Subject: Re: A few here may have an opinion on this
From: "Wendel de Joode, Ruben" <>
Date: Fri, 25 Oct 2002 09:14:11 +0200

 Fri, 25 Oct 2002 09:14:11 +0200
To continue on the subject, I think the following statement is precisely
the problem:

Stephen> This is an inherent flaw in the free software model from the
point of
Stephen> view of policy makers.  Crucial information about the
productivity of
Stephen> the input is simply not available  to society  because of the
mode of
Stephen> distribution.  Contrast that with proprietary software, where
(1) seat
Stephen> counts have a useful level of correlation with usage levels,
and (2)
Stephen> the customers' willingness-to-pay has a useful level of
Stephen> with the "size" of the software in terms of benefits to them
"at the
Stephen> margin".  This flaw is not a reason to avoid free software.
Rather it
Stephen> makes it difficult to evaluate exactly the questions of benefit
Stephen> cost that are crucial to relevant policy decisions.

I would be very conscious to call the issue described above a "flaw" in
the free software model though. You could just as well reason that free
software has uncovered flaws in both policy making and economic models,
because free software doesn't fit into their "usual" formulas.

I also have some problems with the following model:
Stephen> Measure Y = the TFP residual in the GDP series, measure X = the
Stephen> quantity of free software, regress Y on X, productivity is the
Stephen> derivative of the regression equation.

Don't economists use "ceteris paribus" at the end of these formulas to
say that this is only true if everything else stays the same?

In my opinion a formula like the above doesn't say much, because we
can't measure the quantity of free software and we certainly cannot let
everything else remain the same! Of course you could dismiss all of this
as mere measurement problems, but isn't that a little bit easy? 

To base any desicions, whether a policy decision or not, on numbers that
could prove to be very wrong, beacuse of measurement problems, in my
opinion could sometimes be just as bad (or good) as making decisions
based on some gut feeling. The only difference is that with the numbers
a policy maker after the decision can account for his/her decisions. 

I do agree that these numbers can provide some insight, but they should
be used with care and should not be the sole input in decision-making. 

Ruben van Wendel de Joode
Delft, University of Technology