over the weekend … political scientist Ryan Burge. He is an associate professor at Eastern Illinois University.

Here’s an excerpt of their conversation.

CAROLINA LUMETTA, REPORTER: Just to take this very granular, what is the process for polling that we can trust? And then what are some types of polls that you have found that do not lead to accurate data? How do we basically analyze the verity of polls?

RYAN BURGE, GUEST: Yeah. So there's a whole range of polling, right? From very, very bad to very, very good. And almost always that scale runs on very, very cheap to very, very expensive, very good polls are very expensive, very bad polls are very cheap. So the polls that you don't want to trust are what we call convenient samples. And a convenient sample is literally what it sounds like, we got you in this poll, because it was easy to do so. Right. So it's, we I just found 12 of my friends. And I asked them these questions. Those polls are not reliable, though, because they're a convenience and were not random sample of the entire population. Now, on the other side, you've got like Pew Polls, which are really, really expensive. I mean, the General Social Survey is what I use for a lot of my polling data. It costs 10s of millions of dollars to put together a General Social Survey, because it takes in depth interviews. Those polls tend to be more reliable, though, because they spent so much time trying to get them right, and they have a bigger sample that's randomly selected. So where we try to live is in the happy place between more rigorous but less expensive. So what we try to do is find the cheapest sample that we can get that is still academically rigorous. And so what a lot of what we're doing now are what are called panel surveys. And so there are lots of groups that do these. Qualtrics is the one I use, where they basically have recruited a stable of people who they know will answer polls, and they pay them to answer my poll. Is it truly random? No. Is it a convenient sample? No, it's somewhere in between those two things. So those are being used a lot more now, because they kind of meet both criteria being somewhat rigorous, but somewhat, you know, cost effective. The polls you see on, you know, making the news media run the gamut from being you know, very, very rigorous to very, very not rigorous. You've got to figure out who they're working with, an established polling firm or a fly by night organization. You want to stick with once you've been doing it for years and years, and years, Gallup, Marist, Pew. There's a lot of big names in the industry. And those are the ones we really want to focus on.

LUMETTA: Talk to me a bit about polling bias. There's another criticism that, well, someone of a particular political party is never going to get a favorable poll, because the people who answer might not also fall under those same beliefs. That's a very general example. But how can we trust polls from Gallup and stuff like that? How are they collected to make sure that all views are represented?

BURGE: So, one of the long tails of Donald Trump is that he made the polling industry much harder and less accurate, because he cast such doubt on how polls are done, and that they're inherently biased, especially against him. Really, that was sort of a self-fulfilling prophecy thing, though, because Donald Trump was a guy who a lot of people voted for but wouldn't tell you they voted for him. So you know, when you lie to a pollster, we can't, we don't check that. We have no way to check that. So we don't know who you voted for or not. But then he made it worse by saying, “Oh, they're out to get me. And that's what they did in 2016, they were really trying to make me look bad.” We weren't trying to make you look bad. It's just a lot of people have voted for you were not proud of the fact. So they didn't say that. So that way the polls looked like you were doing worse when you weren't doing worse. So it created like this downward spiral effect. Now, I will say this, these organizations are I mean, every two and four years, they have a reckoning, which is are we being accurate? How far are we off from the actual result? And no, I mean, the bias, the bias there is we're biased towards trying to do the best job that we can. If you start putting out poll after poll and election cycle after election, that's far off the reality on election day, no one's going to take you seriously anymore. So all these organizations are working towards getting closer to the right answer. Now it's getting harder because people aren't answering the phone if you call them, people are are inherently shy about talking to polls, they don't have time, all these kinds of things. There's all these impediments to doing so. But that doesn't mean there's not lots of people working on these problems trying to make them better.

LUMETTA: Speaking more specifically now about what we're seeing, looking into 2024. We are as of this week, one year away from the election. How useful is it to be watching polls right now? And if it is, what types of polls will tell us a story?

BURGE: There's a lot that happens in the year before an election. And one is that people start paying attention to politics again, because they basically take some time off between election cycles. And so you know, what we're seeing in these early battleground states is that Trump is winning a lot of them and by significant numbers, but you also got to take into effect. There's a year of politics that has to happen between now and then whether it be international crisis that what's happening the Middle East right now our economic crisis is getting better. Is it getting worse? People have recency bias. So what happens right before an election is more important than what happens a year or two years or four years ago. I mean, remember, when Biden pulled us out of Afghanistan? Everyone goes, Oh, that'd be the end for Joe Biden. When's the last time anyone spoke about Afghanistan? Like we forget things very quickly. So what's happening now is less important than what happens in September and October and early November of next year. So they give us a good baseline to kind of see like what the large landscape looks like. But it's a very fuzzy picture right now. You know, it'll sharpen up as we get closer to election time, especially in the last three months of the election season.

LUMETTA: So overall, nationally, I mean, the main headline is that Trump is far and away the leader. And then there are more stories on, is it DeSantis? Is it Nikki Haley? What is useful to be kind of narrowing in on for those for basically the whole Republican field right now?

BURGE: ​​It's interesting to see who the Republican second option is right now. I think that's a really important story, because we know that Trump basically dominates the Republican Party, he's got 50- 55% of the vote, which is a lot in a crowded field, right. Who's the number two? Is it DeSantis or is it Haley? I think that tells you a lot about how you want the party, what's the party, what are the other Republicans up to? If they like DeSantis, it means they're more in the culture war area, right? He's very much combative. Nikki Haley's not that she's almost like an old school Republican like a George W. Bush Republican. So I think it tells you, who's the not-Trump vote, right, who are never Trump Republicans? Are they more like DeSantis voters? Are they more like Haley voters? And I think as Haley's pulling forward, I think it's telling you that they're coalescing around her as the counterbalance to Trump on the Republican ticket. But I still don't think she has a chance at all.

LUMETTA: Right, does that matter if Trump becomes the nominee?

BURGE: No. I mean, I think honestly, from a polling perspective, this is the most interesting that can happen in the next twelve months is will Biden choose to not run? Will Trump be indicted in such a way that he can't run? Because right now, it's a foregone conclusion that if Biden runs, he'll be the nominee, if Trump runs, he'll be the nominee. And really, it comes down to the general election at that point, which right now, it looks like Trump might have an advantage. But again, we're a year out.

LUMETTA: So talk to me now about the religious vote. Is it true that some of the religious votes are turning into what we would call nones?

BURGE: Absolutely, I mean, the share of Americans who are nones now is above 30% amongst younger people, Gen Z, it's over 40%. But what we're seeing a lot of is people who are Christian, but Democrats are leaving the Christian label behind and keeping the Democrat label. So what's happening is a lot of people who were voting for Obama, let's say, but were Christians, went to church on a regular basis, now have shed the Christian label because they just see incongruence there between their faith and their politics. So I think what we're seeing more and more is people, if you look at the data, 50% of liberals are non-religious. Now it's only 10% of conservatives. So what we call the Pew gap or the God gap has never been bigger than it is right now. Almost half of Biden's voters in 2020 were non-religious voters. Forty-five percent of all votes cast for him in 2020 came from atheists, agnostics, or nothing in particulars. That is the future of the Democratic Party. Meanwhile, the Republican Party is still majority white Christian. So the parties are going in completely opposite directions when it comes to race, and religion and how often you go as well.

LUMETTA: A question to start wrapping up is a lot of people get very tired during an election season with just the horse race. Now, this person's ahead, now this person is ahead. What benefit do polls give us in that sort of an atmosphere?

BURGE: I think they help us understand how close it is. I think that's more important than who's winning, who's losing is if it's close or not, you know, is it gonna be a blowout election? Or do we really need to go out to the polls? You know, for instance, in 2008, what we realized really quickly was that was a blowout election. We got to September and all the polls looked like Obama was way ahead. We look at the last couple elections that we knew it was very close. So I think for a lot of voters who are marginally attached, if they see blowout numbers, they turn out, tune off, don't even worry about it, right? They're not worried about the election. But they see it's close, especially when they're persons down, they start engaging, they start getting involved, they start voting. And so I think that really is more important. Is it close? That's what polls tell us more than anything else than who's winning who's losing is should I care, is it close?

LUMETTA: Great. Well, thank you so much for being here today.

BURGE: Thank you so much.

WORLD Radio transcripts are created on a rush deadline. 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 WORLD Radio programming is the audio record.