NPR logo
Why One Forecaster Doesn't Think The GOP Will Take The Senate
  • Download
  • <iframe src="https://www.npr.org/player/embed/354371670/354371671" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Why One Forecaster Doesn't Think The GOP Will Take The Senate

Politics

Why One Forecaster Doesn't Think The GOP Will Take The Senate

Why One Forecaster Doesn't Think The GOP Will Take The Senate
  • Download
  • <iframe src="https://www.npr.org/player/embed/354371670/354371671" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

Most forecasters predict that Republicans will pick up enough seats in the midterms to control the chamber. But Sam Wang at the Princeton Election Commission predicts the Democrats will hold on.

ROBERT SIEGEL, HOST:

We're less than a month away from the midterm elections. Those elections will determine which party controls the Senate next year. Most prognosticators give the Republicans the edge. That includes the majority of election forecasters who've crunched the poll numbers and other data. But there is one big exception, as NPR's Joel Rose reports.

JOEL ROSE, BYLINE: If you're a Democrat seeking comfort about your party's chances of holding the Senate in November, your number cruncher of choice does not work inside the Beltway or at a high-profile media outlet in New York. He runs a neuroscience lab at Princeton University.

SAM WANG: These are microscopes that allow observation of the brain in a living animal.

ROSE: Sam Wang is a professor of molecular biology and the author of "Welcome To Your Brain." He's used to doing complex math problems at his day job. About 10 years ago, he started applying his expertise to elections by analyzing the polling data for every race he could get his hands on.

WANG: What we do is sort of cook down the data to come up with pure essence of poll and then just turn that into a snapshot that is ideally neutral and doesn't add any judgments.

ROSE: Wang has had his stumbles, like when he called the 2004 presidential election for John Kerry. But he had a very good year in 2012. He called every state in the presidential election right and every single Senate race. This year, when most prognosticators see a good to great cycle for Republicans in the Senate, Wang predicts something different.

WANG: At the beginning of this campaign, analysts and politicos came in expecting the Republicans to have enough of an advantage to take over control of the Senate. Real conditions on the ground are a little bit more favorable towards the Democrats at the moment - but only a little bit.

ROSE: As of today, Wang gives Democrats a better than 60 percent chance of retaining Senate control, a predication that makes him very much an outlier.

NATE SILVER: Yeah, I would like to place a large wager against that guy if I were allowed to do so, which I don't think I am contractually.

(SOUNDBITE OF ARCHIVED RECORDING)

ROSE: That's Nate Silver of FiveThirtyEight, the statistical savant who brought election modeling into the mainstream, speaking last month on WNYC. Unlike Wang, Silver gives the Republican Party a nearly 60 percent chance of taking the Senate, which, as Silver points out, is in line with most of his competitors, including The New York Times and The Washington Post.

SILVER: When you have a result that's outside of consensus, you should be skeptical about that. Almost always the crowd is wiser than any one individual forecaster.

ROSE: To understand why Silver and Wang can reach such different conclusions, you have to look at the differences between their models. For one, Wang only considers polling data, while Silver's model at FiveThirtyEight also factors in what he calls the fundamentals.

SILVER: The polls tell you most of what you'd want to know. But still, looking at what the fundamentals of a state are - meaning how it's voted in past elections, who's raised more money. We look at the generic congressional ballot, which is a sense of the national environment.

ROSE: But Sam Wang at Princeton doesn't include any of that stuff in his model. Wang says adding in more information just makes it more likely that the forecasters will inadvertently skew the results.

WANG: The difficulty with any kind of model based on congressional approval or the economy or what have you is that the researcher has to make judgments about where things ought to be. And if you don't hit it exactly right, then the effect is to put a finger on the scale slightly.

CHARLIE COOK: Intuition is part of what we do. I've been doing this for 30 years.

ROSE: That's Charlie Cook, preeminent political forecaster and editor of The Cook Political Report. He still makes predictions the old-fashioned way, interviewing candidates, talking to people on the ground and ultimately trusting his gut.

COOK: People who just sort of blindly plug polling numbers into a model without necessarily taking into account whether these pollsters have any idea what they're doing, I kind of have a problem with that.

ROSE: Cook, like Silver, gives Republicans a slight edge to control the Senate. But no one thinks it's a slam-dunk. There are eight or nine Senate races that are just too close to call. And someone is going to have egg on their face on November 5. Joel Rose, NPR News, New York.

Copyright © 2014 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.