[EM] A New Multi-winner (PR) Method

[Abd ul-Rahman Lomax]

I continue to be amazed that someone like Forest Simmons, who was an 
early writer about what later was called Asset Voting, as named by 
Warren Smith, and was simply a tweak on STV when invented by Charles 
Dodgson in the 1880s, which can produce *perfect* representation, 
beyond  mere "proportional representation," still works on complex 
single ballot deterministic methods that must compromise and prevent 
true /chosen/ representation in favor of some sort of theory of 
"optimised goodness."

Asset Voting can very simply produce unanimous election of seats, where 
the seat represents a quota of voters who have */unanimously/* 
consented, directly or, probably much more commonly in large-scale 
elections, through chosen "electors," I call them, who becomes a proxy 
for the anonymous voters, for the election of the seat. Asset allows the 
electors --- those who become public voters -- to cooperate and 
collaborate for the selection of seats.

The "Electoral College," fully and accurately, represents *all* the 
voters. But it might be very large, possibly too large to even meet 
directly. But all that is needed is a way for electors to communicate 
and to register their vote assignments for the creation of seats. They 
could use delegable proxy to advise them how to transfer.

Asset was actually tried once, and it produced a result that most would 
have considered impossible. 17 voters, five candidates for a three-seat 
steering committee. A rather sharp division, but the final result was 
the election of the three seats *with every vote represented directly or 
by proxy*. Before the third seat was elected, nevertheless two seats 
were elected, so the steering committee, if needed, could have made any 
decision by unanimous vote (two agreeing) through the representation of 
two-thirds of the electorate. It could, in theory, have decided to elect 
the third seat by using the Droop quota. (the quota had not actually 
been specified, and the one who called the election and created the 
process was not totally sophisticated on Asset, which does not need to 
specify an election deadline, it can, instead, leave that last seat or 
seats open, and consider the rest of the electorate as Robert's rules 
considers unrecognizable ballots: they count for determining "majority" 
but not as a vote for or against any candidate.

So, say, there is a jurisdiction with a million voters. It is decided 
that an optimal assembly would be 49, so the basis for a quota could be 
50, and thus the quota would be 2% of the votes cast. This allows that 
if all electors assign their votes in exact measure to seats, 50 seats 
could be elected, but that outcome is improbable. (But if it happens, 
whoopee!), so, normally, 49 seats might be elected. Instead of electing 
the last seat by plurality, depending on some deadline for vote 
assignments, I have suggesting leaving the election open. Further, those 
electors could be consulted by the Assembly. This is the power of having 
public voters who, collectively, represent the entire electorate.

Unanimous election of seats. Bayesian regret, zero. No losers. Minimal 
damage to the dregs. No expensive or extensive campaigning necessary. (I 
expect that the tradition would develop rapidly that one would only vote 
for people one could meet face-to-face, because how else can they truly 
represent? But this would be voluntary, not coerced. Basically, anyone 
who registers as an elector may participate further, the only 
requirement being a willingness to vote publicly (which is already 
required of elected representatives!)

The U.S. Electoral College was a brilliant invention, knee-capped by the 
party system. It did not, however, represent the people, but 
jurisdictions, specifically states. It could have been reformed to 
represent the people, but it went, instead, toward representing the 
party majority in each state, in effect, becoming a rubber stamp and 
creating warped results.

Asset is simple, close to tradition, but actually revolutionary, a 
dramatic shift from the entire concept of "elections" as contests. The 
voters would be 100% represented in the College, though voluntary 
choices, no need to consider "electability," hence no need for "voting 
strategy." Choose the available person, from a large universe, you most 
trust. As Dodgson noticed, that was relatively easy for ordinary voters, 
much easier than sensibly ranking many candidates. That complexity is 
completely unnecessary with Asset. No votes are wasted.

The only tweak I see as needed involves making it difficult to coerce 
votes, by making it impossible to know that a specific person did *not* 
vote for one. This would not be needed for small NGO elections. In 
public elections, I'd have a set of known candidates who received 
substantial votes in a prior election, or something like that, and when 
electors register, they would assign their own vote to another 
candidate. If they receive less than N votes, their vote would be 
transferred as directed. If they receive N votes, they become an 
elector. Electors would not vote in the election, so, if they receive N 
votes, they get their own declared vote back and so they have N+1 votes 
to transfer. N should be the minimum size necessary to ensure that they 
cannot know that a specific person did not vote for them. If they do not 
receive N votes, their received votes are privately added to the total 
for the candidate they chose when registering. So N might be two, but if 
it is a bit higher, it could provide increased security. 3 might be 
completely adequate, together with it being very illegal to coerce votes.

*No more original content below.*

On 4/10/2019 5:08 PM, Forest Simmons wrote:
> As near as I know the following PR method based on Range/Score style 
> ballots is new.
>
> This method is based on maximizing a measure of "goodness" of 
> representation to be specified later.  Slates of candidates are 
> nominated individually for consideration, because in general there are 
> too many possible slates to consider every one of them (due to 
> combinatorial explosion).  Among the nominated slates, the one with 
> the best measure of "goodness" of PR is elected.
>
> To reduce the abstraction, suppose that there are only 100 candidates 
> and that only five vacancies to be filled. Suppose further, that there 
> are  ten thousand ballots (one for each of ten thousand voters).
>
> Given a subset S of five candidates, we decide how good it is as follows:
>
> Order the set S according to their Range totals, so that the highest 
> to lowest score order is c1, c2, ...c5.  This order only comes into 
> play to determine the cyclic order of play as the candidates "choose 
> up teams" so to speak.
>
> Ballots are assigned to each of the candidates cyclically so that the 
> ballot most favorable to c1 goes to c1's pile, of the remaining the 
> one most favorable to c2, goes to c2's pile, etc. like the way we used 
> to choose teams when we were in grade school.
>
> (Eventually we'll get to how to automate judgment of favorability.  Be 
> patient)
>
> After 2000 times around the circle, each pile will contain exactly 
> 2000 ballots. (Thanks for your patience.)
>
> For our purposes the relative favorability of ballot V for candidate C 
> is the probability that V would elect C if it were drawn in a lottery; 
> i.e. V's rating of C divided by the sum of all of V's ratings for the 
> candidates in S including C.
>
> What happens when one of more of the candidates is not shown any 
> favorability by any of the remaining ballots?  The other candidates 
> continue augmenting their piles until they reach their quotas (two 
> thousand each in this case), and the remaining ballots are assigned by 
> comparing them to the official public ballots of the candidates whose 
> piles are not yet complete. (We won't worry about the details of that 
> for now.)
>
> For each candidate C in S add up all of the ratings over all of the 
> ballots in the pile, but not the ratings for candidates outside of S.  
> Divide this number by the total possible, which in this case is two 
> thousand times five or ten thousand.
>
> We now have five quotients, one for each candidate. Multiply these 
> five numbers together and take the fifth root. This geometric mean is 
> the "goodness" score for the slate.
>
> Among the nominated slates, elect the "best" one, i.e. the one with 
> the highest "goodness."
>
> It is easy to show that this method satisfies proportionality 
> requirements. And (I believe) it takes into account "out-of pile" 
> preferences as much as possible without destroying proportionality.
>
> No time for proofs or examples right now, but first, any questions 
> about the method?
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20190410/f6e2804c/attachment-0001.html>