( Saul Loeb/AFP via Getty Images )
Last November, polls showed--or so it seemed--that Donald Trump held 25 to 30 percent of voter support. Nate Silver of the data news site FiveThirtyEight advised caution in taking those polls at face value: they were probably closer to six or eight percent, and could have been the result of trolling voters. Seven months later, moving out of the primaries and into the next leg of election season, we revisit his reasoning.
BROOKE GLADSTONE: As we noted, you are listening to a repeat of a show we first ran in December, a mythical time when Trump's nomination was unthinkable and, according to FiveThirtyEight’s Nate Silver, unsupported by the numbers. Silver told us that despite Trump's claim at the time that he was ahead in the polls, he was actually polling at about 8 percent, if you figure that he had 25 percent of the 25 percent of self-declared Republicans.
NATE SILVER: It’s not that hard to get 6 percent or 8 percent of people to say much of anything in a poll. I mean, 6 percent of people, I think, told Gallup a few years ago that they thought the moon landings were staged.
BROOKE GLADSTONE: In other words, Silver suggested that Trump's popularity was kind of a lie. He also noted that history shows that the American people are not likely to vote for a person like Trump. In March, after Trump defied the pundits, and history, to clinch the nomination, we spoke to Nate again and asked him what the hell happened.
NATE SILVER: There is not a lot of precedent for something like Donald Trump. I mean, this is something pretty amazing. And I don't know, I guess I don't feel that bad about having been initially skeptical.
BROOKE GLADSTONE: Okay, I hear you but I want to know about what you do feel bad about? [LAUGHS]
[SILVER LAUGHS]
I mean, the thing is, is you’ve gotten numbers, you’re used to reading them a certain way. So is there somewhere where you went wrong, or not? You read numbers for a living.
NATE SILVER: I do, and those numbers still suggest that it's very unusual for a candidate to win the nomination with as much opposition from within his own party as Donald Trump. I mean, there are a couple of issues here. One is extrapolating from history and realizing that we don't have that many examples. We have only one election every four years, and that maybe a lot of things that were true from the period from, say, 1980 through 2012, well, that sounds like a lot but it's also still just one election every four years.
BROOKE GLADSTONE: Mm-hmm.
NATE SILVER: So, you know, we said, well, Donald Trump is someone who’s going to be vigorously opposed by Republican Party elites, so because it's so consequential they’ll find some way to stop him. Well, maybe not.
BROOKE GLADSTONE: You mentioned historical precedent. How many elections are we really talking about to base these predictions on?
NATE SILVER: Well, since 1972, which is when you had the McGovern reforms and you started to have people actually vote in primaries - before it sort of literally was smoke-filled rooms - it’s depending on how you count, somewhere on the order of 10, 12, 15. To say something hasn't happened in 12 tries is much different than when we also make sports predictions. Then we can say, well, this hasn't happened a thousand games. [LAUGHS] Then you’re more existentially certain that something is really unlikely than you can be in the nomination process or, for that matter, a general election too, where there have been, I think, 16 or 17 now elections since World War II in the polling era. That’s not a huge sample size either.
BROOKE GLADSTONE: One thing that historians have noted is that there are political realignments every 40 years or so. This is when the parties kind of break apart and reassemble somehow. Describe a couple of political realignments. What happened in those elections, and whether you think we’re in one of those right now?
NATE SILVER: You had the New Deal coalition in 1932 or so, where you had overwhelming majorities of voters were voting Democrat, both for FDR and for Congress. That was very sudden and very dramatic. The more recent transition was a little slower. It certainly started at about 1968, with Nixon's Southern strategy. You could argue it didn't fully culminate until 1994, when Republicans won back a lot of seats in Congress in the South. I mean, people forget that it was a Democratic South [LAUGHS], right, and the governor of Alabama was sure gonna be a Democrat. He might have been George Wallace or someone who would not have very much in common with a Democrat in New York City but certainly a capital “D” Democrat, still. And that took some time to unravel.
But, you know, I suppose I'm not sure whether the GOP is going through a transition suddenly or it will take a couple of cycles, but I am just saying that as someone who was skeptical – a lot of people were - but we were certainly skeptical about Donald Trump's ability to do what he's doing because the consequences would be so large for the GOP. Now that it's happening or on the verge of happening, we still think the consequences are really enormous. And it's a little bit like looking up and seeing that the sky is no longer blue but orange.
BROOKE GLADSTONE: Last week, you joked on Twitter that, quote, “’With the exception of the 2016 election’ will be a common phrase in PhD dissertations in 2044.” [LAUGHS]
NATE SILVER: We’re pretty good at placing events into context, after we have some window of perspective. Somehow or another, by 2044, the 2016 election will fit into the narrative that we tell about how elections work. I do want people to understand how, how amazing it is, as someone who covers elections. There are a lot of false alarms. There are a lot of elections where prognosticators cry wolf and say something new and different has happened, when it wasn't. But this one, we almost feel like instead of wanting to calm everyone down and saying, no, you probably won’t have a contested convention, no, you probably won't have a third-party candidate running, we have a different role this year. [LAUGHS] We want to be able to point out, like, this kind of breaks so many rules that we have to now consider what rules that we previously held to be true might also be wrong.
BROOKE GLADSTONE: Nate, thank you so much.
NATE SILVER: Uh, thank you.
BROOKE GLADSTONE: [LAUGHS] Weren’t sure about that, were you? [LAUGHS]
NATE SILVER: No, I totally am sure.
[MUSIC UP & UNDER]
BROOKE GLADSTONE: Nate Silver is the founder and editor of the data news site FiveThirtyEight.com. He’s also author of The Signal and the Noise: Why So Many Predictions Fail but Some Don’t. And FiveThirtyEight is our collaborator during this election season.