[EM] reply to Gilmour attack on range voting & social utility; CCd to RangeVoting

[Warren Smith]

>Gilmour:

>What I had in mind was if I vote 1, 2, 3, 4 (1 = most preferred, 
the one I want to see win) for candidates A, B, C, D,
and you vote 100, 99, 2, 1 (1 = most preferred) for the 
same four candidates, it would be fundamentally undemocratic if
your vote counted for more in determining the result just because you expressed your preferences more strongly that I did.

>I can understand that there might well be some difference in the 
social utility assessments of our preferences against
the outcome, but the issue here is democratic representation and "one person, one vote" surely comes above all else.
Once you depart from that (no matter what your opinion of the other electors!), 
you are on a very slippery slope indeed.

--WDS:
Gilmour's attack sounds good at first, but the more I examine it, the less I like
and the more deeply confused it seems to be.

Let me attempt to reply.  First of all, by expressing ABCD in that order you
*already* are expressing a more-strong preference for A over D 
than for, say, C over D.  In many voting systems your vote would
therefore have a stronger effect than some other vote about A versus D.
So you are deluded in thinking that your kind of voting is "more fundamentally
democratic" because it "omits" strength of preference information.

Second, you in range voting are free to express your preferences as strongly
or weakly as you like.  If you express a weak preference which is swamped
by my strong preference then you have only yourself to blame.  
At present you can express a very weak preference by simply not voting at all.
That is fine with me; but it is not fine with me if you then COMPLAIN that
it was fundamentally unfair that you, by not voting and thus expressing a weak
preference, had a smaller impact that I had by actually voting.
It seems to me then that the problem would not be the fundamental unfairness
of the system, but rather the fundamentally whiny nature of certain
complainers.

Third, I happen to consider it "more democratic" if voters CAN express strength
of preference, and "fundamentally undemocratic" when they are prevented from doing
so.  For example, if we are debating the issue "should James Gilmour be tortured to death?" you might have a strong preference which you would want to outweigh
the very weak anti-Gilmour preferences of those slightly annoyed by your posts about voting.  You then might not clamor for restoring the "true fairness"
of "democracy" and might not insist those weak preferences must count as strongly
as yours.

Fourth, at that point you might realize that "social utility" is NOT
some minor notion only of interest to math nerds, and obviously far exceeded
in importance by the demand for fair democracy.

Exactly wrong!  Social utility is THE overriding goal which trumps and
encapsulates all else.  The sole reason democracy is a good idea
(to the extent it is) is that it leads to greater social utility.
Now some of the vague principles you have in mind ("fairness")
again are social-design notions that are important because they tend to
lead to greater social utility than, say, a dictatorship.  And in fact the sole
reason that the human mind ever acquired inbuilt notions about
fairness and morality (to the extent these notions are built into
the human mind - or the minds of other animals - and I think they are)
is BECAUSE having those inbuilt notions led to greater social
utility, i.e. to greater Darwinian fitness for those tribes which had
those genes.

Fifth and finally, consider, say, IRV (Instant Runoff Voting).  IRV often ignores
much of your vote.  For example if you vote ABCD, it is entirely possible that
your preference C>D will be ignored, and thus will be swamped by
my D>C preference in my opposing vote DCBA, which will NOT be ignored
and in fact will affect the course of the IRV election by causing D to
survive un-eliminated in some IRV round in which, say, C is eliminated.

Now, presumably you consider that to be "unfair" and "contrary to
democratic principles."  Eh?  But in range voting, if you express a preference
C>D it will never be ignored.  Ever.   Yes, I grant that if your preference
C>D is expressed more weakly than my D>C then I will outweigh your effect -
but you will not be ignored and will still diminish my effect.

So it seems to me by your own argument, thus turned around, we conclude that
range voting is fundamentlly more fair and more democratic than IRV.

I could also make similar arguments against Condorcet methods, e.g, 
it is fundamentally undemocratic that by refusing to cast my honest A-top vote,
I should cause A to win  (which can happen with Condorcet but never with range).
But all these criteria-based arguments can go on forever with no clear resolution. 
The way out of that eternal morass is SOCIAL UTILITY which measures
EVERY paradox-criterion and weights at appropriately autmoatically for you,
so that you can come to a clear conclusion about which voting methids are better.

wds