The Topline
Two statewide races moving in different directions on the same electorate.
The governor's race is a pure toss-up: Rick Jackson (R) 43.2%, Keisha Lance Bottoms (D) 42.7%, 14.1% undecided. Meanwhile, Jon Ossoff holds a modest but structural lead for Senate, 46.7% to Mike Collins's 42.9%, with only 10.4% undecided. The four-point difference between the two Republicans' margins is not about their bases (both take 83% of Republicans); it is entirely about the persuadable middle, and the trust-on-issues battery explains why.
The environment is rough for the party in the White House: a 57.8% majority says the economy is getting worse, Trump is underwater at 45/53, and cost of living dominates the electorate's stated concerns (Section 4 maps the full issue terrain). Yet the governor's race is tied anyway, and the crosstabs show a Democratic ticket that has won the argument with the middle in one race and not the other.
The Governor’s Race
Jackson 43.2 / Bottoms 42.7 / 14.1 undecided
Leaners barely move it: 65% of undecideds are "truly undecided," and the remainder split evenly (Jackson 12%, Bottoms 12%, another candidate 11%).
The demographic architecture is conventional: a 12-point gender gap (Jackson +12 with men, Bottoms +10 with women), a generational split (Bottoms +16 among 18–34, Jackson +16 among 65+), and a stark density divide, with Bottoms winning urban Georgia 56–30 while Jackson wins small towns and rural areas by roughly 25. Suburbs, at 48% of the sample, lean Bottoms 46.3–40.4: the race is being decided there.
The undecideds live in Bottoms's demographics, but she hasn't closed them.
of independents and moderates are still undecided for governor
Undecideds concentrate among independents (22.1%), moderates (24.3%), 18–34s (20.8%), and high-school-or-less voters (17.3%), all groups that otherwise lean Democratic in this poll. Bottoms leads moderates 51–24, yet a quarter of them are still on the fence. The catch: these same groups report the lowest vote certainty (Section 5). Her ceiling is higher than Jackson's; her floor is shakier.
Jackson is winning a race in which most voters don't know who he is.
have never heard of the Republican nominee for governor
Jackson's favorability is 38.0/37.8 with 24.1% never having heard of him, including 35.8% of moderates and 34.4% of under-35s. He is at 43% on the ballot almost purely on the strength of the R label. Bottoms is far more defined (44.3 fav / 40.3 unfav, 15.5% unknown) but carries the highest "very unfavorable" of any gubernatorial figure tested (29.5%). The first campaign to define Jackson for the persuadable middle (where his image is currently 29/47) likely tips the race; Section 5 compares the definition math for every figure tested, and Section 6 takes a view on which way it cuts.
Jackson's issue trust is concentrated in his base — and the elite-serving concern reaches into his coalition.
Republicans naming Jackson, not Bottoms, as the candidate who serves the elite
This is the most interesting pattern in the governor battery. Statewide, Jackson narrowly leads all three positive traits: strong leader (38.1–37.0), managing the state's economy (41.1–38.4), and keeping communities safe (39.8–37.8). But he also leads, by 10.5 points, on "serves the interests of the wealthy and powerful elite" (38.5–28.0). And that read is not just Democratic projection:
| Group | Jackson (R) | Bottoms (D) | No diff / Not sure |
|---|
Republicans themselves narrowly name their own nominee as the elite-serving candidate, and independents do so by 25 points. Meanwhile, Jackson's positive-trait leads are built on base intensity and rural margins, not the middle: independents give Bottoms the edge on all three positive traits (strong leader 34.1–29.6, the economy 38.0–34.2, community safety 37.8–32.2). In other words, the persuadable middle currently sees Jackson as the elite-serving candidate and doesn't yet credit him on competence; his 43% ballot share is running ahead of his issue trust with the very voters still in play. That leaves a plausible populist opening against him, but it is not yet a settled judgment. His counter-move is the generic Republican competence presumption, which is working statewide but hasn't yet reached the middle. Whichever frame hardens first — "competent manager" or "for the wealthy" — defines him.
The U.S. Senate Race
Ossoff 46.7 / Collins 42.9 / 10.4 undecided
Unlike the governor's race, this one has a clear internal logic: Ossoff leads all four candidate qualities, and his margins with the middle are decisive rather than fuzzy.
Among independents the gaps blow out: Ossoff +20 on strong leader (45.5–25.3), +25 on lowering costs (49.9–24.5), +24 on Georgia-first, +26 on "people like you." Among moderates he leads every trait by 35–40 points.
The same electorate trusts the R on "the economy" and the D on "costs."
Jackson's statewide lead on "managing the economy" vs. Ossoff's on "lowering costs"
Statewide, voters say Jackson would better manage the state's economy (41.1–38.4) while saying Ossoff would better lower costs for Georgia families (45.8–34.7). Same week, same inflation anxiety, opposite verdicts, depending on whether the question is framed as abstract stewardship or kitchen-table relief. And the middle magnifies the gap: among independents, Bottoms's economy edge is narrow (+4) while Ossoff's costs edge is massive (+25). Part of this is definition (Ossoff is 92% known; both Jackson and Bottoms are less so), but part is framing: the Senate battery's language ("lowering costs for Georgia families," "people like you") plays directly to the electorate's #1 stated concern, while "managing the state's economy" defaults toward the generic Republican competence prior. For Bottoms, a similar costs-and-family emphasis is visible in her own party's crosstabs. Compare their standing with independents: Ossoff wins their ballot 52–31; Bottoms only 43–35.
On costs, Ossoff holds the trust advantage.
moderates' verdict on who would lower costs for Georgia families
In a cost-of-living electorate, the Republican Senate candidate trails on costs by 11, including a 56–17 deficit among moderates and a 47–30 deficit among women. Collins is nearly as undefined as Jackson (23% never heard of him; 31% of moderates), and the same consolidation upside applies to him as his unknowns come home to the label (Section 5). The difference is what waits on the other side: a 92%-defined opponent who already owns the trust battery, not a capped one. He matches Jackson among Republicans (83%) but runs 4 points behind him with independents (where his image is 32/43), and in urban Georgia his margin runs about 4 points worse because Ossoff outperforms Bottoms there (61–31 vs 56–30). The roughly 4 points of the electorate splitting Jackson/Ossoff tickets is concentrated precisely in the groups where Ossoff has won trust and Collins hasn't: independents, moderates, suburban women.
The Political Environment and the Issues
The mood of the electorate, and the issue terrain both races will be fought on.
The economy: pessimism is broad, not just partisan
A 57.8% majority says the economy is getting worse, and the gloom is not a partisan artifact: "getting worse" is the majority view among independents (70.5%), moderates (72.4%), women (65.2%), and voters under 35 (68.4%). Even among Republicans, only 39.3% say the economy is improving, and nearly a third of the GOP base (29.7%) concedes it is getting worse. The most optimistic group is seniors (32.9% "better"), the most Republican age cohort, while the youngest voters are the most pessimistic.
Trump: underwater, with a soft middle
Trump's job rating is 45.1% approve / 52.8% disapprove, and the intensity is lopsided: strong disapproval (43.4%) is nearly double strong approval (23.3%), and the approval number rests heavily on somewhat-approvers (21.8%). The middle has largely rendered its verdict: independents disapprove 64–34, with 55% disapproving strongly, and moderates disapprove 70–27.
Cost of living beats national security
Asked which Iran-conflict concern weighs on them personally, voters choose the impact on prices over nuclear risk, 58.9–33.6, and the pocketbook framing wins every group except Republicans and very conservatives. Even somewhat conservative voters break for the cost concern, 49.1–44.8, and the persuadable middle is unambiguous: independents 67–26, moderates 67–22. Whatever the news cycle is doing, the electorate's attention is on what things cost.
Immigration: a wedge that cuts both ways
Immigration divides the Republican base from the persuadable middle. Q22 (Muslim immigration) is the sharpest culture-issue divider in the poll: 44.8% "too many" / 43.6% "about right" / 11.6% "too few." The intensity sits with the GOP base (Republicans 71.4% "too many," very conservative 77.1%), but the middle is elsewhere: moderates say "about right or too few" by 70–30, independents by 63–37, and under-35s are the outlier group where "too few" hits 25.9%. As a base-mobilization theme it works for Republicans; as a persuasion theme with the exact independents and moderates both R candidates need, the numbers run the other way. Suburbs (42.5% too many / 43.7% about right) are, again, the hinge.
The upshot: on paper this is a "wrong-track referendum" electorate: sour on the economy, down on the White House, focused on prices. The races in Sections 2 and 3 show how unevenly that environment is currently priced in; the candidate who owns the costs issue leads, and the one who doesn't is tied.
What to Watch
Two structural metrics will move these numbers between now and November: how the undefined candidates get defined, and who actually shows up.
Definition: three of the five figures are still blank slates
24.1% of likely voters have never heard of Jackson, 23.0% have never heard of Collins, and 15.5% have no read on Bottoms. Ossoff is effectively fully defined (8.4% unknown), and Brian Kemp (5.1% unknown, 60.7/34.2, +26 net favorable) shows what a completed Georgia Republican profile looks like at full strength; neither R nominee inherits that standing automatically.
For both Republican nominees, the blank space is mostly upside, and the logic is identical: their unknowns skew toward voters who default to the party label once it attaches (Jackson is unknown to 35.8% of moderates and 34.4% of under-35s; Collins to 31.3% of moderates), so partisan consolidation alone pulls both numbers up as the races engage. What differs is not the metric but the opponent it runs into. Jackson's definition upside lands against a rival who is 85% defined with the worst intensity score in the survey; consolidating his unknowns can plausibly move him from tied to ahead. Collins's identical upside lands against a rival who is 92% defined, holds a positive net favorable, and leads by 11 on the cycle's top issue; consolidation gets him from behind to competitive, and no further on its own. The next wave should track both nominees' name ID and, more telling, their image among independents — today 29/47 for Jackson and 32/43 for Collins.
Turnout: certainty is not evenly distributed
The voters still up for grabs are the least certain to show up.
vote certainty among moderates and independents, vs. 73.1% of all voters
Overall, 73.1% are "absolutely certain" to vote, but certainty varies widely by group.
The pattern matters because of who sits where. The groups carrying the most undecided voters in both races (independents, moderates, younger voters, and voters without a college degree; Section 2) are the same groups reporting the lowest certainty, while the most certain groups (seniors, postgraduates, and strong ideologues of both stripes) have largely picked a side already. Two practical implications. First, both toplines are likely-voter-screen sensitive: tighten the screen and the undecided pool shrinks before anything else does. Second, the conventional "lower turnout helps Republicans" shorthand deserves care here; self-reported certainty is actually higher among Democrats (79.4%) than Republicans (72.8%), and where a turnout advantage lands depends on which of the low-certainty groups shows up, not on any single demographic. What the certainty gap unambiguously does is make the undecided vote perishable: a voter who never shows up never breaks.
The Pollster’s Read
The data ends here. What follows is a read, not a result: the pattern recognition that comes from decades of measuring public opinion and watching elections confirm the numbers, defy them, and everything in between. We think it’s worth having. We would never call it math.
Jackson has more room to grow than Bottoms does. The race is tied even though nearly a quarter of voters have never heard of him, while Bottoms is already defined and carries the weakest intensity score in the survey. That gives Jackson room to consolidate as voters attach him to the Republican label and the Kemp-era GOP brand. The "serves the elite" number is worth watching, especially if Bottoms can make it stick before Jackson introduces himself. But for now it reads more like a soft impression of an unknown candidate than a settled judgment, so our read is that the governor's race could drift toward a modest Jackson edge as the undecided pool shrinks.
Ossoff is a harder problem. He has less of an executive record to defend, and his repositioning since 2020 is showing up clearly in these numbers. His earlier political profile was more nationalized and more progressive; the senator in this poll is being received as a costs, families, Georgia-first candidate. He wins "would put Georgia ahead of either political party" by 9, wins independents by 20+ on every trait tested, and holds 92% name ID with a positive net favorable. That is not an accident; his paid communication appears disciplined and well matched to the electorate's top concern. Whether that positioning can withstand sustained scrutiny is the campaign question, but the polling says it is working in Georgia, right now. Beating him requires the environment to do some of the work, and likely requires paid contrast that reopens the question of whether his current presentation matches his longer political record.
And the environment will do some work — for both Republicans. Every state with party registration has moved measurably toward the GOP since 2022; registration is a trailing indicator, and the drift is roughly one to two points of partisan lean. Georgia doesn't register by party, so it won't show up in the administrative data, but there is no reason to believe Georgia is exempt from the trend, and this poll's own turnout screen points the same direction: the certainty gap (Section 5) is worth an additional couple of points of Republican lean if June enthusiasm holds to November. That's the Collins case, and it is real as far as it goes. Expect him to make up ground as his 23% unknowns come home to the R column and the environment prices in, tightening the race toward the margin of error. Ossoff still holds the advantage in these internals, especially on costs, but the next wave should measure how much of that environmental drift Collins captures and whether Jackson's introduction lands before Bottoms defines him.