[EM] Re: rank/approval cutoff ballot

[Araucaria Araucana]

On 20 Jul 2005 at 18:51 UTC-0700, Abd ulRahman Lomax wrote:
>At 03:47 PM 7/20/2005, Dan Bishop wrote:
>>[...]I think a good solution would be for elections to have two rounds:
>>
>> 1. A qualifying primary, done entirely with write-in ballots, and
>> counted using Approval.  Candidates with a sufficient number of
>> votes would advance to...
>>
>>2. A runoff election, using ranked ballots.
>
> Not a bad idea. Runoff elections have an additional advantage: an
> opportunity for a reduced field to compete. With fewer candidates,
> there is more ability of the electorate to see who they
> are. (If....)

This now sounds like a primary and general election.  That might be
one way to spin it.  Consider, for example, that Washington State's
top-two runoff was declared unconstitutional last Sunday:

        http://seattletimes.nwsource.com/html/politics/2002384176_webstateprimary15.html

This list was discussing approval-based primaries back in early 2004,
if I recall correctly.

Back then, I was thinking about something like this (borrowed from
Jobst's Imagine Democratic Fair Choice page on electowiki):

                   I     |  I also
                support  |  approve
               directly: |    of:
       ------------------+----------
       Anna        X     |     O
       Bob         O     |     O  
       Cecil       O     |     X
       Deirdre     O     |     X
       Ellen       O     |     O
       ------------------+----------
                 (vote   | (vote for     
                  for    |  as many
                exactly  |   as you 
                  one)   |   want)


The real trick is deciding what to do with the results!

In Washington state, parties objected to the top-two runoff, because
some supporters of a strong candidate might vote for the weaker of the
opposition.

If the approval cutoff is too limiting, the approval primary would
still be susceptible to that strategy.

So there should be some tally method that encourages sincere (and
generous) approval of alternatives.

The best way to do that is to allow the approval winner to win
outright in some circumstances.

What about this primary method:

     Approval ballot (as above)

     A candidate with more than 50% first place votes wins outright.

     Otherwise, if there is at least one candidate with more than 60%
     (75%?  what is the safe cutoff?), the approval winner wins
     outright.

     Otherwise, all candidates with more than 1% approval are advanced
     to the general election and listed in order of first-place vote
     totals, with approval scores also noted.

     The general runoff election would use some ranked ballot method,
     e.g. a ranked approval ballot with DMC tally :-).

     Write-ins would be allowed on the approval-primary ballot, but
     not on the general election ballot.

     The primary would have to be close enough to the general election
     to be meaningful -- within 60 days, for example.

A 1% cutoff is similar to current rules about what parties are allowed
to be listed on the ballot.

The parties would then not be able to protest that their chosen
candidate was being unfairly excluded -- approval less than 1%
indicates that they are on the fringes anyway.  They could still enact
some legislation to differentiate their sanctioned candidate from
mavericks.

Q
-- 
araucaria dot araucana at gmail dot com
http://www.metafilter.com/user/23101
http://wiki.electorama.com/wiki/User:Araucaria
Q = Qoph = "monkey/knot" -- see http://www.ship.edu/~cgboeree/alphabet.html