[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Condorcet Voting and Supermajorities (Re: [CONSTITUTIONAL AMENDMENT] Disambiguation of 4.1.5)

I'm about to leave town for the weekend, so I don't have time to answer 
too many of these in detail.  For now, I'll respond to one comment by 

> I'm uncomfortable saying if I've agreed to this.  The Smith/Condorcet
> criteria that Buddha posted is not something I've agreed to.  So,
> discussion could happen here, to make sure we were using the same
> terminology.

Yes, let's be clear on terminology.  The Smith/Condorcet METHOD I 
posted is not a criterion, but a method of choosing a victor (in 
non-supermajority situations).  It happens to be a method that meets 
both the Condorcet Criterion and the Smith Criterion:

Condorcet Criterion: If there is an undefeated option (in pairwise 
contests), that option should be the winner.

Smith Criterion:  The winner should come from the Smith set.  The Smith 
Criterion implies the Condorcet Criterion, because if there is an 
undefeated option, the Smith set consists solely of that option.

Do you have a problem with these criteria in NON-Supermajority 

The Smith/Condorcet method (as well as Sequential Dropping, Schwartz 
Sequential Dropping, Tideman) all meet both of the above criteria.  
Plurality and IRV limited to the Smith set also meet both criteria.  

In a Supermajority situation, the big question becomes: What does a 
"supermajority" mean in a multi-option election?  In a single-option 
election, it's easy:  More "yeas" than "nays" by a supermajority.  I can even see that 
extended to an Approval voting system:  an option -can- win by a N:M 
supermajority if the ratio of "Approved" votes to "Not approved" votes 
is at least N:M.  This leads to another possible criterion:

Supermajority Criterion:  If any option requires a N:M supermajority, 
then if more than M/N of the ballots disapprove of the option, then it 
should not win.

Does that sound reasonable?  Please keep in mind that I haven't defined 
what "approved" and "disapproved" means on a preferential ballot.

One suggestion I've heard recently says that to vote "sincerely" in an 
Approval election one should vote "approved" on any option one prefers 
to the incumbent (which would be the status quo) and "not approved" on 
any option one prefers the incumbent to.  This is obviously debatable, 
and I don't know if I accept it myself.  It can, however, be turned on its 
head and used to get an Approval election out of a preferential 
election.  It is unclear of the "status quo" option should be "No" or 
"Further Discussion".  This is also debatable.

How about this Supermajority Election proceedure:

1) Find a winner using some method that meets both the Smith and 
Condorcet Criteria (exact method still under debate).

2) If the winner has a supermajority requirement, compare the winner 
with the "status quo" option.  If it defeats the status quo by the 
supermajority requirement, then it wins, otherwise "default" wins.

Now I really must be going...


     Buddha Buck                             bmbuck@14850.com
"Just as the strength of the Internet is chaos, so the strength of our
liberty depends upon the chaos and cacophony of the unfettered speech
the First Amendment protects."  -- A.L.A. v. U.S. Dept. of Justice

Reply to: