At 11:50 PM 6/11/2008, Greg wrote:
The FairVote document that debunks Dopp's claims is available at:
http://www.fairvote.org/dopp
Or, more accurately, "attempts to debunk." Ms.
Dopp is a voting security expert, not an election
methods expert, and some of her statements can be
flawed, especially if one focuses on technical
details. I'm going to break this response into a
series of posts on each separate topic, because
otherwise it gets *way* too large.
This first section is a good example.
De-Bunking Kathy Dopp's 15 Flaws of Instant Runoff Voting
1. Dopp: "Does not solve the "spoiler" problem except in special cases
."
Dopp has her special cases reversed. In fact,
IRV solves the spoiler problem in virtually all likely U.S. partisan elections.
Which does not contradict Dopp's statement. The
problem is that "special cases" implies "only
rare cases," but, technically, IRV *may* reduce
the spoiler effect in ... a special case, i.e.,
when a minor candidate is not in range of
winning, but due to vote splitting, causes the
overall preferred major candidate to lose to one
less preferred. It does not deal well with the
center-squeeze effect, where there are three
candidates in range of winning, and this is, I
believe, what Ms. Dopp refers to.
Notice the "special case" in the anonymous
response: it isn't "IRV" solves the spoiler
problem. There are three hedges: "virtually all,"
"likely," and "partisan." The most operative of
these restrictions is "partisan." Whoever wrote
this -- I suspect Rob Richie, and he is certainly
aware of this interchanges, since he commented on
it on Wikipedia, in Talk for the article on
Instant=-runoff voting -- is quite sophisticated.
IRV, in the absence of the kind of vote transfers
that take place in partisan elections (where the
large majority of vote transfers from one
candidate may favor a single other candidate),
does little to solve the remaining "spoiler
effect," center squeeze. It is debatable whether
or not center squeeze should be called a spoiler
effect. I've done so in the past, but the
phenomenon is different, and I'd be happy with
confining the term "spoiler" to refer to minor
candidates, not candidates who can actually win
an election or come close. In which case, indeed,
Ms.Dopp's statement, as made, would be faulty.
But she is, in fact, talking about center
squeeze, which is a real problem, one of the two
that are so notable that Robert's Rules mentions them.
Further, the response wants us to think of all
U.S. elections when, in fact, IRV is generally
being proposed as a replacement for real runoff
elections. Which do not show the spoiler effect
in anywhere near the same strength as seen with
Plurality. And where that spoiler effect *is* in
effect in runoff voting, IRV may well fail to
resolve the problem; indeed, usually it will.
What has been missed in most discussion of the
issue is that IRV, as proposed in the U.S., is a
plurality method. Don't confuse this with
Majority Criterion compliance. Plurality methods
will elect a candidate even though a majority of
voters haven't voted for that candidate, that's a
simple description of it. Jurisdictions have
runoff voting because they value finding a
majority vote for the winner, and top-two runoff
actually accomplishes this. The truth is that it
is impossible to guarantee a majority vote in a
single ballot; it is even impossible with a
limited series of ballots, except that with
top-two runoff, because of the ballot design and
voter habits, a majority will almost always be
found. (Most top-two implementations allow
write-in votes, so a maintained preference of
voters in the first election for an "eliminated
candidate" can allow that candidate to win. I've
never seen it, but, usually, when the Condorcet
winner is eliminated, the preference strength,
apparently, is not enough to motivate voters to
turn out and vote write-in. In other words,
Top-Two runoff, in real practice, works quite
well, much better than simplistic voting systems theory might predict.
Whenever a third party or independent
candidate is unlikely to be one of the top
vote-getters (true in over 99% of U.S.
elections), IRV eliminates the spoiler problem completely.
Actually, no. If you look at Australia, in the
places where ranking all candidates is optional,
there is a lot of plumping. So there is still a
spoiler effect. FairVote has always implied that
the Australians use a uniform method, but, in
fact, some places use STV with an absolute
majority requirement, and they guarantee that
requirement by voiding all ballots that don't
fully rank the candidates, and other places use
Optional Preferential Voting -- which is what is
proposed for the U.S. -- and plumping -- we call
it bullet voting -- is common and, apparently,
increasing, according to Antony Green of ABC.
Naturally, the majority requirement for OPV is
relaxed to "a majority of ballots containing
votes for remaining candidates." That's a
plurality method. Regular STV, PV, is a majority
method, but it coerces the votes. Take your pick.
(Or require a majority, as jurisdictions which
are using runoff voting do, and then use
preferential voting, of whatever kind, or
Approval Voting, to more efficiently find majorities and avoid *some* runoffs.
If a third party grows to the point that its
candidates out-poll major party candidates,
another issue that is related to the spoiler
problem can occasionally arise. This is where
supporters of a third party candidate may worry
that by supporting their favorite candidate,
they risk causing their less-preferred
compromise choice to be eliminated from the
final runoff, leading to the election of their
least-preferred choice. In other words, the
issue of whether to vote for your favorite
choice, or to rank your compromise choice first
can resurface in this unique circumstance. But
this is extremely rare and no different than a
candidate in a partys political primary
arguing Vote for me because I am more electable in the general election.
This argument presents center-squeeze as if it is
a strategic voting problem, when, in fact,
center-squeeze is the problem, and strategic
voting is how some knowledgeable voters might attempt to fix it.
The problem is not rare if there are three major
candidates. In that situation, there is no
candidate that we would ordinarily think of as a
"third party candidate," which always refers to a
minor candidate. Three major candidates can occur
much more commonly in nonpartisan elections than
in partisan ones, in a two-party system. And,
remember, IRV is mostly being proposed at this
time for nonpartisan elections! That is what it
is being used for in San Francisco.
(Continued with the next point, Dopp: 2.
Requires centralized vote counting procedures at the state-level
")
----
Election-Methods mailing list - see http://electorama.com/em for list info
