Today’s WaPo has an interesting article on the use of instant-runoff voting in San Francisco (þ: PoliBlog). While IRV isn’t exactly perfect, I think it’s better by a mile than plurality voting in multicandidate elections, leaving aside the argument over whether we should have multicandidate elections—which is in essence a debate over whether or not the meaningful policy space is unidimensional.
3 comments:
Chris,
I take it you would prefer Condorcet voting? Based on my limited knowledge, it does seem like the better alternative (the winner has to win a head-to-head race against every candidate) but given the difficulty of the behind-the-scenes calculations, with a matrix and all, it seems unlikely to get adopted. I can see people’s eyes glazing over just at the prospect.
Condorcet seems to work fairly well for us in the Debian project (and we have software that does all the calculations fairly easily), but even IRV would be a massive improvement, even though it has some non-ideal corner cases.
I think there’s a much easier way to explain Condorcet than the usual Poli-Sci matrices. Just say: round-robin tournament. I think people will get that.
But I think IRV is a good way to introduce ranked ballots, and as Chris says, a huge improvement over plurality.