Share this article

print logo

The 'push poll' that wasn't in the Erie County executive's race

Robert J. McCarthy

Back in June, more than 2,500 Erie County voters received one of those automated phone calls asking them about their choice in the November election for county executive.

Democratic incumbent Mark Poloncarz? Or Republican challenger Lynne Dixon?

The Buffalo News then reported on June 16 the results of 1,325 of those calls, noting prominently that they should be viewed in an important context – the poll was sponsored by the Dixon campaign.

The story said the campaign touted close results – 42% to 40% with 18% undecided (Poloncarz won by seven points).

All of this would dwell in the files of an over-and-done-with campaign if not for Geoff Kelly’s July Investigative Post column headlined “Lynne Dixon’s misleading poll.”

The columnist essentially said The News had been duped, because he “tracked down” three people questioned in a “classic push poll” that was being discussed in social media “chatter.” He accused The News of not “investigating the underside of the poll.”

But the Politics Column explained on July 5 that The News had reviewed the questions before even considering publication. No leading questions alleged by Kelly were included.

And it subsequently determined that two polls were conducted – one meant for external publication (reported to the Board of Elections as required) and another that was designed for its own internal research. The Politics Column also noted that the Dixon campaign would be criminally liable had it disclosed results of any poll it failed to file with the Board of Elections.

The Politics Column challenged Investigative Post to back up its suspicions with proof of any illegality. It didn’t. But Kelly wrote he was “confident” his column was correct.

Political consultant Chris Grant of Big Dog Strategies, who commissioned the poll for Dixon, at the time would not release the script of the poll conducted separately from the one reported. With the campaign over and no reporting requirement, Grant unveiled it all a few days ago. Questions posed to 1,214 other voters included:

• Mark Poloncarz has increased county spending by more than $200 million since becoming county executive. Knowing this, are you more or less likely to vote for him?

The results were “less likely” 49%; more likely 27%; no difference, 24%.

• Mark Poloncarz has overseen a social services department overrun by scandal, including the deaths of numerous children and his handpicked commissioner resigning and being convicted of raping an employee. Knowing this, are you more or less likely to vote for him?

Results: Less likely, 54%; more likely, 18%; no difference, 27%.

• Mark Poloncarz has increased the property tax levy by $60 million while advocating for an increase in his personal salary that would ultimately raise his pay by nearly 20%. Knowing this, are you more or less likely to vote for him?

Results: Less likely, 58%; more likely, 19%; no difference, 22%.

Grant explained his separate survey was not intended to fool the public, nor for that matter, the press. It was designed as an advertising message for Dixon. It’s what campaign consultants do.

“The internal poll shows all the message tests we ran against Mark,” he said. “You test messages to see what works and what to put your money behind. It’s not a push poll.”

Indeed, Grant said after testing those questions in its internal poll, the campaign produced hard-hitting TV ads that it knew, from its research, would play with voters.

“I get aggravated at these guys who have never been involved in polling, blasting everyone for lying, and not knowing what they’re talking about,” Grant said, adding that his firm would never risk its reputation by releasing poll results that weren’t filed with Board of Elections.

“We would have been violating election law,” he said, “and we didn’t violate election law.”

There are no comments - be the first to comment