Solving the spectrum crisis: The federal government giveth, then taketh back


Guest post by Larry Downes

In one of the most influential essays on U.S. history ever written, Frederick Jackson Turner announced in 1893 that the American frontier had closed. According to census data, Turner noted, there was no longer any tract of unsettled land large enough to constitute a "frontier line," and the Census Bureau dropped the term in 1890. The Wild West was wild no more. To satisfy its insatiable desire for land, the United States would need to start looking beyond North America--to the Pacific, the Caribbean and elsewhere--a shift that forever changed American foreign policy and our place on the world stage.

Today, we face a similarly transformative moment in the history of commercial communications. After nearly 80 years of licensing bands of radio frequencies in small increments, often at nominal cost to the licensor, the Federal Communications Commission has run out of new spectrum, at least within bands current technology can effectively use. Today, the legacy of the agency's ever-shifting "public interest" criteria for allocation includes some 50,000 licenses, many for now-obsolete technologies and local services that the FCC once felt needed government support.

But like empire builders at the beginning of the 20th century, our demand for bandwidth has only grown more voracious. Fueled by a revolution in mobile computing driven in equal parts by consumer and business applications, existing allocations of wireless spectrum are rapidly being saturated with data. According to the FCC's own 2010 National Broadband Plan, conservative estimates suggest mobile broadband users will need 300 megahertz of additional spectrum by 2015. There's no end in sight to annual increases in traffic, now measured in the thousands of percent.

That means, to put it bluntly, that for the first time in U.S. history, current licensees of a large swath of frequencies will need to be separated one way or the other from their spectrum allocations. Without more spectrum (along with more cell towers, antennae and smarter devices), mobile broadband demand is certain to outstrip spectrum supply. That would mean nothing less than the eventual (perhaps imminent) collapse of the mobile industry, or what the FCC euphemistically refers to as the "spectrum crunch."

Repurposing underutilized allocations isn't easy

Licensees don't actually own their allocations, but canceling existing licenses would be both legally and politically impossible. The easiest solution, rather, would be for those who no longer need some or all of their spectrum to sell it to mobile providers. Unlike 19th century federal land grants, unfortunately, spectrum licenses can't simply be sold to buyers who value them more. FCC licenses can in theory be transferred in a secondary market, but only with the agency's permission. And after approving several transfers within the mobile industry in the last five years, the agency's rejection last year of AT&T's effort to acquire spectrum from T-Mobile USA on supposed antitrust concerns signaled a new unwillingness to let the secondary markets work to head off the crisis.

(Verizon's effort to obtain spectrum from a consortium of cable companies--spectrum that isn't currently being used for anything--hasn't yet been approved. Several self-described consumer advocates are fighting the deal.)

On the western frontier, new and better uses of land were easily accommodated by the market. Farmers and miners could simply sell some or all of their stakes to ranchers, and ranchers to housing and retail developers.  But mobile providers generally can't look outside their industry for solutions.  That's because even when transfers are approved, specific use conditions remain permanently attached to the frequencies.

While it's not impossible to get the FCC to change the terms of a license, the process is expensive, cumbersome and often fraught with political interference. Earlier this year, for example, the agency revoked temporary authority it gave to startup LightSquared to repurpose unused spectrum allocated for satellite applications for a new mobile broadband network. (GPS device manufacturers and the Defense Department complained that the new network could cause interference problems for their receivers.) The company is now mulling a bankruptcy filing; billions of dollars of investment may go nowhere.

And alleged political favoritism to LightSquared in getting the FCC's initial blessing has mired the FCC in a scandal, one result of which is that the appointment of two new commissioners has been held up for months.

To avoid more embarrassment on a similar waiver request by Dish Networks, the agency recently opened a Notice of Proposed Rulemaking, a formal process that could take months or longer to be resolved.

What else can be done? The FCC's "Plan A" to avert a crisis has long been to pry valuable and underutilized spectrum loose from over-the-air television broadcasters operating in the 700 MHz range. The broadcasters, of course, would prefer to keep their spectrum, if only to hold out for higher valuations as the mobile crisis worsens. To break the logjam, Congress authorized the FCC earlier this year to begin proceedings to create "voluntary incentive auctions."  The new auctions will allow participating broadcasters to give the agency authority to bundle and auction off some or all of their current allocations, sharing the proceeds with the Treasury.

Assuming the broadcasters cooperate, the voluntary incentive auctions could free up highly desirable spectrum for mobile broadband users. But it will likely be years before the process yields repurposed frequencies. Even in the best-case scenario, the new auctions won't come close to freeing up 300 MHz. The FCC believes that by 2020 that number will have risen to 500 MHz.

Meanwhile, the mobile revolution continues, with consumers and business users alike largely unaware of the sword dangling just above their heads.

The federal solution looks tantalizingly within reach

Other solutions could come in large part from the federal government itself. Federal agencies, notably the Defense and Justice departments, hold large amounts of extremely valuable spectrum. Many of these licenses were granted years or even decades ago for systems and applications that have long outlived their usefulness to any national priority. Freeing up some of this spectrum and auctioning it for private use could go far in averting the crisis and, not incidentally, generating significant revenue to offset the federal deficit.

To encourage agencies to free up spectrum, the Obama administration issued a memorandum in 2010 directing the National Telecommunications and Information Administration to coordinate government efforts to free up spectrum. (NTIA, part of the Commerce Department, coordinates all federal spectrum, but has no direct authority over other agencies.)  At the end of 2010, the NTIA released a 10-year timetable for meeting White House goals, and, late last month, issued its first substantive report on the "viability of accommodating wireless broadband in the 1755-1850 MHz band." (A good summary of the report by Richard Bennett can be found here.)

At first glance, the NTIA report is encouraging. Since the 1755-1850 MHz band is within the range of frequencies internationally assigned for mobile operations by the U.N. International Telecommunications Union, the NTIA has given top priority to clearing federal users off this 95 MHz swath of spectrum. And the good news is that, according to the 50-page report, "it is possible to repurpose all 95 megahertz of the band" within the 10-year goal set by the White House. 

Twenty different agencies utilizing more than 3,100 individual frequency assignments each submitted timetables and cost estimates to the NTIA for evacuating the 1755-1850 MHz band. They also identified alternate bands to which their existing users could be relocated.

The NTIA also announced that it may be possible for government and private users to share some frequencies within this valuable range even as federal users migrate to other frequencies. That possibility was well-received by mobile industry groups and leading mobile providers, including Verizon and AT&T. 

Overwhelming "challenges" undermine the government's plan

But reading the report carefully, NTIA and the agencies seem more interested in appearing to support the White House goals than actually doing so. The report repeatedly downgrades its optimistic findings of the potential to clear nearly 100 MHz of essential spectrum by reminding readers of significant "challenges" that would need to be overcome. By the end, "challenge," a word that appears 17 times in the report, starts to sound like code for something nearly impossible to achieve.

In the end, it appears as if the NTIA has little expectation the federal government can actually free up the spectrum for the FCC to auction, or even to work out an interim sharing arrangement during the transitional period.

Consider three of the more withering "challenges":

1. Relocation: NTIA asked the agencies to identify alternative spectrum to which it could relocate, or, even better, to "achieve comparable capabilities via alternate means." None of the federal users, however, appeared ready or able to take the simpler step of acknowledging that it no longer needed it current allocation, or suggest that its current uses might become obsolete in the next ten years. Every agency expects its current uses to continue, and prepared detailed analyses of alternative frequencies to which it could migrate. Apparently, each and every one of the 3,100 existing assignments is still essential to its respective agency, and will still be ten years from now.

Assuming each of these applications will really continue to be needed, few if any of the agencies seemed to take seriously the alternative of migrating redundant federal systems to commercial platforms. Thirteen different agencies, for example, operate "video surveillance" systems within the targeted range of frequencies, but there's no suggestion that some of the less-sensitive uses could be served by private providers.

Every federal system appears to need unique hardware and software--at least according to the government. Following similar logic, Congress recently committed more spectrum in the 700 MHz band for a standalone public safety network rather than working to give first responders preferred access to existing and likely more robust private networks. Public safety has been unable to make much progress with significant spectrum already allocated to it, and it remains to be seen if an interoperable network will ever be built, let alone cost-effectively.

In any case, relocation creates a serious and potentially fatal obstacle to the NTIA's plan. Why? All of the frequencies identified as being technically usable for relocation are themselves currently assigned to other government or commercial uses. These frequencies would also have to be cleared, or would require technical and operational solutions for sharing.  "Arrangements, such as sharing, additional relocations, or other forms of accommodation would need to be developed where appropriate," as the report puts it. "Where appropriate," however, seems to mean "in every case."

2. Cost: The total estimated cost to achieve the clearing of the 1755-1850 MHz range comes in at a whopping $18 billion. Or does it? These numbers were simply provided to NTIA by the agencies, and don't appear to have been subjected to any kind of detailed audit or even a sanity check. The report offers no details as to how these estimates were prepared, or what costs they include. They have the appearance, at the very least, of being pulled out of a hat.

In particular, estimates from the Defense and Justice departments seem to have been arrived at by working backwards from an amount that would put off any reasonable intention to relocate. The total cost of relocating video surveillance applications, for example, is pegged at over $5 billion.

And DoD estimates $4.5 billion just to relocate its Air Combat Training Systems to another frequency, a move the department insists cannot happen any sooner than in 10 years. Yet the report also indicates that the airborne operations of the training systems "can be accommodated or are already being performed" in the proposed relocation frequencies and that "these bands have similar or better propagation characteristics." We don't have the details to say for sure that DoD is dragging its feet here, but it certainly looks that way.

The cost estimates are another potential show-stopper. As the report notes, "Current law requires that auction proceeds exceed expected federal relocation costs." In other words, if the FCC doesn't think it can raise more than $18 billion in auctions for the 1755-1850 MHz range, no clearing will take place (at least not for the entire range). Even in the current climate of high valuations for usable spectrum, it isn't clear that FCC auctions for 95 MHz would raise that much money.  No relocation costs, no spectrum.

3. Sharing: On the surface, the offer to share spectrum during a transition period appeared to be a sign that federal agencies understand the crisis and are willing to cooperate with the public to avoid it. But the offer to share comes with considerable strings attached. As the report puts it in no uncertain terms, "sharing during the transition period and possibly thereafter will require establishment of clear regulatory mechanisms prior to any auction to ensure appropriate protection of federal operations." 

Which is to say that if the FCC wants to conduct auctions earlier than 10 years from now, those auctions will be conditioned on enforceable guarantees that any commercial applications take a back seat to remaining federal users. Those conditions, by the way, would likely depress the auction price, making it even harder to satisfy the at-best generous cost estimates for relocation.

Moreover, the winners of any "shared" spectrum will also be subject to "acknowledgement by industry of its status with respect to potential interference from federal operations. "No doubt in response to the conflict that arose over LightSquared and the GPS community, the NTIA appears to be saying that any question of interference in new commercial uses of the shared spectrum will be resolved in favor of the remaining government users. (The potential interference may be the fault of GPS receiver design rather than LightSquared, an argument the company continues to pursue). For auctions during the transitional period, the winning bidder's "status" here will be that of a squatter, subject to eviction on even a pretense of possible interference.

It's also possible that sharing may make it harder, not easier, to clear the federal users. If sharing rules can be worked out, it takes the pressure off the government to ever give up the spectrum outright. In some sense the offer to share may be illusory, taking back the spectrum even as it appeared to be offered.

The doomsday clock keeps ticking

NTIA offers no hint of solutions for any of these "challenges," other than to "establish appropriate fora to encourage communications between federal agencies and industry." Maybe these meetings will generate more realistic relocation plans, cost estimates and sharing arrangements. Or maybe they'll just lead to more talk, more reports and more delay. Without any real requirement (or incentive) for federal users to negotiate, there's no reason to expect any serious effort any time soon.

The FCC's own doomsday clock, meanwhile, keeps ticking. Two years into the National Broadband Plan's five-year benchmark, the FCC has managed to free up only a fraction of the 300 MHz the agency correctly argues is essential for the continued health of the mobile ecosystem. 

With the spectrum frontier now decidedly closed, however, easy solutions to the crisis are all gone. Serious horse trading will have to start, and start soon. But for now, given the inflexible licenses of both public and private holders, horse trading isn't even allowed.

So the real starting point to avoiding catastrophe is radical reform of nearly 100 years of spectrum policy premised on preferred applications and non-transferable licenses. Congress must act to stop the FCC from continuing its legacy practice of picking winners and losers under the undefined and notoriously squishy "public interest" standard, which allows the agency to set aside spectrum for favored applications and bidders and to attach unrelated conditions to spectrum licenses. 

These conditions, along with specific use limitations, travel with the license, unnecessarily constraining the ability of secondary markets to respond to changing technologies, applications and user demands. Congress is currently considering legislation that would rein in the worst abuses that result from the FCC's freedom to attach conditions. A reform bill recently passed the House, but is unlikely to go anywhere in the Senate.

As for usage limits, Congress already amended the law in 1999 to allow the FCC, under specific circumstances, to grant flexible use licenses. That provision, however, is narrow, and doesn't apply retroactively to the 50,000 licenses the agency has already granted. 

The agency, at least, now appears more willing to exercise this partial power. As part of the March proceeding to repurpose Dish Network's spectrum, the FCC cracked open the door to a more flexible approach for some spectrum, offering to open frequencies currently limited to satellite applications for terrestrial use.

That door needs to be flung open, if not removed forever from its hinges. Bold action is needed, not half measures. Congress and the FCC must make clear that, with the closing of the spectrum frontier, American consumers can no longer afford a spectrum management system that elevates the plodding, politicized and transient whims of the FCC over market forces that work faster all the time.

As FCC Commissioner Robert McDowell wrote in March, "The Commission has a checkered past of micromanaging spectrum use only to find years later that technical innovation and market demands have evolved past the government's myopic view."

McDowell, as he so often demonstrates, is a master of understatement.

Larry Downes is a consultant and author, most recently, of "The Laws of Disruption: Harnessing the New Forces that Govern Life and Business in the Digital Age" (Basic Books 2009).