Known Unknowns

The Trump Uncertainty Principle Is Rattling Republican Pollsters

The big question remains whether the GOP front-runner can translate enthusiasm into actual votes.

First GOP Primary Debate Hosted By Fox News And Facebook
Photographer: Andrew Harrer/Bloomberg

The big question about Donald Trump is not—as an endless stream of false predictions for months would have it—will he fade in the 2016 presidential race, but whether Americans who tell pollsters they support him actually show up to vote in Republican caucuses and primaries.

Put another way: are past Republican voters a true measure of the current primary electorate or will Trump bring in new voters with his unconventional brand of politically incorrect nativism and bravado? 

"There's a debate about that—some pollsters think you shouldn't deviate too far from what past electorate has looked like," said Ken Goldstein, a politics professor at the University of San Francisco and polling analyst for Bloomberg Politics. "Most professional campaign pollsters would fall in that camp, that the best predictor of future performance is past performance. Then there are others who say let the data show what they do. If in your data you're picking up more or less enthusiasm, that could be a tell that the electorate is going to be different."

For the sixth consecutive month, the New York businessman continues to lead in nearly every national poll, sometimes by daunting margins. But his leads tend to be higher in surveys of Americans who say they plan to vote than those whom pollsters traditionally consider more likely to vote as they have voted in recent elections.

A Monmouth poll released Monday illustrates the point: Trump had a crushing 27-point lead, clocking in at 41 percent versus Ted Cruz at 14 percent and Marco Rubio at 10 percent. Participants included registered voters who identify as Republican or Republican-leaning, with little clarity as to who among them will actually vote.

In Iowa, Cruz is leading by 5 points in a recent survey of recent voters by Monmouth and by 10 points in a Bloomberg Politics/Des Moines Register survey of registered voters who said they plan to caucus. But Trump led by 13 points in a CNN poll just days earlier, which featured a wider sample of voters whom were deemed "likely" to caucus. 

In New Hampshire, a December poll by Adrian Gray Consulting found that Trump was ahead by 1 point among "likely" Republican voters but trails Marco Rubio by 2 points among "very likely" Republican voters.

The agonizing thing for Republicans is there's no way to know which polls are more likely to be born out when the actual votes are tabulated.

"It’s certainly the case that Trump’s supporters don’t necessarily look like 'likely voters' as we know them. But it’s not impossible to think that Trump could re-shape what it means to be a 'likely voter.' This is, after all, a very strange election indeed," Republican pollster Kristen Soltis Anderson wrote in National Review, a conservative publication.

Some anti-Trump Republican strategists take comfort in the discrepancy, arguing that plenty of his backers won't show up to spend their evening caucusing or wait in line on primary day to cast a ballot. They also note that national polls are of limited value given that there is no "national primary."

Recent history teaches an importance lesson. In 2008, numerous polls showed Democrat Hillary Clinton leading Barack Obama in the days before the Iowa caucuses. She lost by 8 points. The polls had underestimated Obama's success at mobilizing new voters and remaking the electorate, as did the Clinton campaign.

"It's what Hillary Clinton did in 2007 leading into the caucus, and my sense is they were blind—they were blinded by not knowing there was a huge influx of new people coming," said J. Ann Selzer, the president of the Iowa-based Selzer & Co., which conducts polls in the state and nationally for Bloomberg and the Des Moines Register.

Selzer said it "certainly is a possibility" that the 2016 Republican contest will pull in new voters who don't typically caucus or vote in primaries. "There's a huge crop of candidates out there and maybe somebody's struck their fancy," she said.

But as Selzer and Goldstein both noted, an important—and unknown—question is whether Trump can match Obama's impressive ground game in 2008, which methodically took advantage of the enthusiasm behind the candidate by collecting Iowans' information at rallies, encouraging them to sign pledges to caucus for Obama and mobilizing them as foot-soldiers on the campaign's behalf.

If the Republican establishment believes the 2016 primary is going to look like past Republican primaries, "then they shouldn't worry about Trump," said Goldstein. "But the fact that they're worrying suggests, right or wrong, that they believe something different is going on this year."

Before it's here, it's on the Bloomberg Terminal. LEARN MORE