Cobb County is still the place to watch in the Georgia Sixth runoff
A little over a week ago, I wrote up a full analysis on the Georgia Sixth primary results:
Following up on that, I wanted to work with the early turnout to see if we’ve learned anything new. First, here’s an updated interactive map with results from the 2016 general and the primary; turnout from the 2016 general, primary, and early runoff; and current demographic breakdowns. The default style is the early turnout from the runoff.
As in the previous analysis, let’s break this down a bit. The table below looks at each county individually. It’s relatively self explanatory, but the final two columns are averages of the precincts rather than actuals as in the previous columns. The “Percent of Primary Turnout” columns look at the percentage of the actual primary turnout. For example, Fulton county’s early turnout in the runoff is nearly 92% of the total Fulton turnout in the primary.
As we have been, let’s continue looking at Cobb County. Cobb had about 9% early turnout (and for reference, the early turnout numbers are calculated as actual percentages rather than a ratio of early to election day). This was the lowest of the three counties while simultaneously ending with the largest final turnout. Election Day turnout was substantial in Cobb County, and it has been. Take a look at these two maps:
One thing is certain: Cobb County shows up on Election Day. But will they show up enough to make up the early vote gap?
DeKalb increased its early vote by over three times, Fulton by 2.75 times, and Cobb by about 2.4 times. In the primary, Cobb early turnout was only 66.25% of Fulton and DeKalb was 70.69% of Fulton. Moving to the current runoff, DeKalb has shrunk that margin to 86.41%, but Cobb County has dropped to only 57.84%. Of course, this is considering turnout percentages and not raw vote totals.
Fulton obviously pulls a significant weight in this district given that it has roughly the same number of voters as DeKalb and Cobb combined. However, Fulton was essentially the average in the primary with DeKalb and Cobb splitting off in their own directions. Given that, I think it’s important to compare the changes in Cobb and DeKalb.
Cobb shows up on Election Day, but there is even more ground to make up here than in the primary. And as a reminder, Cobb saw massive turnout for candidates other than Handel and Ossoff. For Handel to win, she needed to consolidate the Judson Hill and other voters behind her and show up with the same excitement as they did in the primary.
Is it possible for the ground to be made up on Election Day? Possibly. Let’s look at some ratios. First, in the primary, Cobb’s early vote comprised about 18.86% of the total turnout. This can’t possibly hold for the runoff given that to do so would see a turnout of 115%. In the 2016 general, the percentage of early turnout to total turnout in Cobb County was 44.19%. Let’s say that same ratio applies to the runoff. That would give Cobb 48.65% total turnout. If the same held true for the other counties, then DeKalb would end up at 53.56% and Fulton at 52.99%. Intuitively, this seems low given the pace of the early vote. It’s not unrealistic, but this would put it near the 2014 midterm total turnout. Along with most others, I think the turnout will end up higher.
If we take the average of the primary ratio and the 2016 ratio, then we’d end up with Cobb at 68.19%, DeKalb at 79.92%, and Fulton at 71.85%. With this method, I think we end up with the opposite problem in that this probably overestimates the turnout. Honestly, my gut says that if we took the average of the two again, then I’d think we’d have a more likely scenario (for reference that’d be Cobb-58.42%, DeKalb-66.74%, and Fulton-62.42%). But anyways, this doesn’t necessarily answer the question of whether Cobb can make up the difference.
I’ve said before that I think Cobb needs to have the highest turnout again for Handel to win. Cobb’s early to election day ratio has to be closer to the primary than the Election Day in order for the turnout to even surpass Fulton much less DeKalb. Again, it’s possible for Cobb to make up the ground, but not likely. The question then becomes whether there is other evidence to suggest that Karen Handel has made up the ground in other places. For that, I’ll point to this next table:
Forgive me for the length, but I decided to be thorough here. This breaks down the same early turnout from a variety of categories and then into the three county splits within those categories. For Handel to have made up ground outside of Cobb County, she would want to see growth in early vote in the areas where Ossoff did worse, where she did best, and where the other candidates did best. The caveat here is that there is obviously no guarantee that these precincts vote in similar percentages as they did in the primary. I would say that when you plot %Ossoff against %Clinton, there is a significant positive correlation. Things change, but probably not that fast.
Anyways, the first couple categories are interesting to look at, but not as important as the others. Neither candidate needs 50% in every precinct, and a successful precinct performance doesn’t always mean getting 50%. That’s why I included the third category: Ossoff Benchmarks. These are the benchmarks calculated by FiveThirtyEight eons ago before the primary. They’re the same benchmarks included in the spreadsheet posted for you to follow along.
I think this category is more informative than the previous. Notice that the average primary early turnout was higher in precincts where Ossoff did not meet his benchmark. But the early runoff turnout has flipped and the precincts where Ossoff beat his benchmark now have slightly higher early turnout. Not substantial, but informative.
Next, there’s some good news in Handel’s best precincts. Following the same pattern as the primary, Handel’s best precincts have seen a larger turnout than her worst and her best precincts have already seen about 80% of the total primary turnout (compared to 70% in her below average precincts). You can see on the very first image in this post the north Fulton Handel stronghold with high early turnout. That’s seen in this table where Handel’s best precincts are a full 10 points higher in early turnout than her worst precincts. This is the best chance for her to make up any lost ground from Cobb.
Moving on to the “other” precincts, things aren’t as rosy for Handel. First, the precincts with less than average “Other” performance have increased about 2.75 times the primary, while the above average precincts increased at about 2.32 times the primary. In similar fashion, the below average precincts have already reached about 82% of the primary turnout while the below average precincts are only at 69% of the primary turnout.
Beyond this, I’ll point out another pattern and the best evidence that the Ossoff campaign has really stepped up the game in DeKalb County. The average precinct in DeKalb has increased the early vote by about three times, but there’s a split here. The precincts in DeKalb where Ossoff met his benchmark increased the early vote by about 3.46 times while the precincts where he failed to meet the benchmark increased by only 2.69 times. The same pattern is true for Handel’s above/below (2.68x/3.24x) and Other’s above/below (1.97x/3.23x). Any ground that Handel made up in Fulton may have been cancelled out in DeKalb’s early vote.
In the end, there’s nothing here that makes me believe that anything has really changed from my previous analysis. Cobb County is still the first question to be answered. If Cobb doesn’t at least match the turnout of Fulton, then I don’t think it matters how much Handel pushes in her north Fulton precincts.
But for those of you who want a bit more speculation, I think Cobb does turnout in substantial fashion given how they have in other elections. That’ll make things close between DeKalb’s increased turnout, the North Fulton powerhouse for Handel, and the Cobb turnout numbers. I’ve said to others that I think this stays within a 51–49 race, and I do think it goes to Ossoff (admitting my potential bias here) given the ground needed to make up in Cobb.
We’ll know tomorrow night (hopefully) how this works itself out. Follow along with PeachPod on this spreadsheet:
Follow us on Twitter throughout the night (Austin Wagner, Kyle Hayes, and Luke Boggs), and look for our recap episode when all this is over.
