MARY REICHARD, HOST: It’s Wednesday the 14th of August, 2019. Glad to have you along for today’s edition of The World and Everything in It. Good morning, I’m Mary Reichard.
NICK EICHER, HOST: And I’m Nick Eicher. First up today: Washington Wednesday.
The first recorded straw poll in the United States took place in 1824. Over the next century, as the polling industry developed, market-research firms were born.
Then in 1936, a man named George Gallup founded the American Institute of Public Opinion. That same decade, President Franklin Roosevelt became the first president to use private polling to inform his policy and campaigns.
REICHARD: Today, polling plays a huge role in shaping the conversation during presidential election season. The media constantly report who’s up and who’s down. And we’re inundated with headlines like these: “Kamala surges in 3 polls after strong debate performance.” Or this week: “Biden, Sanders both trounce Trump in head-to-head matchup, new poll finds.”
Of course, it was these same pollsters who predicted Hillary Clinton would “trounce Trump.” And then voters went to the polls.
ABC BROADCAST: That is the race, frankly. Uh, there is no path forward for Hillary Clinton. We’ve just seen Pennsylvania has been called by AP. I think, for me, this is one of the most stunning results of the night. We had seen her up comfortably 5 to 6 points.
EICHER: So as another presidential election season heats up, how much weight should we give to political polls? Here now to discuss that is Karlyn Bowman. She’s a senior fellow with the American Enterprise Institute. She studies opinion polling, U.S. politics, and the media. Karlyn, good to talk with you this morning.
BOWMAN: Good morning. Delighted to be with you.
EICHER: I’d like to start with President Trump’s poll numbers. Where are they, generally speaking, and what do you think they say about his reelection prospects?
BOWMAN: President Trump’s approval ratings have been remarkably steady for a long time, moving in a very narrow range from a low of around 38 percent to a high—in one or two polls—of 45-46 percent overall. He won, I think we’ll both remember, with 46 percent of the vote last time and he’s close to that. He’s holding his base very strongly, but of course, he has to be able to expand beyond that if it ends up being a two-way race.
EICHER: Turning to the Democratic presidential field, what trends of interest are you observing there?
BOWMAN: Well, there are a number of interesting trends, I think. First of all, Biden is holding the lead as he has been since he announced his candidacy overall. There have been a blip or two in that overall number, but he has a pretty substantial lead over most of the other candidates.
Of course, we all look to Iowa where the caucuses are held and in Iowa his lead is not quite as strong as it is nationally overall—with several Democrats coming very close to him Of course the amount of attention you get from those Iowa caucuses tends to catapult you into the next tier of races overall, but it seems that most Democrats feel quite comfortable with Joe Biden.
And following below would be Elizabeth Warren, Kamala Harris, and a few others. But, again, Biden’s strongly in the lead at this point.
At this point, many Americans, many Democrats believe he is the most electable of the candidates. And I think the support we’re seeing for him also reflects the fact that even though many of the candidates call themselves progressive lean to the left part of the Democratic party, but Biden has that middle lane right now virtually to himself.
EICHER: And so in that sense, is it too early to say it’s Biden’s to lose?
BOWMAN: I think it is Biden’s to lose, absolutely, at this point. Anybody who goes through the cauldron of running for president—it’s such a difficult thing to do, and he’s done it more than once before… I think that’s really important to his candidacy going forward that he knows how difficult this is and he’s been through it before. But I do think it’s his to lose.
EICHER: With respect, we talk about these opinion polls—the polls that we’re discussing here—with a level of certainty that maybe we shouldn’t have. Back in 2016 we saw how wrong the polls could be—some by small amounts, some by big amounts.
So talk about your confidence in the polls these days.
BOWMAN: Well, if we look back to 2016, at least at the national election, we know that the national polls—the major ones—got it right. They said that Hillary Clinton would win the popular vote and she did. The state polls had a very bad night and the people using probability estimates like FiveThirtyEight and the others had a very bad night overall. But the national polls did a pretty good job.
Now, as you begin to look at the series of Democratic contest going forward, there’s some—Iowa is going to be particularly difficult to poll in this year. They’ve set it up in a different way. There are fewer caucus states overall in the Democratic contest in 2020, but the three caucus states tend to be important early—Iowa in particular, and Nevada will be important.
And those caucus rules are very different than they’ve been in the past, making it extremely difficult to poll in Iowa. A couple pollsters have already attempted to capture not only those people who would show up on those cold nights of the actual caucuses but also those people who are going to be voting in the virtual part of this contest overall that starts about a week before the overall caucuses.
So that’s going to be just extremely difficult this year. And so they have their challenges and in a field this big there could be a lot of movement that’d be difficult for the polls to capture, say, in the last 48 hours or the last week of the campaign.
EICHER: You use the word challenge and I read a piece fairly recently by Walter Shapiro saying that the polling industry is in crisis. How far would you go here?
BOWMAN: Well, I think there are some really serious concerns for the business going ahead—even the best devised polls, polls from the Pew Research Center, from the Gallup organizations—now have response rates below or around 10 percent. And we don’t know or we’re not sure that we can create a sample that looks like America with response rates that are that low.
I mean, people don’t want to answer the phone in the evening. They don’t want to answer their cell phone if they don’t recognize a number. And so there really are some serious challenges the business has going forward. There’s a lot of crisis rhetoric around, and I’m not sure I’d throw out that term just because everything seems to be in crisis at one time or another. But I would say that the challenges are very serious.
EICHER: Interesting that you mention more or less old-fashioned phone calls as obstacles. Do you see more modern methods that could help the polling industry?
BOWMAN: Well, I’m not sure there’s anything specifically that can be done about that. But pollsters who are doing landline interviews at least will continue to call until they get someone in a household. And those who really want to do more work on, let’s say, developing a sample that looks like America, will do things like send a letter, offer a small financial incentive.
Some of the online polls offer you points almost like airline mileage points to participate in a survey. And so there are a few small things that they could do, but I’m just not sure that it’s going to be enough.
EICHER: Well, listen, we may be guilty of this, too, but there are quite a lot of journalists putting out breathless stories about the horse race of 2020—who’s up and who’s down. Given all of this, what would be your advice to the average American looking at these stories, trying to get a sense of where things are when they read about polling?
BOWMAN: Well, I think one thing that’s helpful is to look at the averages. And a number of organizations are now publishing the averages of all the major polls to give us a sense of where the race stands. And then I’d perhaps follow a couple pollsters and just see how they determine what their sample is—whether they’re interviewing adults, whether they’re interviewing registered voters, whether they’re interviewing, let’s say, people who can vote in certain primaries in caucus states.
And look a little more closely at the methodology. Virtually all of them have that available online and you can see how they’re asking the questions, the dates that the survey was in the field. Again, the way they’re asking the questions tends to be very, very important. So, dig a little deeper and also look at the averages and that will give us some hints about which way the contest is going.
EICHER: And when all else fails, we’ll just call up Karlyn Bowman and ask her.
BOWMAN: [Laughs] Thank you.