p241.png p240 _ -chap- _ toc-1 _ p241w _ toc-2 _ +chap+ _ p242
----- {{llfoip241.png}} || The Future of Ideas ||




/tab\/tab\_Free_Spectrum_/tab\/tab\

The physical layer includes technology upon which, or in which, or
over which the network lives. It includes the computers that connect to the
Net, the wires they connect to, the routers that feed those wires, and the
spectrum that substitutes for the wires.

As I have described, most of these elements are owned -- and with one ex-
ception, I think properly so. Computers are private property, whether the
government's (in the NSA or libraries) or individuals'. The wires, whether
copper or fiber, linking these computers to routers to other computers are
privately owned. Massive investment laid them; even greater investment has
been needed to bring them up to date.

These private investments deserve the reward of private property. With
one qualification that I will offer in the next section, the owners of this prop-
erty should be free to use it as they wish. No one should have the right to sit
at my machine. Access to my machine, and the wires of AT&T, should not
be free. If it were free, then those who buy the machines and those who lay
the wires would lose lots of the reason to buy the machines and lay the
wires. If access were free, the incentives to build the Net out would largely
be lost.

But the same virtues from control can't be said of spectrum. Or at least
they can't be asserted with the same confidence. No one builds spectrum;
no investment from AT&T has made it possible. No entitlement justifies
special control over spectrum, and the owners of Ryder Truck Rental and
Leasing should not have a monopoly right to control the highways just be-
cause they purchased expensive equipment to use it. Thus, any monopoly
control over spectrum should be allowed only if it can be shown that mo-
nopoly control is needed.

My argument is not that exclusive control is not needed. It may well be
that Hazlett is right. The congestion in the airwaves may push us to build
out a property system. Spectrum auctions -- either in advance or in real
time -- may turn out to be needed to use the spectrum in the best possible
way.

But we don't know that yet, and we certainly don't know enough yet to
know how spectrum will be used. Thus, rather than architecting the space
exclusively for control, we should begin, as much as possible, as we be-
gan with the Internet: by building a regime that by design leaves a signifi-
cant part of these resources in the commons. And once we see how that


[[241]]

p240 _ -chap- _ toc-1 _ p241w _ toc-2 _ +chap+ _ p242


v?

name
e-mail

bad

new


or