Subject: Re: Software as a public service
From: "Stephen J. Turnbull" <>
Date: Mon, 16 May 2005 13:02:02 +0900

>>>>> "Russell" == Russell McOrmond <> writes:

    Russell> Stephen J. Turnbull wrote:

    >> The correct statement is that "any given program, once written,
    >> is no longer scarce, under the assumptions of perfect
    >> information about its nature and availability, and zero
    >> communication and storage costs."

    Russell> I think there is yet another problem with language we are
    Russell> bumping up again.  When talking about supply/demand
    Russell> dynamics and pricing, a tangible being "scarce" makes the
    Russell> price go up (demand is high, supply is low).

Not to an economist.  "Scarcity" is defined without reference to
supply and demand (ie, to the market), more or less as I defined it.
And that _is_ the relevant definition of scarcity; it is the one that
all of the leading thinkers on the matter, specifically including RMS,
have in mind.

    Russell> With intangibles like software the dynamic is different.
    Russell> Rather than saying something is or is not scarce, it may
    Russell> help to say that software is non-rivalrous.  The question
    Russell> is not whether it has value to exist, but whether its
    Russell> value is determined by its existence or it
    Russell> number. Counting something that is non-rivalrous to find
    Russell> some number is an artificial construct to support
    Russell> specific business models,

Well, yes, exactly.  While it is rivalrous, equity is also an
artificial intangible construct whose only purpose is to support
specific business models.  I really don't have a problem with
artificial constructs that support business models.

    Russell> not something that is built
    Russell> into either its natural or economic features.

Nothing is built in to the economy; it's all a social construct.

But with respect to software, call it "non-rivalrous" if you like.
Call it "supercalifragilisticexpialidocious" if you like.  Either way,
not-yet-written software isn't that.  I know of one 4-letter financial
firm whose IT development process has some assigned projects, but the
main method for allocating tasks was a list of RFEs from various
departments.  The software engineers picked the ones they wanted to
do.  I was told that the backlog was lengthening at the pace of 30%
per annum in the face of departmental expansion at 15% per annum.  In
that firm, software "at the extensive margin" is scarce, rivalrous,
non-supercalifragilisticexpialidocious, and getting more so rapidly.
Given the incessant moaning about the need to produce more software
engineers, I can only conclude that this is the general case.

    Russell> When taken in that context all knowledge is scarce given
    Russell> no individual can retain all of it, and the vast majority
    Russell> of humans retain so little of it as to not be worth
    Russell> measuring.  I think the context is wrong, and that
    Russell> intangible knowledge isn't scarce the way that tangible
    Russell> fossil fuel reserves are.

The only reasonable economic model I know of that provides a
non-transactions cost argument against IP is that of Boldrin and
Levine, and it specifically depends on the assumption that knowledge
is useful only to the extent that it is embodied.  Ie, even _existing_
knowledge is tightly bound to some tangible resource, and is scarce,
in their model.

This is what is so damn frustrating about the economics of software.
"Everybody knows" that software is intangible and non-scarce, and that
OSS is just plain The Right Thing.  But economic analysis says that's
a contradiction---OSS leads to market failure.  While if you assume
that software is tangible and scarce, then you _can_ get the result
that IP doesn't give the incentives that it claims to give.

School of Systems and Information Engineering
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.