[Election-Methods] elect the compromise

[Abd ul-Rahman Lomax]

At 06:27 PM 9/2/2007, Forest W Simmons wrote:
>I'm not sure about the "most" but it does seem true that at least some
>of the more ego stricken candidates would be less willing to compromise
>than their supporters.  [However, in the challenge, if C voted for
>himself above A or B, it didn't register.]

Well, as stated, there is no C faction, if we name factions by their 
first choice. Yet, contrary to this, the universe of options is narrow.

Negotiationof compensation, as I've come to realize, could shift the 
utilities by essentially modifying the choices. I.e., instead of the 
choices being A, B, or C, they would become

A with compensation to the 45% faction of $X
B with compensation to the 55% faction of $Y.
C with compensation to the 45% faction of $M and the 55% faction of $N.

If the given ratings as given are commensurable utilities such that 
they can be meaningfully summed, we immediately recognize C as the 
efficient compromise, requiring the smallest total compensation. If 
we assume that whatever goes to one faction, net, must come from the 
other faction, then there is a net utility for each faction that is 
equalized, with the proper values of X, Y, M, and N.

Jobst looked at this in a restricted way, and wondered why the B 
voters should compensate the A voters, if the best compromise was 
being chosen. But the benefit of the "best compromise" -- or the loss 
of it, as could be the case -- is not equally distributed. Here the 
majority is losing and may, indeed, be able to force their outcome. 
However, justice is what I was considering, not power as such, even 
though being in the majority is power.

And justice would clearly indicate the most just outcome as the one 
which evenly distributes benefit or loss, other things being equal. 
(i.e., we might assert some distortion if one faction was wealthier 
than another, for example, but that would be noise in this study. 
Indeed, it could shift utilities, but we would have complicated the situation.)

What negotiation would do is to introduce a whole family of other 
options than the raw choices of A, B, or C. If it's a public facility 
being chosen, as the example I used, the proposal could be modified 
such that it was funded by a tax that was different on different 
districts. Or the whole thing could be done voluntarily, outside of 
government or tax funding of compensation. It all, ultimately, comes 
out of the same set of pockets, though.

If the negotiation is voluntary, it should, if done efficiently and 
the factions are well-represented in it (delegable proxy, anyone?), 
shift the utilities so that they are maximized for all, rationally, 
all voters would then support the outcome developed through 
negotiation. The proof of this, of course, would be the election 
results, but the negotiations could indeed set up preset compensation 
that depends on the election outcome.

Folks, this could be done *now*. It does not depend on changes in 
law. It does not buy votes, for the compensation is dependent, not on 
any individual vote, but on the outcome. "If this community passes 
this initiative, I will donate $X to a fund for community benefit," 
would be legal. "If the community center is built at C, I, as trustee 
for the B voters, will pay $X to every citizen who offered to pay $Y 
if A is elected." And the funds have been collected, or pledged on 
security, and then paid according to the outcome. Nobody is required 
to participate in this, and nobody is paid according to how they 
voted. You get paid, however, if you made a binding bid. (It's 
possible that the bids could be individual amounts, and thus even 
where utilities were not evenly distributed in a group, they could 
*all* be equalized. If you bid more for your option, you are implying 
an increased value to you of that option. "Which would you rather 
have, A, or $X?" If you bid inaccurately, you may get what your bid 
indicated! Some people, if they know how the bidding is gone, may 
exaggerate their bid if they are confident that the other option will 
be chosen, but they are running the risk that this will cause their 
bid to be accepted by the electorate. And the negotiators could 
specifically cap the compensation, if they suspected this was going 
on. With a travel distance utility underlying the process, 
exaggerated bids would be pretty visible.)

In any case, I find the idea of shifting preferences through 
compensation based on outcome to be quite interesting. If only a 
small number participate, it would improve election outcomes a small 
amount. If large numbers participate, it could make democracy 
function more efficiently and fairly without changes in law. Those 
kind of solutions truly interest me, because they bypass the 
entrenched interests which can often block changes in law that 
threaten their inequitable power.