[EM] (hopefully with paragraphs now) More complete statement about IRV, and appeal to GPUS for Approval. Proposing a petition to GPUS.

[Michael Ossipoff]

 (This message posted without any paragraph spacings. I'd somehow set
it for "rich formatting". I'm going to try it again with the
"plaintext" setting, in the hopes that that will preserve the
paragraph spacings.)


If I say some things that I've already said, I might say them differently,
&/or in a different order. But I'll also probably say some things that I
haven't said.

A few definitions:

By "Green scenario", I refer to a scenario in which the GPUS has been
elected to the presidency and most of Congress. Among the many resulting
reforms, proposed in the GPUS platform, would be greatly expanded public
decisionmaking via initiative and referendum, and with the power of
swiftly-implemented recall of elected officials. Likely there'd be an
initiative or referendum regarding the choice of voting-system, before the
first candidate election in the new Green government.

"MM" stands for "mutual majority", a set of voters adding up to more than
half of the voters, who all prefer some set of candidates or alternatives to
the other candidates or alternatives.

MM-preferred candidates (or alternatives) are the ones that an MM prefers to
the others.

The Mutual-Majority Criterion (MMC) says that if an MM vote sincerely, then
the winner should be from among the MM-preferred candidates or alternatives.

I'll define other criterion acronyms when referring to them.

For discussing the relative desirability of various voting-systems, it's
necessary to specify the conditions for which we're proposing, evaluating or
choosing the voting-system(s).

Most of our discussion about voting-systems for official public elections
has been about methods for the existing conditions, the current electorate,
with its beliefs and assumptions promoted by media disinformation.

Of course, for such conditions, a voting-system should be chosen that works
adequately for voters under those conditions. I've been arguing,and still
do, that, under such conditions, the Favorite-Betrayal Criterion (FBC) is
absolutely essential. Without it, for voters believing media disinformation,
the optimal voting strategy for progressives would require voting the
Democrat over everyone else, regardless of whether the Democrat is really
favorite. I've talked a lot about why that is, so I won't repeat it here.

For these currently existing conditions, I prefer Approval and Score. If
count-complexity and the resulting fraud-vulnerability, weren't a problem
(and William Waugh has pointed out that maybe computer counting could be
fraud-safe if voting machines printed out paper ballots, to be examined and
deposited by voters,and then imaged by various political parties, and then
securely stored and guarded by locks and video-cameras belonging to those
parties), then I'd especially like Symmetrical ICT, as my favorite method
for existing conditions. ER-Bucklin would be good too, though it wouldn't be
a significant improvement on Approval, because it doesn't automatically
avoid the chicken dilemma.

I consider Approval and Score to be the best proposals, under existing
conditions, partly because of their merit (FBC, Later-No-Help (LNHe), and
their social optimizations)t, partly because of their particularly easy and
secure handcount, and partly because Approval's minimal change from
Plurality, its compliance with all criteria met by Plurality, and its
unique, elegant simplicity.

But the Green scenario would be a completely different situation. The
methods that I prefer for our currently-existing conditions would be good in
the Green scenario too, because those methods are excellent all-purpose
voting-systems of robust merit. Given the current lack of information about
mutual majorities in a Green scenario, I now prefer, for the Green scenario,
the methods that I prefer for existing conditions--named in the two
paragraphs before this one.

Nevertheless, other methods become desirable too, under Green scenario
conditions. There are at least five political parties (all of them
progressive), including GPUS, that offer IRV as the voting-system for the
government offered in their platform. I don't know of any other
single-winner voting system that's offered in a party platform. That makes
it worthwhile to consider the merits of IRV, under Green scenario
conditions.

Later, I'll discuss a few other voting systems that, while not
good now, would be alright under Green scenario conditions (even if not as
good as IRV). But I'll discuss IRV first.

First, some of IRV's criterion-compliances:

Mutual-Majority (MMC)
Later-No-Harm (LNHa)
Clone Independence
Later-No-Help (LNHe)

These are important and valuable criterion-compliances. In particular the
first two are powerful in combination.

Of course every criterion compliance is gotten by trading other criterion
compliances for it. IRV's criterion compliances listed above are gotten at
the cost of certain problems. I claim that those problems, in the Green
scenario, are exaggerated.

What would it be like to vote IRV in the Green scenario?

Well, if you're in a mutual majority, you can be assured that if the members
of your MM rank sincerely, then the winner must come from your MM-preferred
set. Don't underestimate the importance and power of that. It means that
your MM has a guaranteed win, with no strategy required--just sincere
ranking.

And, because of LNHa, there is _no_ disincentive to voting as many choices
as you want to. Your ranking of your 2nd choice can't help hir against your
1st choice. Your ranking of your 3rd choice can't help hir against your 1st
or 2nd choice.

That ensures that IRV's MMC compliance will mean something, because mutual
majorities will really be voted. There's absolutely no reason for anyone not
to. You have no reason to not rank all of your MM preferred candidates over
all the others. No disincentive to support whomever you want to.
So, if you're in a MM, your MM has a guaranteed win, without any strategy
required.

Sure, if you're not in a MM, it isn't so good. Then you have IRV's
favorite-burial need. But if you aren't in a mutual majority would you do
well in any method? Should nonmajority candidates expect to win?

There's nothing wrong with government by a cohesive majority.

IRV's failure of the Condorcet Criterion is particularly flagrant,
particularly frequent. That's one of the abovementioned costs of IRV's
important criterion-compliances.

While the benefits of MM, and the unbenefits of non-MM give IRV a
win-big/lose-big quality, IRV's maybe frequent lack of respect for CWs
further increases that win-big/lose-big attribute, by sometimes eliminating
the compromise, and making the result be more all-the-way, one way or the
other.

So, what about that failure to favor CWs?

First, looking at it objectively, in terms of IRV's justifications:

The CW might not be the CW if we only look at voters in the MM. IRV chooses
within the MM. There, it chooses based on favoriteness, in a way that
admittedly isn't always perfect. But if, in that MM, more people prefer a
wing-candidate to the CW, than vice-versa, then is it wrong to elect the
wing-candidate instead? I say it isn't.

Looking at it personally:

If your favorite is (at least very likely to be) the CW, and if electing
your favorite is the most important thing to you, then of course you should
oppose IRV.

But are you sure you feel that way? Speaking for myself, the reason I'd like
Approval is because I want a progressive candidate to win, and I don't much
care which one. So I'd approve all of the progressive candidates, in
Approval.

Well, if that's how I feel, then why should I care if my favorite, the CW,
doesn't win, and a different MM preferred candidate (progressive) wins. For
the purposes of this discussion, of the personal implications, for me, of
Condorcet-Criterion failure, I'm assuming that the progressives have a MM.

So, if I don't much care which progressive wins in Approval, then why should
I in IRV? The fact that _sometimes_ my CW favorite will get eliminated,
letting a different progressive win, won't bother me. In fact, of course, my
favorite might not even be the CW.

So IRV isn't causing a new drawback for me, when it sometimes eliminates a
CW. And, gained at that cost, IRV is giving me an unequalled _choice_ among
those progressive candidates, those MM-preferred candidates. I can sincerely
rank them all, in order of preference. Any MM found at any level of my
ranking will be enforced.

While letting me fully protect a win for the MM,
IRV also gives me an impressive choice among those MM-preferred candidates.
That's a lot of benefit, for the cost of not doing a good job of electing
CWs. I suggest that it's well worth it.

Yes, the way IRV starts by eliminating small party candidates first, and
thereby coalesces them first, via their transfers, can sometimes result in
an arbitrary choice among a MM. Sometimes, but not always. And if a wing
candidate is preferred to the CW by more MM voters than vice-versa, then the
election of the wing candidate isn't arbitrary or unjustifiable.

IRV's sometime failure to elect the middle compromise, the CW might not be a
good thing, but neither is it a prohibitive disadvantage.

As I said, right now we don't know about MMs in the Green scenario. At that
time we'll know more. If there's an initiative or referendum to choose a
voting system, in the new Green govt, then, based on what is then known
about MM, we might top-rank Approval, or IRV. It depends on whether we're in
a MM.

As I said, based on current lack of information about MM, I now prefer
Approval, even for the Green scenario. That's why I suggest that we e-mail
the Greens, and make sure that they understand why IRV might not be in their
best interest, even in the Green scenario.

Regarding the elections as u/a, and wanting an acceptable to win, I'm not
inclined to gamble. Based on current lack of information, IRV for the Green
scenario would be a gamble. IRV is admittedly an unstable, risky gamble.
But if, at the time of that initiative or referendum, you felt sure that
you're in a MM, then you should top-rank IRV.

As I said, IRV is what the Greens (and at least 4 other parties) offer in
its platform. What if, even if we wouldn't now choose it, IRV were going to
be the next voting system? How bad would that be?

Because we can't discuss
these things in a vacuum, I'll speak for myself, as a progressive:

reasons that I've amply discussed, I claim that, when people look at
platforms, the progressives will be a majority. The progressive parties'
platforms are so nearly identical, in their pro-humans life-quality
improvements, and their criticisms of the rich few who run things for their
own benefit, to the detriment of everyone else, that I suggest that that
majority should be mutual. It would probably voted as a mutual majority.
But, based on platform-similarity, it certainly should be.

The above facts are enough to make the progressive MM a good bet. If IRV is
a gamble, then it looks like a good gamble. As I said, I prefer not
gambling, which is why IRV, at this time, isn't my Green scenario choice,
but, because the gamble looks like a relatively good one, then I suggest
that if IRV must be the next voting system, that isn't a bad thing at all. I
suggest that it would be ok.

Does that sound self-serving? It's just a frank evaluation. And doesn't
every political persuasion believe that it's a MM, or in one? Well, with
IRV, as with any voting system, the winners will be the voters who are right
about such things. That isn't something peculiar to IRV. May the MM win.

Other voting systems for the Green scenario:

I've already said that, as of now, I'd prefer Approval or Score for the
Green scenario. How I'd vote in that voting system referendum, however,
would depend on what be known then about MM. But if IRV must be the next
voting system, that's ok, for the reasons described above.

But, while Beatpath (B) and Ranked-Pairs (RP) are inadequate now, because of
their FBC-failure, that won't be a problem in the Green scenario. RP and B
meet MMC, and that gives them some merit in the Green scenario. But they
fail the Chicken Dilemma Criterion (CD). IRV meets CD, because it meets the
stronger LNHa. B and RP meet the Condorcet Criterion, but, in the Green
scenario, I consider CD to be more important than the Condorcet Criterion.
MMC loses much of its meaning without CD or LNHa.

Bottom line:

Five political parties, including the popular GPUS, offer IRV in their
platforms. I don't know of any other single-winner method offered in a party
platform. There's a good chance that IRV will be the next voting system.
That would be ok. Don't worry about it.

But, in the Green scenario, there'd probably be a voting system referendum,
before the first IRV candidate election, so that the public could choose the
voting system. Your vote in that (IRV) referendum should depend on whether
or not you're in a MM.

In the meantime, now we should be contacting the GPUS, and the other IRV
parties too, to tell them why IRV likely won't be in their best interest,
even in the Green scenario. If GPUS wins, by optimal Plurality voting, then
they have a good chance of being CW. CWs are disfavored in IRV, in
comparison to Approval or Score, which will do a much better job of electing
CWs. GPUS also might not like the idea of small factions, away from the CW,
being relatively favored, in comparison to Approval and Score.

They should be told about these things. I'd say, "I'm not trying to
influence your choice. But honesty and ethics call for someone to tell you
what you'd be getting into, with IRV."

And, of course, additionally, the GPUS should be informed that IRV, now,
like Plurality, would unduly favor the Republocrats, keeping them in power
forever, due to FBC-failure.

I suggest a petition, or, better yet, a combination of a collective e-mail
and a petition, from the EM membership to the national GPUS platform
committee.

Is anyone interested in participating in that effort?

The collective e-mail would consist of a message from everyone who wants to
include his or her message in it. The petition-wording could be arrived at
as an collectively edited and voted-on combination of what different EM
participants have to say on the matter.

Mike Ossipoff