more on regulation/deregulation of communications

[Dave Farber <dave () farber net>]

------ Forwarded Message
From: "Faulhaber, Gerald" <faulhabe () wharton upenn edu>
Date: Sat, 11 Jan 2003 11:00:53 -0500
To: "David P. Reed" <dpreed () reed com>
Cc: dave () farber net
Subject: RE: [IP] regulation/deregulation of communications

This is a way better description of the issue (as I would expect from David
Reed).

David's central point is Schumpeterian: static analysis will not suffice in
a world dominated by technical change.  It is certainly the case that
natural monopoly is a static concept, and as such it properly evokes
regulation as a welfare-improving government intervention...maybe.  When the
technology is relatively constant, it is arguable that a natural monopoly
should be regulated.  But when the technology begins to change, what then?
Regulation has great difficulty accommodating innovation;
bureaucratic/administrative systems break down in the face of innovation, or
even worse, reach out to suppress it.  Not evil people; just inappropriate
institutions.

In fact, I would say that our greatest challenge in telecoms today is how
institutions need to change, and how difficult the political process makes
that change.  We see this in wireline regulation, where the FCC and the
RBOCs are locked in a sweaty embrace of mutual suspicion, inefficiency, and
lack of innovativeness that neither side seems to be able to break.  In
wireless, it is the old-fashioned GOSPLAN model of spectrum allocation we
all hate (in different ways).  But political/regulatory change is much more
difficult than technical change.

History tells us it is subversive technologies that force bureaucracies and
politicians to accommodate the new realities (e.g., banking deregulation in
the '80s, microwave transmission in telcos in  the 1950s).  But it can also
act to suppress innovation (extending regulation from railroads to trucks in
the 1930s, to negative effect).  The Internet is just such a subversive
technology; it holds great hope to undermine the old institutions, including
regulation.  But it is not guaranteed to happen.

I am concerned that those who ask for government intervention to "protect"
something (e.g., end-to-end) don't understand the endpoint of regulation.
Economists use the phrase "the dead hand of regulation" based on long
experience and massive research results; and we see it in the real world as
well.  If you want to see the endpoint of government intervention in the
Internet (e.g., to ensure end-to-end), look at telco  regulation today,
because that's the fate of regulated industries, regardless of how
"enlightened" regulators may be.

Gerry

-----Original Message-----
From: David P. Reed [mailto:dpreed () reed com]
Sent: Saturday, January 11, 2003 8:55 AM
To: dave () farber net; Faulhaber () mail3 atl registeredsite com; Faulhaber,
Gerald
Subject: RE: [IP] regulation/deregulation of communications


At 09:17 AM 1/9/2003 -1000, David Farber wrote:
> David (not djf) is confusing the concept of natural monopoly with its
> application to the telephone business.
>
> To say that "natural monopoly" is a discredited concept in economics is
> simply false.  It is a clearly defined concept dating back to John Stuart
> Mill, and more recently refined by many others (among them, moi).  Most
> generally, a natural monopoly is an industry in which the firm's cost
> function is subadditive.  In the special case of a single product firm,
> the industry is characterized by declining average cost.

Gerry is, of course, correct.   Natural monopoly is not a discredited
concept in the abstract.

What I should have said, which Gerry said more clearly, is that it's far
from clear whether "communications" is a natural monopoly.

The other thing I should have said, which Gerry did not say, is that the
purpose of deciding (empirically)  whether some industry is a natural
monopoly is not to minimize "waste" by duplication as Feldmeier suggests,
but to decide whether that industry will necessarily be dominated by a
single player who will exercise monopoly power to set prices far from
marginal costs, thereby minimizing public welfare.  It's clear that
competition very effectively eliminates wasteful redundancy.

Finally, I didn't mention the deepest flaw in thinking about "natural
monopolies"  which is the implicit assumption that pervades much of
economics - that the entire system is in equilibrium.   In the areas of
technology-based business in particular, one has to account for massive
technology change, based on new scientific knowledge and architectural
redesign.   The Internet shows this clearly.   A very simple idea (that
bits are bits) was finally incorporated into a universal interoperable
network that unified many unrelated and "independent" communications
"markets".   All of a sudden, the system was no longer in equilibrium.   A
copper wire need not be viewed as a "telephone" wire, but as a "bit"
wire.  A "coax" need not be viewed as a "community antenna TV distribution
wire" but as a "bit" wire.

It turns out (in 20-20 hindsight) that the "natural monopoly" in Telephony
disappeared when Shannon invented the idea of the bit (or perhaps
earlier).  But what caused that change.   The inherent flaw in Natural
Monopoly is that it presumes that one can view a system in isolation and in
equilibrium.   Of course this is not possible.

Does this mean that we don't need regulation?   No, I think that is a false
conclusion.   But it does suggest (and I know that economists vary in their
opinion on this) that it is not easy to determine empirically that some
business is a Natural Monopoly.  The fallacy is that one can measure
"costs" and determine whether they are subadditive in the face of
technology uncertainty and market uncertainty.

If we had an Oracle (not Larry Ellison's) that could know in advance what
we will learn in the next 50 years about how communications works, we could
in principle determine whether a business is a natural monopoly.   But that
Oracle is clueless about both technology and market needs.

Who could have predicted the Internet's cost structure, the potential of
fiber optics, Moore's Law (which depended on a mass-market computer
industry's growth potential), or the rise of consumer demand for the WWW
and email, even 25 years ago?

Testing for Natural Monopoly is a bogus idea, wrongly applied.    That's
what is discredited.

A few economists (Carliss Baldwin is one) are trying very hard to get
economics to recognize that its view of technology and architecture is way
to reductionistic.   Knowledge and Design are not at all minor elements of
economic systems, mere coefficients or parameters.    Knowledge and Design
are the big deal - economics (in its narrow definition) has developed by
ignoring these major factors.



------ End of Forwarded Message

-------------------------------------
You are subscribed as interesting-people () lists elistx com
To unsubscribe or update your address, click
  http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/