[Election-Methods] A better IRV (was Re: Range Voting won't eliminate spoilers)

[Steve Eppley]

Hi,

I've no idea what Chris meant below when he claimed plain IRV would 
provide "extra meaningful choices" compared to IRV//Withdrawal.  I 
believe he has that backward, that plain IRV deters candidates from 
competing at compromise positions by making those positions falsely 
appear relatively unpopular.

It would be nice if more IRVings would admit, as Chris seems to have 
done here, that IRV won't reduce polarization and that other methods 
would.  He might be right that "some people" won't consider polarization 
as bad as other voting method flaws, but from what I gather, most people 
do want to greatly reduce polarization.

I don't know where Chris got the idea that my preference for 
IRV//Withdrawal over plain IRV is based on a preference for criteria 
like Condorcet and Minimal Defense over Later No Harm.  Rather, I want 
to reduce spoiling, and I especially want to increase competition to be 
the "best compromise" since (1) that would reduce polarization and (2) 
it would tend to elect less corrupt winners by allowing voters to rank 
less corrupt (compromise) candidates over more corrupt (compromise) 
candidates.

I find Chris' "Later No Harm" example utterly uncompelling.  Candidates 
B & C are clones who appear to be nearly identical on the issues, since 
the supporters of A are indifferent between B & C.  Thus there's no 
reason to believe voters would be significantly harmed by electing B 
instead of C or C instead of B.

Furthermore, if the supporters of A are not truly indifferent between B 
& C, they've apparently harmed themselves by (strategically?) 
truncating.  They've taken away A's ability to improve the outcome for 
them by withdrawing. 

Finally, proper analysis of the scenario must not neglect that the 
incentives induced by Withdrawal begin to act on A, B & C long before 
election day to lead them to compete to be the best compromise. 
(Throughout their political careers, if Withdrawal or another 
compromise-promoting voting method becomes ubiquitous.)  If they are 
near each other on the issues, why should we fret over election day 
voter strategizing?

Regarding Chris' "philosophical" opposition to letting candidates affect 
the outcome after the votes are cast, I'll point out that parliamentary 
systems give them much more of that sort of power.  Members of 
parliament negotiating to pick the prime minister and the rest of the 
cabinet aren't bound by preferences of the voters, whereas with the 
Withdrawal option the only power candidates have is to step out of the 
way of their supporters' next preferred candidate.  With Withdrawal, 
they can't determine the outcome except within certain bounds set by the 
votes.  So I must wonder if Chris' opposition to parliamentary systems 
is greater than his opposition to Withdrawal.

Assuming Chris' opposition to Withdrawal based on his philosophical 
opposition against giving candidates any influence over the result after 
the votes are cast isn't kneejerk, he seems to believe Withdrawal 
negotiations would tend to benefit the candidates or their parties at 
the expense of the voters.  Why? (Philosophy is not a compelling argument.)

Regards,
Steve
---------------------------
Chris Benham wrote:
> Steve Eppley wrote:
> "IRV might cause polarization even worse than what we already have, since 
> the effective plurality rule campaign strategy of shifting toward the 
> centrist swing voters after being nominated--which reduces polarization 
> somewhat--would risk, under IRV, the late entry of an independent 
> candidate competing at a position slightly closer to the party base."
>
> Some people might think that (for many elections) providing voters with extra
> meaningful choices is more important than  "reducing polarisation".
>
> "IRV can be significantly improved by letting candidates withdraw from 
> contention after the votes are cast."
>
> That is debatable. This seems to be based on the assumption that IRV is a
> lousy method because it fails Condorcet and Minimal Defense and when on 
> the sincere preferences there is a Condorcet winner who is not a Dominant
> Mutual Third (DMT) winner then some of the voters have incentive to use the 
> Compromise strategy; and so that any modification that reduces these 
> problems must be a big unambiguous improvement.
>
> On the other hand I consider that it is one of the good methods because it has
> some redeeming positive properties that are incompatible with those that it lacks.
>
> One of its good properties (criterion compliances) that endears it to its supporters
> is that it meets Later-no-Harm.  IRV "improved" in the way Steve suggests doesn't.
>
> 49: A
> 24: B (sincere is B>C)
> 27: C>B
>
> Normal IRV elects A. The voters have no incentive to truncate, including the 24 B
> voters who could have created a preferable result for themselves (the election of C)
> by voting their full sincere rankings.
>
> But with Steve's suggested option of allowing candidates to withdraw after the ballots
> have been cast and analysed, the B supporters' truncation can pay off  if  C plays the
> game and withdraws. If they don't insincerely truncate then B can't win (unless C is 
> somehow induced to withdraw from an unassailable win), so that is a failure of  
> Later-no-Harm.
>
> As I made clear in a previous post, I am philosophically opposed to to having the results 
> of elections determined by the machinations and manoeuvres of  candidates/parties 
> *after* the voters have cast their ballots.
>
> Chris Benham
> -------------------------
> Steve Eppley wrote Mon Mar 17  2008:
>
> ...However, IRV is worse at eliminating spoilers than some other methods.  
> It also undermines candidates who take centrist compromise positions, by 
> defeating them and making them appear unpopular.  As a consequence, we 
> can expect IRV would continue the "two big polarized parties, each 
> nominating one candidate per office" system (including its haphazard 
> primary elections).
>
> IRV might cause polarization even worse than what we already have, since 
> the effective plurality rule campaign strategy of shifting toward the 
> centrist swing voters after being nominated--which reduces polarization 
> somewhat--would risk, under IRV, the late entry of an independent 
> candidate competing at a position slightly closer to the party base.
>
> IRV can be significantly improved by letting candidates withdraw from 
> contention after the votes are cast.  At the end of election day, the 
> votes would be published in a format that candidates (and others) can 
> download.  Then the candidates would be given a few days to decide 
> whether to withdraw.  They could use those days to calculate what the 
> result would be with or without themselves (and/or some other 
> candidates) in the voters' rankings, and to negotiate with supporters 
> and with other candidates about who, if anyone, should withdraw.  The 
> official winner would not be tallied until after the withdrawal period.
>
> --Steve
>
>
>       Get the name you always wanted with the new y7mail email address.
> www.yahoo7.com.au/y7mail
>
>
> ----
> Election-Methods mailing list - see http://electorama.com/em for list info
>