Add-On Multi-Winner Proportional Representation


Which is why I prefer Asset Voting, and why all PR discussions that I am in have to fall back to Asset at some point. Asset is extremely simple. Asset is party agnostic and melts 2PD. Asset does not have weird pathologies like participation failure or anti-proportionality. Asset is precinctable. Asset makes gerrymandering less potent.

But it only really works if there are enough candidates from each party that there will be enough who are not corrupt evil dishonest politicians that the number of CEDPs will be reduced dramatically.


So I appear to need to respond to:

and also to:

First, you may note that I do not prefer to speak in terms of “two-part domination (2PD),” but rather in terms of “elite party capture.” “Elite” overlords can easily capture three parties, or twenty of them. If you have twenty “elite” captured parties you are certainly no better off that if you had only two “elite” captured parties (“2PD”).

It is not at all clear to me that asset voting would effectively disrupt a system with two “elite” captured parties, let alone twenty of them. But I think you could easily use strategic hedge simple score voting to create some form of asset voting procedure, essentially by setting the stronger winners in some great hall to make bargains.

All large political organizations become infiltrated by other adversarial organizations. Highly structured organizations, such as political parties and lobbies are usually easy to infiltrate; less structured systems such as affinity groups, or especially small circles of say three to seven individuals, can be far more difficult to effectively infiltrate. But still, that happens too.

There are different kinds of infiltration. For espiotive infiltration you just send in an “agent” to spy on, and suss out the strategies, etc. of an adversarial organization. This is normal and legitimate so long as you are not purloining state secrets or trade secrets. I have a small circle now that has an espiotive infiltrator hidden within a pro-RCV organization. Then there is disruptive infiltration, where agents are sent in to sow discord and advocate illegitimate actions. This is dirty and low-down, so I would not be involved in it. The worst kind of infiltration is co-optive infiltration, where you send in agents to “take over” an adversary. An FBI source once stated that at one point over 50% of the membership of the U.S. Communist Party was comprised of FBI agents. Of course, all major political parties (including the Greens) have been co-optively infiltrated.

Truly proportional representation is completely useless if relatively large, thus co-optively infiltrated, groups are being “represented.”

Please note that Europe is littered with all sorts of proportional representation, yet except for Russia and Switzerland, governments there are universally despised by their citizens. They don’t even have any democracy at all of late, due to the hegemony of the European Union.

So I hold little hope for truly proportional representation, and would settle for quasi-proportional, add-on multi-winner election with tranche distribution of the winners. I suspect that asset voting would devolve into a melting pot denouement, while tranche distribution would produce a few colorful gadflies, and I would prefer the latter.

At this point, I am sure that RCV is riddled with so many pathologies that it is worse than single selection. And I am convinced that approval voting will do nothing but disappoint, due to the double bind quandary:

So that just leaves simple score. Part of the “simple” is that there is no “averaging” and there exists no zero vote, only an abstentive vote whereby the voter can simply ignore an “on-the-ballot” candidate. A zero vote leaves too much power to the ballot publishers and election officials, who could always be bribed to declare entire ballots on which some candidate has not been scored to be “spoiled.”

Better yet, why not just leave the decision about the voting method up to the voters themselves? Every year they should get to vote on which method they would prefer to use in the next year.


But there are many other PR systems that introduce score ballots without the kinds of flaws that I’ve pointed out in your system. Why promote this, instead of RRV, or the kinds of methods Parker Friedland mentioned?

Also, the approval quandary has a solution. See:


About elites, if a party with enough grassroots support can rise up against the ruling class in a system that has no spoilers, then as long as the elites are not bond fidelity cheating, that party can eventually (assuming voters are tired of the elites) gain enough power through the nursery effect.

Asset would just make it so leftover votes on the anti-elites go to the least evil elite, but if this is happening then we just have not made enough progress yet.

Better yet, if there are several anti-elite parties, then nearly every voter will have some anti-elite party to support and perhaps the elites will fall faster.

Other things like education and an objective news media will help prevent demagogues and fascists from abusing this system.

I tried to write this from the perspective of both defending Asset Voting, and also defending Score Voting (someone might argue that the nursery effect in Score can help demagogues and fascists).


Reweighted range voting (RRV) is far too complicated and, realistically, must be centrally counted.

And there is just one minor drawback, which is not a “flaw in your” (“my”) system. All that is attempted to be done is to provide a small voice, even if it is a somewhat random one, for individuals who have less votes to contribute.

This is certainly no solution. Your alleged solution involves what you should recognize as an absurd strategy for so-called “honest” voters (rectitudinous voters) based on their determining an approvable / unapprovable threshold for candidate approval, and then giving maximum votes to those above it and minimum (or no) votes to those below it. This is really no strategy at all. The only effective strategy is to give maximum votes to truly desired candidates, and somewhat less than maximum votes to lesser evil candidates. This is the strategic hedge strategy, and it maximally reduces the chances for the greater evil candidates, while providing reasonable opportunities for the most-desired candidates to actually win.

Abstract mathematical schemes based on the Condorcet benchmark have proved to be basically useless. What we must strive for is victory for the common everyday voters, and defeat of the “elite” rulers. We must keep everything a simple as possible.

Shortest Splitline Thread

Any such grassroots party would quickly be infiltrated and destroyed or co-opted by “elite” overlords. The people must rise amorphously, and the election system must respond effectively to such amorphous movements.


rkjoyce quoted someone as saying:

Approval voting, whereby voters can grant just one (1) vote to as many candidates as they approve of suffers from a “double bind quandary”. Suppose the voters using the approval method are presented with three candidates: two of whom are ruling establishment supported (presumably a Republican and a Democrat), and one of which is truly desired. The voters will face a quandary. Should they vote for both the desired and the lesser-evil ruling establishment candidate, or just for one or the other of these two, so as to contribute to the optimal, or at least lesser evil outcome? I presume that for most voters this is simply undecidable. Perhaps this should be called a “triple bind quandary”, and it could result in strongly desired, but non-ruling establishment candidates almost never becoming elected. A simple score/range method could effectively thwart this quandary.

I believe the right strategy is to approve your favorite candidates definitely and the lesser evil on a random basis. When many voters do that, the effect is the same as finer-grained Score.


Add this to the new multi-winner election theory category?