The Oregon Consensus Initiative

[Steve Eppley]

Rob Lanphier wrote (in election-methods-list at eskimo.com):
> The Oregon Consensus Initiative that you posted appears to be 
> a form of Approval Voting for the primary.
-snip-
> Approval voting for a primary while still first-past-the-post in
> the main election would have some interesting effects.  It brings
> in some interesting strategy dilemmas that I'm having trouble
> deciding if they would be a net-positive or a net-negative.

Yes, it's a mixed bag.

> I think it's actually an interesting initiative, and I wish them
> well. The only concern that I have is that it doesn't seem to
> limit the final election to only two candidates, and so the
> strategy implications boggle the mind.  

Yes, primary elections exist because parties need some way to winnow 
all but one of their contenders out of the general election, since 
the "vote for only one" method currently used in general elections 
is so badly flawed.

If the Oregon initiative passes, each party will probably find a way
to apply pressure on all but one of their "voter-approved" candidates
to drop out of the general election, in order to avoid the "vote for
only one" method's spoiler effect.  The voters can't force candidates
to run!  Since parties need to winnow their field down to one if the
general election method isn't fixed, they'll just evolve a new way to
winnow.  I predict they'll eventually learn to apply overwhelming
peer pressure to drop out, either against candidates unfavored by the
party elite or against all but the voters' most-approved candidate
(assuming this approval primary system isn't revoked by the Oregon
voters).

> In a predominantly two-party system, it might make sense to vote
> for *all* of the opposition, and only the best in your own party,
> so that the opposition spends all of its time running against
> itself, and your party's sole candidate wins by default. 

But if *none* of your parties' candidates survive the primary,
because voters of other parties vote No on your parties' candidates,
then your party won't have a sole candidate.  (The initiative's 
provision for new candidates who would need a majority of votes 
in the general won't be very helpful.)  So the optimal strategy
will depend heavily on expected voter turnout in the primaries and 
the ability to organize the voters' voting behavior.  (Parties which 
appeal to people who enjoy diversity and tolerate uncertainty will 
probably have a much tougher time organizing than parties which
appeal to people who are less tolerant and more dogmatic.)

To me it looks like the primary election would eventually become the
decisive election, once the voters figure out they need to turn out
at the polls in better numbers and learn how to "properly" strategize.
If there's a party (or coalition of parties) which has a majority of
the voters, they should simply disapprove all the other parties'
candidates.  This wouldn't be as bad as it sounds, since the other 
parties ought to lose anyway, according to majority rule.

But the brief period allowed in this initiative for test-driving 
it looks to be insufficient for the voters to learn how to use it.  
The one year period won't be enough to increase voter turnout
significantly, and they won't learn that they're not really voting
"disapproval"--they're voting to eliminate candidates from the 
general election.  (These are different concepts, and it will take
time for the voters to see through the popular misconception about
approval and disapproval.  The proponents of the initiative also
appear to believe this misconception that "approval" and "disapproval" 
have some absolute meaning, even though they really have meaning 
only as preferences *relative* to some alternative, such as the
alternative of holding a new election with some new candidates.  
If there are possible results besides electing one of the
candidates--such as the result of all the voters voting No on all
the candidates--then these outcomes could and should be explicitly
included on the ballot as choices to be ranked relatively, instead 
of misleading the voters with constructs like "approval and
"disapproval.")  

The proponents of this initiative appear to be responding to a desire
by voters to have more "good" choices to vote on--dissatisfaction 
with the two-party system which evolved in response to flawed
general election voting methods, but they haven't yet caught on that
the best way to do this is to use a voting method in the general
election which doesn't have a spoiler dilemma.  The problem is
solvable with either proportional representation methods (but only 
in legislative elections) or with voter preference orders 
(a.k.a. ranked ballots) tallied by Condorcet's method.  These two
tools are better at electing *majorities on multiple issues*, 
which is the overlooked but important goal of democratic elections 
of people who will win power to decide multiple issues.

The proponents also appear not to have a solid grasp of the meaning
of majority rule: they've freely mixed the concept of a majority
preference for electing one candidate instead of another candidate
with the unrelated concept of a majority preference for electing
some candidate instead of holding a new election, and they don't 
pay attention to the important point that there are majorities on
multiple issues.  It may be possible to elect multiple majorities on
the issues important to the voters, especially when the absence of a
spoiler dilemma in the voting method creates clear opportunities for
new candidates whose platforms match more of the multiple majorities
than other candidates' platforms.  Failure to elect majorities on 
the issues important to the voters (in other words, "submerging" of
majorities) is what drives public dissatisfaction with government.

I don't endorse this initiative, even though I sympathize with the
goals of its advocates.  The test-drive period is too short for the
people to learn how to use it, so it may sour the people on election
reform.  And it doesn't solve the underlying problem as well as
reforming the method used in the general election. 

---Steve     (Steve Eppley    seppley at alumni.caltech.edu)