[EM] Meta-criteria 9 of 9: Conclusion

[robert bristow-johnson]

On May 7, 2010, at 11:48 AM, Kevin Venzke wrote:

> --- En date de : Ven 7.5.10, Jameson Quinn <jameson.quinn at gmail.com>  
> a écrit :
> Thanks for your response. You have some interesting points about  
> utility not necessarily being summed.

well, *some* sense of utility must be summed if all voters are to be  
treated equally and interchangeably.

> As to whether the majority principle is fundamental even though it  
> can contradict utility: in my schema, you'd defend such a position  
> with arguments grounded in legitimacy and/or expressivity, and quite  
> possibly couched in terms of fairness/honesty (that is, strategy).
>
> For me, the argument is that majority usually *is* consistent with  
> utility, and you won't
> often have any better indicator of utility.
>
> But ignoring majorities will tend to create legitimacy and strategy  
> problems as well.


totally agree.  if the majority candidate is properly identified and  
elected, the only strategy i can bring to the poll to help the  
political interest i have is to vote for the candidates i believe hold  
such interests.  but, if somehow, a minority candidate might be  
elected, then i have to think hard if voting for the candidate i like  
best actually helps my political interest.

and the only issue for me (and why i haven't gotten involved in the  
minute discussion) is how to determine such a majority will with  
multiple candidates (or parties).  for example, i now see the real  
problem that the Brits have.  with Labour, Lib Dem, Sinn Fein, Plaid  
Cymru, Social Dem-Lab, and Green (adds to 327 seats, a working  
majority in Commons, Brown should offer the PM to Clegg and see if he  
can get everyone to jump on board) and with just Labour and Lib Dem  
getting 52% of the popular vote, it's pretty clear to me that the  
majority of voting Brits are not Conservative as the Tories would have  
you believe.

but, of course one can ask "How do you *know* that Lib Dem and Labour  
have more in common with each other than with the Tories?", and i  
would say that we don't with FPTP but we can find out with a ranked  
order ballot.  in my town it would be just like asking "How do you  
know the Progressives and Democrats have more in common with each  
other than with the Republicans?"  We know, and a ranked ballot would  
collect that information.

and since everyone agrees how the majority (of equally-franchised  
voters) is measured between two candidates (FPTP's "simple majority" =  
majority), whatever method we come up with for more than two  
candidates should leave the relative majority determination between  
Candidates A and B unchanged when more candidates are included.  if  
there is no cycle, only Condorcet does that.  only Condorcet (again,  
assuming no cycle) defaults directly to "simple majority" when the  
race degenerates to two candidates.

to me, the only issue is how best to resolve a Condorcet paradox, if  
one should rarely pop up.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."