I’m reading through Yochai Benkler’s argument for a spectrum commons in The Wealth of Networks, and so far I’m not impressed. His footnotes pointed to this fantastic working paper by Gerald Faulhaber and David Farber that I think makes the case better than I could:
Establishing property rights in spectrum is often portrayed as eliminating the commons (Benkler (1997), Reed (2002), Ikeda (2002)); this is not the
case. Commons (and more generally sharing) can exist within an ownership regime; our recommended ownership regime with an easement for non-interfering uses establishes such a commons via the easement. Should it be necessary to have a commons for potentially interfering uses, the most obvious avenue is for the Federal government can purchase a block of spectrum (which it then owns) and open the band to general use under terms and conditions similar to Part 15 (for example). In fact, any state or local government can do the same thing, establishing a “park” in which users are completely free to use the spectrum without permission provided they follow the rules laid down by the owner of the “park.” This is perfectly analogous to public lands, such as National and State Parks, National and State Forests, and municipal parks. Further, private foundations could establish such “parks;” for example, there are many horticultural parks open to the public that are maintained by private foundations. Local neighborhood cooperatives could achieve the same end, possibly requiring a one-time or monthly fee for use. Similarly, private firms could establish such “parks,” charging a one-time or monthly fee for use. We would expect that manufacturers of mesh network devices, for example, may choose to “prime the pump” by establishing spectrum parks in various localities to increase their equipment sales.
Any or all of these mechanisms would permit mesh networks to flourish. The authors cited above have alleged that an ownership regime is fundamentally incompatible with the deployment of mesh networks. In the paragraph above, we count at least six ways in which mesh networks can flourish in the ownership regime with non-interfering easement. While we agree with these authors that mesh networking is an exciting new technology that may well shape the future of communications, we have demonstrated that their assertion regarding mesh networking’s incompatibility with an ownership regime is incorrect.
If Benkler is right that new technologies will allow costless sharing of spectrum, then adopting a property regime will simply mean that society bears some unnecessary transaction costs as the manufacturers of equipment have to purchase spectrum rights for their products. The market value of a non-scarce resource tends toward zero, so we would expect the price of obtaining such a license to fall rapidly once efficient sharing technologies came onto the scene. You would still be able to do everything under a property regime that you would under a commons regime, the costs of administering the system would just be somewhat higher.
In contrast, converting the entire spectrum to a commons rules out many traditional uses, such as traditional radio and television broadcasting. If we adopted a commons regime all at once, and commons-based wireless technologies turn out to be inferior to exclusive uses for some purposes, there will be no easy way to reintroduce exclusive uses. Our only alternative will be to lobby Congress to change the rules.
With that said, I think Benkler’s concrete recommendations, laid out here, are fairly sensible. He understands that a “Big Bang” conversion to a commons regime would be disruptive and risky, and so he advocates gradual experiments with unlicensed bands. Benkler also agrees with Farber and Faulhaber that the law should permit anyone to transmit on any frequency provided the transmission does not interfere with the license holder on that frequency. That strikes me as sensible–analogous to allowing anyone to fly their airplane above anyone else’s property provided it doesn’t cause any harm to the property owner below. Most importantly, Benkler also advocates that the FCC “increase the flexibility of current spectrum licensees to experiment with market-based allocation of their spectrum.”
I think it’s important to keep in mind that property and commons regimes are not inherently in conflict. Land in the United States is primarily privately owned for private use, but there are also roads, parks, shopping malls, and other public spaces managed as commons by the government or various private entities. Libertarians and leftists can debate exactly what fraction of spectrum should be propertized and how much should be held as a commons, just as we argue about whether there are too many or too few government-owned parks. But the existence of state parks doesn’t undermine the property rights of private owners.
Most importantly, everyone agrees that flexible property rights would be superior to the command-and-control system we’ve got now. The disagreement seems to be simply a quantitative one, over what fraction of the spectrum should be propertized and what fraction should be unlicensed. Although it’s important to get that question right, it’s far more important to quickly and decisively dismantle our current, Soviet-style system.