Fact Check: Is This US Senate Candidate Polling Higher Than Sen. Thom Tillis?

For a sign that North Carolina's U.S. Senate race is tight, look at the polls.

Incumbent Republican Sen. Thom Tillis at one point last year had the lowest approval rating of all senators, according to one poll.

On the Democratic side, two of the five candidates appear to be getting the most attention.

Cal Cunningham, who was endorsed by the Democratic Senatorial Campaign Committee, raised more than $1 million in the third quarter of 2019 alone. But in the polls, Cunningham hasn't separated from state Sen. Erica Smith, D-Northampton, who raised about $40,000 that quarter.

In fact, Smith has polled quite well. In an interview with Spectrum News host Tim Boyum, Smith described her run as historic. And she made a couple claims that caught our attention.

Around the one-minute mark of the interview video, she said, “There have been five polls that have been conducted since this race began, and in every poll, not only am I ahead of the Democratic challengers, but I'm beating Thom Tillis in all of those polls that show a head-to-head match. I am beating incumbent Senator Thom Tillis, and I am beating him by the widest margins.”

Around the six-minute mark, she added, “There has been no one, no black woman of color, who has been running for U.S. Senate in this nation and beating the incumbent in every poll out the gate. We're making history.”

(Actually, there have been six public polls. But we're not going to focus on that error, since it's likely unimportant to readers.)

Instead, we'll focus on her two much bigger claims: Is Smith beating both Cunningham and Tillis in the polls? And, is she the first black woman to lead an incumbent U.S. senator upon entering a race?

This is a classic case of cherry-picking. Smith cites polls that are either old or inconclusive to make a sweeping claim that gives a misleading impression. No poll that featured all three candidates showed anyone with a lead that cleared the margin of error.

Also, Smith has appeared in only half of the public polls on the race.

Smith poll results

PolitiFact found six public polls on the race, and only three of them mention Smith.

Smith's best performance: Getting 46 percent support to Tillis' 39 percent in an Emerson College poll. This is the only poll where her poll victory was outside the margin of error, which was plus or minus 3.1 percentage points.

The drawback? The poll was conducted between May 31 and June 3, before Cunningham entered the race. Jason Husser, director of Elon University's poll, said voters shouldn't put too much stock into polls conducted so early in a campaign.

“At this early moment, the vast majority of voters tend to rely on party labels when answering survey questions about non-presidential races,” Husser said. “That will change when campaigns do more voter outreach over the next couple of months. However, almost all poll numbers for non-presidential primaries are highly volatile at this time.”

Head-to-head against Cunningham: There has been only one public poll. A Fox News poll conducted between Nov. 10 and 13 found that 18 percent of respondents supported Smith, while 13 percent supported Cunningham.

The hangup? Nearly 50 percent of people polled said they were undecided. In other words, more than twice as many people were undecided as chose Smith. (Note: In a similar fact check from 2017, PolitiFact hesitated to label a U.S. Senate race “neck-and-neck” because a plurality of voters was undecided.)

Only one poll featured Smith, Cunningham and Tillis: It was conducted by Meredith College between Sept. 29 and Oct. 7 and showed Smith polling less than a percentage point ahead of Tillis, and Cunningham polling less than a percentage point behind Tillis.

There are a couple of reasons this poll shouldn't be seen as a reflection of who's winning or leading the race. In terms of counting actual people, Smith had only four more supporters than Tillis, and Tillis had only seven more supporters than Cunningham out of 996 people polled.

Smith's results were also within the Meredith survey's 3 percent margin of error, so poll director David McLennan referred to the race as “neck and neck.” He added, “I would not call anything in the margin of error as a lead for Smith.”

Other polls

Smith can accurately say she has, at one point, polled ahead of each of her opponents. But she's not the only one who can claim to be polling ahead of Tillis. Cunningham also scored better than Tillis in not one, but three head-to-head matchups. However, none of those leads was outside each respective poll's margin of error.

A Public Policy Polling poll conducted on June 17 and 18 found Cunningham to be ahead of Tillis by 1 percentage point. The margin of error was plus or minus 4 percentage points.

An AARP poll conducted between July 29 and July 31 also showed Cunningham ahead of Tillis by 1 percentage point. The margin of error was plus or minus 4 percentage points.

Another PPP poll conducted on Sept. 16 and 17 had Cunningham leading Tillis 45 percent to 43 percent. The margin of error was plus or minus 3.9 percentage points.

Old polling limited

Now let's look at the second part of Smith's claim. To determine its accuracy, we need to know how many black women have run for U.S. Senate and what their polling numbers were.

This part of Smith's claim is more difficult to check because the crux of her claim is about polling, not whether a black woman went on to win. But it's difficult to find a comprehensive list or even a total number of African-American women who have run for Senate in any state in the last 150 years. Polling records are even more spotty, said Kathleen Weldon, director of operations at Cornell University's Roper Center for Public Opinion Research.

“There's no database of people who have campaigned for an office, (and) there are many short-lived campaigns,” Weldon said. The Roper Center has comprehensive records only for national polls, she said.

Comparing state-level polling is even harder to do.

The Rutgers University Center for American Women and Politics tracks minorities who have been elected to high-profile offices. But the group has data only back to 2004, said Chelsea Hill, CAWP's data services manager. At least 31 black women have run for the Senate since 2004, she said.

One possibility

Experts struggled to recall off-hand a black woman who definitely polled well against an incumbent early in a race. But they mentioned one possibility: Carol Moseley Braun.

Braun is one of only two black women to ever serve in the U.S. Senate. The other is current California Sen. Kamala Harris, who won an open seat in 2016.

Braun went on to represent Illinois from 1993 to 1999 after defeating incumbent Sen. Alan Dixon in the Democratic primary.

PolitiFact could find no record of Braun leading in the polls.

The University of North Carolina's H.W. Odum Institute for Research in Social Science, which has a collection of old polls, found records that Dixon polled ahead of Braun throughout the primary.

Author D. Harris also claimed in the book “Black Feminist Politics from Kennedy to Clinton” that Moseley Braun trailed in the polls up until the primary.

“The week before the primary, a Chicago Sun Times/Fox News Chicago poll showed Dixon with the support of 41% of Democrats who planned to vote, compared to 29% for Moseley Braun and 21 percent for Hofeld,” Harris wrote.

At the time of Braun's win, the Chicago Tribune called it a “stunning victory.”

Our ruling

Smith said she's polling “ahead” of Democratic candidates for North Carolina's U.S. Senate seat and is the first black, female Senate candidate to be “beating the incumbent in every poll out of the gate.”

Smith appears to be right about the fact that no other black woman has polled ahead of a sitting U.S. senator.

But it's a stretch for Smith to claim she's beating Cunningham or Tillis. She did poll higher than Tillis “out of the gate,” but that was before Cunningham entered the race. And there's no poll featuring all three candidates that showed Smith with a lead that cleared the margin of error.

Smith's claim is partially accurate but leaves out important details or takes things out of context. We rate it Half True.