ATLANTA — There’s been limited polling before the Jan. 5 Georgia Senate runoffs. But the dearth of data isn’t preventing the candidates from using whatever findings they have to fundraise and get-out-the-vote ahead of the critical contests.
The Senate’s balance of power for the next two years hinges on the runoffs between incumbent Republican Sen. David Perdue and Democratic filmmaker Jon Ossoff and appointed Republican Sen. Kelly Loeffler and Democratic Rev. Raphael Warnock. Republicans currently have a 50-48 seat majority. And Tuesday’s races offer Democrats the opportunity to claw back a 50-50 vote tie, which can be broken by Vice President-elect Kamala Harris.
Yet, despite Georgia’s political importance, the record amount of money being pumped into the state, including more than $100 million apiece for the Democrats between mid-October and mid-December, and the $457 million projected to be spent before Jan. 5, pollsters haven’t crowded the field with surveys.
Pollsters are being overly cautious regarding Georgia after trends and some forecasts made ahead of the Nov. 3 contests were bucked by the actual results. But while pundits have unleashed on the profession, the campaigns have seized on available figures, politicizing them to suit them best.
“The polls have us in a dead heat, and it’s clear that this race is going to come down to the number of folks who turn out to vote by Election Day — that means there are ads that need launching, doors that need knocking, and Georgians that need reaching out to in this final stretch,” Warnock wrote to supporters last week.
And the pastor at Martin Luther King Jr.’s Ebenezer Baptist Church is not wrong.
If the polls are to be believed, Ossoff leads Perdue by an average of 0.8 percentage points, according to 11 surveys analyzed by RealClearPolitics. Warnock is ahead by an average of 1.8 points, the same 11 polls found. FiveThirtyEight also has the preacher with a 1.8-point margin, while the prognosticating outlet has Ossoff with a more generous 1.2-point berth.
With the benefit of hindsight, Georgia’s Nov. 3 polls were more accurate than those of other battleground states, especially Wisconsin. In that state, multiple surveys had President-elect Joe Biden with a double-digit lead over President Trump, including an ABC News and Washington Post poll that put the former two-term vice president ahead by 17 points. Biden claimed Wisconsin’s 11 electoral votes by 0.7 of a point.
Yet at the presidential level in Georgia, Biden beat Trump by less than 1 point after the recounts and legal challenges. Trump had an average edge of 1 point before Nov. 3.
And in the Senate races, Ossoff was polling ahead of Perdue on average by less than a point. The sitting senator won by 2 points but fell short of the 50% threshold required to avoid a runoff. Surveys of Warnock’s 21-candidate special election against Loeffler were less precise. The reverend had an average polling advantage of almost 16 points, yet finished only 7 points in front of the businesswoman, the contest’s second place-getter.
In spite of pollsters’ record in Georgia, a handful of runoff surveys are being treated skeptically. A mid-December Survey USA poll, for instance, had Ossoff leading by 5 points and Warnock ahead by 7 points.
Survey USA’s findings could indicate Republicans “are buying” Trump’s argument that Georgia’s electoral system “is not to be trusted,” including a Jan. 1 tweet in which the president suggested the runoffs were “illegal and invalid,” University of Georgia political science professor Charles Bullock told the Washington Examiner.
“One possibility is that Republicans simply aren’t answering the surveys. That didn’t seem to be a problem in November, but it might well be now,” he said. “If you don’t trust the electoral system, are you going to trust somebody who calls you up and asks you how you’re going to vote?”
Bullock had another warning about the polling: The historical voter participation drop-off between general elections and general election runoffs may not materialize this cycle, skewing the figures. More than 3 million Georgians voted early, roughly 2 million in-person, and another 930,000 cast their ballot by mail.
“You look at the kinds of numbers of people who voted early, and they are tracking very closely with the number of people who had voted early at the same stage in November,” Bullock said.
Low-propensity voters tend to have their say on Election Day, so a drop-off could still eventuate. Yet, Bullock said “the general rule of thumb” was that strong early voting turnout typically precedes a high-participation race.
Lee Miringoff’s Marist College Institute for Public Opinion is one of the polling organizations that isn’t publishing surveys of Georgia before Tuesday. Instead, the group is conducting an experiment to see whether pollsters should have changed the wording of questions concerning who a respondent backed if they had already voted.
Miringoff explained that respondents who had already cast a ballot were asked directly about who they supported rather than a hypothetical query.
“That may have been a little standoffish,” he said. “And most of the pollsters were seeing a higher refusal rate that wasn’t enormous, but there was some drop-off on that question.”
Miringoff added, “We’re trying that, as well as a couple of other things, to try to use this as an opportunity coming off of November to see if things can be fine-tuned.”