Re: [EM] STV+AV.

[Bryan Mills]

>
> > Why STV? The original poster wanted elected representatives to have votes
> > proportional to their electoral support yes? There's no need for
> fractional
> > transfers from elected candidates then.
> > >
> >
> > IRV is a form of STV, but it's not my favorite.  Some of the other STV
> > methods (e.g. Schulze-STV and CPO-STV) tend to produce better
> eliminations.
> >
> > But the question of why not STV is a good one.  Several reasons.
> >
> > STV requires much more work on the part of the voter - ranking all the
> way
> > down to a candidate likely to be elected, instead of just one.  That
> > probably means a much larger ballot and/or an arbitrary cutoff between
> > ballot-candidates and write-in candidates.
> >
> dlw: If the number of possible rankings is the number of seats + 2 then
> it's not too bad.  And nobody would be forced to rank umpteen candidates,
> so the low-info voters could just vote for their favorite candidate.
>

The number of possible rankings is quite a lot larger than S+2.  Even if
you don't transfer votes from elected candidates, there are still C-S
candidates eliminated -- so you'd have (C choose C-S-1)*(C-S-1)!
distinguishable rankings, and even more if you allow equal rankings.  The
only way out seems to be to pre-filter the set of candidates, so you
basically have to drop to approval voting at some point --
candidate-registration petitions and the like -- and then we're back to an
arbitrary cutoff.


Partial rankings might be workable in a weighted-seat STV variant, though.
 If a vote only transfers in case of elimination (and not in case of
surplus), one would only need to rank candidates down to the first
candidate sufficiently likely to be elected, and you could split the ballot
into manageable chunks by party.  Determining a suitable cutoff candidate
still has a cognitive cost, but it probably wouldn't be that bad in
practice.

But if we assume that partial rankings are effective, there's still the
strategy/computation tradeoff to deal with: allowing truncated ballots
still doesn't help with favorite-betrayal, and STV variants less
susceptible to favorite-betrayal are also less susceptible to efficient
counting.


> The STV variants that are less strategy-prone are computationally
> > inefficient, and even those are not strategy-free.
> >
> > And perhaps most importantly, the more resistant an STV method is to
> > strategy, the more complicated it is to explain and understand.
> >
> > As deterministic methods go, I do like STV methods; but DS fixes a lot of
> > the worries I have about them.
> >
> One could also apply the same sort of approach to simplifying STV with the
> initial treatment of all of the rankings as approval votes to get the
> number of candidates down to N+2, where N is the number of seats.
> As with IRV, it's easier to explain STV when there's relatively few
> candidates to eliminate.  And, it'll mitigate the strategy effects, which
> have to be examined more closely.


The initial treatment of rankings as approval votes introduces some other
problems, though.

With an explicit "approval threshold" in the ranking, it induces a
substantial cognitive cost on the voter (determining the approval threshold
strategically).
With an implicit "first-preference" approval, it has the same problem as
traditional STV (i.e. IRV), namely of unduly rewarding favorite-betrayal.
With an implicit "all-ranked" approval, the overall system would likely
violate later-no-harm with much higher frequency; by expressing a
preference between two dispreferred candidates one might unintentionally
put the higher of the two in contention.


It may well be that these issues are all less severe than in the
deterministic alternatives to STV, but I still think they're enough to
merit consideration of nondeterministic alternatives.

Election-Methods mailing list - see http://electorama.com/em for list info

[EM] SODA criteria

[Jameson Quinn]

SODA passes:

Majority
MMC (as voted)
Condorcet (as voted, and in a strong Nash equilibrium as honest)
Condorcet loser (ditto)
Monotone
Participation (with the fix that delegation can be any fraction)
IIA (delegated version - that is, if a new candidate is added, the winner
is either the same, or someone higher on the new candidate's delegation
order.)
Cloneproof
Polytime (there is no guarantee that optimal delegated assignment strategy
is polytime calculable, but it will be in any real case, and anyway,
candidates can just choose some near-optimal strategy.)
Resolvable
Summable
Allows equal rankings
FBC

So, of the criteria in the wikipedia voting systems
table,
the only ones it out-and-out fails are:
Consistency (though it comes damn close)
Later-no-harm and later-no-help (though it does satisfy LNHarm for the one
(two) candidate(s?) with the most voted approvals, and for other
candidates, adding later preferences is probably strategically forced; so
I'd say it fulfills the spirit of both of these. Similarly, it satisfies
LNHelp for the last-to-delegate candidate, and nearly so for other
late-delegating candidates, and the point of LNHelp is to prevent a weak
candidate from winning through clever bottom filling, so again it satisfies
the spirit.)
Allows later preferences (though delegation substitutes for this affordance
in some cases.)

If we could just get some wikipedia-notable mention of SODA, we could put
it in the table, and I think it would graphically stand out as the most
criteria-compliant method there.

I'm working on an academic article on SODA, which would not be focused on
these criteria or even on SODA, but would quickly state the above. But if
anyone can make an article happen in a wikipedia "reliable source", that
would be great.

Jameson

Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] STV+AV.

[robert bristow-johnson]

On 1/31/12 3:23 PM, David L Wetzell wrote:


STV requires much more work on the part of the voter - ranking all
the way down to a candidate likely to be elected, instead of just
one.  That probably means a much larger ballot and/or an arbitrary
cutoff between ballot-candidates and write-in candidates.



this is *not* just STV.  it is ranked-choice voting (i would use the 
acronym RCV, but i think there were some places out west that they 
attached the label "RCV" to what we have been commonly calling IRV.  and 
i don't think it requires any more work on the part of the voter than 
that of deciding whom the voter would vote for in a runoff if their 1st 
choice was not in the runoff.  but it requires the voter to make up 
his/her mind about the candidates by Election Day, and i have never 
understood why that is such a burden (but the voting-reform opponents 
here in Burlington surely bitch about that).


the ballot meaning and rules are exactly the same, with one exception.  
for some ranked-choice methods, such as Condorcet, there is no 
restriction against equal ranking of any candidates.  for STV, equal 
ranking is a problem because we don't know how to promote (or 
"transfer") the pair of equally-ranked votes if some higher-ranked vote 
is eliminated.


 In the FPTP case, it trims the ability of dissenters to move the de 
facto center towards the true center.  In the IRV case, it does the 
opposite, it penalizes the major parties when they do not move enough 
towards the true center.
again, that was not the case in Burlington.  the center party was 
"squeezed" (as Jameson would say).  the candidate in the center received 
nearly all of the 2nd-choice votes from voters who ranked one of either 
the left or right wing candidates 1st.  it was relatively rare that the 
left-wing voter ranked the right-wing candidate as their 2nd choice and 
also vise versa.  but the center candidate did not benefit from that 
because IRV is opaque to your 2nd choice if your 1st choice has not been 
eliminated.  but, under Condorcet-compliant rules, the center candidate 
would have benefited greatly (and would be elected), so it can be said 
that Condorcet tends to favor the center candidate more (than either IRV 
or FPTP) whereas IRV tends to favor the largest subgroup (i.e. the 
Progs, in Burlington in 2009) of the majority group (liberals).  and, we 
know, that FPTP gives the minority candidate the best chance they have 
of winning (they need a 3rd-party or 3rd independent candidate to draw 
votes away from what would be their majority opponent if the spoiler was 
not there).


Most rational choice models implicit here take as fixed the position 
of candidates/parties on the spectrum, when in real life, this can be 
changed somewhat.  This reduces the "badness" of strategic voting.  It 
becomes less important thereby to devise an election rule that doesn't 
give any incentive to anyone to vote strategically.


the reason why i have never agreed with that is because people *resent* 
being saddled with the burden of voting tactically and particularly 
resent finding out ex-post facto that their sincere vote served their 
political interests more poorly than the tactical vote (the most common 
tactic is "compromising").  that resentment has consequences, one of 
which is a cloud hanging over the elected candidate as not being 
entirely "legitimate", not being the "true" choice of the electorate.  
but the worse consequence is that of holding back what would otherwise 
be viable independent or 3rd party candidates, sometimes leaving the 
voters with a choice between Dumb and Dumber.  that is the *main* evil 
we're trying to avoid with voting system and ballot reform.


we are *now* experiencing some of these consequences in Burlington.  the 
Progs have decided not to run a candidate (the current mayor is or was a 
Prog and is not running for re-election).  it looked for a while that 
there would be only two (Dem and GOP), but recently an independent 
candidate emerged and her political appeal is a lot like a Prog 
candidate (the Progs are not ashamed of sticking up for the poor and 
powerless whereas, ever since Reagan, Democrats have modified their 
rhetoric to be for "the middle class" so as not to sound "socialistic" 
or too "liberal", both were bad words and continue to be used 
disparagingly in American politics).  so we are going to have an 
interesting test case for the election coming up March 6.  we might very 
well get an elected candidate with 41% of the vote.


now the Progressive party in Vermont is declining *rapidly*.  Burlington 
is the most populous town in the state (but Vermont has the the least 
populous largest city of all 50 states) with a population of about 
42000.  we have about 9000 voters in a mayoral election.  the Democratic 
caucus had over 1000 valid voters showing up.  the Prog caucus (which i 
attended as an observer, so also did another EM lister, who i just 
discovered has 

[EM] STV+AV.

[David L Wetzell]

> Why STV? The original poster wanted elected representatives to have votes
> proportional to their electoral support yes? There's no need for fractional
> transfers from elected candidates then.
> >
>
> IRV is a form of STV, but it's not my favorite.  Some of the other STV
> methods (e.g. Schulze-STV and CPO-STV) tend to produce better eliminations.
>
> But the question of why not STV is a good one.  Several reasons.
>
> STV requires much more work on the part of the voter - ranking all the way
> down to a candidate likely to be elected, instead of just one.  That
> probably means a much larger ballot and/or an arbitrary cutoff between
> ballot-candidates and write-in candidates.
>

dlw: If the number of possible rankings is the number of seats + 2 then
it's not too bad.  And nobody would be forced to rank umpteen candidates,
so the low-info voters could just vote for their favorite candidate.

> The STV variants that are less strategy-prone are computationally
> inefficient, and even those are not strategy-free.
>
> And perhaps most importantly, the more resistant an STV method is to
> strategy, the more complicated it is to explain and understand.
>
> As deterministic methods go, I do like STV methods; but DS fixes a lot of
> the worries I have about them.
>

One could also apply the same sort of approach to simplifying STV with the
initial treatment of all of the rankings as approval votes to get the
number of candidates down to N+2, where N is the number of seats.

As with IRV, it's easier to explain STV when there's relatively few
candidates to eliminate.  And, it'll mitigate the strategy effects, which
have to be examined more closely.  As I argued before, there's a diff
between making 3rd party dissenters vote strategically and making the
supporters of a major party out of touch with most voters vote
strategically.  The possibility that their voters get pushed to vote
strategically is an incentive for changes in candidate/party positions.  In
the FPTP case, it trims the ability of dissenters to move the de facto
center towards the true center.  In the IRV case, it does the opposite, it
penalizes the major parties when they do not move enough towards the true
center.

Most rational choice models implicit here take as fixed the position of
candidates/parties on the spectrum, when in real life, this can be changed
somewhat.  This reduces the "badness" of strategic voting.  It becomes less
important thereby to devise an election rule that doesn't give any
incentive to anyone to vote strategically.

dlw

>
>
> -- Forwarded message --
> From: Jameson Quinn 
> To: EM 
> Cc:
> Date: Tue, 31 Jan 2012 11:04:40 -0600
> Subject: [EM] SODA is monotonic. Earlier failure/fix was actually for
> participation, not monotononicity.
> A week or two ago, I sent a message to the list with a scenario which I
> claimed was an example of nonmonotonicity in SODA as defined; and mentioned
> a natural fix for this problem (allowing partial assignment of delegated
> votes).
>
> I was mistaken. It was not an example of nonmonotonicity, but rather an
> example of failure of the participation criterion. The rest of what I said,
> including the simple fix for the problem, still applies.
>
> Jameson
>
> ___
> Election-Methods mailing list
> Election-Methods@lists.electorama.com
> http://lists.electorama.com/listinfo.cgi/election-methods-electorama.com
>
>

Election-Methods mailing list - see http://electorama.com/em for list info

[EM] SODA is monotonic. Earlier failure/fix was actually for participation, not monotononicity.

[Jameson Quinn]

A week or two ago, I sent a message to the list with a scenario which I
claimed was an example of nonmonotonicity in SODA as defined; and mentioned
a natural fix for this problem (allowing partial assignment of delegated
votes).

I was mistaken. It was not an example of nonmonotonicity, but rather an
example of failure of the participation criterion. The rest of what I said,
including the simple fix for the problem, still applies.

Jameson

Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] Sortition and the Delegable Proxy system

[Bryan Mills]

On Jan 31, 2012 3:17 AM, "Clinton Mead"  wrote:
>
>
>
> On Tue, Jan 31, 2012 at 7:13 PM, Kristofer Munsterhjelm <
km_el...@lavabit.com> wrote:
>>
>> On 01/31/2012 07:14 AM, Clinton Mead wrote:
>>>
>>> Why not simply IRV until 500 candidates are left.
>>
>>
>> STV would probably be better - or if you want a weighted assembly,
continuous cumulative voting (which is like RV except every ballot's rating
is divided by the sum of the undivided ratings on that ballot).
>>
>
> Why STV? The original poster wanted elected representatives to have votes
proportional to their electoral support yes? There's no need for fractional
transfers from elected candidates then.
>

IRV is a form of STV, but it's not my favorite.  Some of the other STV
methods (e.g. Schulze-STV and CPO-STV) tend to produce better eliminations.

But the question of why not STV is a good one.  Several reasons.

STV requires much more work on the part of the voter - ranking all the way
down to a candidate likely to be elected, instead of just one.  That
probably means a much larger ballot and/or an arbitrary cutoff between
ballot-candidates and write-in candidates.

The STV variants that are less strategy-prone are computationally
inefficient, and even those are not strategy-free.

And perhaps most importantly, the more resistant an STV method is to
strategy, the more complicated it is to explain and understand.

As deterministic methods go, I do like STV methods; but DS fixes a lot of
the worries I have about them.

Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] Sortition and the Delegable Proxy system

[Clinton Mead]

On Tue, Jan 31, 2012 at 7:13 PM, Kristofer Munsterhjelm <
km_el...@lavabit.com> wrote:

> On 01/31/2012 07:14 AM, Clinton Mead wrote:
>
>> Why not simply IRV until 500 candidates are left.
>>
>
> STV would probably be better - or if you want a weighted assembly,
> continuous cumulative voting (which is like RV except every ballot's rating
> is divided by the sum of the undivided ratings on that ballot).
>
>
Why STV? The original poster wanted elected representatives to have votes
proportional to their electoral support yes? There's no need for fractional
transfers from elected candidates then.

Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] Sortition and the Delegable Proxy system

[Kristofer Munsterhjelm]

On 01/31/2012 07:14 AM, Clinton Mead wrote:

Why not simply IRV until 500 candidates are left.


STV would probably be better - or if you want a weighted assembly, 
continuous cumulative voting (which is like RV except every ballot's 
rating is divided by the sum of the undivided ratings on that ballot).



Wouldn't this produce a similar result without the randomness?


Then it would no longer be strategy-proof. And if it's not 
strategy-proof, you could use any conventional PR rule.


Another advantage of the randomness is that it's procedurally simple. 
Instead of doing transfers and re-transfers, you just pick a ballot from 
the heap and elect the candidate named on it.


With randomness, the voters don't have to know who's in the running, 
either, making the rule more sortition-like.



Election-Methods mailing list - see http://electorama.com/em for list info