Russ, I had written regarding your "Ranked Approval" proposal:
I think that with pre-polls and strategy, this would usually give the same result as IRV. In the three-candidate case, I see it only giving a different result when a lot of voters have a big sincere ratings gap between their second and third choice and yet they wouldn't be prepared to order-reverse (Compromise) in IRV. Otherwise, parties/candidates that see themselves as having a chance of winning will advise their supporters to bullet-vote, and the same candidates as in IRV will be eliminated.
To which you responded:
I don't see why that would happen. I could be missing something, but I don't see why any voter would not approve the same candidates he would approve in a pure approval election.
Well, I think I gave a fairly large qualification. I think *normally* pure Approval in 3-candidate public political elections would give the same result as IRV.
The fact that he gets to rank the approved candidates is just "icing on the cake."I agree that all these reasonable versions of "Ranked Approval" are much better than pure Approval.
I had written:
What is your argument against the even simpler: "Voters rank the candidates they approve. Elect the CW if there is one, otherwise elect the Approval winner" ? (or a slightly more complex version that stipulates that the winner must be in the Schwartz or Smith set?).
You responded:
I think that's a good system too. In fact, I independently thought of it and proposed it recently here on EM, stating in the process that I would be very surprised if it hadn't already been proposed. Indeed it had. I think it is preferable to defeat-dropping Condorcet methods, but I don't think it is as good as RAV because it devolves straight back to Approval in the absense of a CW. It would at least need to restrict the winner to the Smith set to make sense to me.
With the voters only ranking candidates they approve, I would say that the chance that the AW wouldn't be in the Smith set would be almost nil.
Even if the voters are allowed to enter an approval cutoff and rank candidates below it, then there is an argument advanced by Adam Tarr in March 2002 that the winner *shouldn't*
be restricted to the Smith set.
http://lists.electorama.com/pipermail/election-methods-electorama.com/2002-March/007699.html
Chris Benham
---- Election-methods mailing list - see http://electorama.com/em for list info
