Tuesday, May 23, 2017

Interview: How can pollsters improve on their 2016 failure? Charlie Cook explains.

This is the final article of my discussion with Charlie Cook. The firstsecond, third, and fourth parts are available. 
Story by Joseph Ford Cotto
"Like I've said before .... polls are only good for strippers and cross-country skiers," Former Alaska Governor Sarah Palin told ex-Fox News host Bill O'Reilly two days before last year’s presidential election.
While I have criticized her more than a fair bit during years gone by, it is undeniable that she turned out to be a sage among pundits. (Read more here)
****

Joseph Ford Cotto: Since it is growing more difficult to take scientific surveys, might the events of 2016 spell an eventual end for the polling industry?

Charlie Cook: As long as people care about the outcome an election and are preoccupied by the horserace aspect of the campaign, there will be a focus on polling, maybe just a bit more skepticism than in the past.

Cotto: Beyond any other factor, what can pollsters improve on relative to their general failure last year?

Cook: For decades the challenge for pollsters was ascertaining who would vote and who would stay home.  Screening questions that work very well in one or more elections suddenly don’t work in another.  The Gallup Organization had a half dozen screening questions that had been quite accurate in determining likelihood of voting suddenly didn’t work in 2012.  As a result, Gallup got out of the horserace polling business after that election, instead polling on candidate and party images and issues.

The Obama campaign knew that there was no way that Clinton could replicate the black vote that our first African American president received in 2012, they estimated that the African American share of the electorate would drop by about five points, it dropped by eight points.  What could Donald Trump have possibly been able to say that would antagonize Latinos more than he did, yet the Latino share of electorate only barely went up from 2012.  Jill Stein and Libertarian Gary Johnson dug deeper into younger voters who were never particularly enthusiastic about Hillary Clinton than anyone anticipated.

Going into Election Day, Nate Silver’s Fivethirtyeight website gave Hillary Clinton a 71 percent chance of winning, many other modelers, some of whom were put her chances at 97 percent or more trashed Silver mercilessly for being too cautious when they ended up having egg on their faces instead.  Silver was intellectually honest, stuck with his data, others were guilty of hubris.

We should all remember that when you are looking at elections, we are talking about human behavior, with humans occasionally behaving in unpredictable ways.  There are many variables that go into these elections and they can change in the final days or even hours before an election.  Those of us of a certain age can remember the 1980 race between Ronald Reagan and President Jimmy Carter was almost universally considered to be ‘too close to call’ a week before the election but exploded over the last weekend.  Carter Strategist Hamilton Jordan and pollster Patrick Caddell called White House Press Secretary Jody Powell who was traveling with the President on Air Force One late Monday night, Jordan telling Powell, “the bottom has dropped out – it’s all over.”  The next day Reagan won a ten point landslide.

No comments:

Post a Comment