I agree that Approval is more real-politick, and (in relation to the other
discussion) that proportional representation will never get off the ground
in the US.
As a last resort (and I mean a really last resort - excuse the grammar),
where IRV is about to take hold, you could try proposing to whack Condorcet
on the front (ie, find a Condorcet winner. Where there is none, do an IRV
count). Not ideal, but hey, even the IRVies might agree to that one.
Mike Ossipoff wrote:
When I said that, in addition to probably being easier to get,
Approval will probably work better in use than Condorcet will,
I didn't mean to imply that Condorcet should never be proposed.
So let me say what I believe should be proposed when:
1. Both Approval & Condorcet should be proposed where IRV is proposed.
That's because the IRVies have already got some people convinced that
they want rank balloting, and, additionally, preventing IRV is the
main goal there. So offer both, to maximize the liklihood of
beating the IRVies. Opposing IRV's money, membership hordes, and
promotional zeal, on the issue of how to count ballots, sounds like
a losing proposition. Therefore, even when proposing both Approval
& Condorcet where IRV is being proposed, Approval should be the main
proposal, the main recommendation under those conditions.
Of course when advising an initiative committee, or a legislator
who asks for all the facts, one should mention both, rather than
withold information, but, except under the previous paragraph's
conditions, it should be recommended that a certain one of those 2
methods be chosen instead of the other.
2. Approval should be be the proposal when it's a special one-time
sort of opportunity to get a proposal considered. In that instance,
we want the one most likely to be adopted, and most likely to be
understood & used well by the voters. Don't take a chance of blowing
a big one-time opportunity.
3. In a few particular experimantal trials, where IRV isn't being
proposed, and it isn't a special one time sort of oppotunity to get
a proposal considered, Condorcetists might want to try for the best,
as they believe it to be. I don't suggest that Condorcet should
always be the proposal whenever IRV isn't involved and it isn't a
special one-time opportunity. Maybe, instead, Condorcet should be
the proposal in a few selected experiments, with Approval proposed more
often.
Of course, when proposing reform to a group that's already using
rank-balloting, Condorcet would be a better proposal, if that group
has shown that it has the courage to vote sincerely with rank-balloting.
Likewise with groups or areas where people are already strongly
interested in rank balloting--if they can be convinced about how
to count the ballots--a _big if_. Rank balloting seems to elicit a
more enthusiastic response than Approval, when mentioned to someone
whose primary concern isn't cost, and who wants reform--but, regrettably,
that person almost invariably thinks you're talking about either Borda
or IRV. When they're told that's no good, and told about pairwise
counting, there sometimes seems to be a lessening of the enthusiasm,
because the proposal isn't what they thought it was. If rank balloting
is what they can get with, and if they accept Condorcet instead of
Borda or IRV, then Condorcet would make sense--if the entire group
is like that. But if how to count rank ballots is a hopeless or
longlasting disagreement, then Approval would be the better proposal
then too.
Mike Ossipoff
