The Airwaves: Open or Owned
� posted by Dan Gillmor 07:30 AM
� permanent link to this item
Heading this morning to what promises to be a remarkable gathering at
Stanford Law School -- a daylong conference called Spectrum Policy:
Property or Commons?.
Thanks to Dave Sifry's Sputnik wireless system, I'm online at the
conference. I gave Stanford the wrong MAC address for my computer's WiFi
card (doh), but Dave created an insta-network that I'm using.
Running notes here.
--------------------------------------------------------------------------------
Joi Ito has set up a trackback page so you can follow weblog postings from
the conference. Turns out that others are doing a much more complete job of
note-taking than I can even attempt.
For instance: Cory Doctorow's great notes...
--------------------------------------------------------------------------------
David Reed is the keynote speaker, and is making the central point on which
this conference is based. A while back I wrote about his thesis, that
there's not really a spectrum shortage -- and that allocating spectrum to
monopolies is entirely counterproductive.
We're going through a viral communications revolution, like transition from
mainframes to PCs, he says. Lower economic barriers to innovative uses.
Because PCs were near the places where they were used you could do things
likie sound and video, things that weren't possible on mainframes.
Central point: "Interference" is a concept we must re-think.
From a regulatory standpoint, interference equals damage. But in terms of
simple physics, radio "interference" equal superposition, an addition of
information.
No information is actually lost, even though receivers may be confused.
What we need is smarter radios.
--------------------------------------------------------------------------------
Yochai Benkler, another seminal thinker in this area, is explaining the old
ownership versus new open models of spectrum. Open spectrum brings
innovation at the edges, not in the middle -- and it places value on
communications, not "spectrum usage."
This is the paper he's using as the basis for his talk.
Cheap information processing is the key, he says, echoing Reed. Adding
users (equipment at the edge) is the way to add capacity -- not adding
traditional "infrastructure" as in the past.
It's not market versus anti-market, but rather what market we push. What we
need is an intermediate-term experiment, he says, some property spectrum
and some commons spectrum, and make it revisable if it didn't work.
Gregory Rosston is looking for the most practical way to guage value. He
accepts Benkler's analysis, though he worries about lock-in by corporate
monopolists that might replace the FCC and certain standards bodies as a
gatekeeper.
Stuart Benjamin says there advantages and disadvantages to private control.
Advantages come in creating and expanding the network, incentives to find
better solutions. FCC has no profit incentive (hmmm, what about the
revolving door?). Is monopoly likely? Probably not, he thinks, because
there would be competing networks.
Howard Shelanski notes that spectrum is about technical abilities of
transmitters and receivers. And that this is about the right to speak -- a
crucial point. The issue is the "displacement cost" of speech over
distance, creating an allocation problem. Both sides want to do the same
thing, to maximize value of speech (but measured in different ways). What
is the value of speech? If we assume people only speak for commercial
purposes, no problem -- whoever bids the most can speak the most. (A narrow
view...) The commons approach doesn't change the need for some kinds of
regulation. Key question: What framework leads to optimizing the flow of
new technology? But who'll invest?
Dave Farber also hears (and celebrates) the commonalities in approach. How
do you figure out what to do? You experiment, but not for 10 years (as
Benkler suggests). Problem: Almost no research left in this country, except
in some small companies. Creates real problems. Universities can't do it
all. Recalls Internet as experiment asking many of the same questions, and
that government almost destroyed it while private sector rescued it in the
end. "We could always have Microsoft buy the whole thing, and I could
testify again against them." Do real experiments, not paper experiments.
"We don't know what works."
Benkler: The Net functions in a commons model in precisely the way some of
the panelists have criticized. Why does market need to give us the signal
which is better? How do we know it's a good thing to have highways and
sidewalks? We don't, and don't place the infrastructure wholly in private
hands. We don't necessarily need the market to teach us that the commons
has value. We do need to think about where investment will come from, but a
credible (long-term) commons will spur the investment we need in end-user
equipment. WiFi market suggests this will work.
Conversation that follows too interesting to spend time taking notes...
FCC's chief engineer, Ed Thomas< says the spectrum "Big Bang" has already
occurred. There's lots and lots of spectrum out there right now for a
commons. The question is what's the economic motivator to do the
experiments people are discussing.
Lessig: How do we create standards/protocols that make this commons
possible? The puzzle is how we can ever know which incentive to choose.
Benkler: I'm the only person here who really believes in the market.
--------------------------------------------------------------------------------
Time for property-rights advocates:
Evan Kwerel from the FCC says there will be some spectrum shortage for the
forseeable future, because there are spectrum uses we cannot tamper with in
some ways, specifically where broad coverage and high mobility is a
requirement.
Lack of interference doesn't mean lack of contention for the resource.
Scalability doesn't mean end of scarcity. Expanding capacity costly no
matter how it's done.
Markets are useful when there's scarcity, he says. (Question: Is the
scarcity artificial?)
Michael Calabrese is bringing the free speech question to the fore. He
wants the bias not to be in favor of permanent rights for licensees.
Again...the discussion is too fascinating to keep up this blog-noting
properly...
http://weblog.siliconvalley.com/column/dangillmor/archives/000823.shtml#000823
