Here's one style of vulnerability of IRV that has not been discussed
that much.
In elections that have many candidates that represent small interest
groups, and when there are so many such candidates that practically
all voters have at least one such candidate that closely reflects
their views, then those candidates that appeal to many voters may not
be the first preferences of practically any voters. Let's assume that
the environment is rather competitive and the candidates with wide
support are therefore considered to be "too compromising" and "not
our men/women" and "having too good relations with the enemy" by most
voters. Maybe we are talking about an environment that is split in
multiple ideological, religious, ethnic, family, work team, business
interest or neighbourhood groups.
Candidates with wide support may thus get get only few or no first
place votes at all and may therefore be eliminated at the first rounds.
It may be typical that after the candidates of one's own small group
most voters do list some candidates with wide support (since the
environment is competitive and they definitely want to rank
candidates of all the competing groups last).
(Note that the fact that Condorcet may elect candidates that don't
have much first place support has been used as an explanation to why
Condrocet is not acceptable and why IRV would perform better.)
If a voter expects all other voters to defend their own small group
candidates then it may be strategically sensible to follow the trend
and rank one's sincere favourites first in order not to eliminate the
chances of one's own (small group) candidates to win the election.
One would thus prepare for the possibility that the candidates with
wide support will be eliminated early. It would be better from this
voter's point of view if other voters than oneself (and other voters
supporting the same small group candidates) would give the compromise
candidates those votes that they need to survive to the last rounds
(since there is a risk that they will not survive to the last rounds).
It is hard to say what the probability of eliminating all the
candidates with wide support early on would be in e.g. in some
typical presidential elections (certainly less probable than some of
the three candidate vulnerability scenarios), but in suitable
circumstances this may anyway well happen. The probability increases
when the number of candidates increases. Also the possible
competitive set-ups that were mentioned at the beginning of this mail
increase the probability.
Many IRV vulnerabilities deal with cases where there are more than
two major candidates, or two major candidates and one or more
centrists with less first place support. In this described case it is
also possible that there is only one candidate with very wide second
(or close to top) support and multiple competing minor candidates
that all are all strongly disliked by large majority.
Saying that IRV allows voters to express their true preferences also
on minor candidates ("avoiding the spoiler effect") doesn't seem to
hold in this kind of environments with many candidates that have
strong local support and that together cover most part of the opinion
space.
Juho
___________________________________________________________
Try the all-new Yahoo! Mail. "The New Version is radically easier to use" The Wall Street Journal
http://uk.docs.yahoo.com/nowyoucan.html
----
Election-Methods mailing list - see http://electorama.com/em for list info
