IP: Two rather long comments re faulhaber/Farber paper READ BOTH PARTS

[Dave Farber <dave () farber net>]

> From: Dave Burstein <dave () dslprime com>
> Date: Mon, 17 Jun 2002 14:44:41 -0400
> To: farber () cis upenn edu
> Subject: Don't auction the future commons
>
> Dave
> (For the list, edited if you like.)
> Respectfully disagreeing with Gerry's proposal to expand auctions. In general,
> I'm reluctant to disagree with you, Dave, because in general you're far more
> likely to be right than I am. But I think your academic friends and their
> peers at the FCC need to consider that competition, information, market power,
> access to capital (crucial change since the bust) and access to the political
> process are so far from perfect the models break down.
>
> Dave Reed and many I respect have made the point that common use of spectrum
> is often the most efficient. Today's chipsets, including those for
> ultrawideband and 802.11 variants, can intelligently adapt to congestion and
> hence work in a common band. They hop to an alternate frequency if data isn't
> getting through. 
>
> Enormous excitement moving to real deployments already in 802 in shared
> spectrum, and as chips become more capable (Moore's Law), shared spectrum will
> be increasingly crucial. This is especially important because innovation is
> much less expensive without a purchase cost for spectrum.
>
>   So my belief, in 2002, is that less, not more, spectrum be "owned." While in
> a perfect market, with perfect competition, perfect information, and no
> political interference, auctions have theoretical advantages, that doesn't
> correspond to the real world of telecom I report on.
>
>    In particular, phone companies (and their cable/TV peers) are intelligent
> profit-maximisers. For the last few decades, they have been trying to corral
> oligopoly positions in every endeavor, knowing that then they can raise
> prices. With the Internet bust, there is virtually no capital for competitors,
> while telcos continue to have a modest cost of capital.
>
>    Result: smart longterm move of telcos is to buy spectrum and freeze out
> competitors. They are trying to do that in retrospect, which is the effect of
> the proposed wireless mergers, aimed at raising prices above a competitive
> level. Any reader of this list know the bitter struggles on broadband
> competition, and also knows the current result: the U.S. prices are 40-60%
> higher than Canada and Japan.
> ----------------------------------------------------------
> Auction and similar "free-market" models are based on assumptions (similar
> capital access, minimal market power, no political power). In addition, they
> assume static technology and competition i.e. do not provide for a Verizon or
> SBC to respond forcefully to any potential competition, and often overwhelm
> it. The actions of companies with market power using that clout to make
> mincemeat of economic theories are legion.
>
>     You write of government "buying" some spectrum, but the government budget
> process isn't likely to correspond to the theories you espouse. Result, my
> best guess, will be stagnation. Textbook economics makes a lot of assumptions,
> but the best economists know they only go so far in the real world.
>
> db
>
> > He and the paper's co-author, David Farber, former chief technologist at
> > the FCC and a telecommunications-systems professor at the University of
> > Pennsylvania, suggest that the FCC have one "big bang" auction.
> >
> > As part of the plan, Faulhaber envisions that the federal government
> > would buy some of the spectrum for anyone to use, similar to the concept
> > of national parks. Local communities or even individuals could buy
> > spectrum as well and create a shared area.
> Again, the theory makes sense on one level, but the likelyhood of practice
> corresponding is slim. Can you imagine the years of effort it would take to
> persuade say New York or San Francisco to budget for an "electronic commons."
> It would be a great idea, but unlikely to happen soon.
>
> Editor, DSL Prime & Telecom Insider; Co-author DSL: A Wiley Tech Brief (Feb
> 2002)
> Special correspondent, The Personal Computer Show, WBAI-99.5FM, 8 p.m.
> Wednesdays
> Three time winner of Best Radio Show from the Computer Press Association;
> "The power of the printing press belongs solely to those who own the presses"
> A.J. Leibling - The Internet is our press

From: "Faulhaber, Gerald" <faulhabe () wharton upenn edu>
Date: Mon, 17 Jun 2002 15:48:35 -0400
To: "'Dave Farber'" <dave () farber net>
Subject: RE: Comments for inclusion in posting?

Dave--  way too long, but Burstein raises many issues.
 
The interest shown in our proposal is heartening; thanks for the opportunity
to respond.
 
The Seattle Times article covered only a part of our proposal; I recently
presented our full proposal at the FCC's Technical Advisory Committee (TAC)
(Dave was recuperating from Japan unfortunately).  The slides of that
presentation are at
http://rider.wharton.upenn.edu/NEW_SPECTRUM_MANAGEMENT.ppt.  The point of
the analysis is to show that economists and technologists, marketeers and
commoners, are much closer together than either group thinks.  Included in
that analysis is a comparison between markets and commons in the context of
spectrum.  Our conclusions are that (1) there are several property right
regimes that are compatible with commons-like services such as wideband and
agile radio; and (2) in the medium term, a market (with new radio
technology) is likely to result in what is effectively a commons.  The key
to understanding when markets work and when the commons work is very simple:
scarcity.  If the good is scarce, then markets work and commons don't; if
not, commons works and markets are unnecessary.  Please check the
presentation slides for the full story.
 
Now about auctions.  Generally speaking, auctions have worked rather well in
the US.  When they have led to less-than-stellar results, it is because the
FCC (following Congress's mandate) tried to get "social" in its auction
design.  For example, when the FCC restricted bidders to small firms and
permitted installment payments (with the FCC as banker), it led to the
Nextwave debacle.  Well, that's easy to avoid: don't try to use auctions for
social engineering.  But generally, the FCC auctions have done very well at
allocating wireless spectrum.  I know of no evidence to suggest otherwise.
 
The European auctions were a case of really bad auction design, fairly silly
expectations of the stock markets (to which the telcos responded slavishly),
and herculean inexperience and stupidity of the Euro telcos.  But they were
not a failure; people make mistakes in markets all the time (I'm always
overpaying for something!) and we don't think it's a market failure.  I
would suggest reading Paul Klemperer's paper on the European auctions, at
http://www.nuff.ox.ac.uk/users/klemperer/hownot.pdf .  He is the world's
leading economics expert in the European auctions.
 
Now about markets generally.  It is a common misperception that markets must
be perfectly competitive in order to work well.  Not true.  The reason that
most of mankind has come around to the view that markets are the way to go
is that they are rather robust; as we get further from the assumptions of
perfect competition, we don't get perfect results, but generally we get
pretty good results, and almost always far better results than government
dictat (which is usually the alternative).  It is also believed by many that
markets are easy to monopolize because business people are
profit-maximizers.  In fact, markets are extremely difficult to monopolize,
precisely because everyone is a profit-maximizer, and will work very hard to
undermine monopolists; after all, that's where the money is!  In the economy
as a whole, monopoly is rather rare, and it's usually there because the
government sustains it!  When it does occur, we have the antitrust folks to
get after them.  I know that many IPers think antitrust is ineffective; this
is possibly because they haven't looked at the record.  The FTC and DoJ deal
with hundreds of cases each year and they are very effective.  If you didn't
like the outcome of the Microsoft case, well, a sample of one is probably
not statistically significant.  While cases like this attract the attention,
the fact is that in most of the economy, including the most important
commodities (food, clothing, shelter), markets work just fine, even if they
are not "perfect."  As far as spectrum is concerned, once all of it gets in
the market and we can deploy the new radio technologies, there will be so
much spectrum available that it will be impossible to "corner the market."
The only way that could happen is if government policy screws it up and
creates a monopoly.
 

For archives see:
http://www.interesting-people.org/archives/interesting-people/