Bernard Reynolds · October 28, 2026
Washington has already decided this is a wave election. Georgia, as usual, is reserving judgment. Both of those things showed up in my reading this morning, and if you work in this state, the difference between them is your whole planning problem for 2027.
Nationally, the wave talk has numbers behind it. Democrats lead the generic congressional ballot by about 9 points in this week's averages, and the forecasters make them heavy favorites to take the House. The Senate is tighter. Democrats need four seats on a hard map, and the models I've seen put their odds between 55 and 75 percent. Favored, not safe.
Now put the same week's Georgia numbers next to that. The AJC and the University of Georgia polled 1,000 likely voters October 7 to 15, margin of error 3.1 points. Ossoff leads Collins 48 to 40, which is outside the margin. Then the wave stops. The governor's race sits at 46 for Jackson, 45 for Bottoms. Lieutenant governor is 42 to 39, secretary of state 43 to 42, attorney general 44 to 44. Three open statewide seats, all inside the margin of error, in a state where Democrats haven't won one since 2006.
One number in that poll explains a lot of the rest: 68 percent of Democrats call themselves very enthusiastic, against 48 percent of Republicans. Enthusiasm is what keeps a candidate who has been badly outspent in a tied race.
Polls are a forecast. Ballots are weather. As of 8:34 a.m. ET this morning, the Secretary of State reports about 2.13 million accepted early ballots. That's more than one active voter in four, and about 179,000 came in since yesterday morning. Mail is the small lane: about 98,000 accepted of about 165,000 requested. If yours is on the counter, it isn't a vote yet.
GeorgiaVotes, which compares both years the same way on its own clock and source, has 2026 running about 9 percent ahead of this point in 2022.
Who are they? Nobody hands you a party card at a Georgia check-in table, so any split is somebody's math. L2's model, stamped 9:25 a.m. today across about 2.1 million early voters, guesses 53 percent Democratic, 38 percent Republican, 9 percent neither. A guess built from voter-file records, not a sign-in sheet, and it estimates who showed up, never how they voted.
Ballots accepted as of 8:34 this morning, the one-day change, and the modeled split of each county's modeled voters:
Here's where I'd keep both hands on the wheel. Early totals measure participation, not preference. And the polls above are built on a model of who votes; the AJC's own analysis finds nearly one early voter in five skipped the 2022 midterm. If the real electorate doesn't match anyone's model, the polls move, in either direction.
The General Assembly convenes January 11, 2027, and these races decide the building. If the lieutenant governor's race flips, a Democrat presides over a Senate Republicans are still expected to hold. Whoever wins governor, there will be a freshman class that doesn't know your issue yet.
So two questions worth sitting with. Who is carrying your issue in that building in January? And if the answer is nobody, how much of this uncertainty are you comfortable carrying alone? The best time to find an ally is before the count is final.
If you've got something coming in the 2027 session, grab a time with me: https://calendly.com/bernardreynolds/true-north-strategy-conversation
Bernard Reynolds is the principal of True North Strategies, a Georgia public affairs firm.
