[EM] Combating the Approval Burr dilemma

[Abd ul-Rahman Lomax]

At 09:46 AM 11/16/2006, raphfrk at netscape.net wrote:
>A1 then wins as A1's supporters defected on the agreement even
>though A2 was the least supported candidate and a minority of
>the A party liked the candidate.
>
>This incentive could result in B winning if both A2 and A1's
>supporters did it, which results in a minority candidate winning.

The "Burr dilemma" is not a voting system problem, it is a party 
politics problem. The voting system, quite properly, is unconcerned 
about internal party politics. Approval Voting is not, in itself, 
party-list PR.

If a group of voters agree with another set of voters to vote in a 
certain way, and some of them fail to perform on the agreement, it is 
absolutely no concern of the election method. It is not an election 
method failure, period.

Secret ballot makes it impossible to enforce such agreements. The 
solution for a political party is to not rely on such agreements, 
rather the party should conduct itself such that its supporters are 
motivated to approve its candidates. If, instead, some of them prefer 
to vote for someone else, this is their complete right. This is why 
we have candidates on the ballot, not parties, per se.

If there is a faction within a party that wants B to be elected, they 
either have enough votes in that faction to do it or they don't.... 
Not our problem, I'd say.