[EM] Voting on matters of pure fact

[Blake Cretney]

Here's another interesting situation.  Lets say you lock a group of
people in a room, and present two proposals to them.  You ask each
person to quantify how much each proposal benefits him or her
personally.  They all vote sincerely.  Maybe they're really honest. 
If the voters choose based on enlightened self-interest, the total
benefit for each proposal tells us which maximizes social utility for
the group.  This is what I refer to as a subjective election.  The
purpose of the election is for each person to reveal how beneficial a
proposal is to him or herself, so that we can maximize social utility.

Now, let's consider another group of locked away voters.  Their job is
to determine the best proposal, not for themselves, but for the first
group.  Presumably, this is the proposal that won the election for the
first group.  Unfortunately, the second group doesn't have access to
the first group's ballots.  They have to guess, based on what they
know about the other group, and the proposals.  They are guessing
about external reality, not revealing the usefulness of the proposals
to themselves.  This is what I call an objective election.

The same is true, even if the first group never actually vote.  The
second group can still guess how the vote would have gone, if the
first group had voted.  This is the best result for the first group.

This example is similar to representative democracy.  The first group,
is the population as a whole.  The second group, is their legislature.
 The subjective vote by the population as a whole is only
hypothetical.

On Fri, 20 Apr 2001 14:31:56
Anthony Simmons <asimmons at krl.org> wrote:

> It's a pretty commonly heard adage that we can't vote on
> matters of objective fact.  Well, obviously, we can, but the
> result carries no weight.
> 
> Yes, we do sometimes have to vote on matters of pure fact,
> but not really.  For example, climatologists might be asked
> to vote on whether CO2 production is turning the whole planet
> into a Devonian swamp.  But we would only do that if it were
> necessary to make a decision of some sort.  It might be
> whether to stop using fossil fuels and start building psychic
> reactors.  Or it might even be an implied election -- in
> which there is no formal vote, but simply a recognition of
> the relative abundance of opinions -- that influences a
> personal decision, such as whether to buy an electric scooter
> or a sport-utility-amphibious-armored-vehicle.
> 
> But to the extent that there is a social decision to abide by
> an election outcome in any meaningful sense, doesn't there
> have to be a consequent action of some sort?  Even if the
> question being voted on is nothing more than the "sense of
> Congress" clause of HB1189, which only expresses an opinion",
> the whole point of that is to influence.

True, but lets say we decide that the best method for matters of pure
fact is method X.  Now, we have a vote on a policy to carry out.  This
decision is based on a number of facts.  In this case, I would expect
that the best method for the pure facts would be the same as the best
for the decision vote, since people will be voting the same way, and
we want the same result.

> Here's a nifty complication:  If we were to vote on a matter
> of objective fact, what would we really be voting on?  On
> whether a particular fact happens to be true?  Or whether we
> should take a particular action -- officially announcing that
> a fact is true.  These aren't necessarily the same thing.  In
> the latter case, I may have reasons other than objective fact
> for voting the way I do.  I'm sure if we thought hard enough,
> we could come up with a hypothetical situation in which, for
> example, legislators might vote to modify the science
> curriculum in Kansas for reasons not entirely confined to
> science in Kansas.  Hypothetically.  In such a case, social
> utility determines perceived fact.

Creationists often argue that acceptance of Darwinism inevitably leads
to social breakdown, and mass murder.  This belief is logically
independent of belief in Darwinism itself, so someone could accept
Darwinism, but still not want it taught, for fear of catastrophic
results.  I think Voltaire once remarked that there is no God, but
don't tell the help, or they'll steal the silverware.

But the vote on curriculum is still based on some objective issues. 
Either Darwinism leads to crime, or it doesn't.  So, just because the
school board base their decision on something other than the truth of
what is being taught, doesn't mean that they aren't basing it on
beliefs about reality.

> It's an interesting question -- is it meaningful to vote on a
> pure matter of objective fact?  Whatever the answer to that
> question, I think a more practical question is:  does it
> affect the choice of election method.

That's a good question.  It seems to me that people who base their
reasoning on the consideration of subjective questions tend to favour
methods for approximating average ratings, and vice versa.

To take a very simple example, consider a group of people deciding
between two options.  If you see this as a subjective election, where
each person rates the options based on personal benefit, then the
majority opinion shouldn't necessarily win.  On the other hand, if
they are voting on a statement of fact, then it doesn't seem that a
minority opinion would be more likely to be correct, simply because it
is more strongly held.  So, if you see an election as primarily a
determination of fact, rather than a revelation of personal utility,
this seems to argue more in favour of probabilistic arguments and
majoritarianism.

---------------------------------------------------------------------
Blake Cretney   http://www.fortunecity.com/meltingpot/harrow/124/path

Ranked Pairs gives the ranking of the options that always reflects 
the majority preference between any two options, except in order to
reflect majority preferences with greater margins. 
(B. T. Zavist & T. Tideman, "Complete independence  of clones in the 
ranked pairs rule", Social choice and welfare, vol 6, 167-173, 1989)