March 14, 2003 · Lessig
There’s a new piece by Eli Noam posted (for free for 2 weeks) at the FT about spectrum policy. This follows the brilliant and much discussed piece by David Weinberger on Salon explaining David Reed’s views about “interference.”
Noam has been in the middle of the debate between the spectrum-as-property types and the spectrum-as-commons types for a long time. As his piece concludes, “spectrum should be free to access but not free of charge.” The reason is an assumption that is at the core of his and the spectrum-as-property school of thought: “Eventually, any resource whose utili[z]ation is of value, yet whose use is without a charge, will be over-utili[z]ed.”
The more I hear these property and quasi-property types talk, the more I believe that this is the core assumption that needs to be attacked. There are two possible lines of attack, both, in my view, true, but only one which is useful.
The true and not useful line argues that new spectrum technologies will eliminate scarcity. Maybe, but that argument will not convince anyone but those closest to the technology (if them). Instead, the better line, I believe, is to grant the assumption that “eventually” this resource could be “over-utilized,” but then ask: when?
For what is most striking about this debate is that at the same time the property types argue that property is necessary, they are also arguing that at present there is no spectrum scarcity. (More accurately, they argue that the actual amount of spectrum that is used is tiny, and that any scarcity is solely a function of stupid allocation decisions by the government). But if there is no scarcity, then why we should build an immensely complicated system for pricing and charging for spectrum? Especially why, when we are in a time when the resource, if set free in a commons, could be “shared” through technology without any constraint. Perhaps someday there will be a need to propertize some spectrum. But just because that’s true someday does not mean it is true today.
Two examples to make the point:
(1) Oxygen in the atmosphere is a resource “whose utilization is of value” and at the moment, humans get to use it “without a charge.” Yet even though it may “eventually” be “over-utilized,” no one would argue that we should auction off the oxygen rights now, or charge user fees for breathing just now, simply because “eventually” the use of oxygen may well become “over-utilized.”
(2) The Internet when born was a resource “whose utilization is of value” and yet which did not include protocols that required that people using it be “charged.” Instead, the network simple facilitated coordination without worrying about a price. Economists criticized that from the start, but it would have been a terrible mistake for designers of the internet to listen to the economists and wait for a functioning pricing system before they deployed the network. Instead, the better solution was to begin with the simpler technology (tcp/ip) and layer on more complicated protocols (like a pricing protocol) as they become necessary.
Here again, we should be following the meta-lesson of the end-to-end network: simple network, smart applications. Open the spectrum to free use with minimal protocol regulation, and let’s see whether and when we need to add long-distance charges to the air.
For other resources, check out the web page on our Spectrum Conference, which will have the video from the conference soon, but does have a great paper by Yochai Benkler as well as a great set of links about the spectrum debate.