[EM] Intriguing proposal & musings on strategy and autonomy

[Michael Rouse]

 ----- Original Message -----
 From: "Alex Small" <asmall at physics.ucsb.edu>
 To: <election-methods-list at eskimo.com>
 Sent: Thursday, June 13, 2002 2:30 PM
 Subject: [EM] Intriguing proposal & musings on strategy and autonomy


 > While surfing voting sites I ran across this Australian site:
>
> http://2mec.freeservers.com/
>
> In brief, they are advocating replacing the current ceremonial head of
> state (the British Crown) with a head of state elected in a more
> democratic
> manner.  The manner they propose is very intriguing.
>
> They propose a "2-member electoral college."  Use a PR method to choose 2
> electors, who must then agree with one another on whom the President will
> be.  This system would elect a President respected by at least 2/3 of the
> population.  (We can debate whether that guaranteed broad appeal outweighs
> the dangers of giving a minority veto to 34% of the population.)

This reminds me of an odd multi-stage voting method I came up with awhile
back, though the Australian version is a highly-truncated form.

1. Hold a general election using a secret ballot. Voters can either hold
onto their vote (by voting for themselves) or transfer their vote by voting
for their favorite candidate.
2. If one candidate has a majority of the remaining votes at any stage, he
or she is declared the winner.
3. Any candidate with 2^n votes (where n is the election number) or higher
advances to the next round, while those with fewer votes are removed from
further consideration. In other words, the cutoff vote is raised to the next
power of two (i.e. 4, 8, 16, 32) votes in each subsequent election.
4. During each round, remaining candidates can either hold onto their vote,
or transfer their votes to another candidate. These votes are made publicly
so that voters know how their candidate spent their vote.
5. Return to step 2 until there is a winner.

One minor problem is if a candidate cannot win but is unwilling to vote for
another, then all of their votes are lost. Those lower on the election
ladder may hate losing their vote even worse than supporting someone they
really don't care for -- in any event, candidates will have an incentive to
compromise in order to gain concessions from another candidate they are
voting for.

Another problem is the number of elections -- but the first election should
trim the possibilities from millions of voters to a few thousand candidates
at most, and each subsequent election cuts the number of possibilities by
half. With one election per day the entire planet  could have a decision
within a month (2^30 ~ 1 billion), even in the most contentious, drawn-out
decision. (The Australian version would jump from the first election to the
second-to-last election, at least in effect).

Still, I think choosing the winner from the mean Kemeny order better
reflects the electorate as a whole (Mean Kemeny winner, followed by binary
comparisons if more than one MKW, followed by Borda count as a tiebreaker if
there is *still* a problem).

Michael Rouse
mrouse at cdsnet.net

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