Tuesday, February 05, 2013

Unwashed Rental Cars and Spectrum Policy

There's an old saying: "Nobody ever washed a rental car."
Have you?
This adage comes to mind as I read some of the reactions and comments to Cecilia Kang's curious front page Washington Post story strongly suggesting that the FCC has a plan for the government to build "Super Wi-Fi" networks across the country using unlicensed spectrum. Access to these airwaves "which the FCC officials want to hand over to the public" would be free of charge.
And according to Kang, “if all goes as planned, free access to the Web would be available in just about every metropolitan area and in many rural areas.”
As George Will would say: "Well."
After a flurry of initial breathless stories repeating and building on the "free Wi-Fi for all" mantra, the Washington Post story now has been pretty much debunked. For a good account of the debunking, see Jon Brodkin's "No Free Wi-Fi Isn't Coming to Every U.S. City" in Ars Technica. Another good debunking piece is Jerry Brito's "All You Need to Know About 'Super Wi-Fi' in One Tweet," on Technology Liberation Front.
I don't want to repeat what Jon Brodkin, Jerry Brito, and others already have said concerning the problems with the Washington Post and follow-on stories. I'll just note that there is a lot of justifiable confusion, chronicled in Jerry's piece, concerning what spectrum bands – guard bands or white spaces from the forthcoming incentive auction, or some other "white spaces" spectrum – Cecilia Kang had in mind in suggesting the FCC has a "Super Wi-Fi" plan.
For my purposes, it doesn't matter. Instead, while people are thinking about the subject of using unlicensed spectrum versus licensed spectrum, I want to take advantage of the "teachable moment" to emphasize a couple of points – fundamental principles, really – that haven't been addressed sufficiently thus far.
They relate back to the unwashed rental car phenomenon.
When the FCC licenses spectrum awarded through an auction process, it creates (to a greater or lesser extent, depending on whether the agency imposes extraneous conditions) a property-like right in the licensee. This property-like right provides an incentive for the licensee to develop the spectrum to its highest and best use. In other words, the licensee, having paid for the spectrum, has an incentive to develop and fully utilize it in a way that will maximize the return on its investment.
This economic incentive created by a property rights regime is the opposite of the disincentive that exists to wash a rental car, or change its oil.
It is true that FCC Chairman Julius Genachowski (and others) are fond of touting the innovations that may take place in unlicensed wireless spectrum bands. Garage door openers are most commonly mentioned by way of example of these innovations. And I like garage door openers too. But much less is usually said by these touters of unlicensed spectrum by way of acknowledging the tremendous innovations – and investments – that occur by virtue of the use of licensed spectrum. For instance, just think about all the innovative ways you use your smartphone, the ongoing development of which has been dependent on the availability of licensed spectrum services. Or your cable, satellite, and broadcasting services. Or wireless medical telemetry services. Or aeronautical navigation and flight testing services. And so on – all of which use licensed spectrum.
This is not to say the FCC should have gone through the process of allocating licensed spectrum to create a "Garage Door Opener Radio Service" instead of allowing garage door openers to be developed, tested, certified, and marketed using unlicensed spectrum under the agency "Part 15 rules." But it is to say it by no means follows that garage door openers would not have been developed and marketed in a licensed spectrum band. And this is especially true to the extent the FCC allows flexible use of spectrum licenses. Indeed, the agency should further reform its spectrum policies to allow considerably more flexible use in the spectrum bands it allocates for licensed use.
In a regime in which license holders win their licenses in unencumbered auctions – and in which the FCC allows licenses to be used flexibly without imposing unnecessary command-and-control-style use restrictions -- licensees will have every incentive to develop their spectrum to its highest and best use, whether such use happens to be, say, for new super-duper garage door openers or for super-duper Wi-Fi applications.
I do not mean to say there is not any role for unlicensed spectrum or that there have not been innovations on unlicensed spectrum, for example, including Wi-Fi mesh networks. And, in any event, current congressional policy favors the FCC setting aside some amount of spectrum recovered through the upcoming incentive auctions for unlicensed use.
But it is important from a public policy perspective, when considering the amount of unlicensed spectrum the Commission ought to set aside in one situation or another, for there to be an appreciation of the trade-offs inherent in the licensed/unlicensed equation. Otherwise, all the loose talk about "free" spectrum access and "free" networks obscures an understanding of the real costs associated with the government setting aside spectrum for unlicensed use.
And speaking of costs, you will find precious little in the stories touting a tilt towards unlicensed spectrum concerning the many billions of dollars of revenue foregone to the U.S. Treasury from failing to auction spectrum for licensed use. This is a real tangible cost to American taxpayers that ought not to be lost in the talk about "free" this and that.
With a nod to Bruce Owen, a member of FSF's Board of Academic Advisors, for driving this home to me, one final related point: The use of unlicensed spectrum has an "opportunity cost" which is the value that would be generated and which would flow through to consumers if the same spectrum were licensed. So, for example, the cost of the spectrum used by garage door openers is the opportunity cost to consumers of foregoing whatever would have been developed for use in that spectrum if it were subject to licensing in a property rights regime.
If the proper policy presumption is that generally private property rights and free markets should govern spectrum use – and, in my view, this is the preferred policy presumption – this does not prevent the government, if it decides that some service should be provided at public expense, from acquiring the necessary spectrum rights in the market and providing a “free” service.
Of course, such service wouldn't really be "free." No government services provided for "free" really are.