The blog · Fall 2026
Nearly 95% of House seats are safe for one party or the other. In those places the general election confirms a decision that was already made, months earlier, by a small fraction of the people who could have made it.
The numbers
That figure is from 2022, reported in a peer-reviewed study of primary turnout published in December 2025. The same study puts midterm primary turnout at about 20% of eligible voters nationwide, and cites the Cook Political Report finding that only 22 of 435 House races were a toss-up in 2024.
So the round that decides most seats draws roughly one voter in five, and in the seats that are safe, that round is the whole election.
The idea
It's free, it's the same drive to the same building, and in a safe district your vote there is worth several in November.
The objections, first
Why this keeps happening
People stick with what they've chosen before, and prefer a known disappointment to an unknown risk. That isn't a flaw. You do it with mechanics and restaurants, and it's usually right.
What's worth noticing is that an election rewards that tendency more than most things do. An incumbent has the name you recognize, in a district often drawn to be safe, decided in a round most people skip. None of that requires anyone to have made a mistake.
The bigger fixes
The reform most often proposed is opening primaries to unaffiliated voters. Around three in four Americans approve of it. But the December 2025 study found open primaries were not associated with higher individual turnout.
What did move turnout was nonpartisan primaries — where candidates of every party run in one contest open to everyone, used in only four states. Turnout rose across all age and income groups, with the largest proportionate gains among younger voters.
That's worth saying plainly even though it complicates the argument. The popular fix didn't do the thing. A rarer one did.
Both are decided at state level. The one federal item in this territory is term limits, which needs an amendment and therefore needs Congress. 82% of Americans favor them — 85% of Republicans, 83% of Democrats. The last House vote was in 1995. See the record.
What this is not
There's a version of this argument that says people vote against their own interests. We're not making it. It assumes we know someone's interests better than they do, and usually it means their economic interests ranked above everything else they might care about. People vote on values and on the kind of country they want, and calling that a mistake is just substituting our priorities for theirs.
The record doesn't need that argument. It shows something simpler: large majorities of both parties want the same fifteen things, and fourteen of them haven't happened. That's true regardless of who anyone voted for.
Where these came from