[OPE-L:7595] Re: Value of information

From: Gil Skillman (gskillman@mail.wesleyan.edu)
Date: Wed Sep 04 2002 - 18:45:56 EDT

Paul writes

>What is the value of information?
>More particularly what is the value of software?

Depends on what you mean by "value."  If you mean it in the traditional 
Marxian sense, then the answer is immediate, as immediate as the 
accompanying inference that embodied labor time has very little if any 
relevance to the determination of market prices for software or other modes 
of commercial information provision.  Here's the problem as I understand 
it:  in a commodity market populated by large numbers of profit-maximizing 
producers facing strictly positive, non-decreasing marginal (and thus 
average) costs of production, the price of a commodity would tend to get 
determined by the interplay of market demand and marginal cost of aggregate 
production.  [N.B.:  the typical assumption of fixed-coefficient technology 
poses no special problem in defining marginal costs, though one does have 
to be careful to specify the "margin" at which capitalist firms are assumed 
to be operating.]   If we add the condition of free entry and exit, then 
this intersection must occur at the point where marginal cost just equals 
average cost of production (where the latter includes a "normal" rate of 
profit, or, in a Marxian context, the average economy-wide rate of profit, 
however determined).

But the problem with information is that, once it is created, the marginal 
cost of supplying it to additional consumers is (virtually) zero.  Given 
some strictly positive cost of creating the info in the first place, 
marginal cost of supply must thus always fall below (declining) average 
cost, implying that the good cannot be supplied under the competitive 
conditions described above; price = marginal cost (= zero, or thereabouts) 
always implies a *negative* rate of profit.

The capitalist solution to this has been to encase information in 
"intellectual property rights" so that what is being bought and sold is not 
the information per se, but the limited *right* to gain access to that 
information.  Thus the market price is not driven by marginal costs of 
production in the standard sense but rather the level of demand for the 
information in question.  I thus agree with Jerry (I just checked his 
incoming, which appeared as I was writing this) that the market price of 
information, or information embodied in software to speak more 
specifically, essentially corresponds to a rent.  Furthermore it can, under 
the logic of capitalist markets, *only* correspond to this, if this market 
is to be reproduced; under the conditions indicated above, (a) marginal 
labor values diverge from average labor values, necessarily; (b) market 
competition in the absence of enforcement of "intellectual property rights" 
would drive prices down to reflect marginal, rather than average, labor 
values (or more generally, costs of production), and thus (c) the market 
would self-destruct under competitive conditions.

In other words, the supply conditions for information land us in the 
territory Marx attempted to cover in K.III, Chapter 10, and as I tried to 
explain in a much much earlier post, he alas made a complete, incoherent 
hash of that analysis.  But hindsight says it wasn't his fault--labor value 
theory just isn't equipped to cope with the case of zero-marginal-cost 

For what it's worth,


This archive was generated by hypermail 2.1.5 : Fri Sep 06 2002 - 10:13:06 EDT