Why Tennessean’s mayoral poll was a failure

Friday, September 2, 2016, Vol. 40, No. 36
By Sam Stockard

When a Nashville newspaper poll showed Megan Barry in a dead heat with David Fox shortly before the 2015 mayoral runoff vote, the Barry campaign refused to panic.

Instead of shifting strategies based on those figures, Barry depended instead on her own polling numbers, enabling her to forge ahead with a consistent message: Keep Nashville moving forward to help the entire city prosper.

That steady hand likely played a role in helping the two-term Metro Council member capture a 10-point victory over Nashville businessman Fox.

“Our polls did indicate we had a fairly good lead throughout the runoff period,” says campaign spokesman Sean Braisted, who now works in the mayor’s office for Barry.

“So when The Tennessean poll came out, certainly there were supporters who were concerned. And, obviously, you have to take every poll like that seriously, and you have to run the race like you’re five votes behind at all times, but it didn’t require us to change our strategy in any specific way.”

While polls conducted by media organizations tend to look at head-to-head numbers, Braisted says, “those aren’t terribly instructive for a campaign.”

Barry, he explains, was more interested in demographic information and message testing to ensure her words were “resonating with voters.”

Jim Williams of Public Policy Polling in Raleigh, North Carolina, who conducted the poll for The Tennessean, points out polls are “kind of a snapshot in time,” adding, “We could see at the time it was very much a partisan race.”

And though the poll called it a virtual tie, 10 percent of Nashvillians were undecided, a group friendlier to Barry, including more Democrats and African Americans, Williams says.

In addition, the polling did not include cell phones, depending on land lines only and requiring the results to be weighted for the age of respondents, Williams says.

“That’s sort of the art in the science,” he notes.

Yet most pollsters say including cell phones is vital to ensure accuracy, including the proper type of screening to determine who will turn out to vote.

In addition, the newspaper poll showed a partisan breakdown of 46 percent Democrat to 44 percent Republican, which didn’t reflect Nashville’s politics in an election clearly defining Barry as a Democrat and Fox as a Republican.

As the race neared its final days, Fox sent out a press statement showing confidence in his campaign and contending early voting turnout was strong in precincts he expected to win.

Barry’s people, who had access to voter scoring profiles based on Democratic National Committee numbers and consumer research information, dismissed Fox’s claims, because they had a “better idea” of who was voting, Braisted says.

Ultimately, the Barry campaign determined its voters were turning out.

And considering the newspaper poll came out shortly before the final vote, Barry’s people hardly had time to overreact, since TV commercials were shot already and mail pieces completed.

“There might have been a few tweaks in terms of what we did,” Braisted says. “But our field campaign strategy was to contact as many likely Barry voters as possible, and I think we did that through volunteers and different folks helping out.”

Sam Stockard can be reached at [email protected].