[EM] Re: Blake's margins arguments

[Bart Ingles]

Just to further muddy the waters on the definition of "majority", note
Duverger's use of the term as apparently synonymous with "plurality":
http://www.janda.org/c24/Readings/Duverger/Duverger.htm

(BTW the "two-ballot majority system" Duverger discusses is NOT the
Runoff method, since there is no fixed elimination rule.



Alex Small wrote:
> 
> On the subject of majorities and Margins vs. Winning Votes, my
> understanding of the matter (which I will deliver without prejudice, since
> I have no strong opinion on the matter) is this:
> 
> Say that two of our pairwise contests in an election are:
> 
> A>B  51-49
> B>C  45-40
> 
> (Assume there are 100 people casting ballots, at least 3 candidates, and a
> cyclic ambiguity to resolve.)
> 
> A majority of the people casting ballots have said "We prefer A to B."  A
> plurality of the people casting ballots have said "We prefer B to C", with
> 15 voters abstaining.
> 
> A winning votes advocate can say "A has majority support relative to B.
> That's a stronger defeat for B than C's defeat, since B only beat C with a
> minority (plurality) of the people casting ballots."
> 
> A margins advocate can say "Yes, A beats B with a majority of the people
> casting ballots.  But, B beats C with a _majority of the people
> participating in the B vs. C contest_.  Since the other people voluntarily
> abstained from that contest, we shouldn't consider them, we should only
> consider who got a larger percentage of the people participating in each
> contest."
> 
> I hope that clears some air on the debate over the word "majority."  It
> all comes down to "majority of whom?  People casting ballots or people
> expressing preferences with respect to a particular pairwise contest?"
> 
> I don't claim to resolve any normative issues of which provides a more
> socially desirable criteria, or technical issues of which method provides
> more perverse strategic incentives.
> 
> Alex
> 
> ----
> For more information about this list (subscribe, unsubscribe, FAQ, etc),
> please see http://www.eskimo.com/~robla/em

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em