I recently noticed that my local food co-op, Seattle's Madison Market
( http://madisonmarket.com/ ) holds an election for their board of
trustees.  I wrote to Webster Walker at Madison Market about using
Range Voting - score the candidates, and elect the highest average.
He said they had considered upgrading their voting method, and a lot
of people had talked about using Instant Runoff Voting.  My letter to
Webster outlines the great benefits Range Voting offers over Instant
Runoff Voting, and other alternative methods.  You can read it here:

http://groups.google.com/group/rangevoting/web/madison-market-letter

Apparently, the current system uses plurality voting to elect 4 people
to the board - but all in a single election where the voters get 4
votes to spend, and cannot spend more than 1 on the same candidate (my
response was, why not let them have as many votes as they want, and
have Approval Voting, the simplest kind of Range Voting?).  The 4
winners elect a board president or supervisor of some sort (if I
recall the general outline correctly from my discussion today with
Webster).  I don't know whether proportional representation would be
worth the effort here, but if the ballots are to be tallied
electronically (on a spreadsheet for instance), Reweighted Range
Voting would be advisable - that is, if you take the view that the
board should represent the ideological variety in the electorate as
closely as possible.  When I brought up the complexity of IRV to
Webster, he made the obvious point that we can simply use a computer
program or online polling system (e.g. betterpolls.com) to do the
tally.  Quite right.  And I could easily use a spreadsheet to do an
election using Reweighted Range Voting.  So if proportional
representation is a good idea, there's no "complexity" argument to
make against it.  As much as I would love to see them use Range
Voting, seeing them use Reweighted Range Voting would be _awesome_.
And when used to elect a single winner, RRV is the same as RV - so the
board could then use regular Range Voting for electing their head.
For more about Reweighted Range Voting, see

http://rangevoting.org/RRV.html

My concern is that things like social utility efficiency,
monotonicity, and independence of irrelevant alternatives all sound so
esoteric, that it may be hard to convey to people why they should
really want Range Voting.  Another point Webster brought up is that so
many people have heard about IRV, and may be dubious of this new thing
that "nobody's heard of".  Well, sure they've heard of it, if they've
ever seen Amazon.com book reviews, or any type of rating poll.  But
the idea of using it for elections may sound foreign.  My thought is
that an open discussion about Range Voting, with a Range Voting
advocate like myself there to field questions, would be an educational
opportunity that would allow science to trump marketing.

Another thing that's cool is...Webster seems like quite a bright guy,
who thinks a lot about these kinds of things, like electoral fraud for
instance.  What if Madison Market could be convinced to try Ron
Rivest's 3ballot anti-fraud scheme, as a social experiment to prove
its feasibility?  That could be an awesome media event, that maybe Bev
Harris would be willing to show up to (she lives in Seattle).

Thoughts and/or opinions folks?  Clearly such an election doesn't need
to worry about two-party domination, and certain other issues where
Range Voting outperforms plurality and IRV.  Nevertheless, social
utility efficiency (expected voter satisfaction) is important, to put
it mildly (it's the whole point of voting).

Clay Shentrup
Seattle, WA
206.203.6619

On Feb 8, 9:02 pm, "BROKEN LADDER" <thebrokenlad...@gmail.com> wrote:
...

I think this is a fairly common way of forming boards, and some cities
have several "at large" seats on their city councils. For example
Denver has two at large seats. Everyone in Denver can "vote for 2"
from among the candidates competing for those two seats.

Whether you limit it to 4, or allow any number, either way I don't
think it's a very good method. Approval Voting (and Range Voting)
isn't very good as a multiwinner method. It is sort of anti-
proportional. If there are two factions of voters and candidates, say
Vegetarians and Carnivores, and say there are slightly more C voters
than V voters, and also the C's run exactly 4 candidates, then the
result is likely to be 4 C winners.

I think the question of whether any kind of PR makes sense depends on
whether the voters are divided into strong factions (say C vs V,
Organic vs. Junk, Suits vs. Blue Collars...). If so, then I guess it
makes sense to use PR. Otherwise, perhaps all that is wanted is to
elect the 4 people with the best board member skills (whatever those
may be), and multiwinner Approval might be OK.

  When I brought up the complexity of IRV to

Sure.

Also remember Asset Voting is a PR method. And you can "Fractional
Approval" ballots to make things easy for the voters. I.e. vote for 3
candidates, and each gets 1/3 of your vote.

Yes, I think it's better for the board members to elect their own
leader than to just declare the first RRV winner to be the head.

I think 3ballot is still sufficiently cumbersome that the voters would
be scratching their heads, saying "Now why are we being asked to go
through all this rigamarole?"

Do we know anything about SUE for multiwinner methods?

Cheers,
- Jan