[EM] Norm Petry's Candidate Sortition/IRV Method

[Instant Runoff Voting supporter]

- - - - - - - - - - - - - - - - - - - - - - - - - 12/20/00
Greetings EM List,

    I have long felt that what you vocal persons on this list really wanted
was a form of candidate sortition. I hinted at it once, but I never went so
far as to accuse you of that. But it now looks like Norm Petry has let the
cat out of the bag and has exposed you by presenting his Candidate
Sortition/IRV method.
    You now have your first sortition method in your stable of Bottom
Methods that subsidize the bottom candidates. We can forget about the IRV
part of his method, it is merely window dressing in an attempt to fool the
voters into thinking they have a say in something.
     Norm pretends his sortition method is for me, but we know better. Norm
gave himself away when he proposed his method. For this is the ideal method
for you and him. A sortition method will give every candidate an equal
chance of winning regardless of the popular vote. This will mean that a
Nader, a Buchcanan, a Browne, etc will each have the same chance to win as
a Gore or a Bush. This will be a much better method for your agenda than
any of the other Bottom Methods, better than Condorcet, even better than
Approval Voting.

     Approval Voting is currently your best method that subsidizes the
bottom candidates. If the first choice count of an AV election is:
      49 A,  48 B,   2 C,   1 D,  and the voters ranked every candidate,
then the results would be:
     100 A, 100 B, 100 C, 100 D,  All the candidates have received extra
votes, but the lowest candidates have been subsidized the most. Giving
every candidate 100% of the vote is really treating every candidate as an
equal, but in order to have a winner we need one voter to not make any
lower choices, that will give us a difference between the candidates and a
winner. Suppose the D voter only made one choice, the results would be:
      99 A,  99 B,  99 C, 100 D,  candidate D wins, that's almost as good
as sortition. Oh, but we have a problem Houston, if the D voter can only
make one choice, so can the C voters only make one choice. Now the results
are:
      97 A,  97 B,  99 C,  98 D,  candidate C wins. But, the same policy
prevails, if the C and D voters can only make one choice then so can the B
voters only make one choice. Now the results are:
      49 A,  97 B,  51 C,  50 D,  (the Approval support of the candidates
seems to be fading away) It is now the A voters turn not to make any lower
choices, and we get:
      49 A,  48 B,   2 C,   1 D,  which is the vote count that we started
this with. (All Gone, is the Approval subsidy for the lowest candidates -
that's too bad)
     Approval Voting tries to treat all the candidates with equality, but
if those darn voters refuse to go along we can't depend on the results.
(sarcasm)
     There will be no such problem with Candidate Sortition, it will give
every candidate an equal chance of winning regardless of what the voters
do. (the voters can go to hell as far as sortition is concerned)
     Candidate Sortition is so bad, so undemocratic that it should be your
number one method of choice, your most preferred method.  (more sarcasm)
     As far as I am concerned, I will rate it the worst method and place it
on the bottom of the list of Bottom Methods, right after Approval Voting,
but I will understand if you embrace it.
     Keep up the good work Norm, your revelations are very educational.

Regards, Donald

Post Script:  By the way Norm, I do support IRV and the Hare quota and I
see no conflict. I also do not believe that a candidate needs to have a
full quota in order to win in STV nor in IRV, as long as he is among the
candidates that can fill a quota size container with the most votes over
half full.
     It is also possible in an IRV election for a candidate to receive 100%
of the votes, provided all the voters rank every candidate. This would be
the best vote of confidence for the winner, but this is common knowledge,
why does it amuse you?? (ask him what it is he is smoking)

- - - - - - - - - - - - - - - - - - - - - - - - -
From: "Norman Petry" <npetry at accesscomm.ca>
To: <election-methods-list at eskimo.com>
Subject: Re: [EM] IRVie majority
Date: Thu, 14 Dec 2000 08:31:03 -0600

What I find particularly amusing about Don Davison's reasoning is that not
only does he now strongly advocate IRV (Single-winner STV), but he is also a
huge fan of the Hare quota.  Of course, since the Hare quota for
single-winner STV is 100% of the voters, I'm surprised that Don limits
himself to claiming the IRV winner is elected by a a mere 'majority', when
he is able to claim that winner achieves *consensus* :-)  Why stop at
transferring up to 50% of the votes to the 'winning' candidate, when you can
transfer them all?

In other words, no matter who wins under IRV(Hare), they are the consensus
winner!  We could hardly do better than that.  Since whoever wins under this
system can claim to be supported by 100% of the voters, perhaps Don should
consider advocating the following system instead, which is similar to IRV
but has fewer strategy problems and offers the same advantage:

1) Collect ranked ballots, as in IRV
2) Choose the winner randomly
3) Transfer ballots as in IRV, except that the winner is protected from
elimination.
4) Once all the ballots have been transferred to the (random) winning
candidate, they are declared the winner (having achieved 100% support).

Of course if Don has some objection to IRV(Hare), he is being inconsistent,
but fortunately the above system works equally well with the Droop quota, if
all he wants is a 'majority'.

--
Norm Petry