Re: [EM] 04/03/02 - Adam needs a method that will win elections for his party:

[Alex Small]

Donald wrote:

>When the people have reason to support a third party they will support a
>third party.

We have plenty of reason to do so.  It is unfathomable that two only two
options, diametrically opposed in rhetoric (if not in actions) can
represent all of America.

>It is understandable that you support Approval Voting, for it is a method
>that will steal an election for a third party.

In the US, stealing elections is the job of Supreme Court Justices, not
election methods (sorry, couldn't resist an irrelevant wise-crack).

>Because the people will not support your party is not the fault of any
>method. It is the fault of your third class party.

Not entirely true.  Plurality restricts choice.  When our choices are
restricted to indicating support for only one party the decisions become as
much about strategy as about sincere evaluation of candidates and parties.

You might say that if most people really wanted a different option then
their strategies would reflect that.  However, there are two problems.  The
first is that people are risk averse.  Unless their true favorite is one of
the top two contenders they will often abandom him if the election method
gives an incentive to do so.  Because plurality (and arguably IRV) give
incentives for rapid defection, a third option never really has the chance
to make his case.

Second, once two parties are entrenched it's difficult for a third option
to attract support.  Anybody with political ambitions is better off joining
a camp that already has the legitimacy of having people in office.  He or
she can then count on endorsements, fund-raising support, brand loyalty,
etc.

Technically this is true of Approval, Condorcet, etc.  However, since those
methods do not give incentives for rapid defection during the campaign they
make it easier for people with distinct viewpoints that don't really fit
either camp.  Those candidates can at least make a case without worrying
about rapid defections.  With plurality (and arguably IRV), however, it's
almost impossible for a third option to advance if he starts in last place.

>It is the public that should have the power to decide which party is to be
>propelled up into the winner's circle. IRVing gives the public that power.
>The other methods, Plurality, Approval, Borda, and Condorcet, will favor
>one faction or another.

I didn't realize that Approval ballots have fine print saying "Votes for
Adam Tarr's favorites shall be counted twice.  Offer not available in
Delaware, Nebraska, or where prohibited by local law.  No purchase
necessary.  See election officials for details."

Seriously, all that Approval and Condorcet do (and, for the sake of
argument, perhaps IRV, although I'm skeptical) is remove the phenomenon of
rapid defections.  If you start out in last place early in the campaign,
you can continue pressing your case without defections.  You might not gain
new supporters, and it might be hard to get press attention, but at least
you don't have to worry about your supporters abandoning you (which only
further erodes credibility), unless of course new information causes them
to change their minds.

Note:  I realize that Condorcet does not satisfy FBC, and Approval
satisfies only a "weak" FBC.  However, since violations of FBC and
monotonicity are leading causes of early defections, it is safe to say that
Approval and Condorcet do not cause the defections that make it so
difficult for a third option to take the stage.  People like Perot and
Ventura, charismatic people who walk in at the right moment with the right
message (leave aside discussions of their merits for now) are rare.

Alex