RCV And The Elite Party Capture Effect


#1

At the risk of violating the rule about: “Don’t cross-post the same thing in multiple topics,” I wish to reiterate the very short “preamble” to my previous article, which engendered numerous disagreements, and ended with rather bizarre comments concerning “Utopian candidates” (so I didn’t know how to continue):

“A Radically Different Approach To Score/Range Voting”

I began studying election methods during the 2004 pre-election season. I blogged on perhaps 50 sites, and may even have originated the term “spoiler effect.” My first attempt to cure the spoiler effect was “Consecutive Runoff Approval Voting”. See:
Consecutive Runoff Approval Voting.

Some of my “criteria” include:

<> There must be simple hand counted paper ballots, with absolutely no casting or counting automation, and any election method must be simple enough to tabulate to enable this. Machines that most voters perceive to be operated via “black magic” can never be trusted. Any proper election method should be simple enough to enable this.

<> Any proposed election method should thoroughly disrupt the spoiler effect, and thus disrupt the “two-party-system,” which essentially becomes a “one-party-system.”

<> It is necessary to take account of the fact that elections are not primarily contests between opposing candidates, or even opposing ideals. Rather they are contests between the common voters and a ruthless ruling establishment. This causes election methods to become very complicated when the support of so-called “honest voters” is assumed as a criteria. Everything will become much simpler if voters are expected to vote strategically. After all, the ruling establishment will always act strategically.

I at first began with the realization that “IRV” could never be implemented without the utilization of automated election devices (e.g. voting machines), whereas score/range, could. So in 2004 I began blogging relentlessly in opposition to “IRV.”

My opponents were “rabid” promoters of “IRV”, and they offered all sorts of spurious, contradictory, and sometimes blatantly dishonest reasons why “IRV” would disrupt the “two party system,” and I began to realize that election automation was not nearly the whole problem with “IRV,” which began to display numerous pathologies. At the end of this article I will illustrate one of them.

Of course, the most pervasive election method circa 2018 is single selection voting (usually absurdly called “plurality voting”). This dreadful method blatantly imposes a “spoiler effect.” This was vividly illustrated in the year 2000 presidential election when it was claimed that Nader spoiled the election for Gore, and G. W. Bush eventually won. The “IRV” promoters began speciously claiming that “IRV” removes the spoiler effect (and most of them still do claim that). And the spoiler effect concept is rather difficult to define when one is dealing with RCV/“IRV” voting methods.

Eventually, it became clear that many of the real pathologies of the various election methods are almost impossible to discern unless one takes into account an “elite party capture effect.” In fact, the perspective on all of the issues regarding election methods transforms radically once one begins analyzing elections as contests between common individual’s interests and elite’s interests. For example, in the year 2000 election Bush and Gore were clearly elite-linked candidates, and Nader was perceived as the common people-linked spoiler for the (perceived lesser evil) Gore.

From my perspective, the truth is that it is not necessary to utilize a great number of arcane concepts in the evaluation of election methods. Many (including modern RCV methods) of the are far too complex to be utilized without automation. Even those that are easy to explain are often not simple to implement, and the notion of “expressiveness” is not a simple concept either. Prime numbers are very simple to describe. The numbers (2), (3), (5), (7), (11), (13)… are prime because they can only be divided, with no resulting additional fractions, by themselves or by (1). And all the other numbers between (2) and (13) are non-prime. This notion of prime numbers that cannot be divided with no resulting additional fractions is “dirt-simple,” yet it leads to many of the most complex challenges in mathematics. Merely because something is easy to describe by no means implies that it will not create infinitely complex structures.

The “expressiveness” notion is not so simple as to imply that the information entered into a system correlates with the degree of control available to the one who enters it. If I give you a lottery ticket with ten digits that will be chosen at random tomorrow, would I be offering anything more If I allowed you to chose five of the digits?

All ranked choice voting methods have been historically shown to have multiple pathologies that always result in elite party capture, which is, in reality, aristocratic dictatorship.

Compare this with strategic hedge simple score voting. With this, you could allot, say, between (1) and (10) votes to each of as many candidates as you wish (but you couldn’t spend an eternity in the voting booth). (No, we are not talking about what has been called “cumulative voting”, whereby each voter is allotted a fixed number of points which they may distribute.) You could allot Nader (10) votes, and allot Gore (7), (8), or (9) votes, and so by voting for Nader, you would still not be sacrificing all of the votes you would prefer to allot to Gore as the “lesser evil.” This is where you would be using the strategic hedge strategy. There would be no (0) vote that ballot publishers could use to waste time looking for mistakes, which might be ruled to result in discarded “spoiled ballots.” There would merely be “abstemious non-votes” whereby candidates would be ignored. There is no point in ceding power to ballot publishers.

This is much better than approval voting, which simply eliminates the option to use the strategic hedge strategy. Everybody could cast one vote for Nader and also one vote for Gore, but the result would simply be a close “toss-up.” Or you could “vote your (foolish) conscience” only for Nader, but the individual next to you might be casting his/her vote only for Gore. This represents the double bind quandary whereby it is a fair likelihood that we would never ever see a Nader win. Approval voting could produce a fair likelihood that a common people-linked candidate might never win.

What is called STAR voting looks like score/range voting with a second consecutive round to produce nothing more than a (fake) “majority winner.”

If you don’t believe all this, take a look at:
Why Greens need range voting
https://www.rangevoting.org/ForGreens.html

Below I have worked out an example of one of the major pathologies of “IRV” voting. As a mere example, it’s more than just a “vignette.” It is not explained in as much detail as I would prefer. And it’s very easy to make silly mistakes in such exercises, so feel free to point out any.

Ranked Choice Voting fails through “bullet voting”. Voters are numbered 1 through 28. Candidates are designated A through Z. Candidates within parentheses are the ones that get eliminated.

=-=-=-=-=-=

1: A>F>B>C>D>E --> 1xA

2: B>(A)>C>D>E
3: B>(A)>C>D>E --> 2xB [A and then F are eliminated.]

4: C>(A)>(B)>D>E
5: C>(A)>(B)>D>E
6: C>(A)>(B)>D>E --> 3xC

7: D>(A)>(B)>©>E
8: D>(A)>(B)>©>E
9: D>(A)>(B)>©>E
10: D>(A)>(B)>©>E --> 4xD

11: E>(A)>(B)>©>(D)
12: E>(A)>(B)>©>(D)
13: E>(A)>(B)>©>(D)
14: E>(A)>(B)>©>(D)
15: E>(A)>(B)>©>(D) --> 5xE

16: G>(A)>(B)>©>(D)>(E)
17: G>(A)>(B)>©>(D)>(E)
18: G>(A)>(B)>©>(D)>(E)
19: G>(A)>(B)>©>(D)>(E)
20: G>(A)>(B)>©>(D)>(E)
21: G>(A)>(B)>©>(D)>(E) --> 6xG

22: T>(U)>(V)>(W)>(X)>(Y)>(Z)
23: T>(U)>(V)>(W)>(X)>(Y)>(Z)
24: T>(U)>(V)>(W)>(X)>(Y)>(Z)
25: T>(U)>(V)>(W)>(X)>(Y)>(Z)
26: T>(U)>(V)>(W)>(Z)
27: T>(U)>(V)>(W)>(X)>(Y)
28: T>(U)>(V)>(W)>(X)>(Z) --> 7xT
=-=-=-=-=-=

First Round eliminates A.
Next Round eliminates F.
Next Round eliminates B.
Next Round eliminates C.
Next Round eliminates D.
Next Round eliminates E.
Next Round eliminates F.
Next Round eliminates G.
Since U~Z are always ranked below T, they are eliminated in various rounds.
T wins.

21 voters voted for “clones” A through G, all of which got eliminated.
7 voters voted for “clones” T through Z.

T wins, with a 7 to 6 “majority”. Note that 28/7 = 4, thus the T~Z minority interest prevailed over the A~G majority interest despite a 1 to 4 ratio.



#2

What do you mean by this?

The second round exists primarily to disincentivize strategic exaggeration in the first round. That it produces the same “majority” as IRV just helps in explaining it as an improved version to people who are familiar with that.


#3

I said this because because of possibly unexpected consequences of the presence of a (0) vote. I believe that votes should be able to be allotted to write-in candidates. This option has long existed for the purpose of allowing candidates to run even If ballot publishers refuse to include their names on ballots. This moves power away from the ballot publishers and toward voters. If (0) votes can be cast, it becomes possible for vandals to write in individuals as candidates and and then to cast (0) votes to them. This is hurtful, and would be likely to cause the ballot publishers to create “special rules” that would diminish the power of the voters.

Presumably the hand ballot counters would be required to count the (0) ballot, so there would be more work for them.

Also, the ballot publishers could attempt to create rules that would throw out whole ballots if any one candidate were not allotted a score (at least a (0) score).

STAR voting requires two consecutive rounds. I attempted to promote “three-consecutive runoff voting” back in 2007 (see the “Consecutive Runoff Approval Voting” link above). The eight candidates who won the most approval votes in the first round would go on to the second round. In the third round, voters would choose between the two of those eight who had won the most votes in the second round, and then one would finally win. So I nearly invented STAR voting.

I have found out the hard way that people have absolutely zero interest in three round consecutive voting, so I have little faith that two round voting will be very popular. Also, is it really so bad if non-majority candidate wins? The result could even be preferable to proportional representation.

The last thing we need is to “disincentivize strategic exaggeration.” We must promote strategic hedge voting, which is almost the opposite of that.

STAR would presumably be quite far from the worst voting method. But I see no significant improvement over score/range.


#4

I don’t understand. You think that on a Score voting ballot, write-in candidates should be given special privileges of not having negative votes cast for them? Wouldn’t everyone run as a write-in?

Have you seen 321 voting?

With Score voting you need to prevent it. What is hedge voting and how do you promote it?


#5

There definitely is something very odd about how election methods thinking so easily slips off the rails. I have to have mattresses covering all the floors in my house for protection when I get dizzy and fall down. Let’s just call your approach “the old way,” and my approach that holds the elite parties as distinct from the true grass roots, non-elite parties “the new way” :slight_smile:

I did not intend to imply that write-in candidates should be treated any differently than those whose names appear expressly on the ballot papers. There should be no option to cast a (0) vote in either case. If voters want to give no votes to a candidate they can simply abstain. In fact I think it would be better if only (5), (6), (7), (8), (9), and (10) votes could be cast per candidate – This would be simpler for both the voters and the vote counters, and it would remind people that they should vote strategically. The elite parties will always do anything to remain in power, and will never ever be “honest.” So why should the common voters be expected to be “honest” (really “rectitudinous” would be a more neutral, scientific term)?

Whoever invented that was likely a couple beers away from conception when I proposed consecutive runoff approval voting 11 years ago.

Well, my main point has been that non-elite voters must vote strategically if they intend to disrupt the elite party capture effect. Please recall that the very strongest insistence upon the tender coddling of so-called “honest voters” has been voiced by the lavishly foundation-funded, misrepresentative lobbying campaign promoting the astonishingly complex intellectual shell game called RCV/“IRV.”

(I’ve started posting a bit again at my website at:
Simple Score Voting — Democracy Was Always Sabotaged.)


#6

I did not intend to imply that write-in candidates should be treated any differently than those whose names appear expressly on the ballot papers.

Write-ins are pretty thorny for averaged Score, though, since you can write in a name and everyone else would have to write-in the same name and vote it down to 0 to prevent it from having an unfair advantage. You’d have people writing in every name they could think of and voting it 0.

I did not intend to imply that write-in candidates should be treated any differently than those whose names appear expressly on the ballot papers. There should be no option to cast a (0) vote in either case.

Score voting needs to have 0 as the lowest score, both because “zero support” is a common thing that people want to express, and to make it very clearly different from ranked-choice voting, where “1” represents your favorite. No one can claim they will be confused into thinking 0 means “my favorite”, but they will claim that if the lowest score is 1.


#7

With simple score voting, instead of a problematic zero vote option there is an abstentive option whereby the voter ignores an on-the-ballot candidate just a though that candidate were not on the ballot at all. There is no averaging of votes that would make it “complicated score voting.”

It should not be too difficult to just tell the voters to simply mark the number of votes they wish to confer to each candidate. If they cannot understand that then it won’t matter at all, since they will not be able to find the voting booths anyway.


#8

We just need to give each candidate 1000 artificial zero-votes (or less in smaller electorates and more in huge electorates) so that you cannot get above 50% score without at least 1k voters scoring you. This will have negligible effect on the front runners.


#9

You could add a ‘default score’ field for what score your ballot should be interpreted as giving people who you don’t explicitly assign a score to (such as write-ins.)


#10

Marylander:

You could add a ‘default score’ field for what score your ballot should be interpreted as giving people who you don’t explicitly assign a score to (such as write-ins.)

Yes, I have been advocating for that. It could be called “all others”.