One of the things that came up on the list a few days ago inspired an off-list discussion where we questioned whether certain "failures to meet criteria" might, in the real world, actually have positive results. (as heretical as it may seem to say so) Here is an example:
Let's say that we decide that parties (i.e. clustering of people into ideological groups so as to have strategic advantage) are not a good thing, as they limit our choices and drive people into opposition and conflict. Any system which is susceptible to vote splitting, then, is bad because it will force people into such clusters. However, there are other reasons -- probably less straightforward to mathematically analyze -- that people will gain strategic advantage by forming such ideological clusters, such as pooling advertising money to promote a single candidate, as opposed to having many similar candidates competing for people's attentions while dividing the money of contributors who have similar ideology.
What if an election method actively counteracted this effect? For instance, it was mentioned recently that a certain method would give a (small?) strategic advantage to clone candidates This could be considered the opposite of vote splitting. The real world effect it might have would be to discourage the removal of similar candidates, balancing out the effect I described above and resulting in even less partisanship.
In any case, I'd think a discussion of this sort of thing -- sort of a "big picture view" of why we care about the various things we are trying to achieve (or avoid) with election reform -- might better help put things in perspective and prioritize things.
Is there a resource out there that discusses this stuff in an accessible way that I haven't found?
-rob
---- election-methods mailing list - see http://electorama.com/em for list info
