[EM] A rant; forgive me.

John B. Hodges jbhodges at usit.net
Sat Jul 5 05:18:01 PDT 2003


Offlist, I received a comment:

>John,
>Interesting  and provocative. I  agree with [snip].  But  Adam 
>Tarr's  IRV  "nightmare" example is conclusively damning, especially 
>since it is political-spectrum based with everyone voting 
>sincerely/plausibly. [snip rest.]

I agree that Adam Tarr's "nightmare" example is impressive, but I 
don't think it is "conclusively damning." Because, I am an old (well, 
50) cynical post-economics-major. (I completed all-but-thesis on a 
Master's degree, many years ago.)

In economics there is a criterion known as "pareto-optimality" that 
some mathematical models of economic systems manage to satisfy. 
Pareto-optimality is on the face of things a modest criterion: that 
when comparing two situations, A is to be judged better than B if in 
A there is at least one person who is better off, in their own 
opinion, than they are in B, and there is no one who is worse off. 
Who could possibly object to that? Economists make a virtue of their 
humility; they are "scientists", and (arguing from a philosophy known 
as logical positivism) science deals with facts, and values cannot be 
derived from facts, so "scientists" therefore, AS SCIENTISTS, can say 
nothing about values. Economists therefore allow themselves the tiny 
presumption of using Pareto-optimality as a value-criterion, as a way 
of judging which proposed economic systems/policies are better than 
others. It is presumptuous because Pareto-optimality IS admittedly a 
value-criterion, and so technically off-limits for scientists, but it 
is such a modest one, such a tiny presumption, that they hope 
everyone else will forgive them.

I believed that for awhile as an undergraduate, and even as a 
graduate student, but later came to regard that whole act as a 
swindle. The trick is the flip side of the coin: not that 
Pareto-optimality is bad, but that ALL OTHER VALUES ARE EXCLUDED FROM 
THE DISCUSSION. So economists cannot say anything about the justice 
of the distribution of income and wealth, or about ecological 
sustainability, or about many other subjects, because the background 
philosophy of logical positivism claims that "scientists" cannot say 
anything about values. And logical positivism, originally developed 
by A.J. Ayer, was destroyed by other philosophers decades ago; in 
1958 even A.J. Ayer admitted that enough holes had been poked in it 
that he could no longer hold it himself.

As Pareto-optimality is a tiny, modest criterion, so the flip side, 
of excluding all other values from the dicussion, is like stealing 
all the gold in Fort Knox. (Where the U.S Govt. kept all its gold, 
back in the days when the dollar was backed by gold; I don't know if 
there is anything left there nowadays. So it may be an anachronistic 
expression.)

Taking both sides together, and reducing Jargon to plain English, 
Pareto-optimality effectively means that the most desirable situation 
is one in which all possible profitable trades are taken advantage of 
(I.E. all possible profits have been made.) So, economists are 
therefore allowed to advocate policies that make it easier for 
corporations to make profits, and forbidden from advocating policies 
that interfere in any way, for any reason, with corporations making 
profits. In fact, they can say "AS SCIENTISTS" that policies 
interfering with Free Trade are Bad, Bad, Bad, but "AS SCIENTISTS" 
they can say nothing else, and no one else can either.

Another little fact about Pareto-optimality is that it can be 
satisfied only in abstract models that make highly unrealistic 
assumptions. The more "realism" you allow into the models, the harder 
it is to achieve pareto-optimality; in the real world nothing does. 
The normal operation of the marketplace violates pareto-optimality 
all the time. If a new store opens up on Main Street, and takes some 
customers away from another store, that other store's owner is made 
worse off. So we cannot say "AS SCIENTISTS" that the new situation is 
better. Economists inhabit a mathematical fantasyland, but they use 
their mumbo-jumbo to argue for policies in the real world, which 
coincidentally favor the interests of the 
already-wealthy-and-powerful.

I suspect that Condorcet-efficiency is playing a similar role in the 
discussion of voting methods. It sounds like a fine criterion. Who 
could possibly object to it?

In plain English, Condorcet-efficiency means "elect the centrist, 
always, always." I'm not sure that's a good idea. Even if it's a good 
idea, ALL OTHER THINGS BEING EQUAL, it's certainly not the only 
quality we might desire in an electoral system. Other things are  not 
equal.

In the real world of electoral politics, with parties, with biased 
and false information broadcast wholesale, with control over 
electoral offices worth billions to particular industries and 
corporations, with voters of all different levels of sagacity, we 
have existing, established electoral systems, and entrenched corrupt 
interests that have evolved methods to use them. There is some 
movement among citizens to reform these electoral systems, in the 
hope of rooting these interests out of their entrenched positions. 
One technique for defeating efforts at reform is to sow confusion 
among the reformers; for example, if there is a proposal on the 
ballot that you wish to defeat, put one or two more proposals on the 
ballot that sound like they are addressing the same problem. The 
voters will split, and nothing will be enacted.

The battle-cry of a small-but-growing set of reformers is 
"Proportional Representation and the Instant-Runoff Ballot!" I am not 
accusing those who favor the Ranked Pairs method and Approval Voting 
of being pawns of the Capitalist Hegemonists. Really, I'm not. Not 
yet, anyway. I'd just like to keep the discussion in the real world.
-- 
------------------------------------
John B. Hodges, jbhodges@   @usit.net
The two-party system is obsolete and dysfunctional.
Better forms of democracy: www.fairvote.org
REAL CHOICES, NEW VOICES, by Douglas J. Amy



More information about the Election-Methods mailing list