Subject: Re: Charging the Charger
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Mon, 16 May 2005 01:31:52 +0900

>>>>> "Laurent" == Laurent GUERBY <laurent@guerby.net> writes:

    Laurent> By lobbying, I mean paying people to show up at all
    Laurent> political meetings where *economists* are telling
    Laurent> lawmakers and the medias that infinite exceptionless
    Laurent> copyright, no reverse engineering, extensive database
    Laurent> rights, ultra broad software patents and monopoly on
    Laurent> support are the best thing since sliced bread for
    Laurent> customers and social welfare, and try to sing a different
    Laurent> song just to avoid immediate death of all OSS (under a
    Laurent> full worldwide software patent regime). And unfortunately
    Laurent> I agree it's the best use of the money for the interest
    Laurent> of OSS right now.

You have an URL quoting an economist saying such dreck?  AFAIK, all
reputable economists consider IP _at best_ a necessary evil from the
point of view of customers and social welfare.  There is a subset who
consider that irrelevant, advocating strong IP based on "creator's
rights" (diametrically opposite but methodologically similar to the
way the FSF advocates free software based on the right to share), but
the competent ones are careful to separate the economic from the moral
arguments.  And I find it hard to believe that the lobbyists don't
have both enough money and enough sense to hire reputable experts.

    >> Doing economics "for OSS" would cost me all credibility in my
    >> own eyes, and in the eyes of my peers.

    Laurent> That's why I'm asking other people more objective and
    Laurent> competent than me to do the economics :).

Welll, unfortunately, as far as I can tell the economics say that
_full_ OSS is a special interest looking for a subsidy.  Eliminating
software patents and reforming the patent system in general look like
good ideas to me at the moment, and I know a lot of economists who
feel the same way or more so.  But in general I think most customers
can do substantially better by having a range of licenses available in
the market, and most of the economic harm from patents etc has nothing
to do with OSS.

There are social benefits to OSS that I'd like to give society the
advantage of, but they're not quantifiable in conventional economic
accounting.  There's hope; the 1998 Nobel Memorial Prize was awarded
to Amartya Sen for his work on incorporating humanistic criteria such
as freedom and "empowerment" into the economic calculus.  But a lot
more work needs to be done before economists are going to be unanimous
enough to get the attention of policymakers.

    Laurent> But I do hope economists do take input from practionners
    Laurent> of the field they study to get some ideas,

They do, but mostly practitioners are functionally equivalent to
lobbyists looking for special treatment, and while they probably
honestly think they're advocating socially beneficial measures, they
rarely have a clue.  And the ideas are rarely new or coherent:

    Laurent> and I'm saying that they should look a bit more (than
    Laurent> exactly zero according to my random readings)

Well, it's not quite as bad as looking for SHA1 collisions, but really
you could do a lot better than reading randomly, you know.

Start with Shapiro and Varian's _Information Rules_, which has a good
bibliography.  Suzanne Scotchmer has a review in the Journal of
Economic Perspectives ("Standing on the Shoulders of Giants", 1991)
covering scores of papers, where one of her main themes is the
importance of treating innovation as incremental.  Boldrin and
Levine's paper has a short bibliography covering some of the topics
you mention.  Others, such as protocols and standards, have been
intensively researched, but I don't have a survey or textbook
treatment offhand.  Efficiency of in-house work is a managerial issue;
economists have very little to say about that ... try the SEI.
Outsourcing is an interesting question; economists still have not much
of a clue as to what determines the boundary of a firm.  But that is
an area of active research, though only indirectly related to the
issues of outsourcing in the software industry.

One of the things that really is sad about the whole discussion is
that economists _have_ looked at all the issues you've brought up.
They may have missed something important, but they are too busy to
listen to cranks who open the conversation by shouting that the
economists have totally missed the point and have evidently never
thought about the relevant issues---on the basis of "random readings".

And that's why Marshall SHOULD (in the sense of RFC 2119 ;-) be doing
what he's doing ... the veins you're talking about have been pretty
much mined out---if there's anything there, it's deeply buried.  Now,
I think that most honest academics have had the experience of talking
to an eager layman who has done his homework and forces the "pro" to
look at things in a new perspective, and sometimes the insights are
deep and/or important.  But the layman _must_ work very hard, because
it's rare that something of great economic importance is sufficiently
unique to some field of endeavor that economists have missed it
entirely.  So for all of the things you mention, a specialist will
cough and say "Yes, that's the Biyesovich model; Dr. Dreckenspiel
wrote the last word on that in 1957" or something like that---you need
to provide a reason to reopen the debate, which will be something the
specialist can translate to the form "in the Biyesovich model the zeta
parameter is assumed to be constant, which is invalid in the case of
software patents for reasons X, Y, and Z."

So Marshall is off doing something that nobody's done before, which is
far more likely to turn up a new proof or an interesting anomoly that
bears on the costs or benefits of OSS or proprietary arrangements.  I
mean, unless you're willing to get a PhD in econ and do the research
(and get the grants :-) yourself, you have to leave it up to his
judgment as to what the best line of inquiry is, no?


-- 
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.