Pollster Who Correctly Called Heitkamp’s 2012 Race Pushes Back (Sort Of) On Latest N.D. Survey

There is history with a Republican U.S. Senate candidate in North Dakota being given a 10-point polling margin over Heidi Heitkamp.

Democratic candidate for North Dakota’s U.S. Senate seat, Heidi Heitkamp, holds up the Oct. 20, 2012, Fargo Forum, Wednesday morning Nov. 7, 2012, in Bismarck, N.D. The October newspaper had Berg leading Heitkamp in the state by ten points. Heitkamp holds a slim lead Wednesday as ballots continue to be counted. Berg has not conceded (AP Photo/Will Kincaid)

It involves my newspaper, The Forum of Fargo-Moorhead, and became the state’s own “Dewey Defeats Truman” moment.

The Forum released a poll in late October 2012, conducted by Essman Research of Des Moines, Iowa, that showed Fargo businessman Rick Berg leading Heitkamp by 10 points, 50 percent to 40 percent with 10 percent undecided. Republicans praised the poll, of course, and Democrats shredded it because they said it undersampled Democratic voters.

Turns out the margin was not 10 percent, nor was it in favor of Berg. Heitkamp won the race by less than a percentage point, 50.2 to 49.3, in one of the great upsets of that political cycle.

In fairness to my employer (I was not at The Forum in 2012, having moved to KFGO in 2009 before returning to the newspaper biz in 2015), Berg’s percentage was spot-on. And, if every single one of the undecideds broke for Heitkamp, that would get her to 50 percent. So … yeah, OK. It wasn’t a good look. Truth is, the two polls Essman did for The Forum in 2012 (a survey done in May showed Berg with a 7-point lead) were the only two outliers among all the public polling done that year. Every other poll fell within the margin of error and showed a tight race.

So here we are in 2018 and there has again been a poll released by a North Dakota media company that shows Heitkamp’s Republican opponent with a 10-point lead as Election Day nears. This time it is NBC North Dakota and its owner, Gray Television of Atlanta, that shows Kevin Cramer leading incumbent Heitkamp by a 51 percent to 41 percent margin, with 8 percent undecided.

The survey was conducted by Strategic Research Associates of Spokane, Wash. — a marketing firm not necessarily known for political polling.

Again, Democrats are apoplectic and are questioning the methodology.

Again, Republicans are smugly not contesting the poll numbers. And why would they, even if their internal polling shows a tight race? Any narrative showing their guy ahead by a wide margin is a good narrative.

But it’s telling that a respected Twin Cities pollster almost immediately took to Twitter after the NBC North Dakota poll was released and, while not openly disputing the survey, offered some numbers of his own that would bring the latest poll into question.

DFM Research, owned by Dean Mitchell, published its own survey numbers of select areas of North Dakota that show Heitkamp outperforming her results in 2012.

Why is this important?

Because in 2012, DFM and Mitchell was the only polling firm to predict the outcome of the 2012 race. He consistently polled North Dakota that year and predicted a narrow victory for Heitkamp based on regional polling he did shortly before Election Day. Mitchell’s statewide poll showed the race tied 47-47 percent and the rest undecided. Mitchell based his prediction on the bellwether Burleigh County, where Heitkamp had better numbers than Democrats historically need to win statewide races.

North Dakota is notoriously difficult to poll on a statewide basis and many experienced polling firms, including DFM, break the state down into specific geographic regions to get a better feel for what is happening. That was the basis of DFM’s tweets following the release of the NBC North Dakota poll.

As of this writing, Strategic Research Associates had not released the crosstabs (comparing sub-groups within survey respondents to the survey group as a whole) so the poll’s methodology is not clear.

Also interesting to note: DFM and Port, the conservative Forum Communications Co. blogger, have teamed up to conduct surveys in the past. Port touted those polls in 2014 (before he was employed by Forum Comm) while bashing polling done by The Forum. Wrote Port:

“I picked DFM Research because they have a solid track record in North Dakota, which Democrats can attest to. One of DFM’s former clients was the North Dakota Democrat party. DFM got the Heitkamp/Berg race right, when many (including this observer) thought Heitkamp didn’t have a chance. I think this survey, which I have seen and approved, will give us a fair look at North Dakota’s races.

“Honestly, given the mess the Forum has made of the polling thus far, I’m glad I made this decision. I’m fairly certain, though we’ll see, that our polling will show a much more accurate picture of races that are relatively competitive.”

Anyway, here are a couple of other tweets DFM posted to give a better idea of where it sees the Heitkamp-Cramer race:

In other words, DFM is saying “maybe” the statewide poll done by Strategic Research Associates accurately reflects the way North Dakota will vote on Nov. 6, but his research of the state shows the race is much more likely to be a tossup (which is much more near what all other public polling has shown).

If Heitkamp ends up winning this race, she can once again pull a Harry Truman. Although my newspaper didn’t provide her with a prop this time.

(UPDATE 8 A.M. 10/3/18: The North Dakota Dem-NPL Party sent out a press release with the numbers from DFM Research.)