Tomorrow (Tuesday) is run-off election day in Georgia, with the control of the Senate hanging in the balance. The polling – such as it is – suggest a razor-thin result in both races. However, the Democrats have been leading more often than the GOP incumbents. But when you look at the error margins, neither candidate has any kind of polling advantage.
A number of polls were released today, but like many of the ones that came earlier they are mostly from unimpressive or unknown pollsters. However, one of them was from a well-regarded Republican pollster National Research Inc. NRI is rated A/B by FiveThirtyEight. Most GOP partisan pollsters are, frankly, lousy. Not NRI; it’s a serious pollster. Here’s the topline results from the new poll:
Jon Ossoff (D) 46%, Sen. David Perdue (R) 45%
Raphael Warnock (D) 46%, Sen. Kelly Loeffler (R) 45%
The survey has a margin of error of 4.38 points. This means that Ossoff’s support ranges from 41.62 to 50.38. Warnock’s support is in the same range. What is important to note here is that – unlike many other Georgia polls – there is a significant number of undecideds: nine percent in both races. As we’ve already seen this year, the undecideds have a tendency to break in the same direction. This means that if NRI’s toplines are correct, the results might not be close at all. This is the one part of this survey I think is likely wrong.
There are other interesting findings from the survey. According to Axios:
The poll cast in stark relief how Democrats have banked early support: Some 54% of respondents said they had already returned their ballots.
Among early voters, Ossoff was up in his race over Perdue 54%-36%; and Warnock led Loeffler by a 54%-37% margin.
Among those who haven't voted, Perdue led 56%-38%; Loeffler had a 55%-36% advantage.
Turnout will be key in determining the winners and the resulting balance of power in the U.S. Senate
NRI’s findings are even closer than the FiveThirtyEight polling average in each race:
Ossoff 49.3%, Perdue 47.9%
Warnock 49.6%, Loeffler 47.6%
As you may recall from my last post, there has not been a lot of polling in this race and many of the polls that we have seen are from inferior (or unknown) polling outfits. NRI might be closer than the FiveThirtyEight polling averages, but just being a good pollster does not mean you will get it right all the time. The races are close, and consequently turnout will matter more than in most elections.
According to Cook Political Report: “Private polling from both parties, in addition to scant public polling, puts both races as a coin toss within the margin of error. It’s impossible to know how much the president’s latest temper tantrum will impact the Election Day turnout Republicans badly need. Still, GOP sources in the state certainly don’t think it will help. Early voting topped 3 million votes — already making this the highest turnout Georgia runoff ever. Democrats are encouraged by that strong turnout, which shows the numbers outpacing even Election Day totals in key demographics, particularly among Black voters and younger voters.”
The early vote, while setting a record for a Georgia run-off, was 23% lower than it was for the general election. If that trend holds up among election day voters, it could spell trouble for the incumbents. In fact, CPR reports that “one GOP source predicts that if the day of vote sees that same drop, it’s not good for them.”
NBC News reports that there were also 112,838 new voters who did not (or could not) vote in the general election on November 3. A significant number of these new voters are young (30% are under 30) and Black (40% are Black). In a state of over ten million people, this might not seem like a lot of votes. However, Biden won the state by less than 12,000 votes. Every vote will matter.
The polls are open from 7:00 AM to 7:00 PM local time tomorrow. Absentee ballots may be dropped off a clerk’s offices until 5:00 PM or in an official drop-box until 7:00 PM. We might not have enough returns counted to make a call on either race before Wednesday or later. The results will come in just like they did on November 3rd. Republican areas of Georgia – which are usually smaller counties with far fewer votes each than Democratic counties – usually report results early. Votes cast on election day are counted before early votes and absentee ballots. Heavily Democratic counties (e.g., Fulton, DeKalb, Chatham) usually take a long time to count all the votes.
This suggests that the count will unfold similarly to the way it did for the general election. The GOP candidates will likely have a strong lead early, but that will narrow as reports come in from early votes, absentee ballots, and Democratic counties. As the AP points out, "In November, Perdue held a lead of about 380,000 votes over Ossoff at 10 p.m. EST on election night. But Perdue’s lead eventually fell below 90,000. In a very tight race, it could take several days to determine a winner. In November, more than 5 percent of Georgia’s votes were counted after noon on the day after Election Day. At that time, Donald Trump led Biden by 100,000 votes in a race that Biden eventually won after all the mail ballots were counted."
What this means is that we probably won’t really know who is going to win until Wednesday or perhaps a little later. However, if a few things occur Tuesday night we can make some educated guesses.
If any of the early-reporting Republican counties report close results, it may be a good night for Democrats.
If Republican candidates get more than 60% of the election day returns, it may be a bad night for Democrats. (If the GOP candidates do better than 70%, it almost certainly will be a bad night for Democrats.)
If the election day results are close overall, the Republicans are going to have a bad night.
If the returns from Clayton and DeKalb Counties go less than 75% for the Democrats, it will be a good night for Republicans.
As you can tell, I am not differentiating the races. Although Nate Silver might disagree, I don’t think there is any chance that the races will split. If they do, it will be because the races are so close that 10,000 votes or so changes the outcome.
Robb, I would have to say that these polling numbers do not inspire confidence. But let's hope that the ground game that Stacy Abrams et al have put together makes the difference. Fingers crossed.