The Des Moines Register Iowa caucus poll has correctly predicted the caucus winner all but a handful of times since its launch in 1984.
So, if history is to be believed, former president Donald Trump is all but guaranteed to win the race this year. In a December NBC News/Des Moines Register poll, Trump notched support from 51 percent of likely Iowa caucus goers — the largest lead ever recorded in a competitive GOP primary five weeks out from the contest. Florida governor Ron DeSantis, Trump’s nearest rival, garnered just 19 percent support.
Support for Trump has risen eight points since October, while DeSantis has seen a three-point increase. Nikki Haley, for her part, came in at 16 percent in both polls.
“Ann Selzer, her Iowa poll, she has some sort of secret sauce in predicting who’s going to actually show up on caucus night,” Rachel Paine Caufield, co-chair of the Department of Political Science at Iowa’s Drake University and director of the Iowa Caucus Project, told National Review.
The winner of the poll taken one month before caucus night each cycle has gone on to win all but three times since 1984.
There are a few notable exceptions, including in 2012, when Rick Santorum pulled off a surprise victory. In the November 2011 poll, former speaker of the House Newt Gingrich was in first place, while Santorum was tied for sixth place. By the next month, Santorum had shot up to third place, behind Ron Paul and Mitt Romney.
Romney was initially declared the winner of the 2012 caucuses, but after the dust cleared, Santorum was later found to be the actual victor.
At the time, Selzer said the results of the survey vastly changed over the course of the four-day polling period.
“On the first day of that poll, we had Ron Paul with a huge lead, and we had Rick Santorum sort of finally breaking into double digits. And by the end of the poll, Ron Paul had fallen 13 points. Rick Santorum had doubled his support. So by the end of our polling period, we ended up with Romney holding steady and holding onto a lead. Ron Paul, Rick Santorum kind of jockeying for that second-place position,” she said.
In 2020, the Iowa poll had Bernie Sanders up by three points one month before the caucus. And while Sanders won the raw vote total — the first time that the raw vote total had ever been announced in Iowa — Pete Buttigieg won the traditional delegate equivalence metric, making him the winner of the Iowa caucus.
In 2004, Dick Gephart led by seven points one month out, though John Kerry won the caucus. However, one last poll released one day before the caucus found Kerry in the lead.
In 2016, the poll one month out correctly predicted Ted Cruz would win. However, a final Des Moines Register/Bloomberg Politics Iowa Poll taken days before caucus night showed Trump in the lead at 38 percent. Cruz sat in second at 23 percent, while Marco Rubio followed at 15 percent.
Cruz ultimately won the caucus, with 27.64 percent of the vote. Trump followed in second place at 24.3 percent, and Rubio came in a close third, at 23.12 percent.
But more often than not, the polling has proven overwhelmingly prescient.
For example: Hillary Clinton led the 2015 and 2016 Iowa polls and ultimately captured a narrow victory over Sanders on caucus night. Mike Huckabee came in first place in the December 2007 Iowa poll at 29 percent — after polling in third place one month earlier — and ultimately won the caucuses with 34.4 percent of the vote.
Barack Obama took the lead in the Iowa poll in November 2007. His lead carried on through December and he won the caucuses by almost eight percentage points.
Other victories have been the “least surprising ever,” Paine Caufield said, like that of Democrat Tom Harkin in 1992.
She expects this year will fall under that category.
“I don’t think there’s any question that Donald Trump will win the Iowa caucuses and I don’t think there ever has been a question that Donald Trump will win the Iowa caucuses. I think that’s been a given,” she said.
Instead, the question is whether Trump will meet the lofty expectations set out for him, and whether the other candidates will exceed expectations.
With Trump polling above 50 percent, the race is his to lose — but it would be “relatively easy” for him to underperform, Paine Caufield said, in which case the story out of Iowa and heading into New Hampshire would become about the second-place winner who surged at the last minute.
“I always tell people part of the mythology of the Iowa caucuses is this last-minute surge by the scrappy campaign that just puts its whole heart into it. And the candidate is out there doing the work and all of a sudden the momentum shifts at just the right moment,” she said, but added that was the story of Santorum more than any other candidate, and it’s a mythology that doesn’t prove true all that often.
But Democrats’ decision to choose their candidate this year solely by mail-in ballot could prove a wild card in the Republican race.
Typically, Democrats would be caucusing as Republicans meet to caucus. Paine Caufield suggested Democrats and independents can potentially register as Republicans to caucus with the GOP and bolster a non-Trump candidate, most likely Haley, since they won’t be busy with their own caucus.
“It would not surprise me if you see a last-minute push to get some independent voters and moderate Democrats to turn out for Republican caucuses,” she said. “And there’s really no way to predict whether that’s going to happen or how effective it’s going to be.”
And even in a more “normal” year, polling for the Iowa caucus is hard, the professor says, because it can be difficult to predict which voters will show up and who they will ultimately support because of the nature of the voting. Unlike in a primary, voters must all arrive at a set time and people in each one of 1,700 different precinct locations can get up and give a persuasive speech about each of the candidates before voting occurs that can make a big difference in voters’ minds.
“The nature of the Iowa caucuses — party-run, local gatherings — makes polling particularly difficult,” Selzer told the Des Moines Register. “Caucus polling is probably the hardest type of election-related polling there is. There is no polling partnership that spends more time or more money to take our best shot.”
DeSantis has traveled to all 99 counties in Iowa. But having attended numerous campaign events for various candidates in recent months, Paine Caufield says while attendees of DeSantis rallies are enthusiastic, she has found that the Florida governor lacks the last-minute momentum that Santorum or Obama enjoyed.
On the other hand, she’s gotten the sense that Haley’s campaign is closest to the “typical scrappy Iowa campaign.”
“I don’t know if she can capitalize on it but it’s there,” she said. “Iowans are known for making decisions in the last days and she may be a beneficiary of some of that momentum.”
The Haley campaign is playing catch-up to the organizational strength of the pro-DeSantis super PAC Never Back Down in Iowa. The PAC says it has knocked on more than 750,000 doors and gotten 30,000 people to commit to caucus for him.
DeSantis has focused significant time and energy on Iowa, holding at least 120 events there since May, according to NBC. Haley held 32 Iowa events in the same time frame, the outlet reported.
“The challenge [for Haley] is that Never Back Down has been on the ground organizing for Governor DeSantis for well over a year and Donald Trump’s apparatus has been operating here for eight years,” Iowa-based Republican strategist Jimmy Centers previously told National Review.
But Drew Klein, a senior adviser for AFP Action, said the group has been collecting data in Iowa since February, knocking on some 350,000 doors and making 700,000 calls to find voters who were open to supporting a non-Trump candidate.
Looking at Haley’s plateau in the NBC/Des Moines Register poll, the question for Centers becomes, “Did she peak too soon? And can AFP activate fast enough to catch up to Never Back Down?”