At 12:50 AM 3/15/2012, Jameson Quinn wrote:
Abd: You are right that C will probably win the chicken dilemma Mike
stated, where C has 49%, under almost any system except SODA. That's
why I usually give a version of the dilemma where C has 40%, not 49%.
You are wrong that this is unrealistic. For instance, see
<http://en.wikipedia.org/wiki/Hawaii's_1st_congressional_district_special_election,_2010>Hawaii's
1st congressional district special election, 2010. I believe that
chicken-like scenarios, while they will be a minority of elections,
are certainly common enough to be worth worrying about; clearly more
common, for instance, than honest Condorcet cycles.
Very interesting election. Special election, very short term.
Djoi (R) 39.4%
Hanabusa (D) 30.8%
Case (D) 27.6%
Case and Hanabusa both urged the other to drop out. I can see why
Case didn't want to, he'd beaten Hanaubusa previously. Case served in
the House before, was a Blue Dog Democrat. Case did drop out of the
next election (also a special election), citing party unity and this
loss, according to Wikipedia.
The national party withdrew from this, conserving campaign funds for
the coming special election, which was going to be for an (almost)
full term. Case is a candidate for the Senate this year, and probably
has a chance. If he'd stayed in that next special election, he might
have been committing political suicide.
I suspect this is center squeeze. With IRV or top-two runoff, it
would might be won by Hanabusa, and she did win in the subdequent
election. However, with a center-squeeze resistant method, Case might
win. Case might be the Condorcet winner. Case would get some
second-rank votes from some Djoi voters, and possibly more Hanabusa voters.
The election, then, is a good example of the problems of plurality.
As a special election, the normal party primary system is bypassed.
There were four other candidates on the ballot, plus (probably)
write-ins, with 2.2% of the vote total.
Bucklin might very well have elected Case, if this is center squeeze,
as it appears. IRV also might have failed to elect Hanabusa, if the
Case supporters were more right than left. I don't find the outcome
predictable. Hanabusa would have to pull ahead a net 8.6% against
Djoi. That's not easy. It's very possible that the fighting between
Case and Hanabusa damaged both; I've pointed out the effect on
campaigning, the division of funds -- and, in this case, the loss of
overall funding or support.
If elections were to use Range ballots (even if it's not used to
determine the winner -- i.e., say, it's Plurality and only the top
vote is counted), it is possible that we'd start to really know what
is happening.
Bucklin was able to handle elections with many candidates. I think
one of the early ones had more than ninety!
----
Election-Methods mailing list - see http://electorama.com/em for list info
