[EM] article on preferential voting
Abd ul-Rahman Lomax
abd at lomaxdesign.com
Sat Dec 30 10:48:33 PST 2006
At 01:05 PM 12/30/2006, James Gilmour wrote:
> > Abd ul-Rahman Lomax> Sent: 30 December 2006 15:42
> >
> > Well, it is a bit irritating to me, and perhaps to some others, that
> > AV is used in the paper to refer to Alternative Vote; we routinely
> > use it for Approval Voting.
>
>Maybe, but I suspect US readers will find that the acromyn "AV" was used
>for "Alternative Vote" many decades before the term "Approval Voting"
>was first coined by Robert J. Weber in 1976 (Wikipedia).
I did not claim that it was wrong, it was irritating. It's irritating.
STV, by the way, could be implemented in a radically different way,
where the transfer is under the control of the candidate receiving
the vote. That would actually be a little closer to the implied
meaning of "single transferable vote." When I first heard the name,
that's what I thought it meant.
What I just described is Asset Voting.... far more flexible than STV
or just about any other election method. It's pretty similar to
Delegable Proxy used as an election method. It's a trick, though.
Asset Voting is not an election method as methods are sometimes
defined, that is, it does not, in itself, necessarily determine the
winner. There is further process involved.
It is similar to the Electoral College method of election, with the
College allowing variable voting power dependent on the vote that
each elector received. Absolutely the most democratic idea
floating.... able to elect a peer assembly with each member holding
almost exactly the same number of votes (i.e, having been elected
with the same number of votes). No wasted votes (or at least a
trivial number of wasted votes, with someone specific being
responsible, and thus accountable, for the waste). Aside from wasted
votes, if the vote assignments are by voting precinct -- which then
requires some variation in vote count per member, but slight --
voters know exactly whom their vote elected. And that member would
quite likely be local to them, where a "party" has many members. A
party with only a few members, but a quota, might have just one
member representing all party voters in the entire state or
jurisdiction. Those voters would be sacrificing locality of
representation for accuracy of representation.
Asset Voting is worth far more attention than it gets.... it is a
solution to most of the problems of proportional representation, it
eliminates the "party list" as an official part of the method. Party
lists may still exist and candidates may promise to follow them in
how they distribute votes, but they are not required to do so, and
voters can choose. With FAAV, voters can decide to award their vote
to a single person or split it among a virtual committee, which might
consist of a number of candidates from a single party....
Asset, single vote, no overvoting, is utterly trivial in ballot
design and counting. It is merely a different way of using standard
ballot information. FAAV, Fractional Asset Approval Voting, is a
little more complicated to count, but, in the end, each precinct
simply transmits vote totals for the precinct. My guess is that most
voters would vote for one and thus counting would be quick. I'd
divide the ballots into piles according to how many candidates the
voter voted for. These voters voted for one. Easy. Count the votes.
These voters voted for two. Easy, count the votes in that pile and
divide them by two before adding them in with the single votes. Etc.
Continue with each pile until all piles have been counted.
More information about the Election-Methods
mailing list