Subject: Re: anti/Law (an attempted explaination)...
From: "Stephen J. Turnbull" <>
Date: Mon, 12 Apr 1999 14:05:31 +0900 (JST)

>>>>> "kms" == Karsten M Self <> writes:

    kms> Discussion redirected to egrok and


    kms> "Stephen J. Turnbull" wrote:

[to someone != kms]

    >> You're missing my point.

And some details of the point.  But Karsten responds:
    kms> Words of a man woefully unaquainted with the SAS Institute's
    kms> licensing policies, which might be summed up as "more blood,
    kms> less turnip".

    kms> Having read through the rest of your post, I find full
    kms> mention of price discrimination.  C.f. branding, generic
    kms> goods, 'store brand' products, Lexus v. Toyota, etc., etc.,
    kms> etc.

I have no idea what your point is.  Of course I mention price
discrimination.  And my example was just that, an example to show
possibility.  The fact that SAS manages finer price discrimination
than in my example doesn't prove that it manages to extract full
perceived value from _any_ of its customers, as my supplier did; far
from it.  For example, I paid $2500 for a simulation and optimization
application called GAMS.  Nothing else would do for my purpose.  I
don't know how much more I would have paid for it, but at least
another $1000 more.  GAMS, Inc. extracted a lot of my gains from the
trade, but nowhere near all.  You need to show that SAS Institute does

Branding (or "versioning" as Shapiro and Varian would have it) is not
necessarily an example of price discrimination.  In fact, one can
argue that it is not, by definition, since anything the consumer
identifies as differing are in fact different goods and therefore in
principle will be priced differently.

This is not a useful definition, of course.  But if you want to argue
that, it is incumbent on you to propose one.

    rn> For that matter, why buy a hardcover when a paperback has the
    rn> same words?  Value perceived.

    >> Well, yes.  In fact, some of my colleagues go to the trouble of
    >> having paperbacks rebound in hard covers (professional
    >> journals, of course).  Measure that value?  Hmmm....

    kms> Depends.  Publishing houses impart a time-value on works by
    kms> not releasing paperback editions until some time after the
    kms> hardcover version.  Factor in other (in)tangibles such as the

_Intangibles_ are by definition hard to measure by mechanical
processes.  That's exactly my point.  Value is perceived, not

Of course you can say the price system "measures" that value.  And
that's exactly what I did (in a different post, I guess) but I called
it "sufficient statistic."  However, the quotation marks are deserved; 
what price measures is the valuation on the marginal[1] unit bought at
that price, and then again only under circumstances with sufficient
continuity.  Inframarginal purchases will necessarily have different

    kms> Arguing that a product has a minimum market price (below
    kms> which it will not be produced), set by its production
    kms> function, is clearly valid.  I suspect this had something to
    kms> do with the 'cost plus' thread on the mailing list previously

Your conjecture is wrong.  That thread was purely and simply about
"just pricing" under conditions of uncertainty about return, despite
the tag attached to it.  Cf. Thomas Aquinas, et al.  Cost-plus was
simply the most familiar of the mechanisms that would more or less
achieve Brian's goals.

And don't bother [re]reading that thread.  It's not worth your time.
I wrote most of it; I should know.  :-)

    kms> not known as egrok.  Aren't legal services a cost-plus
    kms> product?

Define cost-plus.

According to my definition, the one relevant to the above-mentioned
thread, they are not.  Yes, lawyers usually pass on certain ancillary
charges (court costs, courier fees) directly to the client.  But my
contract with my lawyer doesn't say anything about adjusting the price
to give him an appropriate return on equity.  He gets paid a flat
hourly rate, which presumably was computed to put his children through
college and amortize a beautiful second home in the Rockies after
allowing for a certain amount of uncharged overhead.  But there's
nothing that allows it to be adjusted if that overhead changes for
some reason (rent increase on the office, for example).  A constant-
markup pricing _policy_ is not the same thing as a cost-plus
compensation _contract_.

    >> Funny thing, I haven't noticed my university library system
    >> complaining about paying 10 times as much as I do for journal
    >> subscriptions;

    kms> Have you actually _talked_ to a librarian about this?  You'd
    kms> probably find yourself in a whole new world.

Yes.  One.  :-)  He felt that since the user population was about
10,000,[2] he had little objection to paying 10 times the price.  Of
course he wished the price were lower; users were always asking for
more journals, and his budget was finite.

    >> It's even easy to find example of people happily paying double
    >> or triple the price for inferior products as long as they have
    >> snob appeal.

    kms> Evian <=> naivE

    kms> ...but you have to consider externalities and secondary
    kms> characteristics.  Would you serve Bud Light at a black-tie
    kms> occaision, rather than Brut?  It's clear, bubbly, and mildly
    kms> intoxicating.  Any economist can see that they are
    kms> substitutes.  The hostess with the mostest will bore you to
    kms> tears with an alternative interpretation.

I couldn't have put it better myself.  Whose side are you on, here?

Any economist (including the same consultant who told you they're
substitutes) will tell you that the hostess is, of course, right (and
flee, leaving you to be bored).  "Substitutability" is not a
dichotomous property.  Thus this example is completely irrelevant, as
far as I can see.  What do you have in mind?

    kms> Generally, though not always, an ROI calculation of some sort
    kms> is involved in purchase decisions.  The closer a purchase is
    kms> tied to an actual production activity, the more likely, in my
    kms> unresearched opinion.
Sure.  But where it is hard to measure mechanically, the value that
goes into the "R" in "ROI" is evaluated by some "expert's"
_perception_.  And we're back to "plan A".


[1]  The marginal units are the traded unit that provided the trader
the least amount of net value, and the non-traded unit that would have
induced the smallest net value loss to the would-be trader.  Under
conditions of continuity, they will be the same, and the net value is
zero.  Another way to put it is the marginal units are the traded and
non-traded units which require the smallest change in price to get the 
trader to reverse her decision about that unit.

By definition, then, with that minimal change in price, traders will
not change their minds about _inframarginal_ transactions.

[2]  Realistically, of course, even the weekly newsmagazines probably
really have a user population in the hundreds, not thousands, but
that's the way he felt about it.

University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
What are those two straight lines for?  "Free software rules."