[EM] IRV et al v. EPR
jameson.quinn at gmail.com
Mon Jul 16 11:09:02 PDT 2018
This system as described is very close to being an excellent "Proportional
Majority Judgment" method, but has one key flaw. When a candidate is
elected with more than 1 quota of support, all of their supporting ballots
are marked as used. To give a proportional method and to minimize strategic
incentives, only 1 quota of supporting ballots should be marked as used.
This could be done through some ordering criterion (highest support for
winner/lowest support for others), by proportionally reweighting ballots,
or by using up randomly-chosen ballots; the differences between these three
options would be relatively minor.
This reflects the basic way to transform any single-winner method into a
proportional multi-winner method: find single winners sequentially, and
then for each of those winners, "use up" the one quota of ballots that
"contributed most" to making that candidate the winner. There's room for
judgment calls in defining "contributed most", but other than that this is
a general template that IMO gives an optimal combination of good and
practical from methods as varied as IRV (which becomes STV), MJ, STAR,
Score, approval, Condorcet... in short, almost any single-winner method.
Are all good proportional methods examples of this pattern? I wouldn't go
that far. Delegation and (partial or full) biproportionality are extra
ingredients that can be added, giving possibilities like (Bavarian-style)
MMP, DMP, PLACE, and PAD, all worthy alternatives. But "sequential
allocation methods" using the recipe above should definitely be at least
one of the mainstays of the proportional toolkit.
2018-07-15 17:55 GMT-04:00 steve bosworth <stevebosworth at hotmail.com>:
> The recent responses to Sennet’s attempt favorably to report some recent
> successes of RCV (i.e. IRV) prompt me to hope that readers will be willing
> to test the counter claims of a newly developed voting and counting method
> for electing multi-winners. This method is fully described in the
> following published article:
> Legislatures Elected by Evaluative Proportional ...
> This article explains how a new voting method called Evaluative
> Proportional Representation (EPR) fully satisfies the demand that in the
> best representative democracy, no citizen’s vote would be involuntarily
> wasted, quantitatively or qualitatively. 2 EPR is intended for voters who
> are electing members of a legislative body, for example a ...
> This method is called Evaluative Proportional Representation (EPR). EPR
> builds upon the arguments for Majority Judgment (MJ)offered by Balinski and
> Laraki (2010, MIT). For example, when electing all the members of a city
> council, each EPR citizen is asked to evaluate (not rank) as many of the
> candidates in the city as she might wish, i.e. to grade each with regard to
> their fitness for the office: either EXCELLENT, VERY GOOD, GOOD,
> ACCEPTABLE, POOR, or REJECT. The article explains the relatively simple
> step by step method by which all these evaluations are counted by hand (or
> by the algorithm provided). As a result, each citizen’s one vote
> continues fully to count in the deliberations and decisions of the council.
> It does this through the weighted vote earned by one of the elected
> members, i.e. the one winner whom she has helped to elect and judges to be
> the one most qualified for the office.
> Unlike any known variety of plurality, ranking (Condorcet or IRV (RCV,
> STV, etc.), range, or approving voting methods, only EPR allows each
> citizen to guarantee that her one vote will continue to count in the
> council (or legislature). Also, I'd like to offer Abd ul-Rahman Lomax
> some hope: perhaps the extra appeal of this unique feature of EPR would
> prompt enough citizens in a state like California to adopt EPR by one of
> its referenda.
> Again, only EPR allows no vote to be wasted as defined by the second
> paragraph of the article.
> What do you think?
> I look forward to your feedback.
> Today's Topics:
> 1. Re: IRV / RCv advances (Abd ul-Rahman Lomax)
> Message: 1
> Date: Sun, 15 Jul 2018 14:08:24 -0400
> From: Abd ul-Rahman Lomax <abd at lomaxdesign.com>
> To: election-methods at lists.electorama.com
> Subject: Re: [EM] IRV / RCv advances
> Message-ID: <69bc286e-c970-9f6c-e60a-21e7f900f81c at lomaxdesign.com>
> Content-Type: text/plain; charset=windows-1252; format=flowed
> On 7/14/2018 10:08 AM, Richard Lung wrote:
> > [...] Agree completely about score voting. I can't help but feel
> > approval voting is essentially a rebranding of cumulative voting. In
> > about 1867, John Stuart Mill knew it was only a trifling improvement
> > on plurality counting but at least opened peoples minds to alternatives.
> I'd question that simply counting all the votes would be a "trifling
> improvement." It would have flipped the US presidential election in
> 2000, almost certainly, and is a no-cost improvement, simple to
> implement and easy to understand. In a two-round system, it can provide
> substantial flexibility, perhaps even allowing runoffs to have a third
> candidate. Counting all the votes would be, again, an improvement over
> IRV, allowing voters to equal-rank. But the IRV method, by discarding
> votes, is intrinsically flawed, and Bucklin actually worked when used.
> The arguments that it didn't work were based on the fact that it wasn't
> magic pixie dust. The only system that fixes about everything is Asset,
> which hardly gets any consideration at all. My suggestion has been for
> EM reformers to suggest Asset for NGOS, since it's really simple, as
> well, and creates a deliberative structure, which is far more flexible
> -- and functionally democratic -- than pure amalgamation. To be
> fail-safe, it could be used in a nomination process, to be actually
> ratified for a final result.
> Approval -- and most methods -- are plurality methods unless a true
> majority of the votes is required for a result. In Australia, in some
> places they accomplish that by making it illegal to not completely rank.
> In other words, to make a result more "democratic," coerce the voters.
> Ah, the things people do to preserve the way things are!
> > Re 3.
> > Weighted Condorcet pairing arguably offers a back-door that partly
> > gets round the Laplace criticism of Condorcet pairing, that it does
> > not establish the relative importance of higher and lower preferences,
> > in the over-all election count.
> > I see Condorcet pairing primarily as a research tool for
> > cross-referencing the results of an at-large election with the results
> > from sub-elections of one-to-one contests or less minimal partitions.
> > Ideally, we would have an election system that does not have to watch
> > its back for a Condorcet paradox.
> A condorcet paradox is an indication of an incomplete process. Depending
> on preference strength, which some Condorcet methods attempt to
> estimate, it might be meaningless. But to truly analyse election returns
> requires preference strength information. Borda, again, estimates it
> with an assumption of full ranking, such that with many candidates, the
> "rank distance" would approximate a measure of preference strength.
> But the only ballot that actually allows the voters to directly express
> preference strength is a score ballot. And then some advocates of other
> systems point out Condoret failure, as if that matters when preference
> strength is obviously low. It doesn't. And then it is pointed out that
> voters may vote "strategically," as if that is dishonest or bad. In a
> Score system, there is never any incentive to reverse preference. Voters
> decide what preferences matter to them, and will vote accordingly, and a
> good overall system will detect situations where is ambiguity, perhaps
> due to inaccurate perception of probable results, and will then set up a
> Instead of working together to create a system that will actually
> improve and foster full democracy, we don't, it seems, trust the people
> and want them to conform to our own ideas. In other words, same old same
> old. The problem with democracy is the damn people!
> But democracy is still government by consent, and whenever that fails,
> oppression is inevitable. It's only a question of how bad it gets. The
> logic that captured me, so many decades ago, is that we need
> representation by consent and choice, not by "winners" and "losers." And
> that appears to be doable. But who cares enough to try it?
> Some. Not yet enough, but it's possible any day. It might only take one
> person to make a proposal and carry it through, in one place.
> > Even an admittedly crude election like IRV (Alternative Vote),
> > according to this group, has only come-up with the Burlington case.
> > That may have been politically unfortunate. But, if about 150?
> > elections have not suffered the paradox, that incidence is not
> > statistically significant.
> It is possible to show that about one-third of IRV results were not
> optimal. This is done by comparing IRV results with top-two runoff
> results. If a full-information ballot were used with IRV (it could be
> done! Easily!) we would then know. However, actually collecting the data
> to determine, definitively, if a voting system is actually working is
> not a part of any reform proposal I have seen. The collecting of
> information is confused with and considered less important than creating
> a result, and I can easily see the counter-argument: what if this causes
> the result of an election to be called into question? Won't this damage
> our trust in government? I think I have actually seen that argument.
> If our trust in government is based in ignorance, it's worse than a bit
> of trouble. A simple, coarse-score ballot, with explicit approval
> cutoff, would be cheap and easy, and how the actual result is determined
> could be explicity declared on the ballot. Providing the additonal
> information would be optional. One could vote the ballot, if one
> chooses, as vote-for-one. It would all generate useful information, and
> could guide future election method decisions.
> > The real comparison is how many "Bush beats Gore minus Nader" contests
> > are there?
> In a two-party system. Duverger's law and party attempts to corner the
> electorate create many of them. Look at any close election and at the
> participation of minor parties in it. There may be hundreds of these a
> year. As well, the existing system tends to suppress minor party
> participation. New York has Fusion voting, a step in a direction of
> improved democracy. It was proposed for Massachusetts and lost. Who
> opposed it?
> The "Democratic" party, of course! It lost. So why did people vote
> against it? Well, perhaps they trusted their party. After all, isn't our
> party the Good Guys? If it's bad for them, surely we don't want it! It
> all makes sense until and unless one starts to look more closely, which
> most people don't do. And realizing that, Dodgson proposed Asset, which
> actually creates, very simply, representative government, with the good
> stuff associated with that, while allowing low-level decisions, by
> ordinary people who don't want their lives to be about politics, to be
> useful and effective.
> > And how many simple plurality elections make voters act as their own
> > returning officers in an implicit ranked choice election, where the
> > voter excludes his first preference for Nader, and counts it for
> > second preference Gore?
> My sense is that many would vote for a third party candidate if they did
> not know it would be a wasted vote. It could be trivial to fix that
> problem. But we obviously care about something else more. about what?
> Subject: Digest Footer
> Election-Methods mailing list
> Election-Methods at lists.electorama.com
> End of Election-Methods Digest, Vol 169, Issue 8
> Election-Methods mailing list - see http://electorama.com/em for list info
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Election-Methods