As the 2016 campaign comes to a close, the FiveThirtyEight Elections podcast team is recording daily episodes. In our final pre-election installment, we assess the forecast model, discuss defining moments of the campaign and prepare for Election Day.
Here is a portion of our conversation that has been lightly edited for clarity:
Jody Avirgan (producer and podcast host): I think my main question for all of you — maybe we can just go in order — is what your big question heading into Election Day is. Nate, you want to start?
Nate Silver (editor in chief): It seems like we’re seeing a few signs of momentum for Hillary Clinton in the polls, more in national polls than in state polls. I guess my question is does that reflect actual momentum or — because the last poll sometimes isn’t the best poll. The last poll, sometimes pollsters herd and cover their ass a little bit. Now, in this case, there are logical reasons to think that the news cycle has turned to a more favorable position for her. But we see what happens when she goes from having a 2- or 3-point lead to a 3- or 4-point lead. Even that narrow difference in our model “flips” a couple of states. It’s going from 49 percent to 51 percent, so don’t get too excited. But there are a lot of states that are really on that margin. She had a good poll in New Hampshire last night, for example. And so, a little bit more security potentially in the last 24 hours.
Jody: Clare, what’s your big question going into Election Day?
Clare Malone (senior political writer): What I’ll be watching for our live blog is how Election Day unfolds, like the actual procedure and whether or not there are incidents of voter harassment, things like that. Long lines are definitely a thing that people expect, but I think I’m interested to see if there are going to be other, I guess, more ominous things that are happening at polling places. We had a couple early reports of Trump supporters on bullhorns, yelling at people. So, that’s my big thing, is how the actual voting unfolds and what happens.
Jody: And Harry, what’s your big question for Election Day?
Harry Enten (senior political writer and analyst): My big question is the Latino vote. We have seen some evidence in the early vote that Latinos are voting in higher numbers than they did four years ago. We’ve seen some evidence in the early vote to suggest that pollsters may be underestimating their impact on the electorate, and if they are, we shall see that in states like Florida, Nevada, Arizona, a little bit in Colorado and New Mexico. And that, to me, is the question that remains unanswered. If pollsters are missing or underestimating the effect of the Latino vote, Hillary Clinton’s margin in those states will be wider than some pollsters have expected it to be.
Jody: So, let’s talk about that, because that does feel like a big storyline in general. Very briefly, what are some of the reasons why polling might miss a Latino vote?
Harry: Well, there are two reasons that come right to the top of my head. No. 1, Latinos tend to be younger. It’s more difficult to reach younger voters. Many pollsters, or some pollsters, still don’t call cell phones or perhaps don’t call enough cell phones. The other reason why it could be easy to miss is because there are many Latinos who don’t speak English or for whom English is a second language. So, if you only have English only interviewers or someone comes on the phone in English, a Latino voter who speaks primarily Spanish might hang up on them. So, you basically have these converging things. There are probably other things that also go into it, but as we have seen in the past, some polls in states like Nevada underestimate the Democratic vote. We haven’t necessarily seen that in a state like Florida, but it’s potential this time around given a high Latino turnout we’ve seen so far in the early vote, and we should get basically record Latino turnout in the state of Florida.
Jody: Nate, as our most ardent skeptic of emerging conventional wisdom, what do you make of this storyline that, oh, yeah, the Latino vote is going to be huge, the early voting numbers are overwhelming. I mean, Jon Ralston, who we all respect and is in Nevada, he’s going pretty far out there and saying, Nevada’s in the bag, Donald Trump is cooked. Do you buy that?
Nate: I think when you don’t have a systematic way to analyze things and there’s complex and contradictory information, your biases are fairly likely to get the better of you. Now, I think within the universe of Nevada and the universe of Jon Ralston that Jon Ralston has a systematic way to understand how Nevada early vote translates to votes. And you actually have a lot of Democrats who are like, oh, don’t worry about the polls, the polls show a close race, but there’s a lot of wishful thinking. Which is not to say that the value is zero, but … [people] should keep track — how many times when you try and predict the direction in which the polls are biased, how often you come out ahead. And by the way, see how often you predict that the party that you prefer to elect comes out ahead.
Jody: Because we’ve said this many times — if you’re going to say that the polls are biased, then you kind of have to expect that they could be biased in either direction.
Jody: Let’s talk a little bit about ground game, get-out-the-vote, the day of mobilization efforts. Democrats have tended to be better at this the last few cycles. Obviously, Hillary Clinton and Barack Obama are very different candidates, but the Hillary Clinton machine, if you want to phrase it that way, has learned a lot of lessons and has a lot of people who in many ways perfected it for Obama. So, Harry, what does a data driven, very concerted, get out the vote on Election Day effort look like, and how much of a difference maker is it?
Harry: Well, I’ll start at the second part of that. I think what’s most important here is you need to have enthusiasm among your supporters to turn out the vote. That’s something I learned a very long time ago. Get out the vote efforts are only as good as the people you’re trying to get out to vote, and if you don’t have enthusiastic supporters, you can forget about it. In terms of how much it moves the polls, not very much — you know, maybe a point of two here or there.
Jody: When you say moves the polls, you mean moves the results.
Harry: Right, moves the results a point or two, especially in a presidential election. It may be a different story in a down-ballot race. So, we’re not looking for huge effects. However, in a close race, that can make the difference. You know, we could talk about the ways in which that get out the vote effort works — door knocking, calls, apps, getting your friends to try to get to the polls.
Jody: Clare, on the Trump side, his last push, he went to Tampa; Wilmington, North Carolina; Reno, Nevada; Denver — all those were on Saturday. On Sunday, he went to Iowa, Michigan, Pennsylvania, Virginia. I think he’s in Manchester, New Hampshire, today. What do you make of that collection of states, and what do you think is important to watch when it comes to their final push?
Clare: Well, I’ve got to say — the fact that New Hampshire has turned back into, not even a symbolic battle ground, but … it’s like the primaries again. Everyone is in New Hampshire, and it’s a small state, but it’s…
Jody: Remember how cold it was?
Clare: But beautiful. I have to say, New Hampshire was beautiful. Just the loveliest snow.
Jody: That’s probably why Donald Trump went back there.
Clare: I have very fond memories of meeting Dave Wasserman in the middle of a blizzard. You guys had this lunch and this lovely picture…
Jody: We’re already getting so sanguine on this podcast.
Clare: Anyway. Trump was [also] in the Upper Midwest trying to, I would say, shore up that white support. [Then there’s] North Carolina. North Carolina has turned into the real interesting, fraught state. Micah Cohen, our editor, just shared a North Carolina GOP press release where they’re putting out these early voting numbers, and the GOP says “North Carolina Obama Coalition Crumbling.” African-American early voting is down 8.5 percent from this time in 2012. Caucasian voters early voting is up 22.5 percent from this time in 2012, which is quite a blunt…
Harry: They ain’t hiding nothing.
Clare: It’s quite a blunt thing to say. I think that it’s a convergence of, maybe, less enthusiasm for Clinton than Obama, but also, let’s not forget that provisions of the Voting Rights Act are no longer in effect in that state, and that the state legislature made efforts to curtail some early voting days and locations, so I think we might be seeing a double effect there.
Jody: OK, let’s go on to a quick round of one of our favorite games: “Buy, Sell, Hold.” And we’ll do it state by state for the most interesting states on the map. So, I’ve got a list here, and I’ll read the leader’s chance of winning, according to our polls only model. We might as well start with Nevada, which seems to be the trickiest state of the 2016 election so far. Hillary Clinton, according to polls-only, has a 53 percent chance of winning as of this taping right now. Nate: buy, sell, hold?
Nate: I would buy Hillary Rodham Clinton in the state of Nevada at that price.
Jody: So, you’re selling the forecast. You don’t trust our model. You think the chances are higher.
Nate: I do trust our model, and I think that anyone who wants to make subjective calls and tries to beat our model, I think — you know, we’ll see how that goes. But you set up this experiment where I have to do this thing that I don’t like to do.
Jody: Technically, you set up the model and then I just set up the experiment on top of the model.
Nate: Nevada is a state where the combination of historical polling error, the combination of demographics being a little bit hard to reconcile with a tight set of polls there — she has a teeny bit of momentum in some of the polls, in certain of them she went from losing Nevada to winning it, so yeah, I’d buy Clinton.
Jody: So you think her chances are a little higher than 53 percent. Clare?
Clare: I’m going to buy. I’m heavily influenced by Jon Ralston’s reporting.
Jody: Never has a picture of a line outside a grocery store electrified the political community more than that one that Jon Ralston tweeted. I mean, I’ve seen that picture everywhere.
Clare: I mean, it also tracks, I think, with the idea that Latino turnout seems to be a lot higher in Florida, so if we’re talking about demographic surges that could be the story of this election, I would say that I feel pretty good about Nevada.
Jody: OK. And Harry?
Harry: Make it three, three’s a crowd.
Clare: Make it three, bartender.
Harry: I am buying Hillary Clinton’s chances in the state of Nevada for the reasons that have basically been spoken about here. Democrats are voting in high numbers, the unaffiliated voters traditionally in that state lean Democratic, the internal polls suggest that they continue to do so. I am buying Hillary Clinton in the state of Nevada.
Jody: OK. Ohio: 66 percent chance, according to our model, of Donald Trump’s winning Ohio. Nate — buy, sell, hold, what do you make of that evaluation?
Nate: Um, I think I hold. Those seem like appropriate odds to me. I’d buy Clinton if I had to pick, I think just because it’s a little bit hard to explain why Ohio polls have been quite that bad for her. And so, given 2-to-1 odds, and given that she seems to maybe have a half a point of a lift here in the last set of national polls and that you’ve seen the campaign still spending resources there as though it were a competitive state, she could win it by half a point or something.
Jody: Clare, what do you have to say for your home state?
Clare: Um, sell. I think she might have a better chance than we’re giving her.
Jody: So, you’re buying Clinton? You’re selling Trump?
Clare: Yeah. Yes. I don’t know which way I’m supposed to phrase —
Jody: Yeah, no one knows. We’re just confusing our audience.
Jody: Harry, Ohio?
Harry: My tendency is also to match Nate’s, which is to hold. But I have been interested in Clinton believing that that state isn’t gone, that she did do some rallies here at the last minute. We also know that the Republican Party in that state hates Donald Trump, so there could be a GOTV problem for him out in that state. Again, I hold, I think Trump will probably win in the end, but if our model is going to be wrong, it will be in underestimating Hillary Clinton, not overestimating her.
Jody: OK. Let’s go through a couple others on my list, and we’ll just do them real quick here. North Carolina. Basically, a toss-up in our model. Nate, what do you think?
Nate: I think it’s a toss up. I mean, the polling trends haven’t been great for Democrats there, the African American turnout hasn’t been great for Democrats there, so I would be reluctant to put money down in that state.
Jody: Clare?
Clare: I’m putting money down.
Jody: On who?
Clare: I think … oh, this one’s hard.
Jody: How could you not expect that follow-up question?
Clare: No, I know, I know. I think Trump might win North Carolina.
Jody: OK.
Harry: I’m going to go with Nate. I’m holding.
Clare: See, I refuse to hold on this game.
Jody: Holding on a 50/50, it’s no fun. And politics is supposed to be fun.
Clare: I have no statistical reputation to uphold.
Harry: Gary Johnson will lose.
Jody: All right, Gary Johnson will lose. That’s Harry’s bold prediction. Florida, essentially the same: 51 percent, as of this taping, Clinton favorite. Nate, do you also see it as a pure toss-up or do you think one of them has an edge that’s not being represented in our model?
Nate: You know, Florida and North Carolina are the shruggy states for me. I think that early voting looks a little better for Democrats in Florida and North Carolina, so the extent that you believe in that. I’ll put it like this. If I had to call that split, then I would go Florida for Clinton and North Carolina for Trump, which is of course how it went four years ago.
Jody: Clare?
Clare: I think Clinton’s going to win Florida. And I’m going to go to the Latino turnout and say that I think that will be a powerful motivating force.
Jody: Thank you for playing this game, Clare.
Clare: You’re welcome.
Harry: I’m going to agree with Clare on this one. I believe Hillary Clinton will win the state of Florida, and it will probably be apparent as of 8 p.m. Eastern time when the polls in all the parts of the state close because the large early vote —
Nate: I don’t know. I’d be careful with that, because people are going to make inferences from the early vote, and we’re not going to know —
Harry: What I know from watching the past few cycles is that the early vote, and the early vote as of counted as of 8 o’clock in the evening when all the polls close matches fairly well with the final results. That doesn’t mean it’s perfect.
Jody: OK. Pennsylvania. Hillary Clinton has never trailed in the polls there. Is that right? Our model says 76 percent chance for Clinton — does anyone think the chances are lower than 76 percent?
Nate: I mean, relative to the rest of the country? My theory is that it’s more plausible for Trump to win Pennsylvania and some other rust belt states like Michigan than people infer, and that there are these kind of sophomoric arguments like, “Oh, well, she’s been ahead in all the polls.” But that doesn’t really matter much if you happen to be one point ahead or two points ahead and it’s close. And you understand the scenario by which she loses a state like Michigan or Pennsylvania is, you know, those white working class voters, those union workers go further to Trump, and she doesn’t turn out her African-American base. If both of those things happen — so you’re flipping a coin twice, 25 percent chance — then I think she could be in jeopardy in one of those states. People are a little smug about saying, “how dare Trump campaign in Michigan, that shows how desperate he is.”
Clare: I would say on Pennsylvania that I have to, since I gave her better odds of winning Ohio, I have to give her worse odds of winning Pennsylvania, if I’m being consistent. Because I think it’s a similar thing with Ohio, where Trump is really relying on turning out those people who are deleted off the voter rolls for years and years in Western Pennsylvania who are going to turn out, or is she going to get enough people in the Philly area to really overcome it? So, I would say … she’s going to win it, probably, but would I give her a 76 percent chance? I’d say it’s closer.
Harry: Clinton will win in the state of Pennsylvania. I would think Michigan is the one I’d be more worried about. There’s just too big of a vote coming out of Philadelphia and its suburbs, even if Trump has a good turnout operation in the western part of that state.
Jody: Is there any chance to wrap up I can extract a Buy, Sell, Hold on the overall forecast? Hillary Clinton, 67.6 percent right now. This is all just a way to back Nate into predicting the race.
Clare: Listen, I will speak, even if no one else will. I think Hillary Clinton is likely to win this election. I think it’s going to be closer than we thought it was going to be like, mid-October when everyone was writing like, the landslide.
Jody: Including me. Although, I did caveat that with, “this will come back to haunt us.”
Jody: Nate?
Nate: Pass. Hard pass.
For more elections coverage, check out:
Have questions you want answered on the podcast? Get in touch by email, on Twitter or in the comments.
We’ll be recording daily Elections podcasts from now until Nov. 8 and posting new episodes every afternoon. Be sure to subscribe so you don’t miss a new episode.
You can listen to the episode by clicking the “play” button above or by downloading it in iTunes, through the ESPN App or on your favorite podcast platform. If you are new to podcasts, learn how to listen.