Notes mostly on the economics of information and ideas.

As a periodic reminder, you can unsubscribe from this list by sending
a message that looks like this:

  Subject: unsubscribe

Also, Eudora users are once again implored not to use the "redirect"
command.  I am still receiving a steady stream of misdirected mail
caused by people who redirect my messages to their friends and then
add extra notations that mislead their friends about where any reply
to the message will actually be sent.  Qualcomm's decision to keep
the "redirect" feature on an easy-to-find menu in Eudora is akin
to keeping a loaded handgun lying around, disguised as a coffee cup.
The "redirect" comment isn't just an accident waiting to happen --
it's an accident that happens every day.  It's flat-out bad design,
and Qualcomm should be ashamed.

Some Web pages constructed by students involved in the demonstrations
in Serbia are on the Web at:

Thanks very much to the RRE readers and friends who responded to
my call for comments on my undergraduates' term projects.  Most of
the comments were kind and helpful, and we very much appreciate them.
Some other comments, however, were neither kind nor helpful.  These,
as best I can tell, were provoked by my humorous declaration of a
revolution in Web design.  My whole idea in declaring this revolution
was to suggest that we forget about making cool Web pages for awhile
and learn instead to make useful ones.  Looking at these beginners'
draft term projects, however, some people were clearly still wearing
their "cool" eyeglasses.  They saw misplaced links, spelling errors,
and the usual stuff that happens when beginners try something new,
and they got nasty about it.  One of them, a long-time RRE subscriber
and an otherwise nice person, sniped (just to me, thankfully, so far
as I know) that the students' pages were "warmed-over term papers".
Well, duh.  What really struck me is how hard some people found it
to evaluate what the students' pages said, independently of their
surface form.  I don't think those people are dumb, but I do think
that they've been influenced by Internet hype, with its relentless
focus on conduits and surfaces and its relative indifference to the
substantive importance of the ideas that those conduits and surfaces
might convey.

Another problem is that we tend to presuppose that a Web page will
be defined as "information about topic X", whereas my students' draft
project pages are defined as "analysis of the structure and workings
of some community which might or might not find it useful to have
information about topic X organized in some particular way online".
It's a conceptual jump from a focus on "what" (a subject matter) to a
focus on "who" (the lives and needs of a user community).  It's more
complicated, but I think it's a necessary first step.

Despite all this, as I say, most of the comments were helpful, and
the students are hard at work on their revisions as we speak.  We're
in the thinking stages here of a longer-term project -- using simple,
powerful social mapping techniques to motivate clear stories about
good design.  We'll know that these techniques are working when they
start telling us that the best design for some particular purpose
involves 3-by-5 cards, not Web pages.  We'll keep you up to date.

All ten-year-olds on the Internet should aim their Web browsers at
It's a directory of US standard highway signs, including good-sized
color images of every sign.

The book that's blowing my mind right now is Lowell Bryant and Diana
Farrell's "Market Unbound: Unleashing Global Capitalism" (Wiley,
1996).  Don't bother reading Peter Huber's cheesy article about this
book in the current issue of Forbes (the one with the blaring pink
cover about CYBER POWER).  Just read the book.  It's a detailed
account of the transformation of global capital markets.  Economics
changed in early 1994, they argued, as it suddenly became possible
to engage in arbitrage among different countries' interest rates on
a large scale.  Their book is rather technical and absymally edited,
and of course its major goal is to recruit consulting clients for
the authors' company, but its facts are remarkable nonetheless.  They
explain a lot about financial derivatives, for instance, that had
not been clear to me.  They also explains the amazing changes in the
banking industry now being brought on by the securitization of a wide
range of revenue streams.  Above all, they explain why we do not yet
have a truly global capital market, and what it would be like to have

The book also helped me understand why New York investment houses
were blanketing the landscape circa 1992 trying to sell radioactive
derivative products to naifs like Procter and Gamble and Orange
County, California.  To wit, they were selling lucrative risk hedging
products to their financially savvy customers, but this also required
them to go out and sell the complementary risk to someone else.  If
financial markets are working correctly then this "someone else" will
be an equally sophisticated party with a complementary risk profile,
or an even more sophisticated speculator with massive research
capabilities.  But financial markets often don't work correctly, and
so in practice the counterparty was quite often an unsophisticated
investor -- like most finance officers, public or private.

Bryant and Farrell also provide the fullest exposition I've seen of
the now-common thesis that global capital markets are bringing an
end to national sovereignty.  This doesn't seem to bother them very
much, and the only risks they see are the risks that will follow if
governments disobey the orders of the capital markets.  Peter Huber
thinks that's just great.

I've been hearing way too much talk about the end of democracy lately.
The religious conservative magazine First Things, for example, has
been ruminating lately about whether the Church should declare the US
government illegitimate and institute a revolution.  We're not talking
fringe wackos here -- we're talking authors like Chuck Colson, widely
read opinion leader amongst the largest organized constituency of the
country's Congressional majority party.  Now, of course, you'll recall
that Chuck Colson had another career subverting democracy, but he got
caught that time.  Now he's legit, talking about revolution in public
and running an organization whose mission is "Biblical standards in
sentencing" -- by which he means that people like him shouldn't have
to go to jail in the future.  Why isn't this big news?  Is democracy
about to die in its sleep?  Let's wake it up before it's too late.

I read in the Wall Street Journal last week that MTV is trying to
charge online services every time their users access the
Web site.  I laughed at first.  But then it hit me, as the article
pointed out, that MTV is basically trying to institute the cable
television model on the Internet.  Then I stopped laughing.  But
only for a minute.

I received several incredulous messages after my little commentary
at the top of that court decision in Cyber Promotions v AOL.  The
court had reasoned that AOL could block incoming messages from Cyber
Promotions simply because it's a private party and not covered by
the First Amendment.  While I find Cyber Promotions to be just as
obnoxious as anyone else, this decision concerned me because of the
precedent (discussed at length in the decision) of political speech
being regulated in shopping malls.  The incredulous messages came
from people who found it inconceivable that the Internet could evolve
in a direction that made it impractical for dissidents to exercise
meaningful free speech rights just by setting up cheap web sites.

In my opinion, those folks lack imagination.  Let's get clear about
this: standards-wise, the Internet as we know it right now is little
more than a vacant lot.  The services that people actually see are
little more than SMTP and HTTP.  SMTP and HTTP, while indubitably
revolutionary in the things they portend, are toys.  The real action
will come with the next couple generations of standards, and the
basic character of the Internet could change radically depending on
how those standards evolve.  In historical terms, the TCP/IP standards
and HTTP were flukes: open standards that spread like wildfire while
the big players were asleep.  Historically, most important standards
have only been implemented through coordinated action by big players,
either singly or in alliances, and I don't see why the Internet should
be any different.

Let's take ActiveX.  Simson Garfinkel reported last week on basic
security problems with ActiveX.  The problem is simple: whereas Java
applications are interpreted in a safe little sandbox, ActiveX is
just a matter of downloading code onto your computer and running it.
Why can't that code trash your hard drive or ship your personal data
to Anguilla?  The only reason, Simson reports, is that the people
who wrote that code have crossed their hearts and promised not to
do such things.  Furthermore, useful ActiveX implementations are only
available for Windows OS's and Intel processors.  What happens, then,
if ActiveX becomes the dominant standard on the Internet?  If ActiveX
becomes the dominant standard on the Internet then Microsoft may have
another near monopoly.  More to the point, it will only be safe to
download ActiveX controls from highly-reputed, bulletproof corporate
Web sites.  The average citizen will be able to publish controls,
but nobody will download them for fear that they may be destructive.

My point is not that this scenario is inevitable, only that it
is entirely possible.  Nothing is inevitable.  Other scenarios are
possible too, some good and some bad.  And the scenario that actually
happens will be a choice that people have made.  If you choose to
run an ActiveX-enabled browser, then that is a choice.  And it is not
only a choice for yourself.  It is also a choice that affects other
people, inasmuch as network effects tend to reinforce the position
of whatever player first establishes a dominant position in any given
niche.  In a world of network effects, Kant's categorical imperative
takes on a new meaning.  Kant suggested that you refrain from doing
anything that would cause bad consequences if everybody did it.  In
a world of network effects it's actually worse than that, since if
you do it, everybody else is more likely to have to do it.

You might have noticed that I have been evolving a new genre of book
advertisement on RRE.  When somebody I know publishes a book that I
like, and the book seems relevant to RRE, and it occurs to me, I send
the author a note suggesting that they assemble an advertisement for
my list.  I use the advertisement for Vinny Mosco's book as a model:
title, publisher, ordering information, blurbs, etc, plus a few pages
from the book so people can get the idea.  The idea is to create a
win-win situation: the author sells more books and RRE readers get a
capsule summary whether they want to read the book or not.

When I send such suggestions to my author friends, one of two things
happens.  Sometimes the author just sits down and whacks out the ad
without ado.  Other times, and this is more common case, the author
passes my suggestion along to the publisher, who just sits on it
and does nothing.  Why?  The publishers, it seems, cannot get their
minds around the idea of sending out a few pages of their intellectual
property for free to a large Internet mailing list.  "Think of it as
a loss-leader", I say.  "You couldn't buy a mailing list as good
as mine for the kind of book you're marketing."  But no.  It's not as
though they actually ever say no.  They just get confused and put it
off until tomorrow.  The publishing industry knows in a vague, diffuse
way that it needs to wake up and comprehend the new technology, but
it sure has a long way to go.  The economics of information is really
starting to bite, and the economics is shouting in your ear: "Give
stuff away for free!"  We'll see if they get it.

When I encountered the December issue of Wired on a newsstand in
Los Angeles, the first thing I noticed was its 310 pages.  I quickly
checked the December issues of all the fashion magazines, hoping
that Wired had more pages than any of them.  Alas, Vogue (I believe
it was) had 346 pages.  The interminable cover story about laying
fiber optic cable in exotic locales was actually engrossing in a very
masculine sort of way.  "Hacker tourism", it's called, and yet this
new genre of travel writing isn't that much different from travelers'
tales across three hundred years.  I hope that Edward Said (author of
an influential book on orientalism) writes an article about it.  I'll
probably have more to say later on Todd Lappin's informative article
about Fedex, which I recommend to everyone.  In the meantime, I'll
just assume that Todd wrote the part about Fedex's desire to be left
alone by the government before Fedex launched its awesome lobbying
campaign in Congress to prevent its employees from forming a union
like those of its main competitor, UPS.

Sometimes people write to me to explain how I should run my mailing
list.  Last month, for example, one person wrote to explain that I
was abusing my mailing list by including my own opinions on it.  The
most recent of these back-seat drivers rebuked me for failing to offer
Borders equal time for Michael Moore's article on anti-union activity
at that company, and for failing to send out a New York Times article
in which the Borders people told their side of the story.  I don't
tend to reply as nicely to such messages as I would like.  They're
easy to write, but they would take a lot of work to debunk.  First of
all, no corporate PR person has ever returned one of my phone calls.
Second, of course, New York Times stories are copyrighted.  Third
and most importantly, it is perfectly weird, is it not?, to suggest
that my little mailing list has an obligation to provide equal time
to anyone the size of Borders.  The doctrine of equal time, which has
long been flouted by organizations with considerably more power to
harm than I have, was based on a premise of scarcity: if you occupy
a big chunk of the nation's broadcast bandwidth, then you have an
obligation to serve the public interest.  The Internet doesn't work
like that.

Now, I do send out rebuttals to RRE items when the rebuttals are
themselves of broad interest.  I have also retracted a very small
number of RRE messages that have turned out to be serious turkeys.
And we do need to understand what the rules are in this new medium
-- or, if we don't want to say "rules", what values should guide
us.  My point is, it's way too easy to simply carry over phrases from
other media to reason about this medium.  It's inevitable, of course,
that people will think about unfamiliar situations using analogies
to familiar situations.  That has happened repeatedly in the history
of media -- after all, the analogy to radio and television is what
gave us the Communications Decency Act.  But analogies are only useful
if we unpack them and determine whether the reasoning really carries
over.  In this case, it doesn't.

Much of the remainder of these "notes" consists of thoughts about
economics that have been accumulating in my head lately.  I am not
an economist, and so it is entirely possible that some of my thoughts
aren't original.  Some of them are probably found in Coase, for
example.  If you know of materials I ought to be reading or citing,
by all means let me know.  In the meantime, I won't claim originality
for any of this.  I just find it interesting.

Lately I have been hearing a lot about the "marketplace of ideas".
This phrase, which comes and goes historically, is one more example of
the importation of economic ideas into places where they don't belong.
The logic seems clear enough on the surface: an argument is like a
sales pitch, agreement constitutes sale, and it's in people's interest
to make wise purchases, so the market will discern truth if we simply
leave it alone.  But like most of these economic analogies, this
one falls apart as soon as we take it seriously as economics.  Let's
suppose that we were founding an economics of ideas.  We would first
need to reckon with the fact that ideas aren't goods like cars and
steaks.  They are public goods, for one thing, and so the burden is
really on the economically minded to explain why our ideas shouldn't
all be provided by the government.  We'd also have to get past the
fact that ideas aren't property, nor is it easy to imagine how they
could be.  Research people do have an semiformal system of norms and
practices for allocating credit for ideas, but that's not the same
thing as property.  Credit for ideas cannot be sold, for example.

Perhaps even more importantly, much -- if not most -- of the utility
that people derive from ideas comes from *other people* believing
them.  If you believe that my products are safe, then other things
being equal I will sell more of my products.  If you believe that
my theories predict imminent changes for your life, then you are
more likely to buy my book.  If you believe that capitalism is a just
social order, then you are less likely to organize a revolution and
expropriate me.  Notice that this kind of utility from ideas depends
only tangentially on whether the ideas are true or not.  That is, the
utility that I derive from your faith in capitalism is independent
of whether capitalism really is a just social order, or even whether
there is a fact of the matter.

This observation could set us on a long and potentially interesting
path of economic inquiry.  It would help us to explore in economic
terms what Oscar Gandy calls "information subsidies" -- research and
story ideas that PR people provide for free to help reporters write
their stories.  The economics of ideas is particularly important
as we contemplate the future of the Internet as a commercial medium.

Consider the situation of an Internet content provider -- let's say,
one that distributes information to users' machines to be displayed
on a screen saver.  That company faces a quandry: where to obtain
the content?  One option would be to pay reporters.  Another option
would simply be to plug itself into PR Newswire, a free service that
consists of very little except press releases.  Which option will
enable that company to compete and thrive?  Well, it could be that
its customers actually want a steady stream of press releases on
their screens.  Or perhaps their customers don't care what they
get on their screens.  Or perhaps their customers would prefer to
get news articles written by reporters, but they aren't willing to
pay what it would cost to get those articles.  It all turns on the
customers' ability to judge the utility that they are deriving from
the stuff that appears on their screens.  The problem is, information
commodities are inherently hard to evaluate, precisely because their
utility consists in being previously unknown.  Content providers
therefore have a strong incentive to provide content with a high
apparent utility and a low actual cost of production.  One way to
reconcile that tension is to provide consumers with materials obtained
from the PR departments of interested parties but dressed up to look
like the results of serious reporting.

In practice, of course, this tension is negotiated in an extremely
complicated variety of ways.  Sometimes you'll get raw press releases
blasted onto your screen.  Other times you'll get something written
by a reporter who has consulted several sources, but whose basic story
idea, core facts, and referrals to experts were all provided by a PR
person.  Opinions differ widely about whether newspapers and TV news
serve the public well in their existing configuration.  As technology
changes, however, those questions will be negotiated all over again.
Understanding those negotiations will indeed require an economics
of ideas.  And that economics is wildly incomplete until takes due
account of the incentives that I have to convince you of some idea.

How to judge those incentives?  If the traditional picture of economic
rationality is to believed then everybody judges those incentives all
the time, so that their investments in persuasion are guided by the
same utility-maximizing calculus as any other investments.  How does
new technology change those judgements?  When it becomes cheaper for
me to conduct research that selectively supports my views, do I have a
greater incentive to persuade people of them?  When I gain additional
channels to communicate my views to the people whose beliefs I want
to affect, do I have a greater incentive to formulate those views
and chase down the facts that would seem to support them?  Do those
same conclusions hold when other parties also have greater incentives
to promote conflicting views?  Is the result a rough parity in the
funds supporting advocacy of different views, or is the advocacy of
differing views funded in proportion to the utility that could be
obtained if people believed them?  The 12/3/96 Wall Street Journal
includes an article about Wall Street's campaign to dissolve Social
Security; it emphasizes that a big public splash won't be as helpful
to their cause as quiet funding for think tanks.  Who has how much
of an interest in funding think tanks to support the opposite agenda?
Greenpeace can raise $10M extra by convincing you that dioxin is
poisoning you, let's say, but chemical manufacturers can save $100M
by swaying public opinion against further regulations on dioxin.
Or maybe it's the other way around.  In either case, let's find out.

The 12/2/96 issue of The New Yorker provides an interesting datum on
this issue.  It concerns the fate of one very consequential idea, the
rational expectations theory of macroeconomics.  Rational expectations
is an idea; how can we understand the economics of this idea.  The
facts are that rational expectations is an alternative to traditional
Keynsian macroeconomics, and that its novel premise is that everyone
understands the economy perfectly well and can predict the economic
consequences of all government actions.  You might say that this
is completely implausible on its face, and in fact it is basically
false, as its originator now admits.  But then, given this, we need
some explanation of why it has largely taken over the relevant parts
of the academic economic world.  Two reasons suggest themselves: (1)
it is a convenient theory for people who wish to minimize government
intervention in the economy because it predicts that such intervention
is necessarily futile; and (2) it is a convenient theory for academic
economists because it permits them to revisit every last one of the
traditional questions of macroeconomics using new mathematics.

Which of these theories is true?  The economics of ideas would suggest
(1); the economics of academic credit would suggest (2).  The truth,
from what I can determine, is a combination of these.  Foundations
established to promote conservative political views have invested
heavily in supporting the work of economists whose views they have
found congenial; this trend took its modern form forty ago throught
efforts of the conservative economist Friedrich Hayek, who worked to
persuade wealthy people to throw their resources into the war of ideas
in society.  On the other hand, I can attest that all academic fields
exult whenever a new vocabulary comes along that requires hordes of
old questions to be revisited.  Academics are swidden agriculturalists:
the old ground having been plowed, its credit having been fully
extracted and converted into tenure files and reputations, the forest
is allowed to grow back while the young and the hustlers move on to
clear fresh ground somewhere else.  The forest has fully grown back
over the ground that Keynes first plowed, notwithstanding that those
old Keynsian models are still the ones that track the economy best
among actual practicing economists who need to make money for their

This leads me to some disagreements with the New Yorker article,
which verges on anti-intellectualism in two of its presuppositions:
that the academic field of economics should be measured primarily
by its usefulness to private sector employers of economists, and that
abstract mathematical modeling is necessarily detached from practical
reality.  Economics should be measured in a whole variety of ways, not
the least of which is its usefulness to other fields of human inquiry
such as history and epidemiology.  And abstract mathematical modeling
is a powerful tool that can be used rightly or abused.  The danger
of mathematical modeling is indeed that it will become detached from
reality.  That danger stems from two sources: (1) on an ideological
level, the capacity of mathematics to hide counterfactual premises,
such as those of rational expectations theory, beneath layers of
esoteric codes; and (2) on an institutional level, the tendency of
mathematical formalisms to take on an independent life when research
are judged on the internal logic of its "results" and not on the rigor
of its engagement with reality.  The fact is, problems of information
are central to economic organization, and they are becoming even more
central as information technology becomes central to the economy.

The economics of information brings out another, even more fundamental
problem with the conventional formalisms of economics.  A premise of
these formalisms is that people act to maximize utility.  This sounds
very technical, but it threatens to become meaningless in practice
unless we can find some way to define and measure people's utilities
except by inference from their actions.  The matter has been debated
endlessly, but in practice things almost invariably return to the same
point: using the prices that people will pay for things as proxies
for their utilities.  The technical way of saying this is that prices
stand as a proxy for subjective or expected utilities -- they serve,
in other words, as "revealed preferences".

The problem here is not that their utilities are subjective.  The
problem, quite the contrary, is an altogether objective matter of
credit assignment.  If I purchase a given shirt, what utility will
I enjoy?  I can estimate the likely utility in various ways: by
trying it on in the store, by learning about good shirt construction
and then looking at the stitches, by following fashions to ensure
that this shirt won't mark me as a loser in the coming season, and
so forth.  However, as Ed Baker (whose work on the economic aspects
of First Amendment law I strenuously recommend) pointed out to me,
other utility consequences of my purchase lie beyond my capacity to
estimate.  For all I know, that paisley shirt that I put back on the
rack in 1991 would have caught the eye of the woman of my dreams, and
she is now married to someone else.  The real utilities associated
with a market choice, then, necessarily fade beyond some point into
the counterfactual mists of life.

Now, most people won't take this problem seriously in the case of a
shirt.  But what about a product that contains an invisible carcinogen
whose effects on me will never be pinned down?  Eventually, of course,
it becomes impossible to reason about such matters.  Some cases,
moreover, are uncontroversial matters of fraud and misrepresentation.
In other cases, notwithstanding the prevailing assumptions of perfect
information in the marketplace, we are inclined to blame people
for not taking more effort to anticipate the consequences of their
purchases, for example by learning from the experiences of others.

In between these extremes, however, is a large middle ground in which
it is inherently difficult to anticipate the utilities associated
with a purchase, and in which sellers consequently have a competitive
imperative to exploit the difference between anticipation and reality.
(This is the microeconomic analog to the failure of the rational
expectations theory in macroeconomics.)  And information commodities
are such a case: since it is hard to evaluate whether information is
true, it is consequently hard to anticipate the utilities that would
be associated with believing it.  Of course, one can conduct a great
deal of additional research, but this would create an exponential
explosion of extra costs, not the least of which is the opportunity
cost associated with no longer having a life.  In practice, no matter
how skeptical we are, we have little alternative to believing most of
what we hear.

(Economists describe some of these situations in terms of asymmetries
of information: the seller knows things that the buyer does not.  But
that formulation is only adequate to our purposes if the seller has
perfect information about the buyer's likely objective utilities in
the future, when in fact the seller is probably much more interested
in developing information about the buyer's likely subjective
utilities in the present.  This is fortunate, because otherwise we
would have to get into a long, confusing discussion of information
asymmetries in markets for information commodities.)

The bottom line, in my opinion, is that the economics of information
works fine so long as we neglect what the information actually says.
But if we believe that the health of our society depends on whether
people are able to get adequate supplies of unadulterated information,
we're on our own.