re:[EM] Majority Choice Approval

Alex Small asmall at physics.ucsb.edu
Wed May 8 10:44:14 PDT 2002


Joe-

I like your analysis!

I want to point two one other things about allowing three levels instead of
two:

First, one problem in approval strategy is that in a close 3-way race
(which I believe will be more common with better election methods) voters
have a tough call when deciding whether or not to approve the compromise as
well.  It's easy enough if we simply have a tolerance threshold and vote
for all who meet it.  However, there are also strategic considerations to
hold your cards close to the vest and protect your favorite by not voting
for a compromise (even if you really do like him).  Allowing a distinction
will make their choice easier.

Second, those strategic incentives make Demorep's suggestion of a new
election (be it by the people or a legislature) when nobody gets a "yes"
majority questionable.  If everybody votes based on a tolerance threshold
then Demorep's idea is sound.  If, however, they vote strategically it may
be that in fact a majority actually finds one of the candidates acceptable,
but they held their cards close to the vest to help their favorite.

Demorep's idea is much more practicable if we allow a favorite/compromise
distinction.  There are fewer incentives to hold the cards close to the
vest, and so true yes majorities will be easier to identify.  In that case
Demorep's idea of electing "none of the above" is very practical and
worthwhile.

Please understand, I'm not defending minority rule.  I'm saying that the
majority may not identify itself for reasons of strategy, even though there
is in fact a candidate acceptable to the majority.  Given that very
plausible situation, demanding a majority in 2-level approval may lead to
plenty of inconclusive elections that failed to identify the majority
preference.  3-level approval makes it much easier to identify the yes
majority advocated by Demorep.  So, I would oppose Demorep's idea in 2-
level approval elections (as well as plurality elections) but support it
enthusiastically in 3-level approval elections.

Alex

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em



More information about the Election-Methods mailing list