A few days ago, the big news in the telecom world was that President Obama again ordered federal agencies to share and sell their spectrum to expand commercial mobile broadband use. This effort is premised on the fact that agencies use their gifted airwaves poorly while demand for mobile broadband is surging. While the presidential memorandum half-heartedly supports clearing out agencies from some bands and selling it off, the focus of the memo is shared access, whereby federal agencies agree to allow non-federal users to use the same spectrum bands with non-interfering technologies.
The good news is that there is no mention of PCAST’s 2012 recommendation to the president to create a 1000 MHz “superhighway” of unlicensed federal spectrum accessed by sensing devices. This radical proposal would replace the conventional clearing-and-auction process with a spectrum commons framework reliant on unproven sensing technologies. Instead of consumers relying on carriers’ spectrum for mobile broadband, this plan would crudely imitate (in theory) wifi on steroids, where devices would search out access over a huge portion of valuable spectrum, avoiding federal users. Its omission in the recent memo likely means the unlicensed superhighway won’t be pursued.
Still, this doubling-down on other forms of dynamic spectrum sharing is unfortunate for several reasons. First, it mostly entrenches the disastrous status quo by acceding to federal agencies’ claims that they can’t be safely moved. Giving federal agencies free spectrum decades ago was a costly mistake that needs to be corrected through pricing and through clearing. By throwing their hands up and saying that clearing and auctioning federal spectrum is too difficult and sharing is the best alternative, the administration forces us to suffer for the mistakes of the past.
Second, sharing, as envisioned in the memo, will not be accomplished quickly or extensively. Whatever technologies come out of this–there are several options, which only adds research delays–will be constrained by what interference risks the agencies accept. Engineering tests and simulations cannot answer this question; it is an economic and political question, and the economics is very distorted as it is. Federal agencies and particularly the military are very jealous of their spectrum. And who can blame them, since their wireless systems are often used for communications and training exercises that, if not directly protecting the lives of civilians, employees, and soldiers, are an important component of preparation for combat. But this jealousy means agencies are not good at sharing wireless bandwidth.
For “sharing skeptics,” UWB’s experiences illuminates our concerns. Ultrawideband (UWB) is a wireless low-power technology used for radar and data services and, beginning in 1989, its proponents sought regulatory approval to share federal spectrum for UWB commercial applications. UWB uses huge portions of spectrum but is very low power–transmissions from a cellphone are millions of times more powerful than UWB transmissions. Even then, UWB applicants were subjected to a process that can only be described as Kafka-esque as it went–for 13 years–agency to agency, submitting filings and completing interference tests, attempting to show that the technology would not threaten federal operations, before it finally got approval. Indicative of agency foot-dragging, a UWB manufacturer noted,
It took NTIA nearly a year to obtain internal sign off by government users of spectrum to approve with conditions the requests for waivers submitted by [UWB] companies. This despite the fact that the devices . . . were lifesaving instruments for public safety and law enforcement personnel, and all 2500 devices requested, if operating together in a single room, would emit less than one quarter the power of a cell phone.
That same UWB applicant made over 100 trips to DC in 6 years and spent millions of dollars to push his technology. Another large UWB company backed by Intel went out of business in the meantime. To be clear, the technologies contemplated in the memo are different from UWB, but UWB is not alone and the institutional resistance will be the same for future sharing technologies. There will be extensive tests, frequent denials, delays, and billions of dollars of continued waste of underused federal spectrum.
I have no doubt the heads of NTIA and DoD favor making mobile broadband more available to consumers. But it is also their duty to ensure that military and federal systems work well all the time. Given these two priorities (faster mobile downloads of cat videos versus public safety and military training), guess which one the NTIA and agencies will favor? What probability of service disruption will federal agencies tolerate? The answer–as we’ve seen in previous sharing attempts–is vanishingly small. That means if any technologies are approved for sharing on federal bands–a process that will take years–they will be likely constrained by very conservative technical criteria and low-power operations.
The memo’s best recommendation is exploring “incentives” (that is, pricing) for federal agencies to relinquish spectrum. Blair Levin–who worked on the FCC’s 2010 National Broadband Plan–voiced support for creating a “GSA for spectrum” at a Washington Post forum this week, and hopefully this sentiment will become a priority. Until agencies are paying market prices for this valuable resource, attempts to force agencies to share are bound to run into these problems since there is no way to analyze the economic tradeoffs.
But a GSA for spectrum is a long ways off and I suspect the regulatory risks and delays in the interim, combined with the poor economics of the permitted technologies, will scare away most investment. Whatever does emerge will be a poor substitute for the robust wireless networks we see everyday on our smartphones using exclusively licensed commercial spectrum, which is why the memo’s focus on sharing–not clearing and auctioning–is sorry news.
For more on proposals for reclaiming federal spectrum through clearing and auctioning, please see my hot-off-the-presses Mercatus working paper.