[EM] I need an example of Condorcet method being subjected tothe spoiler effect if any

[robert bristow-johnson]

On Jan 21, 2010, at 2:07 PM, Terry Bouricius wrote:

> Jonathan makes an important point. The term "spoiler"  means a minor
> candidate with a small percentage of the vote, who changes which of  
> the
> other candidates wins by running. But Kathy and some others wish to  
> expand
> the definition to include a front-runner. (Note that these IRV  
> opponents
> refer to the top plurality vote-getter who narrowly lost the runoff in
> Burlington, Kurt Wright, as a "spoiler" who prevented the candidate in
> third place from winning. This is a dynamic worthy of analysis, but  
> the
> word "spoiler" is never used by the media or political scientists when
> describing the plurality leader.

Terry, i am just as guilty of that as anyone.

i know that, strictly speaking, the "spoiler" is supposed to be a  
candidate with essentially no chance of winning.  and it might be  
their very intent (usually for personal, not policy reasons) to run  
*specifically* to spoil the election, to defeat the otherwise winner.

if you look up "spoiler effect" in Wikipedia (such an authoritative  
reference) they say it's "the effect a minor party candidate with  
little chance of winning can have on a close election, in which their  
candidacy results in the election being won by a candidate dissimilar  
to them rather than a candidate similar to them. The minor candidate  
is often referred to as a 'spoiler.' "

true Kurt Wright was no "minor candidate".  the fact that the top  
three candidates were *all* significant in their 1st-choice base and  
that the Condorcet candidate was less so than the other two, is at  
the root to how the IRV failed in 2009 to elect the CW.  that, plus  
the fact that the two non-plurality candidates were more similar to  
each other than the plurality leader, so no matter who it was that  
got to the final round, the plurality leader would lose the final  
round, those two facts is what is at the core of all the controversy  
in Burlington regarding IRV.  i believe the first fact (IRV failing  
to elect Condorcet winner) is a real problem and that the second fact  
(IRV failing to elect the Plurality winner) is precisely why we  
adopted IRV in the first place.  to argue against the second is to re- 
visit the main argument we might have had in 2005 (but IRV was pretty  
well adopted without a fuss in 2005).

but there *are* three new issues that are legit arguments for how  
well IRV did last March in Burlington:

1.  Thwarted Majority.  what happens when a CW exists and the  
election method in force does not elect the CW.

2.  Spoiler.  what you brought up now, more discussion below.

3.  Encouraging strategic voting.  what happens when one votes  
sincerely and finds that that sincere vote has harmed their political  
interest.

those are about the "Principles 1, 2, and 3" of my little paper  
you're familiar with.

we know that the IRV opponents in Burlington (that wish to return to  
FPTP) are mistaken in thinking that their candidate, the Plurality  
winner, should have been elected.  that would thwart the majority in  
Burlington to even a greater degree.  this is exactly why we adopted  
IRV in the first place and to revert to the "old law" is to ignore or  
reject those very real concerns (the minority candidate splitting the  
majority and getting elected without majority support).  but IRV also  
failed about that, just not as much as FPTP would or Delayed-Runoff  
might have.

as for Principle 3, what IRV did was transfer the burden, of having  
to strategically consider one's first choice to avoid electing one's  
worst choice, from the majority to a minority.  it is closely related  
to the spoiler effect.  with FPTP, one has to consider expressing  
their sincere "protest vote" for a non-credible candidate, if they  
think that their credible fall-back candidate is in danger of losing  
to their worst choice.  in last year's IRV, a large contingent of  
supporter of the Plurality winner had marked their ballots that the  
IRV winner was their worst choice.  but marking their ballots  
sincerely for their 1st choice actually caused their worst choice to  
win.  they cannot be happy about that and, if IRV survives, must  
consider that problem in 2005 and may feel called upon to vote  
strategically for their fall-back candidate as their first choice.

as for the "spoiler", evaluating whether or not some candidate is  
credible or not can be a subjective evaluation.  with that in mind,  
defining "spoiler" as "a candidate with 'with little chance of  
winning' that, if removed from the election, changes who the winner  
is", that definition is not as objective as defining a "spoiler" as  
"a candidate who *loses* that, if removed from the election, changes  
who the winner is."  Kurt Wright *does* satisfy the latter  
definition, objectively.  whether he satisfies the earlier definition  
is a matter of judgment.

but we know that the whole point in having IRV (or some other ranked- 
ballot election) vs. FPTP was to *avoid* electing the Plurality  
leader when that leader lacks majority support.  That's the whole  
point.  and for the anti-IRVers to complain that it worked in 2009 to  
that end is for them to just forget the reasons we adopted it in the  
first place.  people who oppose IRV because it didn't elect the  
Condorcet candidate should understand that FPTP is even worse in that  
regard.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."