Republicans have another year to right the ship before 2026, but they will not do so if they do not take any useful lessons from 2025.
Prices, as economists like to say, are information. They tell you what price people will pay, and what prices people will accept in order to sell. In a democracy, elections are supposed to be the same way. Polls are information of a sort, but nobody is putting their money where their mouth is; they are just samples. Votes, like prices, are real information. When elections are held frequently, they are supposed to provide the same sort of real-world, bottom-line feedback for politics that prices provide to markets.
Even without final results, I’m skeptical that this will be how tonight’s outcomes and margins will be received. This is partly because we have moved out of an era of being obsessed with poll-tested and focus-grouped politics into an era of populism that is, in a great irony, much less interested in how the people vote in general elections. But this era is highly attuned to what moves small-dollar donors.
Tonight looks likely to be a Democratic sweep or something close to it. But what to make of that? Abigail Spanberger (D-Generic) will be the next governor of Virginia — a win for moderate Democrats and a loss for fire-breathing social-conservative Republicans in slightly blue-of-center states. The most hotly contested race of the cycle, for Virginia attorney general, pits a constitutional conservative Republican (Jason Miyares) against a Democrat (Jay Jones) who is an emblem of wanting your Republican neighbors murdered. Is a Spanberger victory a win for sane, centrist Democrats — or if she carries Jones with her, is it proof that a vote for sane, centrist Democrats gives cover to violent radicals?
New Jersey, at this writing, looks most likely to be a narrow victory for Mikie Sherrill over Jack Ciattarelli (although Ciattarelli, like Miyares, cannot be counted out yet at this hour). Sherill’s closing argument in TV ads was that Ciattarelli was a soft-on-crime tax-hiker, so a win on those grounds is not exactly a ringing endorsement for progressive governance. And while a win is a win is a win, after eight years of Phil Murphy as governor and with Donald Trump in the White House, a margin of a few points in a state that hasn’t gone Republican at the national level in 37 years is hardly anything to crow about.
Then there’s New York City. If Spanberger and Sherrill are a point in favor of running generic Democrats in light-blue states, the likely victory for Zohran Mamdani in the mayoral race would seem like a counter-case for harnessing the energy of the youth vote for progressive zealots. But then, New York City (with its vast immigrant population and huge turnover in the electorate every decade or so) is massively atypical, and Andrew Cuomo (the default Republican candidate, with Trump’s endorsement, in a race that also has an actual Republican nominee who made his name as a crime-fighting vigilante in the 1980s) is very much not the normal opponent for a Democrat, what with being both Democratic royalty (son of Mario Cuomo, husband of a Kennedy, three-term governor of New York) and with being so scandal-tarred that even Democrats ran him out of office. Cuomo’s problem is that he could beat Mamdani only if (imagine this as a condition to winning elections) more people voted for him than voted for Mamdani. Unfortunately for Cuomo, almost nobody wanted him as mayor; he has gotten as far as he did only because Mamdani is such a horrible alternative. But as the old saying goes, you can’t beat somebody with nobody. In the final insult after decades ruling New York’s political roost, Andrew Cuomo is nobody.
Meanwhile, the people talking up the Mamdani model and the people talking up the Spanberger model will completely tune one another out.
Finally, there’s the true swing-state race in Pennsylvania, with Democrats bidding to keep control of the state Supreme Court for another decade. But Republicans were massively outspent and outhustled. They will make that their excuse, rather than reflecting upon how five-year-old stolen-election theories are useless in turning out voters against the precise people who actually tried to steal Pennsylvania for Joe Biden in 2020 (and, if you buy the theory, succeeded).
Republicans have another year to right the ship before 2026, but they will not do so if they do not take any useful lessons from 2025. Democrats have three years to figure out what their party is and stands for before 2028, but they will not do so if their contending factions continue to live in parallel factual universes. The lesson of 2025 is likely to be that nobody learns anything at all.