So after a long hiatus I’ve finally gotten back to reading The Wealth of Networks, and am nearing completion. I continue to find Benkler’s treatment of “spectrum commons” frustrating. Here’s how he describes the issue on page 403:
As chapter 3 explains, by the time that legislatures in the United States and around the world had begun to accede to the wisdom of the economists’ [arguments for spectrum property rights], it had been rendered obsolete by technology. In particular, it had been rendered obsolete by the fact that the declining cost of computation and the increasing sophistication of communications protocols among end-user devices in a network made possible new, sharing-based solutions to the problem of how to allow users to communicate without wires. Instead of having a regulation-determined exclusive right to transmit, which may or may not be subject, to market reallocation, it is possible to have a market in smart radio equi9pment owned by individuals. These devices have the technical ability to share capacity and cooperate in the creation of wireless carriage capacity. These radios can, for example, cooperate by relaying each other’s messages or temporarily “lending” their antennae to help neighbors to help them decipher messages of senders, without anyone having exclusive use of the spectrum.
This rather surprised me, since I didn’t remember chapter 3 explaining any such thing. Looking back, I found a brief discussion of the economics of spectrum commons that (as I’ll explain below the fold) falls far short of justifying the stark claim that the need for spectrum commons “had been rendered obsolete” by technological developments by the late 1990s.
At the end of chapter 3 (p. 87) Benkler states that “By [the late 1990s], however, the century-old engineering assumptions that underlay the regulation-versus-property conceptualization of the possibilities open for the institutional framework of wireless communications had been rendered obsolete by new computation and network technologies.” This sentence is supported by a page-long endnote (#11) that points to more than a dozen works about spectrum commons. But there’s something strange about them: every one of the authors is either a lawyer or an economist. As far as I can see, not a single one is by a computer scientist or electrical engineer of any description. Nor do any of the works cited seem to point to real-world systems that embody the principles of a spectrum commons. It cites this paper, for example, which was written back in 1995 and proposes a new, entirely speculative framework for non-exclusive spectrum sharing. The author doesn’t appear to have even attempted to work out any of the engineering challenges presented by his paper, perhaps assuming that those were implementation details to be worked out later.
The only real-world example that Benkler is able to point to is WiFi. Although WiFi is clearly an important development, and there are certainly lessons to be learned from its success, it is a poor example for Benkler’s purposes. Benkler is trying to argue that a genuinely decentralized network infrastructure can solve the last mile problem by eliminating the need for a corporation to come in and make large, up-front capital expenditures. Yet municipal WiFi, as it’s deployed in the real world, is neither a solution to the last mile problem nor an alternative to capital-intensive infrastructure development. WiFi’s range is far too short to solve the last mile problem directly. And to my knowledge, no one has attempted to solve the last mile problem indirectly by building a WiFi-based mesh network. To the contrary, all of the municipal WiFi efforts I’ve heard about use a more traditional, capital-intensive structure, in which traditional wired connections are run to within a few hundred feet of each resident’s house, and WiFi is simply used to bridge the gap from the telephone pole on the corner and the customer’s home. That will undoubtedly save some money by reducing the amount of fiber that must be laid, but bringing connectivity to thousands of access points sprinkled all over the city will still require significant capital expenditures.
Now, this isn’t to say that mesh networks aren’t possible, or that we might someday have the kind of network Benkler describes. But if Benkler wants to establish that mesh network technology has already rendered the need for spectrum auctions obsolete, the way to do that isn’t to cite papers by himself and Larry Lessig. The way to demonstrate the viability of mesh networks is to point to prototypes of actual working mesh networks. Or at the very least Benkler should have some citations to papers by network engineers arguing that mesh networks are a technologically feasible replacement for traditional telecom infrastructure.
Simply linking to a bunch of his law professor buddies making the abstract arguments that mesh networks are a swell idea doesn’t cut it. A few weeks back, I linked to a devastating critique of the Hahn/Litan network neutrality paper. In that post, “Cog” pointed out that Hahn and Litan made errors that would have been obvious to anyone with minimal technical competence, and suggesting that Hahn and Litan had allowed their ideological commitments to cloud their judgment about the limits of their technical understanding. I wonder if Benkler’s not making a similar mistake here. It would obviously be very convenient for Bankler’s argument if mesh networks could solve the last mile problem, because then spectrum policy would fit in nicely with the broader themes of his book. But the fact that he and his law professor buddies all think mesh networks are a great idea doesn’t necessarily mean they’ll actually work the way he imagines they will.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.