[EM] Actual Approval Elections:

Craig Carey research at ijs.co.nz
Tue Apr 10 13:39:39 PDT 2001

The groups named as using the Approval Vote are not using the method in
government or local government run large scale public elections where a
lot of the public is uninterested in the Approval Vote method or
not well uneducated (or does not like voting methods that no nation's
government seems to use).

If there is a problem naming a nation that uses the [newly named]
"Approval variant of the Block/Cumlative Vote" then perhaps a more
thorough search is needed.

At 4/10/01 19:45 +0100 Tuesday, Martin Harper wrote:
 >I Like Irving wrote:
 >> - - - - - - - - - - - - - - - - - - - - - - - - - - 04/10/01
 >> Harper,
 >> You wrote:  "... actual approval elections do not appear to have these
 >> problems you claim they suffer from."
 >> Davison:  It would be nice if we could see that for ourselves.
 >> It appears you know of actual approval elections.  If so, would you be so
 >> kind as to post to this list the results of an actual Approval election?
 >> We would not need copies of the actual ballots.  We only need the different
 >> candidate combination sums.
 >> If you have more than one election race to pick from, pick the one that has
 >> a history of Approval elections.
 >> Donald,
 >*whips out copy of "Approval Voting", Robert J. Weber, Journal of Economic
 >Perspectives, Vol9, Issue1, Winter 1995*
 >I don't have the results of an actual Approval election on me
 >storing such information. I recommend you write to one of the above-mentioned
 >groups if you are interested - I'm sure they'd be delighted to help.

I'd prefer an answer to my question, which is Mr Davison's but with a small
restriction that the election be public.

Let's consider the failure of Mr Davison to restrict the election to be quite
public election. Can the error be repaired. That might be a dead of a topic.

Was Mr Davison thinking of public elections. Note that Mr Harper's message
that Mr Daviso replied to was all about the so called 'IRV' method. This is a
method that was being promoted by the CVD and was it Mr Ossipoff (it allows
voters too much power)?. Whatever.

The CVD seems to only really recommend that IRV method for US local government

At 4/5/01 16:05 +0100 Thursday, Martin Harper wrote:
 >I Like Irving wrote:
 >> Harper, you wrote: "A ranking of A>B>C>D>E does not equate to an approval
 >> vote which approves all five candidates."
 >> Davison: I know this and you know this, but if and when there is ever an
 >> Approval election
 >There have been plenty already.
 >> most of the voters will cast their ballots as: A>B>C>D>E and you Approval
 >> promoters will have no qualms about tabulating these
 >> ballots as A=B=C=D=E.
 >The mirror problem is that some people voting in an IRV election (with 3+
 >candidates) will cast their ballots as A=B, and IRV promoters will have no
 >qualms about tabulating these ballots as 50% A>B and 50% B>A. Both problems 
 >be solved either by the use of voting machinery which does not permit 
voters to
 >spoil their votes in this way, and/or by the use of education to inform 
 >of how to vote properly.

The comment about voting machinery indicates that Mr Harper and Mr Davison
were perhaps thinking of larse elections. When the IRV factor is factored in
and then it seems that Mr Harper was not thinking of corporate board room
elections where methods that allow 1.5-3 times the proper power to a manager's
faithful could be handy in tricky board room voting situations.

Some of us want to have this question answered.

At 10/10/00 16:18 -0700 Tuesday, JanetRAnderson wrote:
 >The League of Women Voters of Washington funded their study entirely from
 >their  501C3 Education Fund.  CVD was not involved in any of the content
 >decisions we made.  Errors of omission were entirely due to our ignorance
 >and/or space limitations.  Our description of approval voting was omitted
 >for the latter reason and because we knew of no country where it was in use.
 >I don't know why the link from the California League didn't work.  I'm
 >getting responses from all over the country.  Try dialing directly.  I have
 >downloaded from the web sites you suggested but haven't had time to digest
 >the material.  Thank you for the patient explanations.  I really do want to
 >be promoting the best systems.

Are or are we not in some sort of close contact with the supremos of Approval
variant advocacy?.

Alternatively you could opt for maximising the utility of your reply without
doing research. You could help me and Mr Davison out with our researches
into voting theory by actually genuinely simply answering the question
(something that Mike did not do because he did not get the question).

Maybe there is something wrong with the utility arguments too: could it be
that councillors hear utility arguments but do not find them true and so
still reject the Approval sub-variant of the Block Vote (ASVB method [I am
still not sure if this is Mike's method or not: I never got to see any text
on ballot papers).

You (Mr Harper) were the person that sent me some valueless maths on something
you called utility. Complete with the appearance of number 0.5 (a probability).
Never never divulges probability distributions and I end up never even getting
close to being able to calculate a 1st, 2nd, or 3rd moment or a variance, a
kurtosis, any integral product with the real value replaced with quaternions
or complex numbers and integrated over a unit ball. Unfortunately probabilities
associated with utility theory are a well guarged secret. So all those who
wanted to see the kurtosis with function wrong integrated inside a unit circle
will have to miss out. Take it up with Mike is that is not satisfactory.

Is the ASVB method used in Greenland and Finland. I have no confidence in the
ability of the Washington Women's League to search competently inside of
Greenland. Mike has been writing about what I have now renamed as the ASVB or
AVB variant method and while I do not recommend it (any more than maintaining
insects), if it is good then in which public elections in which nation is
the ASVB method being actually used. If you do not know then please say so.

Does anybody here want to have a debate. I saw that Rob used that word.
I thought that what happened here was the list had gone to seed due to
deep set problems that really I am not writing on, and mere subscribers
wrote their thoughts and got ignored and ONE particular person here will tend
lecture alarmingly is if trying to communicate understanding and at the end of
it, the question was unanswered.


Mr Harper sent to me a question

At 4/10/01 19:35 +0100 Tuesday, Martin Harper wrote:
 >Craig Carey wrote:
 >Ok. So you must have some kind of definition of "effective weight" which
 >refer to utilities then? Would you be generous enough to share it with the

The "effective weight" is defined by me so that it ends up with the meaning
that people would think it ought have: it is the ratio x/y, where x is the
number of FPTP votes that are required to offset y of another vote. Both are
made to tend to 0, and this is a statement about a slope along a particular 
direction. The coordinates are barycentric rather than Euclidean. A point
I have not fully figured out is how to replace the reference FPTP votes with
votes having more than one preference.

Martin has raised an interesting question: does any government use my new
Variant Block Vote method?. Unlike Mike Ossipoff's, in the Approval Variant
Block Vote, small text on the ballot paper specifically declares that
there should not be a cross for all candidates. That makes it a better method
that Mike Ossipoff's own Approval Vote method. I also asked Mike in a very
recent private message. It doesn't seem easy to avoid getting no answer but
a statement that an answer was given many months ago without the month being
stated. How many know Mike Ossipoff's own view on the presence or absence of
the last preference?. Is that fine text there or not. Mr Davison has posted
up the view that he didn't think that he needed to know that. I ask Donald
to reconsider.

More information about the Election-Methods mailing list