[EM] Low SU CWs?

MIKE OSSIPOFF nkklrp at hotmail.com
Sat Jan 24 00:01:01 PST 2004

Some methods are advocated for maximizing social utility (SU). With the 
assumption of sincere voting, Borda & CR are suggested for that purpose. 
Sincerely voted CR would do the best job. Sincerely voted Borda would do a 
better job than any other rank method, if we assume that utilities vary 
linearly in a ranking (maybe they don't).

Some methods are advocated for majority rule, or getting rid of the 
lesser-of-2-evils problem. Condorcet for example. Or Approval.

When Condorcet chooses a candidate who has low SU, that isn't a wrong answer 
if the goal is to protect majority rule or avoid a lesser-of-2-evils 
problem.  No method can always do everything, so we choose what we want.

As I said, spatial models show Approval and Condorcet doing much better than 
IRV in terms of SU. So IRVists would do better to not bring up SU.

That means that though strategic CR doesn't do as well as sincere CR, by SU, 
strategic CR still does significantly better than IRV.

I've pointed out that in a real Condorcet election it would be very unlikely 
for an accross-the-board despised candidate to have the voter median 
position all to himself. Someone less despicable would be there too, and 
would win instead.

But that objection can be answered in other ways too:

Some say that maybe the middle candidate is unknown to everyone, and so the 
extreme voters all rank Unknown Middle over their rivals at the other 
extreme, just because they don't know Unknown Middle. Mr. U.M. has very low 
SU, because he'd just a mediocre space-holder for voting against someone 

But are we forgetting something?: A majority of Right voters have said they 
like U.M. better than the Left candidates. A  majority of the Left voters 
have said that they like U.M. better than the Right candidates. They 
wouldn't rank him over someone else unless they llked him better, or 
disliked him less. If they're order-reversing, that's a separate problem, 
Condorcet's only problem, which isn't as bad as IRV's problems, for the 
reasons I've already discussed.

Who are we to take it upon ourselves to overrule those majorities?

The voters are adults. They know whom they like more than whom. Who are we 
to judge their reasons and disregard their stated preferences?

Another thing: If U.M. is really bad in some way, then someone will most 
likely find out and tell people. It's unusual for any candidate to be free 
of negative campaigning. If someone would rather their candidate win instead 
of U.M., then they'll find out the dirt on U.M. and will tell people.

Or at least someone, maybe a rival candidate, maybe someone else, will say, 
"Hey, we don't know anything about this guy--don't be so sure that he's 
better." Then the voters will judge for themselves, because it's their 

Mike Ossipoff

Check out the coupons and bargains on MSN Offers! 

More information about the Election-Methods mailing list