[EM] DMC the greatest Condorcet? I wish, but doubt it...

[RLSuter at aol.com]

Warren,

You need to come down to earth. It is impossible to mathematically
prove a sociological law such that "x election method leads to
2-party domination." I challenge you to find a single political
scientist or sociologist who thinks you can. Societies are simply
too complex to make such proofs possible. For beginners, you
seem to have totally neglected the possibility that parties can
evolve in unpredictable ways in response to all kinds of things,
including election laws and the methods incorporated in them.
The reason this hasn't happened very much until now is that
party leaders have rarely understood election methods and
their consequences very well and therefore haven't done a good
job of taking them into account in developing their strategies.
Nevertheless, any election method you can think of, including
plurality, could be overridden by a currently minor party with
large potential support using a strategy that does a good
job of taking election laws and methods into account. The
result could be that party displacing one of the currently major
parties (and thus, still 2-party domination), but it could also
be three or more parties each with substantial political power
and many elected officeholders -- enough to prevent any one
or two parties from being dominant for more than a few
election cycles.

This is certainly not to say that election methods and systems
of representation are not important. They obviously are and
I have long been in favor of improving on plurality voting and
single member district representation. But they are not nearly
as mathematically determinative as you claim to have proved.

-Ralph Suter

In a message dated 8/31/05, Warren Smith wrote:

<< Earlier on EM I basically constructed a mathematical proof that all
 Condorcet methods lead to 2-party domination.  (It may be seen at
 http://math.temple.edu/~wds/crv/CondStratPf .) >>