State of the Polls: Which Numbers Matter?

State of the Polls: Which Numbers Matter?

Republican presidential candidate Donald Trump greets supporters after a rally at Ladd-Peebles Stadium in Mobile, Alabama, on Aug. 21, 2015.

Why you should care

Because, sometimes, polls are no substitute for a handy crystal ball.

To many in the body politic, the biggest losers of the 2016 election season have been mainstream media pundits. Their heads are now in constant swivel, their talking points wrong almost as soon as they are intoned — especially in re: the Donald — and yet they must keep talking. Polls show that while the Trump juggernaut has slowed some, it’s still a juggernaut, and how many pundits missed that?

Last week, of course, the prognosticators were all over numbers showing a Ben Carson surge, but here’s the thing: Polls are not infallible either. We’ve got that straight from the mouth of polling don Cliff Zukin, past president of the American Association for Public Opinion Research and a political scientist at Rutgers University. Here in an interview excerpt, Zukin talks about Trump’s long-term viability, what’s wrong with traditional polling in a mobile age and why early stage polls matter at all.

OZY: Should we buy the Trump lead in early primary states like New Hampshire?

Cliff Zukin: A lot of people believe that Donald Trump’s lead is overstated, because the people who tend to vote in Iowa and New Hampshire tend to be very, very conservative voters, very religious and fundamentalist. And that’s not Trump. He’s had three wives and he’s not very religious. You’ve got to take polls with a grain of salt. The polls are not very good at predicting results, especially in small turnout elections.

OZY: So if the numbers can’t tell us the winner, what can they tell us?

C.Z.: With polls, you look for change over time. I can tell you who is visible and who is popular, and then those things do, in fact, play out in politics. They help with fundraising, endorsements and setting an agenda on a campaign. Now that candidates like Carly Fiorina and Ben Carson are higher in the polls, they get more media coverage and they can talk about the issues that are important to them. And I think the polls are sort of accurate when they show that Chris Christie, Scott Walker (who dropped out of the race in September) and Jeb Bush have not caught on with potential voters. And I think they’re probably accurate when they suggest that there are a lot of Democrats who are not enthusiastic about Hillary Clinton.

OZY: What do you think we’ll see in the polls as we get closer to primary season in early February and the field of candidates starts to narrow?

C.Z.: The polling will affect fundraising a lot. If your support is going down in the polls, it’s very difficult to raise money. That will give an advantage to someone like Jeb Bush, who has already raised $100 million and can afford to hang in there until the primaries. Some will have to drop out because of a lack of funding. With only two debates left in the calendar year, I don’t know how candidates are going to change perceptions about themselves — and that will be reflected in the polls.

OZY: Which polls can we trust, and which shouldn’t we?

C.Z.: Journalists tend to treat every poll like the same poll. Everything is equal, bad sample or good sample, early state or not. But some are better than others; Pew and Gallup are probably the best out there.

Before cellphones, we could do random digit dialing, where we would just call sample area codes and make up the rest of the numbers. The sampling on those geographic areas got us a pretty good representative sample of telephone households in the U.S. As recently as eight years ago, in the 2008 election, just 6 percent of people in the country used only a cellphone. Now it’s at 60 percent. So if you did a random poll of landlines, you would miss 60 percent of the public. Coupled with that, the number of people willing to respond to a poll has fallen. To do a survey of 1,500 now, you could have to make up to 30,000 telephone calls.

Polls can provide useful information about favorability and visibility, but it’s like having a scale that can do pounds pretty well but can’t do ounces. There’s a level of precision needed to do stuff about primary election polling that we really don’t have the resources to do anymore.

Comment

OZYPOV

Interviews, op-eds, and analysis to help you make sense of the news of the day and the news of the future.