On Fri, 01 Sep 2000, QueeneMUSE@aol.com wrote:
> In a message dated 9/1/2000 9:30:36 AM Pacific Daylight Time,
> alexboko@umich.edu writes:
>
> > In other words, I'm trying to address this at the level of 'How best to
> > divide up finite resource XXX between agents A, B, and C under
> > conditions YYY?' rather than 'Should ruthless sleazy factory owners be
> > allowed to steal food from the mouths of unemployed single mothers with
> > AIDS?'.
>
> Quite literally, that's the biggest problem.
>
> Is it really appropriate to eliminate the humanitarian factors and emotions,
> when the decisions cleary impact humanity?
This is a matter of values and of mathematics.
True free markets offer maximum *average* utility and growth to agents.
Perfectly controlled markets offer a guaranteed utility(?) floor to agents.
In the real world, both systems have problems because they are run by
people, not machines.
It becomes a values argument because the question is not whether either of
these systems work, rather it is a question of whether we should allow for
the possibility of individuals to suffer so that the population on average
can gain the maximum value from their efforts. Proponents of free markets
believe that average well-being of the population is more important than
the well-being of any specific individual. Proponents of controlled
markets believe that the well-being of any single individual is more
important than the average well-being of the population.
So in a sense, it all comes down to design goals, which is where most
people differ. This is very analogous to network protocol design, which
are essentially markets for communication bandwidth. Free-for-all
protocols offer the maximum possible throughput but no guarantees
regarding the performance of individual packets. Controlled, negotiated
protocols offer guarantees of performance for individual packets, but it is
mathematically impossible to meet the average performance of free-for-all
protocols on the same hardware. Both types of protocols are in use today,
with the selection criteria being dependent on the application.
As an additional note, I believe it is quite possible to demonstrate that
in most cases hybrid markets are theoretically less efficient than either
perfectly free or perfectly controlled markets. The only markets for
which hybrid protocols *are* optimally efficient are pretty artificial and
don't map well to the real world in the vast majority of cases.
-James Rogers
jamesr@best.com
This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:36:54 MDT