[Election-Methods] RELEASE: Instant Runoff Voting

Sun Jul 27 14:34:09 PDT 2008

On Jul 27, 2008, at 22:26 , Terry Bouricius wrote:

> While I agree that "core support" is not always well measured by first
> choices (multiple clones can make all these clones appear to have  
> little
> "core" support, where any of them would appear to have massive core
> support running alone). However, I think the concept of "core  
> support" is
> still an important factor in multi-candidate elections.

Any justification to why core support would be important? I can  
imagine some specific cases where that could be a positive thing. But  
there are also ones where there is no need to have core support.

It is also possible that if core support is needed, voters will  
include that criterion in their preferences (i.e. if there is a very  
good candidate but who is known not to have very much core support,  
then most voters would simply rank that candidate lower than they  
would in elections where they think that core support is not needed).

With respect to multi-winner methods there has been discussion on if  
having a representative body where the representatives represent the  
average opinion of the voters is enough or if the representatives  
should be such that all sufficiently large opinion groupings will get  
their own representative (such representatives would have "close core  
support"). I think here it is a good default rule to elect  
representatives that are close to their voters in their opinions.

> While I think
> Condorcet methods are much better than most methods, I remain  
> concerned
> that it may, in fact, reward inoffensive candidates who  
> successfully hide
> their policy positions, rather than just true "compromise" candidates.

I'm not sure if the alternatives would be much better, e.g. to elect  
a candidate that gets support only from some strong faction and not  
from the whole community.

The practice of considering unlisted candidates to be ranked equal at  
bottom (not neutral) already gets rid of at least some types of  
"candidates that hide their position".

>
> Voters tend to have clear opinions of candidates at the top and  
> bottom of
> their preference rankings, leaving the door open for inoffensive
> candidates who have avoided revealing any controversial views, to  
> become
> EVERY voter's second choice ("at least he must be better than X") and
> likely Condorcet winner.

Are you referring to a case where the voter 1) intentionally says  
that the "compromise candidate" is better than the "bad candidate" or  
2) ranks the "bad candidate" below all the unknowns in the hope of  
causing harm to him or maybe 3) sincerely thinks that any John Doe  
would be better than the "bad candidate"? In first case maybe the  
compromise candidate indeed is the most liked winner. In the second  
case the voters seem irrational (or their strategy has failed). In  
the third case maybe John Doe should win.

The third case is interesting in the sense that all the candidates  
seem to be either very disliked or unknown/incompetent. In such  
elections there are no good winners. Maybe one could expect that some  
of the compromise candidates would be at least a good John Doe and  
would deserve to win.

> The voting method will cause candidates to tailor
> their campaigns accordingly, and I fear Condorcet encourages  
> candidates to
> limit voter information and instead campaign with slogans like "I  
> am the
> candidate who listen to you" and policy will become even LESS  
> discussed in
> campaigns than is already the case in the U.S.

In many elections candidates do their best to avoid offending any  
potential voters. I don't know if Condorcet would increase this  
behaviour. One viewpoint is that in Condorcet a candidate should try  
to please the supporters of all the parties and segments of the  
society (to get higher rankings on their ballots), not only those  
that are close to considering himself as their most preferred  
candidate. This may be easier if one has a clear message to all those  
groups. And the messages of one candidate should not be conflicting  
since that might mean loss of credibility. Based on this viewpoint  
one could also argue that in Condorcet the candidates just need to  
plan their messages more carefully and keep them such that they  
reflect the opinions of the whole community.

One can of course ask if it is good to elect e.g. presidents so that  
they always represent the interests of most people or if they should  
represent some smaller groupings (and more biased opinions) and in  
time the policies of several presidents would average to representing  
all the groupings (i.e. steady vs. zigzag).

Note that also IRV encourages candidates to appeal to many voter  
groups although IRV puts more weight on the first preferences.

Juho

>
> So an important caveat to the assertion that "if a majority prefers  
> Brad
> over Carter, this preference exists whether the voting system does
> anything with it"...is that the voting method in use will affect  
> candidate
> behavior, and thus may result in Brad NOT being actually preferred  
> over
> Carter under a different voting method. In other words, voter  
> preferences
> among candidates (whether scores, or rankings) are not actually  
> "given,"
> as we assume when working out models (A>B>C), but will change  
> depending on
> what campaign style is rewarded by the voting method in use.
>
> Terry Bouricius
>
>
> ----- Original Message -----
> From: "Aaron Armitage" <eutychus_slept at yahoo.com>
> To: <election-methods at lists.electorama.com>
> Sent: Sunday, July 27, 2008 2:45 PM
> Subject: Re: [Election-Methods] RELEASE: Instant Runoff Voting
>
>
> --- On Sun, 7/27/08, James Gilmour <jgilmour at globalnet.co.uk> wrote:
>
>>
>> That is a consequence of your interpretation of how the
>> voting system is supposed to work and what the voting
>> system is supposed to
>> be doing.  But that's not what IRV is about.  As I said
>> in the previous message, the origins of IRV are in the
>> Exhaustive Ballot,
>> and in the Exhaustive Ballot there is no possibility of
>> looking "at the entire ballot".  IRV is not about
>> satisfying a set of
>> criteria derived from social choice philosophy.
>>
>
> Of course every reason you might offer for choosing one system over
> another is based on an idea of what a reasonable decision rule for  
> making
> collective decisions in very large groups should look like. This is  
> true
> for IRV advocate no less than advocates for other systems; where the
> system came from is beside the point, especially since most  
> jurisdictions
> have never used the Exhaustive Ballot.
>
> In an important respect, Condorcet is more natural than IRV: if a  
> majority
> prefers Brad over Carter, this preference exists whether the voting  
> system
> does anything with it, or even elicits enough information to determine
> that it exists. Condorcet simply discovers and applies this  
> preference.
> IRV, on the other hand, elicits enough information to discover it  
> exists,
> but may decide to ignore it based purely on procedural grounds.  
> There are
> no good reasons for this, ever. "Core support" is a bogus reason:  
> every
> time IRV chooses someone other than the plurality winner you're  
> letting an
> overall majority trump a comparison of core supporters. But other  
> times
> IRV will fail to do this, for reasons that simply don't exist apart  
> from
> the system itself.
>
>
>
> ----
> Election-Methods mailing list - see http://electorama.com/em for  
> list info
>
> ----
> Election-Methods mailing list - see http://electorama.com/em for  
> list info

___________________________________________________________ 
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine 
http://uk.docs.yahoo.com/nowyoucan.html