This isn’t unheard of, certainly. Several states have popular Republican governors but otherwise elect Democrats statewide, including Massachusetts and Maryland. But “not unheard of” is not the same as “common.” And given the recent trend away from split-ticket voting at the state level, it’s worth exploring the implications of what might unfold in Georgia.
The Journal-Constitution poll found that 4 percent of Kemp voters also planned to vote for Warnock. That’s 2 percent of the respondents overall. In other words, the two contests look like this, with those Kemp-Warnock voters keeping the senator just ahead of Walker.
This is one poll, certainly, and the state of play will change by the general election in November. But it suggests there may be either enough Democrats willing to back Kemp (who famously opposed Donald Trump’s efforts to overturn the state’s presidential results) or Republicans willing to back Warnock (perhaps given Walker’s … liabilities) that Georgia could split its statewide preferences between the parties.
In recent years, that’s become less common. You can see the evolution of the relationship below. The vote margin in the Senate races closest to the indicated year are shown from left to right. The vote margin in the gubernatorial races are shown from top to bottom. Each state is represented by a dot. Dots in red- or blue-shaded zones are ones in which the state voted for a governor and senator of the same party. Dots in the white areas split between the parties.
We can measure how strongly correlated the vote for Senate and governor were in each year. That’s shown in the gray graph: Dots closer to the right indicate that the margins in both races were more similar. So, in the races considered in the 2016 graph, there was a relatively strong correlation between those margins.
You can see this visually: The more the dots align along an imaginary diagonal from bottom-left to top-right, the more the margins are correlated. That’s obvious in 2016, and a bit less so in 2020.
Compare that with the increasing alignment of Senate votes with presidential ones. Again, these are not all Senate races from the presidential election years themselves. But the correlation between the two results has grown significantly.
Why the difference? Presumably in large part because voters are still able to separate gubernatorial races from national politics. When you’re voting for president, you’re voting for the leader of the country; when you’re voting for a senator, you’re voting to increase your party’s position in a significant legislative body. When you’re voting for governor, the stakes are high but different. We nonetheless still see gubernatorial results aligning with Senate ones, a mark of the increased nationalization of politics.
There are two questions that the results in Georgia might address.
The first is the extent to which Kemp will benefit from sympathetic Democratic voters. The Journal-Constitution poll found that Republican Secretary of State Brad Raffensperger, the primary target of Trump, was getting 16 percent of the Democratic vote in that race.
The second, and more important, is the extent to which Republican partisans line up behind Walker simply to increase the GOP’s chances of taking over the Senate despite his well-documented political shortcomings. Should the results in Georgia align by party — Kemp wins by a similar amount as Walker, for example — it could hint at the further erosion of split-ticket voting in the United States. It could also just be a side effect of a strong Republican year.
Governors still can win even as their states elect senators of the other party. But the trendline is not toward that happening more often.