[wplug-plan] Election rule change proposal

[David Ostroske]

Hey Everyone,

On Sun, Jun 7, 2009 at 9:49 PM, Vance Kochenderfer<vkochend at nyx.net> wrote:
> Hello all-
>
> This seemed like the best place to kick off this discussion.  At
> the July meeting, I intend to make a motion to substitute the
> following language for the current Election Special Rule of Order:

I've got a few words to say about this, and not simply because I
co-wrote the original language, with Mike O'Connor, in the current
Election Special Rule of Order.

Vance, nothing personal, but the proposal as it stands needs work. And
even if it's revised, while I can't say I'd sleep any less comfortably
at night if we went to a block voting method, I'm nevertheless
compelled to oppose it.

>  A plurality vote shall elect Directors.  Any ties shall be
>  decided by random selection.

This is a multi-seat election, so you need to add some details.

How many votes does each voter get?
- Just one vote? (There are real-world elections that use this method.
I think one house of Japan's legislature is chosen this way.
- Five, i.e. one for each seat? (That's called block voting, which is
what we use in PA for school board elections, and that's probably what
Vance is thinking of.)
- One for each candidate? (That's called approval voting.)

You'd probably need to add language to allow tied candidates to
withdraw, as you suggest below. Otherwise, the tellers could break the
tie at random before announcing the results. Withdrawing to break ties
after the results are announced could be done informally, certainly.
But I don't think election methods should be so informal.

> This would replace the current wording, available at:
> http://www.wplug.org/wiki/Special_Rules_of_Order#Election_Special_Rule_of_Order

> While I think the current system was well-intended, in practice I
> feel it is unwieldy and confusing to voters.  I believe that in
> every election we've had, at least one ballot had to be thrown out
> because it was filled in incorrectly.  Last year, we had to hold a
> second round of balloting because we did not get five people who
> received enough votes the first time.

The reasons for the current voting method will be explained below.
I'll address the problems evident in current method first.

It's true that we've had a few invalid ballots each year. From the
times I've served as teller, I can say that the invalid ballots were
all caused because the voter did not read the instructions and just
put X's next to people's names. This can be fixed by putting a 1, 2,
and 3 next to the nominees' names and instructing the voters to circle
numbers, instead of putting in a blank next to the names and crossing
our fingers.

All of the sample ballots that I created before we instituted this
rule had numbers for circling, and though I didn't fuss about it when
the numbers were replaced by blanks, I now think blanks are a mistake
waiting to happen.

As for last year's problems, we did not have enough nominees after
Bill Moran and I dropped out. To their credit, Brian Seklecki and Bob
Turley mounted write-in campaigns, but the majority-consent restraints
kicked in when neither one got their names on over half of the ballots
in the first round.

The solution to this problem is NOT to drop the majority-consent
conditions, which would be easy but foolish. We don't want vote
packing to occur, and majority consent does prevent this, despite the
extra complexity it may cause.

The solution is simply to nominate more candidates. We've always had
more than enough nominees each year, except for last year. And
frankly, that competition is healthy. We've also had good candidates
who didn't take it too personally whether they won or lost. I think
the current voting method helps keep the elections amiable, because
voters can consent to candidates and still favor others.

In fact, one of the conditions that Beth Lynn gave Mike O'Connor and
me when we wrote the rules was that every candidate must get an
up/down vote. I think the current method does this well without undue
complexity. If we moved to an approval voting method, voters could
still give each candidate their due, but could not weight their
favorites. We'd likely end up with a lot more ties using that method,
and resolving those ties would throw more randomness into the method
than we ought to tolerate.

> In addition, vote-counting has been time-consuming because tellers
> have had to be trained in the system each time and learn on the
> job.

This would happen regardless of the voting method we use. The tellers
must make sure all ballots are valid and are cast by members in good
standing, tally the votes, put together a report and announce the
results. They need training for that anyway. As far as the voting
method goes, the only complexity added to the tallying comes from
measuring majority consent and breaking ties.

> The new language simply means the top five vote-getters would be
> elected.  Ties would be broken randomly.  (Also, someone in a tie
> could choose to withdraw and hand the election to the other
> candidate.)

> Any suggestions for improvement are welcome.

I've also got to add that Robert's Rules discourages the use of
plurality voting. According to Wikipedia (I don't have a copy of
Robert's readily available), Robert's recommends a preferential system
of block voting. I personally wouldn't recommend it, but it's an
option.
http://en.wikipedia.org/wiki/Preferential_block_voting

Other Wikipedia links:
http://en.wikipedia.org/wiki/Range_voting (a variant of which we currently use)
http://en.wikipedia.org/wiki/Plurality-at-large_voting (Vance's
proposal, I think)
http://www.rangevoting.org/ (because we could get way more complex if
we wanted!)

Anyway, there's my two cents.

tl;dr, right?

-- Dave O