A Belated Critique of Jerry’s Commons Paper

by on August 2, 2008 · 5 comments

A couple of years ago I plugged Jerry Brito’s spectrum commons paper. What I said in that post is still true:it’s a great paper that highlights the central challenge of the commons approach. Specifically, a commons will typically require a controller, that controller will almost always be the government, and there’s therefore a danger of re-introducing all the maladies that have traditionally afflicted command-and-control regulation of spectrum.

I’m re-reading the paper after having read the FCC’s spectrum task force report, and while I still agree with the general thrust of Jerry’s paper, I think he overstates his case in a few places. In particular:

Only if spectrum is first allocated for flexible use, with few if any conditions on its use, can a commons or a property rights regime help overcome the inefficiencies of command-and-control spectrum management. For example, if spectrum is allocated for flexible use, a property rights regime will allow the owner of spectrum to put it to the most valuable use or sell it to someone who will. Similarly, if there are no restrictions on use, a commons will allow anyone to use the spectrum however she sees fit, thus overcoming command-and-control misallocation.

However, while title to spectrum could theoretically be auctioned off in fee simple with no strings attached, a government-created and -managed commons will always have its usage rules set through a command-and-control process. Users of a government commons might not be explicitly restricted in the applications they can deploy over the spectrum, but they will have to comply with the sharing rules that govern the commons. Sharing rules, which will be established through regulation, will in turn limit the types and number of applications that can be deployed.

I think the difficulty here is that just as Benkler and Lessig over-idealize the commons by ignoring the inevitable role for government in setting standards, so this over-idealizes the spectrum property regime. It’s not true that spectrum “could theoretically be auctioned off in fee simple with no strings attached.” The key thing to remember here is that electromagnetic waves don’t respect boundaries established by the legal system. There will always be a need for technical rules to prevent interference between adjacent rights holders. If you hold a spectrum right in a geographic territory adjacent to mine, the government is going to have to have some rules about how much of your transmissions can “leak” onto my property before it counts as a trespass.

Nor would auctioning off nationwide licenses—rather than geographically limited ones—really solve the problem, for two reasons. First, we’re still going to have to set power limits at national borders to avoid pissing off Mexico and Canada. Second, electromagnetic transmissions bleed not only into adjacent geographic territories but also into adjacent frequency bands. Even in a “pure” property system, then, the government is going to have a role to play in setting out-of-band transmission limits for transmitters.

This standard-setting process can easily become same kind of entering wedge for regulation that Jerry warns about in the context of a commons system. In both cases, the ideal FCC policy is one that is focused exclusively on minimizing interference without placing any restrictions on technical standards or the use to which the bandwidth is put. But in both cases, there will be a temptation for the FCC to impose additional regulations designed to promote other goals. We saw this, for example, with the most recent FCC auction, which was mostly organized around a property model, but saw the FCC imposing “open access” restrictions to some of the spectrum. The temptation for the FCC to meddle is going to exist as long as it has a role in setting technical standards, and I don’t see how any feasible system—property, commons, or otherwise, could completely do away with government regulation of transmitter properties.

The big difference between property and commons regimes is that the latter are unlikely to work well for long distance, high power applications due to the risk of a “tragedy of the commons.” When the FCC tries to extend the “commons” model to distances and power levels much above those found in WiFi devices and cordless phones, it inevitably finds that it needs to impose more specific technical restrictions (such as “listen before talk” rules) on the use of the spectrum. Jerry wants to say that this proves a commons model is ultimately just a trendier version of command-and-control, and that therefore, a commons is not a third way between private ownership and command-and-control. But I’m more inclined to say that a commons is a genuine third way that just happens to only work at lower power levels.

The WiFi band really does operate almost entirely free of government regulation, beyond some very straightforward technical limitations. Where I think the Benklers and Lessigs of the world err is not in declaring this to be a new regulatory model, but in radically overestimating the range of applications to which it can be usefully applied. The vast majority of the spectrum should be allocated for exclusive use because that’s likely to work better for the vast majority of applications. But given the success of WiFi and other applications of unlicensed spectrum, I think we should expand that experiment as well and see what the engineers can come up with.

Previous post:

Next post: