I agree with the predicted behavior to determine where to candidates will run. But I disagree with your last conclusion, maybe because I did not express myself clearly. Every candidate in every "astrological" district
would come up with his/her personal matter. So the referendum are not about one subject but about setting priorities between subjects. Let me show a simplified example: parties \ districts 1 2 3 4 5 ... Ancestral health (public) gay wedding (against) drug legalization (against) abortion (pro-life) health(public) Bienheureux free-trade(yes) drug legalization (against) 35 hours (against) Kyoto (against) nationalizations (against) Classical health (mixt) Kyoto (against) nuclear power (for) prop represent (against) free-trade(yes) Dynamic drug legalization (for) gay wedding (for) 35 hours (for) Kyoto (for) gay wedding (for) Evolutive health (private) drug legalization (for) nuclear power (for) health (private) nuclear power (against) Imagine 100 districts, some with direct confrontations, others not. Every district is a poll, with equivalent population samples as input. Suppose voters like "me": health (public), gay wedding (indifferent), drug legalization (for), abortion (pro-choice), free-trade (yes), 35 hours (for), Kyoto (for), nationalizations (for), nuclear power (against), prop representation (for). Many people may think like "me" but with a different order of priority... In district 1: question is "what is my priority ?" A, B or D? In district 2: E!!! In district 3: D!!! In district 4: D!!! In district 5: A or E ? and so on... Suppose now the major stream (a sufficient plurality) has this order of priorities: free-trade (yes) > prop representation (for) > health (mixt) > drug legalization (for) > abortion (pro-choice) > Kyoto (for) > nationalizations (for) > 35 hours (against) > nuclear power (against) > gay wedding (against) Thus, even if no party agrees with the major stream, SPPA can capture it. It would elect B1, C5, E2, D4 and maybe C1 instead of B3 depending on how much the health debate is more enthousisastic than the 35 hours debate. Thus, I think SPPA is equivalent to numerous small referendums to order priorities among different sets of subjects and (positions). This is how I see PR goal. Steph Markus Schulze a �crit : > Dear Stephane, > > > The candidature process you describe happens already > > for any other model. No? > > When a given specialist of the party in power is very > popular, then the opposition party will not spend much > money trying to defeat this person. It will rather run > its own specialist in a different district and run an > unimportant straw person against the specialist of the > party in power. It is a matter of fact that when you > want to win an election then it doesn't make much sense > to attack the strong candidates of the party in power; > it makes more sense to concentrate on the swing > districts (i.e. those districts where the candidate of > the party in power is neither too strong nor too weak). > I don't believe that when SPPA is being used then > elections become "equivalent to numerous small > referendums". > > Markus Schulze ---- Election-methods mailing list - see http://electorama.com/em for list info
