Role Of Polls During The Election Season

Nov 3, 2018
Originally published on November 3, 2018 10:46 am
Copyright 2018 NPR. To see more, visit http://www.npr.org/.

SCOTT SIMON, HOST:

Midterm elections are always a little tricky for pollsters. What do polls about national issues and inclinations mean when control the House and Senate will be decided district by district and state by state? But this time around, there's even less data and fewer polls to analyze. NPR's Domenico Montanaro joins us. Thanks very much for being with us, Domenico.

DOMENICO MONTANARO, BYLINE: Always a pleasure. Thank you.

SIMON: Fewer polls?

MONTANARO: Yeah. There are fewer polls this time around, and that is a real problem. I mean, when you look at the fact that newspaper readership at a local level is dwindling, their revenue is being cut, they were the ones who used to fund a lot of these polls, especially in congressional districts. And on a statewide level, there are dozens of fewer polls. So take them all with a grain of salt.

SIMON: Aren't a lot of people, though, also still put out about the polls from 2016?

MONTANARO: Absolutely. But, you know, there's a lot of myths that have to be unpacked about 2016. And for me, you know, national polls were actually pretty good. Hillary Clinton wound up in an average of the polls of about 3 percent before Election Day. And guess what? She won the popular vote by 2 percentage points. That translated to 3 million votes. It did not translate to her winning the presidential election. And the real problem came in the upper Midwest in states like Wisconsin and Michigan, which were showing us that there were fewer and fewer polls. Leading into Election Day, there was a tightening of polls in Wisconsin and Michigan, but because there wasn't a preponderance, you couldn't really tell what to make of those things there. There were a lot of pieces of evidence and I think a lot of people are - have rejiggered how to make analyses based on fewer and fewer good polls.

SIMON: What do fewer polls mean for the campaigns? Because they kind of steer by those things, don't they?

MONTANARO: Well, the campaigns have their own polls, and the most competitive races, the national committees, will go in and they poll - they track day to day every day. So the real kind of juice, the good stuff, that they base all of their advertising on, all of who they're going to try to boost or pull away from in these final days, are all based on their internal polling. And that's why some of us like to talk to them, compare what they've got in their internal polls to what we're seeing in the public polls as well.

SIMON: Have fewer polls been in a way good for journalists? Does it mean we have to talk about issues?

MONTANARO: I think at NPR we've done a pretty good job going out into the country. We did it in 2016 as well, and we had a particular focus on it this time around. We always do try to look at demographic groups, try to get a sense of the energy and the feeling that's going on in places because, frankly, when you look at who's a strong candidate, who's got that sort of X factor, a lot of times those candidates win. You don't need science for that. At the same time, that kind of anecdotal reporting doesn't necessarily mean it's representative of what's going on everywhere because people have so sorted themselves, you could talk to someone who goes out door knocking and they think their candidate is going to win because of the block they're on and not realizing what the rest of the district is like.

SIMON: Yeah. Well, that raises the question, don't - I don't - perhaps there's been research done on this. We know that people have been known to not tell the truth to pollsters. Probably do the same to reporters, don't they?

MONTANARO: To an extent, and I would say that that's an even bigger potential problem in this campaign when you've had a year and a half plus of the president, who has the biggest bully pulpit, essentially going out and saying that the press is the enemy of the people, that we are fake news and continuing to purposefully knock down our credibility. And when someone goes to somebody and calls them up and says I'm from big media organization X, that can have an effect for sure, one that a lot of the pollsters have been talking to each other about how to try to mitigate.

SIMON: So Tuesday night, millions of Americans are hunkered down in front of their radios. I understand a few other platforms are also presenting (ph) news. There are going to be exit polls. But the AP, Fox and National Opinion Research Center from the University of Chicago are trying to gather data in a different way, right?

MONTANARO: Absolutely. This is actually a huge, fundamental, ground-shifting thing that's happening because this is the first time in American history since exit polling began in the 1960s and 1970s that we're not going to have a shared understanding of the shape of the electorate. You're going to have on Fox News one set of what the shape of the electorate looks like and on all the other networks, a very different look at what those are - potentially very different. It could also be very good. I will say I know the pollsters who are doing this for Fox, AP and NORC, and they're all very good. They're very smart and they're non-biased. So just because it has Fox News on it doesn't mean that it's looking to get a more Republican-leaning result. In fact, a lot of the pollsters I've talked to they think it's a good thing. They think that this competition will lead to the exit polls having to be more rigorous about their models.

SIMON: Domenico Montanaro is lead political editor of our Washington desk. Thanks so much for being with us.

MONTANARO: Thank you. Transcript provided by NPR, Copyright NPR.