Let me restate the problem I addressed by this: How can we answer the question of how much a candidate is worth for society (="social utility") by means of the answers the voters gave to the question of how much that candidate is worth for them (="individual utility").
Maybe the difference is that most of us aren't concerned with "how much a candidates is worth for society" (no matter what definition seems implied by "social utility"), but rather with how well a voting system responds to changes in individual utilities.
Also, I wouldn't consider the "answers the voters gave" to be utilities.
I would call them "ratings" or at best "estimates of utility". As I believe you have alresdy stated, utilities aren't directly measurable, or at best only partially so. Any aggregate of these ratings would not be a "measure of social utility", but rather a "voting system".
> [...]
There are many good arguments to use such a definition of social utility instead of the one which involved the sum: 1. The median is more robust, that is, depends less on extreme values of few voters than the sum. It is not possible for few voters to increase or decrease the median arbitrarily by increasing or decreasing their individual utility for that candidate.
In a voting system, making the median voter the dictator would certainly solve a few problems (but create others).
As a standard for measuring the hypothetical performance of voting systems, this would be of limited value-- e.g. what would it find that would be substantially different from Condorcet efficiency?
Usually Social Utility Efficiency is used as something complementary to Condorcet Efficiency. The whole point of SUE is to measure changes in individual utilities, not ignore them.
Bart Ingles ---- Election-methods mailing list - see http://electorama.com/em for list info
