📷 Key players Meteor shower up next 📷 Leaders at the dais 20 years till the next one
WASHINGTON
Election polls

Why Iowa polls were wrong

Deirdre Shesgreen
USA TODAY

WASHINGTON — If Donald Trump was feeling cocky on the eve of Monday’s Iowa caucus, he had good reason: 13 polls showed him winning that all-important presidential contest.

Ted Cruz greets supporters in Des Moines after winning the Iowa caucus on Feb. 1, 2016.

They were all wrong — as Texas GOP Sen. Ted Cruz trumped the pollsters, and his rival, to come out on top.

On the Democratic side, the majority of recent polls gave Clinton an edge over her main rival, Vermont Sen. Bernie Sanders. But Clinton eked out a very narrow win.

USA TODAY's 2016 Presidential Poll Tracker

Anyone who predicted that outcome deserves a “special pundit medal of honor,” Amy Walter wrote in a post-election analysis for the nonpartisan Cook Political Report.

Prep for the polls: See who is running for president and compare where they stand on key issues in our Voter Guide

Political experts pointed to three reasons the polls were off base:

• This is an extremely volatile political climate, driven by an angry electorate whose voting preferences are difficult to gauge;

• Pollsters low-balled turnout among evangelical voters and underestimated Cruz’s get-out-the-vote operation;

• The Iowa caucuses are uniquely tough to predict, with a quirky process and lots of last-minute deciders.

“It is really difficult to predict (the outcome) in a caucus state when there are so many candidates,” said Darren W. Davis, a political science professor at Notre Dame University who specializes in public opinion and political behavior.

The widespread voter angst, Davis said, also added to the unpredictability, making it harder to estimate how many new voters would turn out and how reliable voters would cast their ballots.

Cruz, GOP foes grab for Reagan legacy in New Hampshire

Amalie Nash, executive editor and vice president of audience engagement at The Des Moines Register (owned by Gannett Co. Inc., which also publishes USA TODAY), said this presidential race has been full of surprises — starting with the appeal of Trump, the New York real estate mogul who gained in the polls even as he made incendiary comments and outlandish proposals.

“A lot of things have been surprising about Donald Trump,” Nash said. “The fact that he was shown to be ahead and then didn’t get as much caucus support is another surprise along the way.”

The Register’s last poll, released late Saturday, showed Trump leading the GOP field in Iowa with 28% and Ted Cruz coming in second, with 23%. On Monday, Cruz won with 27.65%, while Trump ended up with 24.31%.

Donald Trump reclaims lead in latest Iowa Poll

Nash noted that even as the poll was underway, Trump made fresh waves in the race by skipping the final GOP debate — a move that may have skewed the results. Plus, she said, “Ted Cruz put on a really strong ground game.”

Peter Brown, assistant director of the Quinnipiac University Poll, defended his organization’s work, saying the last poll was only “very slightly off” on the Republican side and within the margin of error on the Democratic side.

But he agreed this election presented unique challenges. “You have an unusually large field on the Republican side,” he said. “The more candidates you have, the more you are cutting up the pie and the tougher it is to get a clear picture.”

As for Trump, he said, “he may have been a major factor bringing new people to the polls, but when they got there they didn’t behave the way they were expected to.”

Donald Trump speaks at his Iowa Caucus night gathering on Feb. 1, 2016, in Des Moines.

Walter echoed that assessment. She said experts predicted high turnout would favor Trump, but "his polarizing nature also helped turn out people who don’t want to see him win."

On the Democratic side, the polls were closer to reality — showing Clinton ahead in a tight race. The Register poll, for example, found Clinton winning 45 percent of likely Democratic caucusgoers, and Sanders snagging 42 percent.

With all but one Democratic precinct reporting in Iowa early Tuesday, Clinton's lead over the Vermont senator was less than 1 percentage point. Sanders appeared to get a boost from Iowa voters who initially supported former Maryland Gov. Martin O’Malley.

In precincts where too few O’Malley supporters showed up to make him a viable candidate, caucus rules allowed them to choose a different candidate in a second round of voting.

Ann Selzer, The Des Moines Register’s pollster, said Sanders appeared to pick those voters up, along with other late-deciding voters. “The unviable O’Malley supporters and the undecideds appear to have chosen the non-Clinton candidate,” Selzer said in a note to Register editors Tuesday morning.

Brad Todd, a GOP media consultant, said pollsters face new hurdles in producing accurate results, as more Americans disconnect their landlines and use only cell phones, among other changes.

“Polling as a whole is becoming more challenging as it gets more difficult to reach voters in a random manner and keep them engaged for a full survey,” Todd said. “Most media and university surveys are cheap and use methodological shortcuts that make them even more error-prone.”

He said some outlets still spend enough money to “get it right,” but a lot of surveys are churned out quickly “just to get cable TV eyeballs.”

But perhaps the biggest mistake, some experts said, was not the accuracy of the polls, but how seriously the candidates and the media were taking them.

“I think polls still are a good predictor, but obviously not something etched in stone,” said Nash.

Davis agreed, saying the Iowa results are not a reflection of poor polling methods. “It’s just the difficulty of trying to predict politics, and the way in which the candidates and the media interpret polls,” he said.

Because polls are so hyped in the press, and candidates get so juiced about incremental changes, “it really sets up pollsters to fail.”

As the presidential contest moves to New Hampshire, it’s unclear if the Granite State polls will be any more precise than the Iowa polls. What is clear, said Davis, is the public and the press need to be “a little less excitable and a little more discerning” in how we digest such surveys.

Indeed, Walter said New Hampshire polls may be even more unreliable.

"This is a state that is notorious for breaking late and people deciding late in the game," she said.

Featured Weekly Ad