This will necessarily be long. Perhaps someone
who is interested could compile an executive
summary with the most important points extracted and presented briefly.
http://www.fairvote.org/response-to-faulty-analysis-of-burlington-irv-election
The subject page is one in a series of pages
authored by Mr. Bouricius, and my intention is to examine them all.
Response to Faulty Analysis of Burlington IRV Election
by
<http://www.fairvote.org/list/author/Terry_Bouricius>Terry
Bouricius // Published March 17, 2009
FairVote unabashedly believes that the second
IRV election for mayor in Burlington this month was a great success.
There are fascinating implications to this position.
Of course in a partisan election some people
will be disappointed by the result, but objectively IRV:
It is a common FairVote tactic to point to the
face that "some people" will be dissatisfied with
any result, and to thereby discount specific
dissatisfaction with any single result. But,
remember "unabashedly believes." This would seem
to imply that any dissatisfaction with the
Burlington result is merely a product of the
natural disagreement over results, mere partisan sour grapes.
But, remember, FairVote previously claimed that
the pathologies allegedly present in the 2009
Burlington result were rare and could therefore be disregarded.
1) successfully prevented the election of the
candidate who would likely have won under
plurality rules, but would have lost to either
of the other top finishers in a runoff;
The implication is that, in the subject election,
IRV saved the day. However, under the prior
rules, the outcome would quite likely have been
the same as with IRV. The prior rules were not
"plurality," as such, but vote-for-one with a 40%
victory requirement, with a runoff if that is not
met, and it is highly unlikely that anyone would
have gotten that margin in the primary, and
therefore there would have been a runoff, and,
assuming no major preference changes, the same
winner. It can then be argued that IRV avoided a
runoff, but not that it saved the city from the harm of "plurality rules."
I do not know at this writing if write-in votes
would have been allowed in the runoff, but, if
so, it cannot even be said with certainty that
the result would have been any one of the top
two. Further, people do vote a little differently
under vote-for-one rules. (But I do assume that
the IRV first-preference results are based on
sincere votes; the question is what would have
happened with the minor candidate and write-in votes.)
2) saw voters handle the system remarkably
well, with 99.99% valid ballots and equally
effective use of rankings in low-income wards
and high-income wards despite a minimal voter education effort by the city;
I have looked at some of the actual voting
patterns, which lead me to suspect some
misunderstandings, but that's largely not
relevant. Once the nature of the problems with
the 2009 election becomes widespread, we may see
one of two things: Burlington will dump IRV,
based on results, not on supposed "ease of use,"
or strategic voting will start to increase, to
improve the result from the point of view of
those who are motivated to so vote.
and 3) contributed to producing a campaign
among four serious candidates that was widely
praised for its substantive nature.
This is purely speculative, but Burlington is not
a normal town. It is a three-major-party town,
with two other parties with some strength. Even
if the election were plurality, I'd expect more attention to substance.
However, there are opponents of IRV who have
taken to the Internet to tout their "analysis"
of the Burlington election in hopes of derailing the expanding use of IRV.
This has been a FairVote tactic for many years:
dismiss critics of IRV as no-hope fanatics with
no substance, who are forced to use the
"internet" to publish. Of course, this essay by
Bouricius is published in a blog.
Further, criticism of IRV is dismissed as merely
politics, and analysis, no matter how neutral and
cogent, is dismissed as, at best, "flawed," even
with counterarguments that are preposterous to those who understand the issues.
Interestingly these IRV opponents can unite in
criticizing IRV from opposite perspectives.
Interesting? Boring.
Some people are against IRV because they
understand voting systems thoroughly and deeply,
and the criticism of IRV goes back well over a
hundred years. It did not have a good reputation
as a single-winner system, among political scientists.
And some people are against IRV because they have
a political agenda that it harms. For example,
Republicans in Ann Arbor experienced an IRV
election loss in Ann Arbor, Michigan, whereas
previously they had been benefiting from
vote-splitting between Democratic and Human
Rights Party supporters there. So they scheduled
a referendum on IRV, at a politically opportune
time, due to HRP voters, students, being largely
out of town, and prevailed. Naturally, they
presented whatever arguments they could find.
Does this make the arguments defective? Does this
mean that anyone who criticizes IRV would be supporting the Republicans?
Not necessarily!
In other words, some opponents oppose IRV
because they feel IRV's accommodation of
multiple parties undercuts the two-party
system, while others feel IRV doesn't do enough
to break the two party "duopoly."
That's right. Both arguments are correct, but in
different ways. IRV "accomomdates" multiple
parties, but in such a way as to make it nearly
impossible for minor parties to win. (Remember,
in Burlington, the Progressive Party is a major
party.) Thus IRV appears to allow minor parties,
but, in the long run, they disappear -- or become
appendages of a major party, a faction, so to
speak, that doesn't run independent competing
candidacies. There are other voting systems that
will accurately measure support for minor
parties, and, in particular, there is one that
actually provides serious opportunities for minor
parties to win, and it is the system that IRV is
most commonly replacing, which is not Plurality,
as Bouricius implies, but top-two runoff. To win
under TTR, a candidate must claw his or her way
up to second place, not all the way up to the
top. Then, in the runoff, the candidate has a
serious opportunity to make a case.
This is true for dark horse candidates as well as minor party candidates.
Some feel IRV places too much value on voters'
first choices, while others feel IRV places too
little value on first choices. And so on.
I'm not aware of any serious voting systems
critic who believes that IRV places too little
value on first choices. But the issue is far more
complex than Bouricius will let on. The system
that places more value than IRV on what are
allegedly first choices is Plurality. But, once
again, the real choice with IRV is most commonly
(any exceptions) not Plurality, but top-two
runoff. Which places about the same value on
first choices as IRV, unless a better method were
used for the primary election. Which is where
real voting reform lies, not in substituting the
highly defective canvassing method of IRV for the runoff system.
In this post I will briefly address one faulty
analysis that is posted on the web site of
Warren Smith who advocates a novel voting system
called "range voting" that is not used in any public elections.
The advocacy of Smith for range voting is based
on a deep analysis of the purposes of voting, but
the criticism of IRV is not based on a partisan
favoritism, and whether or not range voting is
used in public elections are not is irrelevant.
It's a debate tactic, more of the attempt to
dismiss criticism as coming from mere malcontents, to be disregarded.
Is the analysis there faulty? Surely that has
nothing to do with advocacy of other methods!
Smith could be advocating random ballot and the
analysis still be correct or faulty, or, indeed,
he could be advocating IRV and the analysis could
be either way. But, we will note, Bouricius has
been a long-term advocate of IRV, and is
committed in writing to *many* faulty analyses.
Let's see if he repeats the errors of the past.
This site devotes a lot of space to tearing
down other election reform proposals (and
thereby, effectively propping up the status quo).
It devotes a great deal of space to election
science. A partisan advocate may indeed consider
that to be "tearing down other election reform
proposals." But Warren Smith is first of all a
mathematician and only supports Range Voting out
of that analysis. I personally do not *advocate*
Range Voting in the near term for public
elections, except in the most primitive forms,
being Approval Voting, which boils down to
Counting All the Votes, and American Preferential
Voting, which, being the most widely-implemented
advanced voting system ever attempted in the
U.S., with far more "success" than IRV, should definitely be considered!
Yes. Approval voting and American Preferential
Voting (also known as Bucklin) are forms of Range
Voting. They just don't look like it. Top Two
Runoff also simulates some aspects of Range. But
enough about other alternatives, this is about the 2009 Burlington election.
The analysis opens with a personal attack
against me as an individual, which is rather odd.
Why odd? First of all, it isn't a personal
attack, and, second, didn't Bouricius just make
some ad-hominem arguments? I'll quote the
relevant section of the analysis, we'll see if it is a "personal attack."
Here is the only personal material in the
introduction to the Range Voting page:
The Propaganda
Instant Runoff Voting
(<http://rangevoting.org/rangeVirv.html>IRV)
advocates, especially FairVote's Terrill G.
Bouricius (who lives in Burlington, formerly
served there as alderman, also formerly served
as a Vermont state legislator, calls himself a
"political scientist," was instrumental in
making IRV happen in Burlington starting in
2006, is denoted a "senior analyst" by FairVote,
and in 2005 received a contract to design
Burlington's IRV voter education program), often
hail Burlington's adoption of IRV for its
mayoral election as a "great success." Bouricius
has also contended in various online posts,
print media, and interviews that IRV always elects a "majority winner."
Is this a personal attack? Does Bouricius really
believe that it is, or does he imagine we can't read?
I am used to personal attacks having served a
decade on the city council and another decade as
a state legislator as a member of the
Progressive Party. I have also worked as an
election administrator for non-profit
organizations when not working as an analyst for FairVote.
Would it be out of line to point out that we
might not also be surprised to see a Progressive
politician hail the victory of a Progressive in a
Burlington election? Personally, I don't care
about who wins in Burlington. I don't live there.
I also believe that the people of Burlington have
the right to choose their own election method,
but I simply don't like them being misled as part of the process.
Now turning to the defective analysis...
The irony is, that this anti-IRV-biased
<http://rangevoting.org/Burlington.html>analysis
is co-signed by a Vermont professor named
Anthony Gierzynski. Gierzynski is a strong
defender of plurality voting for statewide
elections in Vermont and a two-party system
where third parties don't run candidates. In
2002 he in fact ran for state legislature as a
Democrat against incumbent state legislator, the
Progressive Party's Bob Kiss, who later went on
to be the winner in Burlington's two IRV
elections for mayor. Professor Gierzynski does
not disclose this obvious source of potential
bias when offering analysis of Bob Kiss'
victories in the 2006 and 2009 Burlington elections.
In other words, he's not turning to the analysis,
but to a consideration of the personality of the
critic. Why am I not surpised. Should Bierzynski
disclose his affiliations? Probably. But ... I've
seen editorials and opinion pieces aplenty by
FairVote activists that don't disclose their
affiliations. It's all irrelevant, actually,
except as a bit of political polemic. Which is
sometimes in order. It has nothing to do with
whether the analysis is cogent or "defective."
The new analysis also is being cited by other
IRV opponents who defend current plurality
elections (vote-for-one-top-vote-getter) and
two-party system, even though it states: "IRV
still seems to have performed better in this
election than plain plurality voting, which
(based on top-preference votes) would have elected Wright.
This is a common and pernicious argument. If
evidence and analysis is used for bad purposes,
it must be bad. The comment, though, is true, so
this isn't at all a criticism of the analysis,
except that the implication isn't true, i.e.,
that IRV improved the situation in Burlington. It
didn't. The previous method, which was not simple
Plurality, would have elected Kiss as well, quite likely.
That would have been even worse, since Wright
actually was a "lose-to-all loser" among the
Big Three, i.e. would have lost head-to-head
races versus either Kiss or Montroll."
Where is the "defective analysis" promised?
Here are just a few misconceptions in the analysis:
1. The author, incorrectly states that
Burlington's IRV election suffered from a
monotonicity failure. You can read about the
<http://www.fairvote.org/monotonicity>monotonicity
criterion here. In fact, no such failure occurred.
The page cited does not exist today. However, what did the authors state?
6. Finally and probably craziest of all this
election also featured
<http://rangevoting.org/Monotone.html>non-monotonicity.
If 753 of the W-voters (specifically, all 495 of
the W>K>M voters plus 258 of the 1289 W-only
voters) had instead decided to vote for K, then
W would have been eliminated (not M) and then M
would have beaten K in the final IRV round by
4067 to 3755. In other words, Kiss won, but if
753 Wright-voters had switched their vote to
Kiss, that would have made Kiss lose!
The "non-monotonicity" page does exist. This is the definition there:
Monotonicity is the property of a voting system that both
* If somebody increases their vote for
candidate C (leaving the rest of their vote
unchanged) that should not worsen C's chances of winning the election.
* If somebody decreases their vote for
candidate B (leaving the rest of their vote
unchanged) that should not improve B's chances of winning the election.
Bouricius went on:
To understand the context, here is the
situation...Republican Wright was ahead in
first choices in the initial tally with 2951
votes, next came Progressive Kiss at 2,585, and
the Democrat, Montroll, was in third with 2063.
There were also two other candidates with 1,306
and 35 votes respectively (and some write-ins).
In the final runoff tally, Republican Wright
had 48.5% and Progressive Kiss won with 51.5%
(with some voters sitting out the runoff by not
ranking either of the finalists).
What his analysis actually shows is that
non-monotonicity could have affected the election, but did not.
Which is what they said. This is pure semantics,
devoid of substance. What is "monotonicity
failure"? We know that IRV is non-monotonic, that
is not controversial. Monotonicity failure is
demonstrated in this case by showing that voters
could have changed their vote in one of the ways
described and with the associated consequence.
The authors claimed that the election "featured
non-monotonicity." I agree with Bouricius, in
that the wording is not fully accurate and could
be misleading, but, in fact, the meaning was
explained in detail, and the explanation was
accurate. The hypothetical votes do show the
vulnerability of the method. I understand,
however, that there is a better example than the
Burlington election, where voters did cause their
candidate to fail by voting for the candidate,
being the election cited by Kathy Dopp recently:
"In the recent Aspen election if 75 fewer voters
had voted for one of the city council members, he
would have won instead of losing."
...If just over 25% of the supporters of
Republican Wright had abandoned their true
first choice and instead voted for any other
candidate (although to meet the
non-monotonicity definition they would need to
switch to Kiss), they could have kept their
favorite candidate, Wright, from making it into
the runoff, and allowed the Democrat to face
off against Kiss in the final runoff, where the
Democrat would beat Kiss. But this did not
happen, and there is nobody who thinks it was a
sensible strategy for any voters or candidates
to advocate. Certainly it had no impact on how
candidates campaigned nor ever would have.
Slippery. Monotonicity failure does not require
that voters be following some strategy, and, in
fact, the big problem with monotonicity failure
is that the voter's vote has an unexpected effect.
Is Bouricius claiming that it would make no sense
for the Republicans to decide to elect Montroll
instead of Kiss? They could, and that IRV is
being used would allow them to do this, as would
plurality. But they would have a more
straightforward way of doing it, and easier: Vote
M>W>K, which only flips their first preference,
same as would be done with Plurality. What's
defective about the analysis? That IRV functioned
poorly when the Republicans sincerely voted their
first preference is a very bad sign. One of the
biggest arguments presented for IRV is that it
supposedly allows sincere voting. It does that
only in the two-party situation (or only two major candidates).
Another way that non-monotonicity could have
occurred would be if Wright or Montroll lost the
election because they got too many first choices
that might have gone to some other candidate
instead. Since Montroll didn't even make it into
the final runoff there is no way this could
apply to him (any fewer first choices would just
confirm his elimination). That leaves Wright.
Again, even if some of his supporters had voted
for any other candidate first, Wright would
still lose the runoff between either Kiss or
Montroll. Smith would need 1,279 first choices
that actually went to Wright, to get Smith into
the final runoff. But at that point Wright would
be in fourth place and have no chance of
advancing to the final runoff. So it is
mathematically impossible for a switch of first
choices away from Wright to have made him a
winner. Thus, despite the Smith-Gierzynsnki
analysis, there was, in fact, no
non-monotonicity event in the Burlington election.
But they did not claim that there was, when we
look closely. Bouricious is seizing on an
appearance rather than the substance. They
actually wrote about a possibility, when the
details are examined. Bouricius doesn't
contradict the actual analysis, only a detail of
how it was presented. Sure, that could be improved.
It is also worth noting that this same kind of
non-monotonic strategy (Republicans conspiring
to help elect the Democrat to block the
Progressive) could be pursued under the old
separate runoff system just as well. In fact,
non-monotonicity is a much bigger risk under
two-election runoffs because with a separate
runoff, strategic manipulators can change their
first choice on their ballot between rounds, which can't be done with IRV.
This has nothing to do with analysis of the
RangeVoting.org paper. The whole issue is a bit
off. "Strategy" is a way that voters vote that
improves the outcome from their point of view.
Generally, strategy is a response or potential
response to a defect in a voting system, but
sometimes it is simply part of the system. If
strategy is desirable but impractical, it would
show a serious defect in the voting system.
Strategy in Plurality is how voters make
Plurality work.... and the extent to which this
is difficult is just the extent to which Plurality is defective.
2. In another point in the analysis, Smith and
Gierzynski attack IRV for failing to elect the
apparent compromise Condorcet-winner.
"Attack." That's Bouricius' favorite word
describing critics. What's the substance here? It
did fail to do that, that's clear from the votes.
Bouricius can't deny that the analysis is
correct, so he attacks the writers as
hypocritical or "disengenous," with a thoroughly defective argument:
This is disingenuous because Warren Smith
himself dismisses the Condorcet-criterion,
since his favored method also fails to meet
this criterion -- in fact, range voting can
elect the Condorcet-loser, which IRV never can,
and could quite possibly allow the defeat of a
candidate who won an absolutely majority of 51%
or more of voters' first choices.
Bouricious either does not understand the
criticism of the Condorcet Criterion, which only
applies and can be understood if there is
preference strength information collected (or
simulated in studies), or he is himself
consciously attacking Smith. Or both. If there is
no preference strength information, the Condorcet
Criterion is practically an absolute among voting
systems analysts. Including Smith. Smith does not
"dismiss" the Condorcet Criterion, rather he notes its limitations, as do I.
And Bouricious gives a handy example. Suppose we
can, indeed, collect preference strength
information, so that we can know if a preference
is trivial or not. Suppose that, voting sincerely
and accurately, 51% of voters rate A over B, but
only by one point out of 100, indicating a
preference that is barely detectable.Say that the
ratings were 100% and 99%, respectively, and that
ratings that high represent explicit approval.
And then that 49% of voters rate B over A by 100
points. In a pure Range voting system, B wins,
violating the Majority and Condorcet Criteria.
Obviously, and in analogous real-world decisions,
the majority, informed of the result, are very
likely to accept it, for their votes of 100 and 99 make no sense otherwise.
However, I have proposed Range/runoff, which
does, in simulations, improve outcome slightly in
the presence of strategic voting, and it is
possible that such an anomaly would be tested
with a runoff. It's debatable either way.
They observe that in the Burlington election,
the candidate who came in third in the initial
tally was a compromise choice who could have
beat either Wright or Kiss in a head-to-head election.
Which is the criticism of IRV that is found in
Robert's Rules of Order, Newly Revised. It is
also a flaw of top-two runoff, in general, though
the situation might be somewhat ameliorated if
write-in votes are allowed in the runoff.
This is called a Condorcet winner, and is a
mathematically valid calculation, though little regarded in American politics.
The Condorcet winner, if one exists, is one that
would beat all others in a direct contest, and
most people have very little difficulty
understanding the significance of that. "American
politics," though, has given little attention in
recent years to voting systems in general, but
political scientists certainly have.
The point is that this compromise candidate
would come in third and lose badly under the
plurality system most IRV opponents support,
and would also be eliminated in a traditional two-round runoff election system.
That's partially correct, but something is
overlooked. By limiting alternatives to
Plurality, traditional two-round runoff, and IRV,
Bouricius is plastering over an artificial
restriction. It's ironic that FairVote has
continually touted the supposed recommendation of
IRV in Robert's Rules of Order, when, in fact,
the method suggested for possible use is not one
of these three methods, exactly. It is IRV, but
with a majority requirement, and repeated election if no majority is found.
Hence, if IRV were used as Robert's Rules
actually recommends, as the first round in an
election process, with, perhaps, the second round
being, then, a compromise, electing by a
plurality if there are more than two candidates
and voting patterns don't create a majority, the
same possible winners would exist as with IRV,
but a defect of IRV would appear, and there is an obvious solution.
American Preferential Voting uses the same
preferential ballot, three-rank, as IRV often
uses. But it is canvassed differently. It would
find the Condorcet winner, more often; but if
enough voters don't rank the frontrunners, there
can still be majority failure. In that event the
possibility that the Condorcet winner is not
among the top two is truly remote. Basically, if
you are going to have a two-round system (i.e,
for use when a majority is not found in the first
round), then using IRV for the first round
preserves the major defect of plurality voting in
the first round: the inability to detect a compromise candidate.
For this reason, the repeated voting of Robert's
Rules of Order does not allow candidate eliminations at all.
But with a different method of determining the
top two, IRV would actually work for a first
round method. Simple. Follow the IRV process, but
don't modify the "quota," the majority needed to
elect. Use, instead, the counting process in
Robert's Rules. If no majority is found after
*all* eliminations (i.e., down to only one
candidate being counted), then count all the
votes for all the candidates and pick the top one
from those as the one to face the IRV winner in
the runoff. If they are the same, then pick the top two as to all votes.
Let the voters choose.
In fact, such compromise candidates have a
better chance under IRV than any other voting
system used by any government anywhere in the world.
Slightly misleading. It may be better than any
voting system used for public elections, but
voting in parliaments is far more sophisticated,
and a Condorcet winner can't lose in standard
deliberative process -- unless not even nominated!
But some anti-IRV, pro-plurality activists
illogically use IRV's failure to elect the
third-place candidate in this election as an
excuse for attacking IRV and supporting plurality elections.
Bouricious defines the criticism of IRV in his
political terms. The argument is highly
misleading. Almost all IRV implementations have
been with nonpartisan elections, which behave
very differently from partisan elections. In
non-partisan elections, IRV behaves almost
exactly like plurality. Only in very close
elections is it at all likely to produce a
different result, and I haven't seen any examples.
But, in Burlington, an advanced voting system
should be used, for sure. But IRV simply isn't
the one. There is a much better choice, with
wider experience behind it than IRV (in the
United States), simpler to canvass, and which
performs much better with respect to both voting
systems criteria and in simulations which study
overall voter satisfaction with results. American
Preferential Voting, it was called by the
political scientists of the time. Bucklin Voting.
This criticism of IRV is legitimate when coming
from Condorcet advocates, (which Warren Smith is
not, since his favored system also fails the Condorcet criterion).
I explain above why Bouricius's criticism is off.
Range Voting fails the Condorcet Criterion only
in certain cases where that Criterion does not
produce optimal results. This is was not the
situation in Burlington, probably. The Condorcet
Criterion is universally recognized as one of the
most important voting systems criteria, and only
with the introduction of utility analysis did it
become possible to wideliy recognize the defects
in the criterion. To make it clear, most of the
time the Condorcet winner will win a Range
election, it takes unusual conditions for the Range winner to be different.
And it is easy to show that, under those
conditions, there is indeed a better winner than
the Condorcet winner. In Burlington, though, the
Range winner, I'm sure, and the Condorcet winner
would have been identical, unless voters bullet
voted, in which case Range defaults to plurality, as does IRV.
It comes down to a matter of what values one
feels are most important in an election process
- -both in who should win and what kind of
campaign you want to see run. I and many other
experts feel the "mutual-majority" and
"later-no-harm" criteria are far more important
than the Condorcet criterion, for example.
Later-No-Harm is such a highly questionable
criterion that the reviewer of the paper which
proposed it said it was disgusting. Basically,
Later-No-Harm guarantees that a voting systems
method cannot find an optimal compromise.
Where is the criticism of the analysis that we
were led to expect, showing that it was
"defective"? I'm seeing arguments made up to
promote IRV, but not a showing of defect.
Condorcet voting methods discount the relative
importance of first choices, to the extent that
a candidate who came in last place in terms of
first choices, or even wins no first choices at
all, but who is a broadly acceptable second
choice, can win a Condorcet election.
Sure. And should, most people would agree, as
long as those second preferences represented
approval and satisfaction with the result.
Because of the habits of partisan politics, we
often think of elections as contests, with each
side pouring itself into the campaign and trying
to win, rather than as devices whereby a society
finds broadly acceptable solution. But in small
towns and cooperative environments, that's what
elections actually are, and, as well, other ways of making collective choices.
Later-No-Harm and the associated strength
accorded to first preferences largely assumes
this kind of context, and, then, encourages and
maintains it, and that is very much what is wrong
with American politics, it easily becomes the
politics of division. IRV does nothing to fix
this, and perpetuates it, from what we can see of
IRV experience in other countries.
One concern is how that might affect candidate
policy discussion, where the avoidance of
alienating any voters becomes more important
than the earning of any first choices.
This is entirely speculative; however, it applies
as well to IRV, doesn't it? It is claimed that
IRV reduces negative campaigning -- without
evidence -- but, if this were true, wouldn't it
mean that candidates would be avoiding offending
those who have other first choices, hoping to get
lower-ranked choices from them?
This has nothing to do with response to the RangeVoting.org paper.
But this, at least is an area of legitimate
disagreement over what values to reward in an
election process, where reasonable and honest
people can hope to resolve their differences through open discussion.
Hey, great! Why not use a voting system that
simulates what would actually happen in a series
of plurality votes, repeated until one candidate
has a majority, without eliminations. There is a
method which does that, quite well. American Preferential Voting.
It is an approval method, instant runoff
approval, and approval is the simplest Range
method. Range balloting could be used instead of
the ranked ballot of standard Bucklin, to quite
accurately simulated the lowering of approval
cutoff that takes place as voters compromise in a
series of elections. IRV eliminates candidates,
standard repeated ballot does not. Thus the
compromise winner is not eliminated, but quickly
rises in approval, until a majority have approved.
But Bucklin, like IRV, can fail to find a
majority; unlike IRV, when it has done this, it
has counted all the votes, it's done the most
that could be done. Bucklin, like IRV, was
apparently oversold as finding majorities without
runoffs. That's not possible to guarantee without
coercing voters. Hence my suggestion: Bucklin as a primary election method.
So the Smith-Gierzysnki analysis shows that IRV
is better than plurality (and in fact better than two-election runoffs),
I thought Bouricius was claiming that the analysis was defective.
I agree that IRV is better than raw plurality in
partisan elections. In nonpartisan elections, it
is not. Effectively, it *is* plurality.
yet plurality voting advocates twist the story
to claim it shows how bad IRV is.
But it also does that.
The lead author of the report favors other
un-tested voting methods that he naively thinks
are better than any other system, and constantly
attacks IRV in hopes of winning support for his
favorite theoretical system. The net result is to help maintain the status quo.
Bouricius, initially complaining about a
non-existent personal attack, devoted much of his
response to impugning the motives of the authors
of the paper, and to promoting IRV based on
peripheral and speculative arguments with no
basis in fact (such as the supposed effect on campaigning).
IRV is a faux reform that produces, under some
conditions, a slight improvement over plurality.
It fixes what I've called the first-order spoiler
effect, where a minor, no-hope candidate splits
the vote for one of two major candidates, causing the ostensibly
For more defense of IRV in Burlington, including
my critique of Gierzsynki's unfounded claim that
IRV creates a bias against less well-educated
voters, see my
<http://vermontdailybriefing.com/?p=1215>post at Vermont Daily Briefing.
There is some evidence about that, from polls
that have been done. But to my mind, it's largely off the point.
----
Election-Methods mailing list - see http://electorama.com/em for list info
