Pioneer Press poll raises surprising ethics questions
Tom Horner let loose with a jubilant message from his Twitter account on June 8.
"Must be bad Emmer poll coming," the Independence Party candidate for governor crowed. "GOP stepping up attacks. Must be spinning their candidate's collapse. Can we get a preview? How bad is it?"
The next day, the Pioneer Press made Horner look prescient, running a story about a new poll by Decision Resources Ltd. that showed Horner doing much better than anyone had predicted. For a third-party candidate lagging significantly behind the Democrats and Republicans, it was a bonanza of free publicity that gave the campaign serious momentum.
But over the next few days, a different story emerged, raising the question of whether the Pioneer Press got suckered by the candidate's message machine.
Bill Morris, the pollster behind Decision Resources Ltd., admits that he's personally supporting Horner's campaign, and that his company gets about 20 percent of its business from Horner's consulting firm, Himle Horner.
The poll has also emerged as a total outlier—nothing before or since shows Horner performing anywhere near as well.
Bill Salisbury, a veteran political reporter for the Pioneer Press, picked up a copy of the poll from Morris the night of June 6, but was only getting around to looking at it the next afternoon when he saw Horner's triumphant Twitter message.
"I was surprised when I saw the tweet," Salisbury says. "I thought my deal with Bill Morris was that I had an exclusive."
Salisbury has had a deal like that with Morris for decades.
"I've known him for about 30 years. I covered him back when he was the state Republican chairman back in the 1980s," Salisbury says. "I always ask him if I can get the polls first."
The deal has worked well for both sides: Morris gets a big stage, Salisbury gets an exclusive, and the Pioneer Press gets to run a poll story without having to pay for a poll. The Pioneer Press used to commission its own polls from Mason-Dixon, but those days are long gone.
"It's a tough situation for the Pioneer Press, because if you can't afford your own polls, where you're in control, you have to rely on other people's," says Steven Schier, a professor of political science at Carleton. "And that's when quality control tends to slip."
Morris didn't give Salisbury internal cross-tabs—the detailed breakdown of results that allow for close scrutiny of a poll. Nor did he provide the whole poll. Morris had told Salisbury that he was personally supporting Horner's campaign, and that Horner's company accounted for a substantial chunk of his business.
"But he gave me his assurance that that hadn't affected his results," Salisbury says.
Salisbury continued to take Morris at his word even when he noticed that the poll showed Horner doing far better than in any other poll—Decision Resources had Horner at 17 to 19 percent when all other polls had him between 9 and 12 percent.
"I was surprised to see that," Salisbury says. "But I trust Morris. He's got a track record."
Salisbury wrote the story and handed it in to Maria Reeve, the paper's politics team leader. The story went on Twincities.com at 9:48 p.m. and ran on page 10-A in the next morning's edition. No mention was made of the pollster's conflict of interest.
Thom Fladung, the editor of the Pioneer Press, says it was a mistake not to let readers know about Morris's ties to Horner, but says the story is hardly a big deal.
"It ran on page 10-A," he says. "You can see from where we put it that it's not exactly top news."
That's not good enough for Kelly McBride, the lead faculty on ethics at the Poynter Institute. "If you have information about a conflict of interest in a poll like that, you have to pass it on," she says. "Running a poll with that kind of problem with no disclosure? I can't believe an editor would agree to it."
Horner's opponents are incensed that the Pioneer Press lent legitimacy to what they see as a Horner ploy.
"This whole episode was designed to give Horner a bump early on in the campaign," says Thomas Harens, the running mate of Rob Hahn, who's running against Horner for the Independence Party nomination. "I mean, come on: The guy had a poll placed in one of the major dailies."
Republicans also jumped on Horner's inside access to the poll, as evidenced by a later tweet that made reference to internal cross-tabs that hadn't been made public. The GOP filed a complaint with the Office of Administrative Hearings alleging that the inside information amounted to an illegal corporate in-kind donation.
"The question is what it would have cost Horner if he'd gone out and gotten this information himself," says Republican Party lawyer Matt Haapoja. "And the answer to that question is 'a lot.' A poll like that generally costs tens of thousands of dollars."
The judge threw out the GOP's complaint last week, ruling that since the poll had already been released to the press when Horner got it, it didn't have any monetary value.
Documents filed in relation to the complaint have revealed even more entanglements, though: Morris's wife is a volunteer with Horner's campaign, and Horner and his campaign manager met with Morris to discuss the poll the day before Salisbury's story ran.
Since the Pioneer Press ran its story, the poll has looked more and more like an outlier. A June 18 KSTP-Survey USA poll showed Horner's numbers back at 12 percent, making Morris's poll a total anomaly.
"The results of that poll are widely at variance with all the other surveys, and there's no good reason for it," says political science professor Schier. "The worst of it is, it's at variance in a way that benefits a client of his. That's bad."
Morris insists his poll was good, though.
"That was an accurate portrait of a moment in time," he says. "If Horner's support has gone down since then, it's because of the attacks the Republicans have been making in the press."
As news about the conflict of interest began to circulate, the Pioneer Press retroactively added a paragraph to Salisbury's story online acknowledging Morris's connections to Horner.
But the new version is behind a pay wall, and the paper has never printed a correction. Readers of the print edition have never been told there was anything fishy about the poll they read on 10-A of their paper.
Get the This Week's Top Stories Newsletter
Every week we collect the latest news, music and arts stories — along with film and food reviews and the best things to do this week — so that you’ll never miss City Pages' biggest stories.