cess. But if congestion is a problem, then we should allocate access to the spectrum to
minimize that congestion and possibly, if necessary, adopt a structure like Noam's. See
also Benkler, "From Consumers to Users," 561; Benkler, "Viacom-CBS Merger: From
Consumers to Users: Shifting the Deeper Structures of Regulation Toward Sustainable
Commons and User Access," _Federal_Communications_Law_Journal_ 52 (2000): 561; Ben-
kler, "Free as the Air to Common Use: First Amendment Constraints on Enclosure of
the Public Domain," _New_York_University_Law_Review_ 74 (1999): 354; Benkler, "Com-
munications Infrastructure Regulation and the Distribution of Control over Content,"
_Telecommunications_Policy_ 22 (1998): 183; Benkler, "Overcoming Agoraphobia," 287.

These are important differences, though in the end they matter in only one sense. No
one believes it has been proven that spectrum as it is presently used is unlimited. David
Reed points to the research of Tim Shepard and others demonstrating that a wireless
network could be structured so that an increase in the number of users actually _in-_
_creases_ total capacity. Tim Shepard, _Decentralized_Channel_Management_in_Scalable_
_Multihop_Spread-Spectrum_Packet_Radio_Networks,_ MIT, EECS thesis, 1995. See also
Timothy J. Shepard, "A Channel Access Scheme for Large Dense Packet Radio Net-
works," at http://www.acm.org/pubs/articles/proceedings/comm/248156/p219-shepard/
p219-shepard.pdf. Reed argues that the "'capacity' of a free space radio network is not
fixed, but instead is an increasing function of the density of user 'terminals' in that space."
Telephone interview with David Reed, February 7, 2001. This means that as more peo-
ple enter the shared spectrum space (as the number of terminals, that is, increases), the
available spectrum _increases,_ not decreases. The more nodes there are on the network,
the closer these nodes are; the closer they are, the weaker the signal connecting these
nodes must be; the weaker the signal, the more signals there can be. A network of wire-
less nodes could expand spectrum capacity as the number of nodes increases.

But even without this increasing capacity, Baran and Hughes both argue that given ex-
isting capacity, properly deployed, we could fulfill all the need we have for delivering
data across the ether without any constraint at all.

The differences between those who see bandwidth as essentially unlimited and those
who see it as scarce obscure a more fundamental agreement: All would agree that spec-
trum use could undergo a radical shift -- a paradigm change, in Eli Noam's terms. All
would agree that this change would fundamentally alter the nature of our use of spec-
trum and would lead to an explosion of innovation in the use of spectrum that would not
otherwise, under either the government or the market model, exist. Everyone now con-
cedes that state-licensed spectrum has stifled innovation -- the most glaring example was
the government's stalling, because of the pressure of industry, the development of FM
radio. See _Edwin_R._Armstrong:_A_Man_and_His_Invention_ (Eli Noam, ed., forthcoming).
All agree that the alternative of allowing spectrum to be sold is like "having the old
AT&T auction off the right to compete against itself." Noam, "Beyond Spectrum Auc-
tions," 473.

Noam and Benkler have an even stronger (from a legal perspective) argument against
the sale of spectrum. In a world where the control of spectrum was not "necessary," as the
Supreme Court said it was in the NBC case, why was control of spectrum constitution-
ally permitted? Spectrum is speech, and the regulation of spectrum is the regulation of
speech. The constitutional status of spectrum auctions is rendered problematic by the
emergence of this alternative technology. Control is not "necessary" anymore, any more
than control of newspapers is necessary. It would certainly be unconstitutional to force
newspapers to buy a license to print (the way taxi drivers have to buy a medallion to drive

[[289]]

p288 _ -chap- _ toc-1 _ p289w _ toc-2 _ +chap+ _ p290

v?