On Aug 30, 2005, at 03:49, Warren Smith replied to Jobst Heitzig:
So you suggest that when candidate A gives $200000 to 1 voter and
nothing to the other 99 voters, but candidate B gives $1000 to each of
the 100 voters, then candidate A should be considered best for society.
--YES!! (at least, if utility=money.)
I agree that money to some extent correlates with utility, satisfaction
and voting behaviour. Doesn't e.g. range voting however respect more
the one (wo)man one vote principle. I mean that the votes and also
utility might to some extent reflect the following formula.
Let's say that the largest sum of money the candidate A's favourite
voter (a1) can imagine to get (from various sources in her normal life)
is $1,000,000 (she is a billionaire). The 100 B candidate favourites
(b1-b100) are poorer and can imagine only $2000. Voter a1's utility can
now be estimated to be $200,000/$1,000,000 which is 0.2. In range
[0,100] she would vote A:20, B:0. Voters b1 to b100 would vote B:50,
A:0 (following the same formula). Candidate B will get 5000 points
which is better than the 20 points candidate A will get.
Btw, if a1 would imagine only $2000, she would be surprised of getting
$200000, and would change her understanding so that it is really
possible to get $200000 somewhere. After quickly getting used to this,
she would now vote A:100, B:0. The result of the election would stay
the same (B elected).
In short, what I mean is that the (simplified estimate of) utility
_from_the_voting_method_perspective_ does not correspond to the actual
sum of money that is donated to the society but to the average impact
it has on the voters. Warren (as a range voting proponent) should have
answered NO. Right? (I didn't check all the history of the debate
though)
BR, Juho
----
Election-methods mailing list - see http://electorama.com/em for list info
