The Alaskan Top Four Model & iEBs

Jack Waugh

@gregw said in The Alaskan Top Four Model & iEBs:

small but real governments

Smaller than the State of Rhode Island And Providence Plantatations?

GregW

I was thinking in terms of a small municipality, but if Rhode Island wants to go for it, great!

Toby Pereira

Before I look into the rest of the stuff in this thread, I notice that RRV has been mentioned a few times. This is a score conversion of Thiele's Proportional Approval Voting, or PAV. However, I do not think it is the best conversion. The method known as Sequential Proportional Score Voting (SPSV) has better criterion compliance (it passes the additive and multiplicative versions of scale invariance*), and is no more complex.

*Basically if you multiply all scores by a constant and/or add a constant to all scores, the result remains unchanged.

Jack Waugh

@toby-pereira said in The Alaskan Top Four Model & iEBs:

score conversion

might not be necessary. Approval ballots might suffice. So, everywhere that I mentioned RRV (out of admitted ignorance), consider PAV instead.

Jack Waugh

But now, I see that PAV is hard to calculate. So, I don't know. I'm just suggesting that rather than top-four or top-five, maybe some calculation can be used so as to reduce the number of likely clones that make it through. Of course, one candidate from each clone family should be permitted, with sufficient support.

Toby Pereira

I've read through this thread a bit more now. So we first want an election (with potentially quite a lot of candidates standing), where we select four candidates to go into round 2 for the election to determine the winner.

If there might be a lot of candidates standing in round 1, then to keep it simple, I would definitely recommend some form approval voting. However, under normal approval voting there is the possibility of the top four being blocked out by clones, so it might make sense to use a sequential proportional method. SPAV, the sequential version of PAV, I think would be ideal for this. It shouldn't be hard to calculate when done sequentially (in response to Jack).

PAV-based methods often aren't recommended for proportional elections because PAV isn't exactly proportional and has some failure cases. However, I don't see that as a problem here. This isn't meant to be an exactly proportional election. It's just to prevent multiple clones appearing on the final ballot. (S)PAV has better monotonicity properties than most other proportional methods - e.g. Phragmén or quota-based methods. SPAV also has a simple algorithm. Phragmén wouldn't be awful for this either, but I wouldn't recommend a method that deals with quotas.

Then for the final ballot, it's a case of what's your favourite single-winner method. And that's the big debate.

Lime

@gregw said in The Alaskan Top Four Model & iEBs:

Thank you for the help! Especially for telling me about RRV. Much better STV. I think I might be able to simplify the explanation.

You're welcome! You can find other proportional cardinal systems by looking up
approval-based committee voting. (The extension to score voting is usually very simple.)

I have been working on a page that explains a MMP system that elects representatives from single member districts with a pretty accurate proportionality and a minimum number of at-large seats. The downsides of course are not easy to explain and the legislature would have a different number of seats after each election. But people, especially incumbents, do like their single member districts. Should I post it on this forum when I am finished?

This would be great!

I'd suggest taking a look at biproportional representation as well; it's the current state-of-the-art for proportionally allocating seats between parties. It fixes most of the issues associated with mixed-member proportional as well: voters choose candidates instead of parties, all candidates have their own district(s) instead of having 2 separate classes, etc.

I do like Ranked Robin but it will be a bit of a challenge to explain. It needs a video and a name change to Round Robin, I saw that somewhere I think they are the same, yes? Should not let it be confused with Ranked Choice Voting.

Would be nice to see a better explanation for sure! The traditional name is "Condorcet methods", but really they should just be called "majority vote" (since the candidate with a majority over every other wins).

Would you say that the strategy concerns with Score are not so serious that an Automatic Runoff is needed?

Probably not, because strategic score ends up being equivalent to approval (which is already well-regarded as a system).

I did notice a chart that said Score does not have a clone problem but STAR does. Is that correct?

In theory, yes. In practice, STAR is interesting because it intentionally fails clone-independence to encourage a better outcome. The optimal strategy for STAR is to have every party run at least two candidates, so they can lock up both spots in the runoff. This gives each voter multiple choices they can feel comfortable with.

As an example, I'd prefer a situation where both Biden and Kamala Harris were listed separately on the ballot so I could rank Harris higher (and help her win the runoff). Right now I'm not very happy with any of the candidates in the race. On a simple left-right scale I'm close to Biden, but I would much prefer to vote for someone younger. With STAR, every voter has at least two choices they think are tolerable.

Generally I think of STAR as just reversing the primary-then-general order: we have a general election to choose the best party (the score round), and then a "primary" where we pick the best nominee by a simple majority.

Will the Alaskan Top Four model produce a lot of clones?

Not sure what you mean. The Alaska top-4 model has a disadvantage of failing clone-independence, because clones split the vote in the first round. (Which kinda sucks, because one of the few advantages of IRV over FPP is it's immune to clones in theory.)

GregW

@toby-pereira
Great job of reading my mind, I was thinking about Approval leading to block voting last night.

SPAV is very interesting, the KP transformation is ingenious.

I am trying to figure it out, towards that goal I visited Sequential proportional approval voting

and downloaded
The Excess Method: A Multiwinner Approval Voting Procedure to Allocate Wasted Votes

Don’t be surprised If I have more questions

I should clarify what I am trying to do.
I want to propose an alternative to the nonpartisan Top four systems currently being promoted by the friends of independent candidates. This was also inspired by a 17 candidate mayoral election I recently voted in.

It would work like this:

A state run party primary with four options for each party: open, semi-open, closed, or do it yourself.
If there eight or more candidates a round to narrow the field to five candidates.
A final round with a good voting system.

GregW

@lime
I do appreciate all your comments.

I'd suggest taking a look at biproportional representation as well; it's the current state-of-the-art for proportionally allocating seats between parties. It fixes most of the issues associated with mixed-member proportional as well: voters choose candidates instead of parties, all candidates have their own district(s) instead of having 2 separate classes, etc.

I have looked at biproportional representation, I think it would be great, but I don’t think I can sell it. It would be beyond my powers of creating simple, concise explanations.

Round Robin Majority Vote might be a good name. Descriptive, all four words are friendly and/or positive. RRMV is a pretty good acronym.

In theory, yes. In practice, STAR is interesting because it intentionally fails clone-independence to encourage a better outcome. The optimal strategy for STAR is to have every party run at least two candidates, so they can lock up both spots in the runoff. This gives each voter multiple choices they can feel comfortable with.
As an example, I'd prefer a situation where both Biden and Kamala Harris were listed separately on the ballot so I could rank Harris higher (and help her win the runoff). Right now I'm not very happy with any of the candidates in the race. On a simple left-right scale I'm close to Biden, but I would much prefer to vote for someone younger. With STAR, every voter has at least two choices they think are tolerable.
Generally I think of STAR as just reversing the primary-then-general order: we have a general election to choose the best party (the score round), and then a "primary" where we pick the best nominee by a simple majority.

Would California be wise to replace their two round top two system with STAR?

Does STAR work well in elections with many candidates? Like the recent 17 candidate Denver mayoral election?

Will the Alaskan Top Four model produce a lot of clones?

Not sure what you mean. The Alaska top-4 model has a disadvantage of failing clone-independence, because clones split the vote in the first round. (Which kinda sucks, because one of the few advantages of IRV over FPP is it's immune to clones in theory.)

I am trying to decide what would be a good voting method to use in the first round of a two round system, with the first round having 8 or more candidates. The big issue is how to deal with clones.

Lime

@gregw said in The Alaskan Top Four Model & iEBs:

Would California be wise to replace their two round top two system with STAR?
Does STAR work well in elections with many candidates? Like the recent 17 candidate Denver mayoral election?

Yes and yes. STAR has no vote-splitting effect, and it actually works better when there are clones.

@gregw said in The Alaskan Top Four Model & iEBs:

I am trying to decide what would be a good voting method to use in the first round of a two round system, with the first round having 8 or more candidates. The big issue is how to deal with clones.

Then the best system is probably some kind of proportional representation algorithm. That said, I'm not sure why having separate primaries is necessary (most countries and don't have anything like that). I think it would be better to just set a requirement for a fixed number of signatures, donors, and/or endorsement(s) from registered parties to get on the ballot.

(Honestly, I think the main reason IRV uses a 4-candidate primary is that failures like nonmonotonic elections become more common the more candidates you have. Using a top-4 primary makes it easy to sweep those failures under the rug.)

GregW

@lime

I'm not sure why having separate primaries is necessary (most countries and don't have anything like that)

You are right, primaries are not necessary, however they are traditional.

I also have practical reasons:
Eliminating primaries would generate opposition and negative publicity.
By giving political parties primary options, (closed, semi-open, open & do it yourself) I hope that parties will be more receptive to my proposal, at least in comparison to the non-partisan proposals.
Primaries serve to bring public attention and participation to parties' nominating processes. This could be a big help for third-parties if there are reasonably low barriers to primary participation.

Parties are inevitable in a democratic system, and serve important purposes. We need competition between more strong parties.

(Honestly, I think the main reason IRV uses a 4-candidate primary is that failures like nonmonotonic elections become more common the more candidates you have. Using a top-4 primary makes it easy to sweep those failures under the rug.)

I bet you are right.

GregW

@toby-pereira

Both SPAV Sequential proportional approval voting and the Excess Method
The Excess Method: A Multiwinner Approval Voting Procedure to Allocate Wasted Votes
would work well in narrow down the field in large elections.

SPAV is a lot easier to count and explain and should work quite well.

The Excess Method adds calculations after the election of each candidate to distribute the seat winning candidate's excess approvals. This will, at least in theory, better represent the will of the voters. With the Excess Method, voters could approve all the candidates they like with no concern for strategy or temptation to tactical vote.

A difficult decision. Do you have a dog in this race?

Jack Waugh

@gregw said in The Alaskan Top Four Model & iEBs:

voters could approve all the candidates they like with no concern for strategy or temptation to tactical vote.

https://en.wikipedia.org/wiki/Gibbard's_theorem#:~:text=Gibbard's theorem states that a,to two possible options only.

GregW

@jack-waugh
First, let me thank (in order of appearance) Lime, cfrank, Jack Waugh, and Toby Pereira for all the help!

In my last post I made the statement regard the Excess method of allocating wasted votes:
Voters could approve all the candidates they like with no concern for strategy or temptation to tactical vote.

Jack Waugh suggested that I consider Gibbard's theorem.

Perhaps I was too enthusiastic in my praise of the Excess method. Or perhaps I am too generous as to what I consider to be good strategy resistance for a voting system.

I am impressed if a voting system has a voter strategy that that is reasonable and simple enough that all or most voters will be inclined to follow, and if they follow it the system produces fair results or if the calculation of a successful strategy for individual voters or organized voters is so difficult voters will resort to voting their true preferences.

I admit both, especially the latter, leave open the possibility of strategic regret.

Still, I think that the Excess method, though cumbersome, would produce fair results.

Latest news from Colorado: (where I live) Top Four proponents have over the past few months have filed 41 ballot initiative proposals; one remains # 188. The title has been set so only three hurdles to the ballot remain:

Approval of the petition format, a procedural issue.

Signature collection. This is not the first rodeo for the chief funder, Kent Thiry, former CEO of the Devita kidney dialysis company. He has backed successful Colorado ballot initiatives before, including one that established semi-open primaries in Colorado. Yes, he is trying to redo his own work.

Possible legal challenges.

Many of the filings were rejected for violating the one subject rule, however #188 includes both rounds.

What I plan to propose could be considered three separate subjects:
A state-run primary with four options for parties: open, semi-open, closed and do it yourself.
A first-round election if there are eight or more candidates.
A better voting system for the general election.

If #188 passes I might talk to state legislators about my ideas. If Kent Thiry can redo Kent Thiry's work, why shouldn't I be able to redo Kent Thiry's work? Of course, I'm not Kent Thiry, nor do I have his money.

GregW

@gregw

In a previous post I said:
I should clarify what I am trying to do.
I want to propose an alternative to the nonpartisan Top four systems currently being promoted by the friends of independent candidates. This was also inspired by a 17-candidate mayoral election I recently voted in.

It would work like this:

1, A state run party primary with four options for each party: open, semi-open, closed, or do it yourself.
2, If there eight or more candidates a round to narrow the field to five candidates.
3, A final round with a good voting system.

Concerning the purpose of Round 1
(After the primary and before the general election. Does Round 1 need a spiffy, descriptive name? Open to suggestions.)

Round 1 serves to narrow the field and give a publicity platform for third-party and unaffiliated candidates.

By advancing past Round 1 into the general election a third party or unaffiliated candidate gains political credibility and publicity. Very much like a 15th seed beating a 2 seed in in the NCAA basketball tournament.

The party friendly state run primaries, hopefully with low barriers to entry, also serve to generate publicity for third parties.

I am more interested (I have my prejudices) in promoting third parties, I suspect the leaders or the Top Four movement are more interested in promoting independent candidates.

I know all this would be expensive to the taxpayers. I believe that a fair and effective election system is worth the money, much like a good accounting system is worth the cost to a business.

Governments have a great deal of power and money; we need to put the voters in charge even if it is expensive.

Lime

@gregw said in The Alaskan Top Four Model & iEBs:

Eliminating primaries would generate opposition and negative publicity.

Would it? I feel like people complain about the low turnout and cost of primaries all the time. I think it might actually be a big advantage.

@gregw said in The Alaskan Top Four Model & iEBs:

Parties are inevitable in a democratic system, and serve important purposes. We need competition between more strong parties.

I definitely don't disagree! But typically other countries just handle this with some kind of election system that's (weakly) independent of clones, so everyone can run (including multiple candidates from the same party or faction).

Toby Pereira

@gregw said in The Alaskan Top Four Model & iEBs:

@toby-pereira

Both SPAV Sequential proportional approval voting and the Excess Method
The Excess Method: A Multiwinner Approval Voting Procedure to Allocate Wasted Votes
would work well in narrow down the field in large elections.

SPAV is a lot easier to count and explain and should work quite well.

The Excess Method adds calculations after the election of each candidate to distribute the seat winning candidate's excess approvals. This will, at least in theory, better represent the will of the voters. With the Excess Method, voters could approve all the candidates they like with no concern for strategy or temptation to tactical vote.

A difficult decision. Do you have a dog in this race?

I'm not sure in practice how the results of the Excess Method would differ from SPAV, but I think its complexity counts against it. And considering this is just to narrow down the field, rather than get the best proportional result, I don't think it's probably necessary.

Jack Waugh

@gregw said in The Alaskan Top Four Model & iEBs:

Governments have a great deal of power and money; we need to put the voters in charge even if it is expensive.

Agreeing in spades. Thank you for your activism.

GregW

@toby-pereira said in The Alaskan Top Four Model & iEBs:

I'm not sure in practice how the results of the Excess Method would differ from SPAV, but I think its complexity counts against it. And considering this is just to narrow down the field, rather than get the best proportional result, I don't think it's probably necessary.

Thank you for the advice, your point is well taken.

For a Round 1 vote, I am leaning to a Webster version of SPAV and STAR for Round 2, the general election.

I saw that SPAV and STAR are listed as failing Participation on the Comparison of Electoral Systems wikipedia page.
Could those failures be a problem in these two applications?
Am I paying too much attention to the criterion lists?

GregW

@lime said in The Alaskan Top Four Model & iEBs

I feel like people complain about the low turnout and cost of primaries all the time. I think it might actually be a big advantage.

(regarding the elimination of primaries in a two round voting system generating helpful or harmful publicity)

You might be right, we should see how this goes in Top Four ballot initiative votes. There are plenty of folks who are tired of primaries, and primaries get a lot of blame these days. On the other side, political parties and their supporters may howl.

I want to give parties, especially third-parties, opportunities and options for how they choose their members and how they gain publicity.