James--
Yes, it seems to me that the candidate withdrawal option would help any voting system. It would avoid the worst of Plurality & IRV, and it would tend to complete the perfection of Condorcet(wv).
When the subject was being discussed on EM a few years ago, I was saying that the candidate withdrawal option would give Condorcet an additional protection against offensive order-reversal, for the reasons that you�ve described.
It seems to me that Condorcet(wv) has good built-in protection against offensive order-reversal, but the candidate withdrawal would be a good anti-reversal enhancement. We�ve discussed a number of anti-order-reversal enhancements for Condorcet, and the candidate withdrawal option is a good one.
I believe that it was Steve Eppley who pointed out that the CWO, for Plurality, would be very helpful for getting rid of strategy problems, without asking for a change in the actual voting system.
When CWO is suggested to IRV promoters, they always emphatically reject it, saying that it would take control away from the voter. As if it somehow violates your rights when your favorite withdraws when he can�t win, and lets your vote go immediately to your next choice.
IRVists have never been willing to propose ERIRV instead of IRV. They don�t seriously object to it, but only say they don�t understand the need for it. One, for instance, dismissed it by saying that he asked his wife about it and she said that she didn�t know why one would want to use it. Their objection to proposing ERIRV instead of IRV is that it�s simpler to propose the simpler IRV. But that�s a case where the simpler of the 2 proposals is no good at all, and the slightly less simple one is the only one of the two that is adequate.
That same person said that the CWO fails the "laugh test". I asked him if he could define that test or criterion, but he had no reply. Presumably he meant that letting a losing candidate withdraw would be too preposterous to be suggested, so preposterous that it isn�t even necessary to say what�s preposterous about it. And, presumably, irrevocably eliminating candidates based on 1st choices, a tiny fraction of the election�s information; and calling a majority-beaten candidate a majority winner because there�s one candidate he has a majority against; and causing a candidate to lose because someone voted him higher; and saying that IRV lets people safely rank their favorite in 1st place--as long as he�s a sure loser--passes his laugh test :-)
I never replied on-list to your most recent posting about Cardinal Pairwise, AERLO & ATLO. So let me just say here that it seems to me that CP probably accomplishes the same gains as AERLO & ATLO. The advantage of AERLO & ATLO is that they�re enhancements that can be added later, without changing the voting system itself. It seems to me that they�re a more proposable route to enhanced Condorcet. CP would be good too, but AERLO & ATLO offer a more explicitly, transparently obvious way to accomplish the enhancement goals than CP does, it seems to me.
These enhancements are things to suggest _after_ Condorcet has been adopted.
Anyway, now, when Condorcet has yet to be widely-known, much less adopted, and enhanced Condorcet is farther away still, this isn�t the time to let ourselves be distracted about the relative advantages of two Condorcet enhancements (and there aren�t only two of them). That energy would be better spent doing what we can to stop IRV, and proposing Approval, CR, & Condorcet. There are enough advocates of Condorcet, CR & Approval that, with a little unity, we could probably stop IRV and get one of those better methods enacted somwhere.
Mike Ossipoff
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
---- Election-methods mailing list - see http://electorama.com/em for list info
