Iterated Condorcet & Criteria

[Mike Ossipoff]

It seems to me that it could be said that Iterated Condorcet doesn't
meet GMC, if "has a majority against him" is taken to mean "had
a majority against him before the ballot modifications". But
if a candidate who initially had a majority against him is
no longer considered to have a majority against him if some
of that majority have opted to raise him to 1st choice, then
it seems to me that Iterated Condorcet does meet GMC.

That latter interpretation of the language makes sense to me.
If you've opted to raise Clinton to 1st choice, then you are
no longer choosing to make Clinton have a majority against him.

So, what would happen if everyone extended that option to every
candidate doesn't bother me, since no one has to do so. Anyone
who wants to can refuse to use that option, or, if they use
the option, could be allowed to choose exactly how far down their
ranking they want to apply the option.

So if the 1st place promotion is optional, then the method retains
Condorcet's advantages, as far as your & my right to LO2E protections
is concerned.

I suspect that that method could be said to fail LO2E-1, in
the same way that it could be said to fail GMC, but that
if "majority against" refers to _current_ majority against,
then, as with GMC, Iterative Condorcet is ok by LO2E-1.

But it seems to me that, in a way that doesn't depend on
wording interpretation, that method meets LO2E-2 in the same
way that ordinary Condorcet does. In fact, I'd like to change
that from a conjecture to a statement.

***

I've heard some people say that Approval violates 1-person-1-vote,
but I'm not sure whether I've heard that from electoral reform
organization leaders. They don't like Approval, but they usually
give other reasons. But if they say that IR-1 violates 1-person-
1-vote, then they'd pretty much have to say that Approval violates
it too.

That's true, that we use a form of Approval now, to choose from
among several competing initiatives. I guess the thing that
distracts people from the Approval nature of that initiative
system is the fact that it's worded in terms that cause 
people to regard it as entirely separate y/n elections, and
seems to make people forget that the various initiatives' "yes"
votes count against eachother, just as in Approval. But I guess
it's felt to bo ok. In the current initiative system, or in
Approval, there's nothing wrong with letting people votge for
more than 1, and maybe the initiative system could be used
to help show people that that's ok.

When it comes to the most modest sw reforms, maybe the
initiative system--an Approval count of "y" votes between
all the alternatives that got more "y"s than "n"s, would
be the easiest & quickest to get enacted, since it's already
in use. But of course, predictably, there would be plenty
of columnists & politicians, and surely a few professors too,
who'd be claiming that what's good for competing initiatives
isn't good for candidates. So I don't know if it would really
be easier to enact than just proposing rank-balloting, counted
by Condorcet. People seem to easily understand the desirability
of rank-balloting. And people's obsessive need for negative
voting should make it obvious to them that counting "votes-against"
will give them full negative voting powers without compromising
their favorite.

I shouldn't criticize the need for negative voting, because
I too would like to be able to fully vote against someone
without selling-out my favorite. Those of us to whom LO2E voting
isn't as important would still like to have a voting system
that gets rid of the LO2E problem. For ourselves, as voters,
but maybe especially for those other voters who are so
completely cowed by the LO2E problem.


Mike


 


 











--