[EM] Re: ruminations on ordinal and cardinal information

[Araucaria Araucana]

On 21 Mar 2005 at 12:52 PST, Jobst Heitzig wrote:
>
>> Second point: In the USA, at least, it may be more desirable for the
>> pairwise winner to have lower approval.  This seems paradoxical, but
>> it tends to keep the winner from making overly radical changes.  
> Sorry, but I think some radical changes in the US are just what the US
> (and the world) need :-)

Beware of starting fires.  You never know which way the wind will
blow.  If I may point out, Bush believes he received majority support
and is "spending his political capital".

If 90% of the US voting population had voted, instead of something in
the 60's, Bush would have no 'majority' approval (if even elected!),
and might correspondingly be a bit more cautious and conciliatory in
his agenda.

The one-third of the US electorate that doesn't vote is sitting there
(partly) because the two other parties avoid substantive issues, but
can't be dislodged under the current system.

>
>> The US founders distrusted government enough that they put in
>> checks and balances to make the process *less* efficient.
> There is something similar in Germany: the federal government often
> depends on the agreement of a majority of the federal states'
> representatives, and this often leads to nothing happening at all...

Inaction is not always bad, especially when a behemoth like the US is
moving around.  But a Condorcet winner won't necessarily be the
blandest candidate.  In fact, I would expect the Approval Winner to be
even less controversial or confrontational.

Central support can actually lead to more action, not less.  A
centrist agenda can appeal to all sides for support, not just the
majority coalition.

In times of clear need and the right candidate to lead in that
direction, I would expect the Approval Winner to be the Condorcet
Winner.

>
> But now the main point:
>> Consider your DMC tie problem:
>> 
>>        1 A>>B>C
>>        1 B>>C>A
>>        1 C>>A>B
>>        3 A=B=C
>> 
>> With DMC, however, a fourth candidate will see the opportunity and
>> step in to fill the center -- if not in that election, then in a
>> future one:
>> 
>>        1 A>>D>B>C
>>        1 B>>D>C>A
>>        1 C>>D>A>B
>>        3 A=B=C=D
>> 
>
> Well, thank you for giving this example. Since it shows perfectly
> why I think that the Condorcet Winner (in this case D) is sometimes
> NOT a good choice at all! Most probably this D is just someone who
> has no program and says nothing but empty phrases which oppose
> noone. I at least don't think D should be elected here since s/he
> has too few approval and/or direct support!
>

This is a good demonstration of Arrow's theorem ;-).  But it is not
necessarily true that D would do nothing or have no program.  If that
were the case, no block would cast a near-top compromise vote for D.

I'm just saying that *this* worst case (no above-cutoff support for D)
results in 50% approval for D.  There's no reason (with more than 6
voters) why D might not actually garner more approval, potentially
100%.

But to get strong compromise-candidate support, D has to have a
centrist platform.  The center is not a vacuum.  Maybe I'm just being
optimistic, but I don't think you can get into the center with empty
phrases.  You actually have to stand for something, more like
'moderation in all things, all things in moderation'.  Then if you
don't waste time bickering on the extremes, you can actually
accomplish more of substance.

In any case, we're probably rehashing Approval vs. Condorcet arguments
that go back many years on this list.

I'm just trying to look beyond a single election case.

Somebody always loses a single-winner race, so if you want good
candidates to run, it behooves you to give the runners-up a chance to
look as good as possible.

By choosing the pairwise winner from set P, losing candidates in that
set (assuming a close race) still look pretty darn good.  All they
have to do to win the next time is move up one or two ranks.  And to
do that they have to broaden their appeal.

Ted
-- 
araucaria dot araucana at gmail dot com