On May 8, 2010, at 3:09 AM, Jameson Quinn wrote:
2010/5/7 Juho <[email protected]>
On May 7, 2010, at 6:40 PM, Jameson Quinn wrote:
2010/5/7 Juho <[email protected]>
My intent was to propose a nonbinding poll which would be conducted
on the same ballots as, and thus simultaneously with, the actual
election. Any valid vote would be interpreted as an answer in the
poll, but this answer could be overridden by the voter, either to
add subtlety, or effectively to say what part of the actual vote
was strategic. This proposal flows from the realization that
expressivity and outcome utility are separate values, and that
forcing them together sometimes brings them into conflict.
I understand that some might argue that this proposal would hurt
legitimacy. What if the election winner was not the poll winner?
Personally, I'd argue that if this is true, it's better to know it.
Either way, the society would get a better understanding of the
true legitimacy of the winners. With a good system, disagreements
would be rare, and so legitimacy overall would increase; and when
they occurred, they could be an important, though symbolic, check
on the mandate of a winner who's true legitimacy is weak.
One could claim that TTR is a better system than IRV since TTR
collects less information and therefore there is less basis for
complaining about strange end results (e.g. not electing the
Condorcet winner). :-)
I'm sorry... are you seriously arguing this, or doing a reductio ad
absurdum? I'm too thick right now to tell. And I don't want to
offend you by responding to the wrong intent.
JQ
Yes, that was written with some smiley. (I could as well and somewhat
more seriously have written that IRV is good because it is at least in
principle a natural path to Condorcet because people will sooner or
later notice how it differs from the Condorcet methods.)
Juho
----
Election-Methods mailing list - see http://electorama.com/em for list info
