NPR logo
Why Did The Polls Fail To Predict Sanders' Win In Michigan?
  • Download
  • <iframe src="https://www.npr.org/player/embed/469837001/469837002" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Why Did The Polls Fail To Predict Sanders' Win In Michigan?

Political Data & Technology

Why Did The Polls Fail To Predict Sanders' Win In Michigan?

Why Did The Polls Fail To Predict Sanders' Win In Michigan?
  • Download
  • <iframe src="https://www.npr.org/player/embed/469837001/469837002" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

Every major poll of the Michigan Democratic primary showed Hillary Clinton with a big lead over Bernie Sanders. NPR's Kelly McEvers speaks to Harry Enten, senior analyst at FiveThirtyEight, about why those polls proved to be so wrong on primary day.

KELLY MCEVERS, HOST:

As we saw last night in the Democratic primary in Michigan, the polling can be way off. In almost every survey, Hillary Clinton outpolled Bernie Sanders by 10 or even 20 points in Michigan. Not one single major poll had Sanders ahead, but then Bernie Sanders surprised all the pollsters and won. To talk about what this means, we called up Harry Enten. He's senior analyst at FiveThirtyEight. Thanks for coming on the show today.

HARRY ENTEN: Thanks for having me.

MCEVERS: And, Harry, one of the reasons we called you specifically is because FiveThirtyEight and your boss, Nate Silver, have a really good track record of forecasting elections. So yesterday, what was the prediction that you all had for Clinton to win Michigan?

ENTEN: Well, we had her chance of winning was greater than 99 percent. And that was based upon a whole bunch of polls. And what we saw yesterday was very, very shocking based upon the polls that you just illuminated upon, where Clinton was leading not just in one, not just in two, but all of the polls. And the polling average, in fact, had her winning by somewhere around 20 percentage points, so this was a real shocker.

MCEVERS: So at this point, what is your educated guess as to why the polls and even your own prediction was so off?

ENTEN: So, I would say, you know, three things that come to mind. Number one, I think there were a number of Democrats who might have decided, hey, you know what? Hillary Clinton has got this in the bag. It's an open primary. We can go vote on the Republican side, and maybe we'll vote to stop Trump by voting for John Kasich, or maybe we think Donald Trump is such a weak general election candidate, as a Democrat, we want him to face Hillary Clinton, so we're going to go vote in a Republican primary for Donald trump, so that's one reason. The second reason is there may have been a miscalculation of how many young people were actually going to vote in the Democratic primary on Tuesday night. It may have been an underestimation. And obviously, young people love Bernie Sander. And then the third thing I would point out - African-Americans, although they supported Hillary Clinton, they supported her by a smaller margin in the exit polls than the pre-election polls suggested.

MCEVERS: And talking about those young people, I mean, is it possible that people in their 20s and 30s aren't using landlines, and if a pollster calls someone's cell phone, they won't necessarily answer the number? I mean, is that just a harder demographic to poll?

ENTEN: Yeah. I mean, response rates among younger people tend to be lower than they are among older people. I mean, I know a few older people in my lifetime. Some of them wait around the phone just begging to talk with somebody, and they're more likely to pick up a phone and talk to a pollster, but it's even more than that. Some of these posters didn't even call cell phones. And if you're not doing that, then you're clearly going to miss the Sanders vote, and that was a big, big problem. But not one of these things explains completely what happens. It's probably a whole bunch of - a confluence of events.

MCEVERS: Do you think the failure of polling last night was a fluke, or do you expect similar discrepancies in upcoming, you know, Illinois and Ohio primaries?

ENTEN: You know, that's the real question, right? Was it a fluke? If it was a fluke, then no one's ever going to - you know, people will remember this as sort of - oh, my God, Bernie came from behind, but then Clinton, you know, wins in Illinois and wins in Ohio and no one will remember it for anything more than just a fluke. But if it's something systematic in open primaries, specifically in the Midwest, then we could be into some shady ground. If you look at the polls, Clinton's up by even more in Illinois than she was in Michigan, so that, perhaps, is not a place where Bernie Sanders could win. But in Ohio, the polls are actually closer than they are - or were in Michigan, so it could be a place where you may have some Democrats deciding, you know what? I'm going to vote in the Republican primary and vote for our home-state Gov. John Kasich, where a polling error could, in fact, happen.

MCEVERS: That's Harry Enten. You can hear more of him on the FiveThirtyEight elections podcast. Thanks so much.

ENTEN: Thank you.

Copyright © 2016 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.

Comments

 

Please keep your community civil. All comments must follow the NPR.org Community rules and terms of use, and will be moderated prior to posting. NPR reserves the right to use the comments we receive, in whole or in part, and to use the commenter's name and location, in any medium. See also the Terms of Use, Privacy Policy and Community FAQ.