clock menu more-arrow no yes

Filed under:

Why It’s Not Time to Overreact to the Polls

Getty Images
Getty Images

Hillary Clinton’s poll numbers have continued to slip, with her chances of winning the election, as calculated by FiveThirtyEight, dipping below 60 percent Monday for the first time since August. But longtime Democrats Dan Pfeiffer and Jon Favreau aren’t worried — yet. On the latest Keepin’ It 1600 podcast, the two explained why it’s not yet time for Hillary supporters to panic.

Listen to the full podcast here. This transcript has been edited and condensed.

Trump Is Still a Long Way From Victory

Dan Pfeiffer: Trump has to draw an inside straight. If you give [Hillary Clinton] Colorado, Virginia, Wisconsin, Michigan, Pennsylvania, then he has to win all the remaining states. And that’s pretty hard to do.

Jon Favreau: Now there’s a set of Fox News polls yesterday, state polls, and we don’t want to shit on Fox polls because the pollster’s independent of the network. It’s actually a pretty good pollster. So [Fox] has Trump up three in Nevada, up five in Ohio, and up five in North Carolina. Nevada, I think, is a weird polling state, and the Democratic margin has been underestimated in Nevada for the last however many elections. There’s something weird going on with the polling there. North Carolina has been sort of close.

Why Hillary Is Underperforming in Ohio

J.F.: The one that I can’t figure out what’s happening is fucking Ohio, man.

D.P.: Lot of old white people.

J.F.: Yeah, like poll after poll in Ohio for the last couple weeks have not [been] good for her. Now again, like you said, it’s not like she has to win Ohio. Trump has to win Ohio to win.

D.P.: So I think in Ohio, it was pretty close in ’12. And one of the reasons it was close was Romney was a particularly poor candidate to get white working-class voters.

Romney looked, acted, and talked like the guy who fired people from their jobs, right? The guy who bought their companies and then shipped the jobs to China or Mexico or whatever else, and so Trump is obviously a better candidate for that. Ohio is also a tough state for Democrats in the sense that there is not some giant pocket [where you can gather support]. It’s been so organized and worked over. Because it was at the center of the ’04, ’08, and ’12 elections, there’s not some giant pocket of unregistered Democratic votes like [there is] in Florida, right. There’s 725,000, I think, unregistered or ineligible Latino voters who did not vote in Florida. There was a very, very large number of African American voters that were registered in ’08, and ’12 in Florida. That doesn’t exist in Ohio. So it’s going to come down to the wire. Even in a race that Clinton wins by five points nationally, is my guess.

J.F.: Well, what explains a lot of the discrepancy between Clinton-Trump and Obama-Romney: In some of these states, [they are] showing the percentage of the electorate of registered voters who are non-college-educated, white voters, and the states with the highest percentage are [where] she’s having a tougher time. So that’s why [in] the second district of Maine, she’s not doing as well. That’s why Nevada is a little off, even though there’s a big Latino population. Ohio falls into that category, and stuff like that. And then in states where there’s actually more, a bigger percentage of college-educated whites, plus Latinos and African Americans, she’s doing, in some places, better even than Obama did. And that’s the North Carolinas, Florida, Arizona, Georgia, places like that. And I think there’s something to that. I don’t think that’s the difference between Hillary and Obama so much, but the difference between the fact that Trump has so polarized college-educated whites and non-college-educated whites.

Hillary’s Underperformance With Latinos Gives Her Room for Growth

D.P.: The other challenge for Hillary is she’s currently underperforming Obama’s 2012 election day numbers among Latinos in states like Florida. So there’s a big article in The New York Times.

And [there is] a lot of the concern in the community about it. And there are a lot of quotes from people saying they waited too long to go on the air with Spanish language ads. I don’t know if any of that is true. I actually read that as a quasi-positive sign, because that is a pocket of available growth. If she were losing or tied in at Obama’s 2012 win number with the Obama coalition, then you’re in big trouble, because you have run out of places to go get votes. So it’s going to take a shitload of work, and she’s going to have to be really smart to have to go get those votes in places like Florida, but it is an achievable thing. Like you can look at it and say, you have 49 days or whatever it is, and there’s a way to go do that.

Election Models Don’t Matter — Votes Do

J.F.: Now some people are annoyed with the FiveThirtyEight model because it’s showing her at like 57, 58 percent chance of winning, and Trump’s at like 41 percent. It’s not the narrowest margin that they’ve had all year, [because] right before the Democratic Convention, it was a bit narrower. But you know, it’s close. And some of the other models don’t have that.

I don’t know. I think this is a bit of a silly argument. My thing is, like, if you’re looking at all these models and you’re focused on what percent chance she has to win, I don’t think that necessarily [is] the most helpful thing. These models also have what’s her average number, or what’s the poll number. Is it 46–43? 44–42? Look at those numbers and then look at the electoral votes that they would win if the election were held today. Or what the polls average. I think that’s going to give you a much better idea of the outcome than like, projecting ahead on percentages.

D.P.: Everyone has their model. You can’t really know whose is right and whose is wrong. So “The Upshot,” the New York Times folks, who I think are also very good, their model was like somewhere in the 70s, depending on the day. FiveThirtyEight’s is much lower. If Hillary wins, who knows who is right? Like, what’s the 70 percent? What’s the 50 percent? Who the fuck knows?

As long as Hillary wins, they’re both right. And we’re not going to know. I mean, both predicted individual states and then you can look at that, but it really doesn’t make a difference. The other thing is, it doesn’t really matter if there’s a 30 percent chance, a 50 percent chance, a 5 percent chance. Go vote. If you’re really concerned about it, go vote.

J.F.: Then there’s the backlash of that, like, how can so many people, smart people, think that Donald Trump doesn’t have a chance of winning? No. Of course he has a chance of winning. It’s a very real thing. Yes, Donald Trump could win. So go vote. Don’t argue about whether he has a chance to win or not. Don’t argue about whether he should win or not. Go do something about it. That’s the message, you know.