page image

CHAPTER III - What Is General Purpose Spectrum and Is It a Good Idea?

What would a general purpose spectrum regime look like if it existed, and is it a worthwhile goal to pursue? Many of the participant comments focused initially on the existing impediments to efficient spectrum usage and technology innovation, i.e., the problems to which general purpose spectrum is seen as the solution. In describing the ideal of a general purpose spectrum regime, participants emphasized flexible use and decentralized control, and they envisioned a “polyglot” that could accommodate the three competing models for spectrum management (exclusive, unlicensed and shared use). Roundtable participants were unanimous in embracing a general-purpose spectrum regime as a long-term, “aspirational” goal, although they emphasized the importance of continued incremental gains, and individual participants viewed their preferred spectrum management model as providing the best transition path.

Existing Impediments to Efficient Spectrum Usage and Technology Innovation

There are a number of ways in which the existing spectrum management regime impedes efficient spectrum usage and technology innovation. A significant one is the FCC’s longstanding practice of designating bands for dedicated, narrowly defined uses, together with the Commission’s failure to revisit its historical allocations. As Intel’s Peter Pitsch put it, “If we’ve learned anything about spectrum policy over the last 30 years, it’s that…old technologies and uses get locked in, and the…process slows innovation to the detriment of society.”

Single-Use Allocation. Preston Marshall from Google identified three dedicated, single-use allocations that account for more than one gigahertz of spectrum, some of it quite valuable: Broadcast Auxiliary Service (BAS), Cable Antenna Relay Service (CARS) and the C-Band spectrum used for FSS. The FCC authorized these allocations at a time when wireless spectrum represented the only option for delivery of (principally) video programming. Although this content is now distributed largely through fiber optic networks and the Internet (and some of it could be delivered using commercial wireless services), the FCC still protects the allocations.

Although not everyone agreed with all three examples, no one disputed the broader critique. The FCC’s practice of single-use allocation masks the opportunity cost of the designated usage. The incumbent operator faces an easy choice: “Do I use the spectrum for this narrow purpose, or do I give my license back to the FCC?” However, the choice is flawed from a societal perspective because it creates a classic externality by ignoring the value of alternative uses of the spectrum.

The practice of dedicated, single-use allocation is made worse by the FCC’s failure to revisit its historical allocations. Incumbents who are protected by these allocations have a litany of politically persuasive arguments for preserving the status quo, including the investments they have made (sunk costs), the high cost to transition to another band or a different service delivery mode, and the threat that the alternative arrangement allegedly would pose to public safety or some other public good. The result is a fossilized system in which old, inferior technologies tend to crowd out new, superior ones.

Fossilization generates waste, as evidenced by the large amount of unused or underused capacity, and waste creates scarcity. Scarcity is a triple threat. It impedes innovation because entrepreneurs cannot get access to the spectrum they need at an affordable price. This lack of access limits competition, including competition from fundamentally different business models. And scarcity begets scarcity, as incumbents hoard spectrum that they do not really need in the short-term for fear of losing their longer-term rights to use it.

Fragmentation of Usage Rights. A second impediment to efficient and innovative spectrum usage that the group identified is the fragmentation of usage rights. For land mobile radio and other services, the FCC has sliced up small blocks of spectrum into scores of channels and then assigned the rights to use each channel to one or more licenses in each of hundreds of geographic markets. This process creates hundreds of borders along which licensees must coordinate spectrum usage and raises the transaction costs to aggregate their interests—through private negotiation with one another or with an outside entrepreneur.

The dispute between Nextel Communications and public safety radio users over interference in the 800 MHz band illustrates the problem. In 1991, FCC lawyer-turned-entrepreneur Morgan O’Brien succeeded in getting the FCC to relax its little-used dispatch (Specialized Mobile Radio, or SMR) licenses to allow for the provision of mobile phone services to the public. As millions of mobile handsets began to operate in frequencies that had been used largely to dispatch taxis and delivery trucks, they created interference in the adjacent SMR bands, many of which were used by public safety agencies such as police and fire departments. Although the FCC eventually embraced a spectrum swap proposed by Nextel, the process took a decade to complete in large part because of the FCC’s highly fragmented SMR band plan.i

The LightSquared-GPS (global positioning satellite) conflict is another example of the high cost of fragmented spectrum rights. LightSquared planned to use L-Band (satellite) frequencies to provide a nationwide 4G terrestrial wireless network that could compete with carriers such as Verizon and AT&T. After the company had invested a reported $4 billion (out of a planned $14 billion), the FCC suspended its authorization for the network because of concerns that GPS receivers would suffer diminished performance due to emissions from LightSquared signals in the LightSquared L-Band spectrum. Because the L-Band spectrum had historically been lightly used, some legacy GPS equipment effectively treated the GPS Band and the adjacent L Band as a single band. Thus, LightSquared’s plan to make intensive use of its L-Band spectrum would have adversely affected some GPS devices. The social benefits of an additional nationwide wireless network (LightSquared) arguably outweighed the cost to GPS users, some of whom LightSquared offered to compensate. However, because the GPS interests were so diffuse, it was impossible to reach an accommodation, and LightSquared went bankrupt.ii

Interference Standards. Another interpretation of the LightSquared-GPS conflict places the blame principally on the FCC’s approach to interference harm, which represents a third major impediment to efficient use of the spectrum. As discussed earlier, although interference is a reciprocal harm, the FCC has traditionally viewed transmitters as the sole source of interference. No less important, the FCC has never set clear expectations, or requirements, for receivers. From this perspective, the LightSquared-GPS conflict was a result of the FCC’s failure to clearly define the border between the L-Band and the GPS Band or to require better performance from GPS receivers.iii

These two perspectives on the LightSquared-GPS conflict reflect somewhat different views of the root cause of interference problems. For those who subscribe to the second interpretation, the fundamental problem is that spectrum usage rights are poorly defined from an engineering standpoint. For those who subscribe to the first interpretation, the problem is that spectrum usage rights are poorly assigned from a legal and economic standpoint. The two perspectives are not mutually exclusive (in fact, they can be seen as linked, in that the technical definition of rights can shape the options for the legal/economic assignment of rights). That said, as discussed in Section VI, they imply different fixes to the FCC’s rights-creation process.

Negotiation and Adjudication Procedures. Despite their differences, proponents of both perspectives agreed that spectrum users should have wide latitude to negotiate with their neighbors over usage and interference, as well as avenues to adjudicate differences that cannot be resolved through negotiation. In fact, if the processes for negotiation and adjudication were sufficiently expedient and fair, they could obviate the need to better define and assign spectrum rights.iv However, the FCC limits both of these activities, which Roundtable participants identified as a major impediment to efficient use of the spectrum.

The best evidence for the value of private negotiations is the CMRS bands, where operators are exempt from many (although not all) of the FCC limits. Roundtable participant and Verizon official Charla Rath described the continuous process of “Coasian bargaining” in which the CMRS operators engage:

You [can] negotiate rights at the borders—that you can use your neighbor’s spectrum, [that] they can use yours.... If you talk to our engineers, they’re doing [this] constantly.... We rarely go to the FCC [for] help.... It’s just part of a normal negotiation we do.

Although the FCC requires the CMRS operators to keep a record of the negotiated agreements, the Commission does not need to approve them. By contrast, in the bands subject to legacy command-and-control regulation, spectrum “neighbors” generally are not allowed to negotiate their rights (if they have any incentive to do so). As a result, they must rely on regulators to resolve any conflicts.

Participants were especially critical of the FCC’s reliance on rulemaking, a process designed to make policy, to resolve disputes that should be subjected to fact-based adjudication. One concern is efficiency. Compared to adjudication, rulemaking is slow and inefficient, and companies sometimes go bankrupt while waiting for the FCC to resolve what amounts to a make-or-break issue for them. A related concern is discretion. As an adjudicator, the FCC is limited to the facts of the case at hand and subject to a “substantial evidence” standard of review on appeal. By contrast, in a rulemaking proceeding, the record can be broad and conflicting, there are no sworn affidavits or opportunities for cross-examination, and appeals are guided by the more discretionary “arbitrary and capricious” standard of review.

Clemson University Professor Thomas Hazlett used the term “vertical integration of government” to refer to the FCC’s tendency to use a policy process (rulemaking) to resolve issues that should be decided based on objective criteria (e.g., technical performance standards). As Hazlett put it, we do not want a public interest fight over “whether or not the Forest Service is going to use a fiber optic line versus a wireless connection. We really want that to be an efficiency decision.”

Transferability. A final problem that the group flagged is the inability of certain classes of spectrum users to transfer their spectrum usage rights. Nextel’s challenge was made harder by the fact that spectrum allocated for public safety cannot be traded and can only be used for public safety communications. As a result, O’Brien was unable to buy out public safety licensees individually as he had done with the similarly fragmented private SMR licensees (O’Brien acquired some 40,000 SMR licenses).

Federal agencies face similarly flawed incentives. The National Telecommunications and Information Administration (NTIA) in the Department of Commerce assigns spectrum to federal agencies, and those assignments (they are not licenses) are not transferable. Moreover, under the Miscellaneous Receipts Act, any money received for the United States must be deposited in the U.S. Treasury. A closely related constraint is the Antideficiency Act, which prohibits an agency from spending money that has not been appropriated by Congress. Thus, even if a federal agency had the legal authority to sell or lease its spectrum, under current law, it could not directly benefit financially from doing so by retaining or spending the proceeds.

Some believe that the NTIA’s use of assignments to convey spectrum rights creates an even more basic problem. FCC licenses and NTIA assignments represent two different “languages,” and those entities that are eligible to hold a license typically are not eligible to hold an assignment and vice versa. As a result, even mundane transactions between the two groups, such as when a commercial operator leases some spectrum to a federal agency, become complicated because the two parties must use an intermediary to “translate” for them.

General Purpose Spectrum Described

When participants described what a general purpose spectrum regime should look like, their watchwords were “flexibility” and “decentralized control.” At a “content” level, spectrum users/operators should have maximum flexibility to determine what service to provide and with which technology to provide it—that is, the regime should be “service and technology neutral.” At a procedural level, operators should be able to change the way they use the spectrum, including leasing or transferring their usage rights, without going to the FCC for approval.

Participants emphasized the importance of procedural flexibility, in particular. As Coleman Bazelon from The Brattle Group cautioned, “general purpose spectrum” sounds like an FCC allocation, and we know from experience that the allocation process tends to limit rather than expand opportunities. Underscoring that point, the FCC’s John Leibovitz noted that it is impossible to write allocation and service rules that are flexible enough to completely anticipate the future: “You don’t really bump into limitations until something new [that you hadn’t thought of] comes along.”

Carnegie Mellon professor Jon Peha questioned whether flexibility was the right goal, as opposed to “using the spectrum efficiently” or “providing the right [spectrum-based] products and services.” However, others argued that flexibility offers the best tool for determining the cost/benefit tradeoff of alternative spectrum-use choices. Returning to the theme of opportunity costs, Peter Pitsch observed that the more flexibility the operator has, the better its opportunity-cost calculation from a societal standpoint. At the point where an operator has complete flexibility, it faces something approaching the true social costs and benefits associated with the use of its spectrum.

Participants discussed other desirable features of a general purpose spectrum regime. One participant proposed a policy of automatic approval of any non-interfering spectrum usages, based on the logic that led the FCC to look at using the noise temperature concept to permit unlicensed devices to underlay the signals of existing (licensed) services. Several participants endorsed having an output-oriented interference parameter that would recognize the reciprocal role of receivers in causing interference. (Sections V and VI discuss this concept further.)

Finally, the group concluded that the concept of general purpose spectrum was itself flexible: like a “polyglot” that can converse in multiple languages, a general purpose spectrum regime should accommodate the three basic models of spectrum management: licensed, shared and unlicensed use. Under the first model, spectrum would be licensed for flexible use, as it is now, but the use in any given band or geographic location would be determined solely by the market. Under the second model, a more agile newcomer could share a given band with a less agile incumbent. (Although PCAST proposed sharing largely as a way to exploit underutilized federal spectrum, participants felt the model would work just as well in underutilized commercial spectrum.) Under the third model, based on the traditional concept of unlicensed access, users could operate any qualified device subject to limits on power.

The Transition to a General Purpose Spectrum Regime

Although Roundtable participants were unanimous in embracing a general purpose spectrum regime as an aspirational goal, they were emphatic about the need to continue making incremental gains. Not surprisingly, individual participants tended to view their preferred spectrum management model as providing the best transition path.

The advocates of shared use argued forcefully for that approach as an ideal transition path. First, from a strategic perspective, it is designed to accommodate hard-to-move incumbents. Thus, it offers an alternative to the “brute-force tool” of complete reallocation (“forced relocation”), which can require enormous time and expense to accomplish. Second, from a substantive perspective, permitting agile newcomers to operate in an encumbered band will help to establish the opportunity cost of protecting the legacy use, i.e., the subsidy going to that use. Once that subsidy cost is known, regulators can compare it to the cost of providing the same service using a different technology (e.g., distributing video programming via fiber optic cable instead of spectrum-enabled satellites or fixed microwave links). Over time, as the FCC and NTIA assign more flexible usage rights to the incumbent users, the incumbents will have an incentive to make spectrum available for other, more valuable uses.

Spectrum sharing advocates underscored the ability of their approach to be highly targeted and incremental, which can facilitate the transition process. To return to an earlier example, the allocation for FSS downlink (space-to-earth) signals covers 500 megahertz of contiguous spectrum in a highly desirable part of the C-Band (3700–4200 MHz). Every ground station set up to receive those C-band signals excludes usage by others, and a ground station located in the middle of a large city excludes more usage (that is, its opportunity cost is higher) than one located in a rural area. If the FSS spectrum were opened up to sharing, an agile newcomer could negotiate with the incumbent FSS operator to reach a mutually advantageous deal. For example, the newcomer might pay the FSS operator to use fewer ground stations or to rely disproportionately on the ground stations whose opportunity costs are lower.

Advocates for (flexible) licensed use made some of the same arguments for their model as a superior transition path to a general purpose spectrum regime. The CMRS bands have most of the features needed for flexible use of spectrum. (Although many of these features, or authorities, were part of the original allocation, the FCC approved others later as part of what has been a gradual process of liberalization.) Thus, these bands offer a proven approach that can be extended to other bands, as the FCC did when it relaxed the restrictions on the 800 MHz SMR bands. Precisely because it permits so much flexibility, the flexible licensed use approach would do a better job of clarifying the opportunity costs, according to its supporters.

The licensed use model can also accommodate hard-to-move incumbents, in the view of these supporters. One tool is the spectrum overlay. The FCC’s auction of Personal Communications Service (PCS) spectrum in the 1990s illustrates one type of overlay. Rather than clear the PCS bands in advance, the FCC issued licenses with overlay rights: Licensees could use the PCS bands as long as they did not degrade the transmissions of existing users during a multi-year transition period. Many licensees paid incumbents to vacate the frequencies early, thus expediting the process by which bandwidth was freed up for wireless voice and data communications.v Under an alternative approach to overlays, incumbents are not required to vacate the spectrum.

A second tool for accommodating hard-to-move incumbents is the incentive auction, such as the one the FCC is planning to hold in 2016 to repurpose broadcast spectrum for mobile broadband use. Under the FCC’s proposed plan, broadcasters will be able to relinquish some or all of their spectrum usage rights in exchange for incentive payments.

A third tool is the assignment of additional rights to incumbents. For example, the FCC has granted certain Mobile Satellite Service (MSS) operators the right to use their spectrum for terrestrial mobile broadband while preserving sufficient MSS capability to serve key national needs. (The FCC also allowed the MSS operators to transfer their MSS/terrestrial spectrum usage rights in the secondary market.) Although the FCC considered using an auction process to assign those new rights, in part because there were so few licensees remaining, it opted to grant the rights to incumbents at no cost.

In contrast to supporters of exclusive use and shared use, proponents of unlicensed use did not advocate for their model as the preferred transition path to a general purpose spectrum regime. In fact, proponents of unlicensed use generally expressed support for the shared use model, at least insofar as it accommodates opportunistic users. However, they argued for giving unlicensed and opportunistic devices maximum access to unused capacity in shared use bands as well as bands that have yet to transition to a flexible regime.

In a different vein, John Leibovitz urged the participants to think about the transition to general purpose spectrum—and the goal of flexibility, in particular—more incrementally, with an eye to actions that would not require reallocation by the FCC or NTIA. For example, he asked them to consider whether there were categories of like services (e.g., airborne systems) that could be combined into one “uber-service.” If that were done, the relevant bands would still be constrained by service-level stovepipes, but the stovepipes would be larger and thus more flexible. He suggested that this approach might have applicability in federal as well as non-federal bands. As another example, he suggested that they consider how to make “flexible (licensed) use” even more flexible.

i Thomas W. Hazlett, “The Wireless Craze, the Unlimited Bandwidth Myth, the Spectrum Auction Faux Pas, and the Punchline to Ronald Coase’s ‘Big Joke’: An Essay on Airwave Allocation Policy,” Harvard Journal of Law & Technology, 14 no. 2 (2001): 387–388.
ii Thomas W. Hazlett and Brent Skorup, “Tragedy of the Regulatory Commons: LightSquared and the Missing Spectrum Rights,” Duke Law & Technology Review, 13 no. 1 (2014). Available online: For one source of the $4 billion figure (a figure that Hazlett and Skorup cite), see Michael Bathon, “LightSquared Files Bankruptcy After Network Blocked,” Bloomberg Business, May 15, 2012. Available online:
iii Jean Pierre de Vries, “Optimizing Receiver Performance Using Interference Limits.” Paper presented at 2012 TPRC, Arlington, VA, November 2012. Available online:
iv Peter A. Tenhula, “Enforcement of Spectrum Usage Rights: Fair and Expedient Resolution of ‘Interference’ Disputes.” Paper presented at 2012 TPRC, Arlington, VA, August 2012. Available online:
v See Thomas W. Hazlett, “Efficient Spectrum Reallocation with Hold-Ups and Without Nirvana,” George Mason Law & Economics Research Paper No. 14–16, June 2014. Available online: For additional analysis of transition mechanisms, generally, and overlays, specifically, see FCC Spectrum Policy Task Force, “Report,” ET Docket No. 02-135 (November 2002): 46–51. Available online:
Title Goes Here
Close [X]