I sympathize with Rob's complaint about the meaning of approval versus disapproval in Approval. There is a trade-off, a price for the simplicity of Approval. However, DMC takes the pressure off this question because in DMC, approval is only used to eliminate enough of the Smith set for an unbeaten candidate to emerge. Rob and others would like a definite question that you could answer yes/no to for each candidate. In the zero info case this question could be, "Is this candidate better than Gerald Ford?" Also, there could be elections where the status quo is the default, meaning if no candidate gets more than 50% approval, we continue with the current administration. In that contxt the question (for candidate X) is, "Do you like candidate X better than the default?" Another possible question is, "Would you rather have this candidate elected than having to return to the polls for a runoff?" Another alternative that I have suggested is comparison with the strongest lottery: (1) Proposed lotteries (a priori winning probabilities) are submitted by anybody who wants to. Certain standard lotteries are thrown into the pool to make sure that the set is highly competitive. (2) Ordinal ballots are collected from a truly random official sample of the registered voters. (3) Joe Weinstein's method is used to intercalate each of the lotteries into each ordinal ballot. (4) The lottery whose smallest score against any of the real candidates is the greatest, is set up as the comparison lottery L. (5) The approval question for the complete electorate is (for each candidate) would you rather have this candidate elected than have the winner chosen by the lottery L? (6) If no candidate gets more 50% approval in response to this question, then the lottery is used to pick the winner. This is likely to happen only in highly fragmented electorates, where factions are disinclined to compromise or cooperate with each other. I would prefer this variant of Approval over almost any other method, except that not even election methods aficionados (let alone the voters) are ready for it. Forest
<<winmail.dat>>
---- Election-methods mailing list - see http://electorama.com/em for list info
