Subject: Re: "a better world" (was Re: GIF/LZW patent)
From: <stephen@xemacs.org>
Date: Fri, 29 Sep 2006 11:14:20 +0900

 Fri, 29 Sep 2006 11:14:20 +0900
Bernard Lang writes:

 > I am certainly naive in all this :
 > 
 > what is the marginal cost of producing software ?

The marginal cost of producing software is high, and increases rapidly
with increasing functionality and complexity.

The marginal cost of reproducing software is low, (at worst) linear in
size and constant in other parameters, for all practical purposes.

Those who provide the capital focus on the former, since they're the
ones paying it.  Those without capital focus on the latter, because
that's the way to justify not paying for the former.  (For those who
are naive about economics, it *is in fact a justification*.
"Sufficient or not?" is the question at issue.)

 > what is the role of marginal production cost in economics ?

In a competitive industry, it determines the supply price.  In a
monopolistic industry, the supply price is determined by multiplying
the marginal production cost by a factor which is related to the price
elasticity of demand.

In welfare analysis, the so-called "first-best allocation" is
determined by the intersection of the marginal production cost curve
and the demand curve.

 > many (most?) laws are based on approximation of real phenomena.  These
 > approximations simply fail to make any sense when the relative orders
 > of magnitude that justified them are no longer in the same
 > relation.

True.  Unfortunately for your analogy, a constant marginal cost of
zero is a standard feature in textbook models and has been so since
Augustin Cournot's  Recherches sur les Principes Mathématiques de la
Théorie des Richesses  (1803).  It's not extreme at all!

 > I suspect that the geocentric theory of Ptolemy is probably better
 > grounded in actual observation and more accurate than anything
 > economists have come up with to justify patents.

Ho-hum.