[EM] Richie, FairVote, IRV

[Michael Ossipoff]

The organization now called "FairVote" was calling itself "Center for
Voting and Democracy" for a long time. "FairVote" is probably quite a
bit more recent.

When Richie said that CVD wasn't a "membership organization", and that
dues-payers had no say,and that it was governed by a self-appointing,
non-democratic board,  I later began suggesting on EM that the
organization should change its name to CVA, Center for Voting and
Autocracy.

The name "Majority Preferential Voting" goes at least back to George
Hallett, Richie's great uncle, who wrote _Proportional
Representation_, in 1926, in which he recommended Unbeaten//IRV (I
don't know what he called it) as his best recommendation for a
single-winner method. He recommended IRV ("Majority Preferential
Voting) as an intermediate proposal, a first proposal, because it's
simpler to define, describe, and offer.

At least it seems to me that someone (maybe Hallett himself) said that
his best recommendation amounted to Unbeaten//IRV. Actually, as he
proposed the method, it was a complicated, involved elaborate count
rule, but it was said that it was equivalent to what I all
"Unbeaten//IRV". The complication and elaborateness was probably for
the purpose of reducing the handcount labor of that Condorcet method.

That voting system was used to elect the "Spoon-Man", at Hallett's
university. That's the only use of it that I know of. But
Unbeaten//IRV, Benham, and Woodall, would be the best methods for many
non-political votes and polls.  ...and of course they'd be good
proposals for official political elections in the Green scenario.

I've been told that Richie objected to Condorcet-IRV, or
Unbeaten//IRV, on the grounds that a CW could be a candidate who isn't
the favorite of many people, and that the election of someone who
isn't favorite could bother people. But even people to whom the CW
isn't favorite would rather elect the CW than elect someone worse.
They'd  prefer the CW compromise to someone worse. FairVote makes much
of "majority", but whenever the CV isn't elected, majority wishes are
contravened.

There's a good case for electing the CW, and I haven't denied that,
even when advocating (for the Green scenario) IRV, MM//Benham, and
MM//Woodall, which violate CC.

Under current conditions, FBC is essential, but CC would be good too.
Symmetrical ICT meets both criteria, when CC is defined in a more
meaningful, general way.

For many instances of non-political voting or polling, CC would be desirable.

Maybe FBC might not be necessary even for some political polls, when
it's felt that media-disinformation-induced favorite-burial wouldn't
be a problem (because it's only a poll), and that the precaution of
Approval strategy (available in Approval, Score, ICT, and Symmetrical
ICT) therefore isn't needed--and if it's desired and believed that
people will rank sincerely, instead of needing Approval strategy.

The reason why I don't consider FBC, or even CC, necessary for the
Green scenario is because I trust majority judgement under those
conditions. I've told why: A population who've elected Greens aren't
media-disinformation-cowed. And the new Green America would have media
whose free exchange of information would produce a population whose
judgment we can have confidence in.

Therefore, it's ok to only consider the strategy-freedom of the
innermost mutual majority (MM), because they're the voting population
who matter the most. IRV gives them very good freedom from strategy.

If I'm not in the MM, then I'd want that CW compromise, and would want
to get it without favorite-burial. But if I have confidence in a
mutual majority's judgment, then, since, like anyone, I expect that
people with good judgment would agree somewhat with me, then I'd
expect to be in a MM. Likewise, someone whose political beliefs and
preferences differ from mine would, like me, expect to likely be in a
MM.

Sure, when the CW isn't elected, the wishes of the majority preferring
the CW to the winner is violated. All sorts of different majoriities
can be demarcated among the population. But a majority who all prefer
a certain set of candidates (their majority-preferred set) to all the
other candidates--there's something special about that _cohesive_
majority. Though the wishes of a differently demarcated majority is
violated if the CW doesn't win,they aren't a cohesive majority, and
the fact still remains that (if the method meets MMC, and the MM vote
sincerely) the winner is one of the candidates preferred by that MM. I
suggest that the innermost MM is the most important majority.

As I was saying before,the majority preferring the CW to the winner
could later successfully vote to replace IRV with a CC-complying
method, or at least with a method that does better at electing CWs
than IRV does. It depends on the CW-preferrers and how strongly they
prefer their CW.

Anyway, FairVote's objection that people wouldn't want a
not-very-favorite CW isn't valid. I'd want to elect the CW under
certain common conditions, and no doubt many others would. So, for
many purposes, Unbeaten//IRV, Benham and Woodall would be preferable
to IRV.

Comparing Beatpath with Benham and Woodall:

All 3 methods fail FBC. All 3 methods meet Smith. Of course that means
that they all meet CC and MMC. Condorcet-Loser too, but that's just an
avoidance of aesthetic embarrassment.

So how do they differ? Beatpath has the chicken dilemma, fully. Benham
and Woodall don't have the chicken dilemma. As I've said, MMC
compliance becomes meaningless when there's a chicken dilemma.

What's that? Beatpath meets Reversal-Symmetry?  :-)

Michael Ossipoff