Scientists Prove a 130-Year-Old Voting Rule Is Fairer Than We Knew

Imagine you and your neighbors are electing a five-person neighborhood council. You don't vote for one person — you tick off everyone you'd be happy with. Now the tricky part: how do you hand out the five seats so that small groups of voters don't end up with nobody representing them? That's exactly the problem this new paper tackles.

The method in question is called sequential Phragmén, named after a Swedish mathematician. It works a bit like a fair-share app: seats get handed out one at a time, always to the candidate whose supporters are currently the least represented. Sweden has actually used a version of it in real elections, and it's also popular in online communities that vote on budgets and projects.

What the researchers proved is a guarantee about fairness. In voting theory, a result is "stable" if no group of voters could point at a smaller set of candidates and say, "we'd have been better off with just them." Ideally you'd want a perfect score of 1. Nobody has proven that's always possible — it remains one of the big open questions in the field. The next best thing is getting a small number, and these authors showed sequential Phragmén always lands at 2 or better. The previous record, from a 2026 paper, was 3.651.

Why does a number like 2 versus 3.651 matter? Smaller means fewer voters with a legitimate gripe. It's the difference between a system where an unhappy minority can be four times worse off than their fair share, and one where they can only be twice as bad. The authors also prove that a fair outcome of this quality always exists — you don't have to get lucky. For anyone designing elections for a club, a company board, a city, or an online community, this is a real-world recipe with a proven floor under it.