The NPR Politics Podcast - NPR Poll: Midwest Abandons Trump, Democrats Chances Balloon Ahead of Midterms

Episode Date: September 12, 2018

In a troubling sign for Republicans less than two months before November's elections, Democrats' advantage on the question of which party Americans are more likely to vote for in November is ballooning, according to a new NPR/Marist poll. Plus, how effective are polls at gauging voter outcomes? This episode: Congressional correspondent Scott Detrow, political reporter Danielle Kurtzleben, and political editor Domenico Montanaro. Email the show at [e-mail]. Find and support your local public radio station at npr.org/stations.Learn more about sponsor message choices: podcastchoices.com/adchoicesNPR Privacy Policy

Hey, this is Nick, and I'm at the Ware Middle School in Ware, New Hampshire, where I just voted in the New Hampshire primary. This podcast was recorded at 1-59 Eastern on Wednesday, September 12th. Things may have changed since it was recorded. Okay, here's the show. Hey there, it's the NPR Politics Podcast. NPR is out with a new poll, a lot of information about the midterms, and we're going to tell you about it. I'm Scott Detrow, I cover Congress.

I'm Danielle Kurtzleben, political reporter. And I'm Domenico Montanaro, political editor. Let's talk polls. Domenico, NPR and Marist have a new poll that came out today, and the story you wrote about it has been getting a lot of buzz on the internets. One of the big headlines from our poll was the fact that the so-called congressional ballot, who people would prefer to see controlling Congress, has now started to balloon in favor of Democrats. So you had 50 percent of people saying that they would vote for the Democrat in their district, only 38% saying that they would vote for the Republican in their district. That's 12 points, right? That's up from seven in July when we last asked the question. And that's a long way to get to the really big headline, which is all of this, almost all of this

is moving from people in the Midwest. There was a 13 point swing from July to now in favor of Democrats with voters in the Midwest. And of course, that is a part of the country that was key to President Trump's big win in 2016. It's a part of the country he goes to frequently and that he sees as kind of like the cultural key to his presidency. That sounds pretty significant if there's a big drop there. Also an area of the country that Trump's policies may be heavily impacting, for example, tariffs. Who lives in the Midwest? A whole bunch of farmers. And also auto companies happen to be in the Midwest. So those are areas

that tariffs that the Trump administration has imposed also might be hitting pretty hard. That could be one reason for the swing. We don't actually know from this poll, but we are extrapolating here. Well, look, I mean, that's a big part of it. I mean, when I talk to the pollster here who conducts the poll, Lee Miringoff, who's the director of the Marist Institute for Public Opinion, he said the Midwest is an area that's getting restless about what they hoped was going to occur and what they feel is not occurring with the president now. So, Domingo, just to underscore that gap that you were talking about at first, it seems like that is the type of gap, 12 points or so, that would lead to Democrats winning a lot of seats, probably the amount of in 2006, when Democrats took back the House, a lot of the polls sort of leading into Election Day started to be about this size. You know, there's this debate that goes on on whether or not it should be seven points or higher or whatever it is. But this number for sure is outside of that margin that Democratic and Republican

strategists when you talk to them, this is something that they look at as a warning sign for the Republicans. That said, this is a registered voter poll. This is not a likely voter poll. If you compare registered and likely voter polls, this is in any year, midterm or not, registered voter polls tend to be a little friendlier towards Democrats. When you get down to likely voters, you tend to get somewhat more Republican-friendly polls. I did look at that gap today, at least based on averages. Likely voter polls do still tend to lean Democratic, but the gap is not quite as big as unregistered. Are you saying it comes down to turnout?

I have no idea what you're talking about here. I do know that elections have consequences, though. They do have consequences, and it all comes down to turnout because it really does matter who the people are that are going to vote in this election. Because I like to also say, as this new cliche that I've created, that, you know, if we were Australia, I could tell you pretty well who's going to win elections because Australia mandates that everyone vote or face a fine. We don't have that in this country. And the fact is, when it comes to minority voters, young women in particular, the voters that are seemingly most tied to Democrats, they're the ones who vote least. Danielle, what else is in this poll? Speaking of who's going to turn out, let me just get it very quickly at two other big points this

poll pointed to, not just the Midwest. One is the gender gap. There is a massive gender gap in this poll. Women this year are very heavily Democratic, and the gap between them and men is simply huge, especially when I compare it to 2014. Likewise, the gap between white college and non-college voters is also massive compared to 2014 exit polls. My point here is that those are two big groups you're going to want to watch on election night. The gender gap in this poll is enormous. It's huge. I mean, when you look at who's approving of President Trump and not, men overall are approving of the job the president is doing. Women, only 28% of women approve of the job the president is doing, 62% disapprove. That's

important because while people might say, well, you know, of course, women don't like the president as much and they also lean Democratic, you know, in general. Yes. But President Trump won 41 percent of women on election night in 2016, according to the exit polls. So if he's down at 28 percent in approval with women right now, that is a giant problem for him. Women are not just leaning Democratic. They are absolutely way on the side of Democrats. All right, we're going to take a quick break here.

But when we come back, we're going to put this into the context of a lot of other polls that we've been seeing that have characterized a pretty similar shift. We'll be right back. Support for NPR and the following message come from Newsy, the TV news channel with honest, in-depth context on the stories that matter. Newsy is for people who aren't satisfied with getting only the loudest part of the story. Newsy delivers more, more context, more solutions, and greater understanding of the people and events that shape our world.

Learn more at newsy.com slash watch. Hey, it's Guy Raz here, and on the latest TED Radio Hour, how to talk about death candidly. Being able to accept that someday I will decompose, there is something comforting to me about that. You can find the TED Radio Hour wherever you listen to podcasts. All right, we are back. And I think one reason why the poll that we just did with Marist really stuck out to me is because we've been seeing a lot of headlines over the last week or so that have marked pretty similar shifts, whether it's President Trump's approval rating or whether it's this question of our voters favoring Democrats or Republicans in Congress.

Domenico, what's the best way to characterize the shift we've been seeing in, let's just take President Trump's approval rating. So President Trump's approval rating in our poll is at 39%. And when a president is below 40%, it is really not a good number. When a president is below 50%, it's generally not a very good number. So at below 40 is really bad. And it's not a change in our poll. Frankly, from back in July, Trump was at 39%. And the mid to high 30s to low 40s is kind of where Trump has been throughout his political career. At the same time, we're a lot closer to Election Day, and a lot of other polls are also showing the president below 40, and that can portend really bad things for the president's party when they're in power. And Danielle, we've seen a CNN poll, we've seen a Washington Post poll, we've seen a Quinnipiac poll, several others, all noting,

their actual numbers are a little different, all noting a pretty significant drop recently in his approval rating. Yeah, somewhat at least. I mean, we don't want to overstate this. I was looking at the polling average of Donald Trump's approval just before we came in. And over the last two weeks, Donald Trump's approval has dropped, according to RealClearPolitics, around three percentage points. When you average polls together, yeah, that sure seems significant. But it's not that it's absolutely plummeting. But to pivot off of what Domenico was saying, this stuff does matter. The races that I've been out covering, the districts I've been out talking to voters, especially on the Democratic side, what you get the sense of is that Donald Trump

really is looming over a lot of congressional races. A lot of these races are not just about things that are happening in X district. A lot of Democratic voters will be casting votes not only for Democrats this year, but against not just Republicans, but against the party of Trump. And Domenico, you sent a note around the newsroom yesterday that I thought was pretty interesting. Taking Trump out of it even, I mean, one of the most important factors to look at when you're looking at how a midterm could shape up is how popular that president is. Right. Well, just to put it in context, when we talk about being below 40%, there's only been one president since polling began who in their first midterm, in the last poll before the first midterm was below 40%.

That was Harry Truman back in 1946. That's how rare it is. There have been plenty of presidents below 50, but below 40, not very good. And Truman wound up losing 45 seats in the House, 12 seats in the Senate, which is the most for any first midterm for a president. That's not to say- He won re-election two years later, which means politics can be hard to predict sometimes. Right. He did win re-election. That's also not to say that all of the circumstances in those elections are the same. You know, this is the worst landscape that

any party has ever faced since the direct election of senators in the Senate for the Senate map when it comes to Democrats are defending more seats than anyone since then. You know, we just went through a huge round of redistricting in 2010 that has insulated a lot of Republican seats and Democrats have moved and sorted themselves far more towards cities so that that the Democratic vote is not spread out as much as it could be. Obvious case in point, Hillary Clinton won the election in the popular vote by three million votes, but still lost the Electoral College and the election. What's also rare is for the president's party to gain seats in a midterm. The president,

the sitting president's party almost always loses seats. So should Republicans lose seats in Congress this year? That in and of itself won't be a great big flipping deal. What could be a big deal is the number of seats that they lose. That's what to watch for. And the key number, of course, is 23. That's the amount of seats that Democrats need to pick up here to take control of the House. So, Danielle, let me get to the elephant in the room here, though. Donald Trump won the 2016 presidential election. He sure did. So why should we trust these polls? Well, he didn't win the popular vote, first things first. I mean, this is, look, people came out of 2016 with a great big kind of freak out about, oh my God, polls are

wrong. Well, there are a few things to say about that. One is that most national polls, I think pretty much every national poll, gave Hillary Clinton a slight edge. And I mean, as we've heard over and over, Hillary Clinton did win the popular vote by roughly that much. National polls seemed fine. Now, state level polls did seem to underestimate Trump's support, at least in some states. Now, the American Association of Public Opinion Research, big consortium of pollsters, after 2016 put out a report saying, OK, here's what went wrong. And one of the things they said is, yeah, polls in states seem to underestimate Trump's support, especially in the upper Midwest. I don't think polls shouldn't be trusted. I'm just asking the question here and being that guy in the podcast, because if you take 2016 aside, there's

been several key primary upsets this year where the public polls and based on reporting, the private polls were way off. I'm thinking of that race in Boston a couple of weeks ago. I'm thinking of the Florida governor's race. And I'm thinking of that big upset in New York when Joe Crowley was defeated by Alexandria Ocasio-Cortez. You are absolutely right that there have been those cases. I mean, first of all, the general rule of thumb with polling is that you have to take the trend line. Right. And, you know, the fact is, in a midterm, midterms are very difficult to poll in anyway, because, again, it all comes down to turnout and the kinds of people who do that every time someone says that, you know, the kinds of people that do wind up voting because of, you know, they have to be fired up to actually get out to the polls to go do so. And we're seeing more and more that these are turning into base elections. Now, specifically in some of these states, one of the biggest problems with polling has been in, first of all, congressional districts are very difficult to poll in because you don't have a lot of outfits who are spending a lot of money doing it over a long period of time and doing it very close to election day. That makes

things really difficult to try to predict. That's why, you know, when you see in New York, for example, where Joe Crowley was up by 30 points or something like that, you know, leading into his election against Alexandria Ocasio-Cortez. That poll was done like a couple months before the election. And frankly, what we're seeing, and a lot of pollsters have been talking about this behind the scenes who I've talked to, they'll say that we can't measure who's going to decide when you're talking even a week out, because a lot of people are deciding very close to election day, if not on Election Day, because they don't even know who these candidates are. Right. And Domenico was saying that midterm polling is really hard. Well, what's even harder is primary polling. Right.

Because Domenico is absolutely right. A lot of these polls are done, you know, a month before primary day, well, big deal, right? Because someone like Ocasio-Cortez or whoever might have their army of canvassers out, you know, the week or the few days before the election. Aside from that, a big thing that polls are based on, that they're weighted on, is who is expected to turn out, who usually turns out. Well, if you have a candidate who turns out a bunch of people who don't usually come out, for example, young voters, for example, less educated voters, that sort of thing, then, well, the polls just might not account for that or expect that. And by the way, these polls are not trying to predict what's going to happen. That's a huge problem with the way people interpret polling is that someone hears, oh, Democrats have a 12 point edge on the generic ballot like we're talking about in our poll in September. Why didn't Democrats take back the House in November if that's what turns out to be the case?

Right now is what the sentiment is. Right now, if the election were held today, then Democrats probably would make significant gains. But you don't know what happens between now and Election Day. Totally, yeah. All right, that's it. Quick episode today, but we'll be back tomorrow with our weekly roundup talking about this and the rest of the week's political news. I'm Scott Detrow. I cover Congress. I'm Danielle Kurtzleben, political reporter. And I'm Domenico Montanaro, political editor. Thank you for listening to the NPR

Politics Podcast.