Chris Benham chrisbenham-at-bigpond.com |EMlist| wrote:
Russ,
I agree with Kevin V.'s last post in this thread. You wrote:
The same considerations apply to the "top-two Approval pairwise
runoff" method I suggested a few days ago. If the "turkey raisers" or
cloners succeed, they have simply forced an effective reversion to
basic Approval.
In terms of strategy and possible results, no, because in "basic
Approval" no-one has any incentive to vote for a "turkey". Assuming the
"turkey raisers" would prefer to elect a turkey
rather than one of the candidates from the rival faction of
turkey-raisers, then they have two levels of success: (1) to prevent the
election of a candidate from there rival faction of
turkey-raisers, and (2) to elect one of their own candidates. It is
possible that the net effect of the turkey-raising is that the final
runoff will be between two turkeys, a complete
fiasco.
Also it is unfair that parties with the resources to run two candidates
should have an (extra) advantage (in terms of the voting system maths)
over parties who can only afford to run one.
I agree. The same consideration applies to two-round runoff, of course,
which is in widespread use.
And if we are going to use ranked ballots with an Approval
cutoff, why not something reasonable to good like
Smith//Approval, DMC or AM?
Yes, I think those first two are solid (but I forgot what AM stands
for). I think I actually lean toward Smith/Approval. Its one slight
disadvantage is the requirement to explicitly define and determine the
Smith set, which DMC avoids. Not that that's difficult to do, of course.
--Russ
----
Election-methods mailing list - see http://electorama.com/em for list info
