Anat Shenker-Osorio

And keep in mind, for both Brown & Baldwin, you have incumbency effect mixed in.

For sure. But if it were true that people are just 100% D or R to their core, we wouldn't see these results. Or we'd see them confined only to the 5-7% you're calling "swing."

90% R and 90% D. Those are the baselines in the polarized era, and definitely the modals. I've seen several creeping up the last couple of cycle to 95%ish. But anything that registers below that 90% split is notable. Less so though in non-competitive elections, like the 2 you're

citing, bc in the absence of competition and spending to reinforce latent partisanship predispositions in both soft-partisans and Indie leaners, some lower info, lower interest voters will move down to their next heuristic, which is incumbency. So long as the economy is stable,

that likely benefits the incumbent but in an environment like the 2008 economic crisis it would not.

Partisanship, neg partisanship esp., a helluva drug. Consistently see Message X test great but add "The Dem says" or "The Rep says" & it drops. But...that's NOT same as persuasion does NOT work. Elections are about tiny margins. We have moved people enough to win & must do so.