IV. Various Techniques for Improving Spectrum Management

Some of the actual or perceived problems with the traditional, centralized administrative approach to spectrum management in the United States were discussed in section III. This section of the report provides a review and critique of some past and current attempts to reform the process. The following discussion of attempted reforms is not intended to be exhaustive. Rather, the examples chosen were selected on the basis of the author's knowledge or experience, with a bias for those topics that may be less well known abroad.

A. The Chicago Experiment

One of the problems inherent in a centralized, administrative approach to spectrum management is that in a very large country like the United States, it is extremely difficult to account for regional differences in spectrum demand. In 1970, when the FCC established the Spectrum Management Task Force in the Office of the Chief Engineer, this entity was charged with responsibility for developing decentralized spectrum management techniques that would address these regional differences and remedy other problems revealed in earlier studies by high-level experts.8 The focus at the time was on private land mobile radio services that were experiencing high levels of growth.

The objectives of the task force included

1. facilitating wider and more equitable sharing of channels in view of the dramatic variations in demand across and within regions, and
2. promoting more intensive use of the spectrum resource by conducting more detailed engineering analyses of applications for land mobile radio licenses.

The location selected for testing these new techniques on a pilot basis was Chicago. By October 1974, the FCC was using the new methods to select frequencies and issue licenses to land mobile radio users in the Chicago region. The design and testing of the regional spectrum engineering system involved a multi-million dollar investment on the part of the FCC.

However, in June 1976, the agency reversed its course and abandoned the notion of trying to micro-manage land mobile spectrum on a regional basis using advanced engineering techniques. Essentially, the agency resumed its use of private sector frequency coordination (described in more detail below). The reasons for this change in direction included the high costs associated with operating regional offices in a time of budget austerity, concerns that some of the analytical techniques were overly ambitious in view of uncertainties in the computer models and the data that supported them, and, significantly, the possibility that the agency was becoming too deeply involved in technical decisions that seemed more appropriate for the private sector.

Although different conclusions may be drawn from the demise of the Chicago experiment, the author believes that it illustrates the difficulty--perhaps, more accurately, the futility--of trying to solve the problems inherent in the traditional, centrally managed, administrative approach to spectrum management. Improved engineering analysis has a genuine role to play in establishing overall regulations and other incentives to ensure efficient spectrum utilization and in the private sector in responding to those rules and incentives, although bigger computers and bigger databases cannot provide the ultimate solution to basic resource allocation problems.

B. Private Frequency Coordinators

From a broad perspective, spectrum management within a particular country includes, in addition to the spectrum allocation, allotment, and assignment responsibilities, representing the country internationally on a global, regional, or bilateral basis; establishing equipment standards to protect against interference and certifying equipment that meets those standards; establishing standards and issuing operator licenses to ensure the competency of users of radio equipment; defining operating conditions and developing such other rules of a general nature as to ensure efficient use of the resource; and enforcing rules established in association with the foregoing activities. Except for conducting international negotiations and establishing broad spectrum allocations, all these responsibilities can be partially delegated to the private sector.

The use of private sector frequency coordinators in the private land mobile radio bands in the United States offers an interesting example of the delegation of spectrum management responsibility to the private sector. Private sector frequency coordination began informally in the private land mobile radio services because, by and large, the available channels had to be shared among many unaffiliated users.9

The basic task of frequency coordination was--and still is--to identify and recommend the most appropriate radio frequencies for a particular applicant. Or, as the FCC formally defines it, "Frequency coordination is the process by which a private organization recommends to the Commission the most appropriate frequencies for applicants in the designated radio service." Choosing the most appropriate frequency in a shared environment involves a multitude of considerations, from trying to minimize interference for existing users to ensuring that competitors are not sharing the same channel.

Until 1958, the FCC's rules did not recognize the need for formal frequency coordination by applicants, although they required licensees to cooperate in the selection and use of frequencies to minimize interference.10 To facilitate this cooperation, groups of users in the various private land mobile radio services formed voluntary committees for recommending frequencies to applicants.

In 1958 the FCC amended its rules to acknowledge the role of voluntary coordinating committees in the licensing process. A decade later, in 1969, the FCC went still further by laying out a set of principles to govern the work of these coordinating committees. These rules provided, for example, that the committees must be representative of all users in the particular service, that they must not discriminate, that they must honor all requests for coordination, and that the fees charged can only represent the cost of the process.

These private coordinating committees benefitted users because they were widely regarded as impartial, professional, and extremely knowledgeable about the particular radio service. The government gained by shifting a significant regulatory burden and by improving the efficiency of the remaining process. For example, the number of license applications with errors was often an order of magnitude less for coordinated applications than for license applications that bypassed the process. The FCC itself conceded that "these coordinators play a pivotal role in helping the Commission develop and manage the radio spectrum."

Private frequency coordination developed because it made sense to the users and to the FCC; somewhat surprisingly, it was not explicitly provided for in the Communications Act. It was not until 1982 that Congress amended the Communications Act to recognize officially the role frequency coordinators play in the spectrum management process. Consequently, the role of private frequency coordinators is now well established under U.S. law.

In spite of the virtues of this cooperative arrangement between the government and the private sector, controversy arises periodically. There have been honest disagreements over the power of the frequency coordinators in the process of frequency management. For example, if a particular application meets all the applicable rules, but the applicant is asking for what seems to be excessive power, should the committee reject it? Or, on the other hand, should the committee simply perform what amounts to the clerical task of determining whether the application meets the rules and then simply recommend a frequency that minimizes any interference that might result? There also have been complaints that the private frequency coordinators have abused what appears to be monopoly power in the process. Despite these problems, the overall success of private frequency coordination in the United States suggests that it could serve as a model for similar types of efforts in spectrum management.11

C. The Use of Quasi-Property Rights in Allocation and Assignments

In administering the radio spectrum resource in a centralized manner the FCC must make a series of decisions, ranging from broad allocation policies at the highest level down to, in the case of private land mobile radio, assigning a particular subcategory of user (for example, a taxicab company) to a particular channel within an allotment for a particular category or class of users (for example, for land transportation users, which includes taxicab companies). Mistakes and inefficiencies can occur at each step, or good decisions taken initially can be eroded over time by technical and marketplace changes. Continuing the land mobile radio example, the FCC can fail to allocate the optimum amount of spectrum to land mobile radio, it can fail to allocate the spectrum equitably between public and private land mobile radio systems, it can fail to allot the optimum amount of spectrum to a particular category of users (for example, land transportation), it can fail to sub-allot the proper amount of spectrum to the particular subcategory of users (for example, to taxicab users), and it can fail to assign a particular taxicab applicant to the optimum channel within the allotment. Moreover, as discussed above, the system chosen by the applicant may waste spectrum, and the agency's technical rules regarding channel spacing, modulation type, and maximum power may preclude the use of a more efficient or effective system.

Even in the unlikely event that the allocations, allotments, and assignments are made in a relatively optimal fashion from a national perspective, they may be far from optimal at a regional or local level. Imbalance invariably arises in major urban areas where forestry frequencies are lightly used and taxicab channels are heavily congested (to use the example given earlier). This rigidity in the traditional spectrum management process was one of the major problems that prompted the FCC's attempt to manage spectrum on a regional basis, that is, the Chicago experiment. The FCC also has tried to minimize this type of misallocation by reducing fractionalization within the allotments and by encouraging what it describes as intercategory sharing.

Nevertheless, different classes of users, as well as different companies within the same category, may have different peak usage times during a typical 24-hour period; for example, plumbers and taxicabs may have different peak usage periods. In the traditional system, economic efficiency suffers from two factors. The first of these factors involves different users loaded onto channels, each of whom has an equal right to access the shared spectrum. This often occurs even though one might be willing to pay more for a less congested channel and another might be willing to pay less and experience more congestion (that is, wait longer to access the channel during peak periods). The other inefficiency stems from the fact that in a shared environment without exclusive assignments, there is little incentive to conserve usage during congested periods because the benefits of utilizing the spectrum more efficiently are not experienced by any one individual, they are dispersed equally to all users sharing the channel. In economic literature this is sometimes described as the "tragedy of the commons" because it leads to excess consumption of the resource, for example, overfishing of common fishing grounds or overgrazing of common pastures.

When confronted with these problems in the 1970s, the FCC created a new class or type of service known as the Specialized Mobile Radio (SMR) service during proceedings that reallocated a substantial amount of spectrum from UHF television broadcasting to land mobile radio. Although this was not stated explicitly at the time, the creation of the SMR actually endowed its licensees with quasi-property rights. That is, although licensees were not granted ownership of the radio spectrum as such, they were given

the ability to voluntarily transfer these rights to others.12 In establishing the SMR service, the agency set aside sufficient spectrum for multiple providers and competitive supply throughout the country. Licenses were awarded on a first-come, first-served basis.

In its consideration of specific technology for the SMR service, the FCC sought to promote the introduction of trunked radio technology. In modern trunked land mobile radio systems, access to a large number of channels (5 to 20, or more) is controlled by computer logic that gives the customer use of a channel for the duration of his or her message and then returns it to the pool for use by other customers. This system facilitates efficient sharing of a number of channels because particular users or groups of users that have no heavy demands during a particular period of time can, in effect, loan their spectrum capacity to others who may be experiencing a peak in their demand and vice versa.

Although the FCC initially imposed significant restrictions on the types of services that could be offered, SMR operators have been allowed to serve broad classes of users. Because they have quasi-property rights in the spectrum to which they are assigned, the SMR operators have a strong incentive to introduce efficiencies in order to maximize their own income. Thus, in urban areas, SMR operators charge for airtime on a usage basis, with higher charges during peak usage times. Some SMR operators go a step further by establishing different priority levels, thus customers with the highest priority can gain immediate access to channels for a premium price. Some SMR operators also try to maximize usage of their channels by targeting customers--security firms, for example--who will use their systems during off-peak hours. By taking such steps in their own economic self-interest, the SMR operators simultaneously discourage frivolous or nonessential use during peak periods, greatly increase the number of mobile units that can be accommodated on a given number of channels in a congested area, and give end-users a choice in terms of price-quality trade-offs in mobile radio services.

Restating this approach, by adopting a system of quasi-property rights, the FCC (a) broke down the inevitable rigidities produced by the traditional methods of allotment and assignment of private land mobile radio, and (b) reduced the inefficiencies (spectrum waste) caused by the traditional sharing of channels (the tragedy of the commons).

In the author's opinion, the creation of the private carrier SMR concept was one of the most successful spectrum initiatives ever undertaken by the FCC. Its initial success was apparent by the mid- to late-1980s, but the wisdom of this series of decisions has become even more apparent in the last few years. From the beginning, the FCC gave SMR licensees near total flexibility in choosing the type of technology they would employ, for example, the type of modulation used. Major technical restrictions applied to maximum power and out-of-band emissions. In addition, over the years, the FCC expanded the types of services that SMR operators could provide. In other words, they could offer interconnected mobile telephone services and data services without seeking special permission from the agency. This flexibility as to technologies and services, coupled with the licensee's quasi-property rights, gave the SMR operator, in effect, the ability to implement a voluntary reallocation of the spectrum to a higher-value use.

This form of voluntary reallocation is precisely what is going on in the United States at this moment. As in other parts of the world, cellular mobile radio has become immensely popular in the United States, and it now serves more than 13 million mobile telephone customers. The SMR service has also been successful, but the channels devoted to trunked SMR systems have been valued in the marketplace at prices significantly below those realized for cellular services.13 Recognizing this disparity in value, one company (FleetCall, Inc., now called Nextel) quietly bought SMR channels in all the major markets, then announced its intentions

1. to convert the systems from analog to digital transmission and from wide coverage (high power) configuration to low power, cellular configuration, and
2. to use the dramatically increased capacity to compete with the two incumbent cellular radio carriers.

Because of the flexibility built into the FCC's rules governing the SMR service, the industry has accomplished what amounts to a reallocation (that is, from ordinary two-way radio services to cellular mobile telephone services) in a very short time without significant regulatory intervention from its entrenched competitors.

Although the regulatory approaches used by the FCC for the land mobile services in the United States and for the creation of the SMR services may not be directly applicable to other countries, the results do suggest that increased reliance on the equivalent of property rights and marketplace forces could alleviate some of the increasingly difficult problems engendered by centralized control of spectrum allocations.

D. Lotteries

Some of the benefits of establishing quasi-property rights in spectrum management were demonstrated above, but the problem of how those rights are to be rewarded initially remains. The problems stemming from the use of time-consuming, costly, and often arbitrary comparative hearings, at least in the context of the United States, were discussed earlier. In 1982, growing frustration with the comparative hearing process led to the passage of legislation permitting the FCC to award licenses by random selection or lotteries. This prompted the FCC to make use of its new authority to award cellular licenses outside the largest markets. It is probably fair to say that, initially at least, lotteries were regarded as a success because they were faster and less burdensome on the FCC and the applicants.

However, lotteries to a large extent have become the victims of their own success. When the FCC tried to use lotteries for other services and the general public became aware of the opportunity to win a potentially valuable license for a relatively nominal application fee, the FCC was inundated with applications. Moreover, "application mills," some of questionable reputation, got into the business of marketing "cookie-cutter" applications for a fee on a broad scale, further increasing the number of applications, which in some instances amounted to tens of thousands. Because of the high volume of applications, the ostensible benefits of lotteries, that is, timeliness and lower costs to applicants and the FCC, were lost.14

In addition to delays and costs brought about by the use of lotteries, questions of fairness or equity also arose. For example, certain cellular lottery winners who had no intention of operating a cellular system were able to garner windfall profits simply by selling their licenses to firms that actually did want to enter the business. To many people, it seemed patently unfair that enormous profits were being randomly bestowed on applicants through the sale of the "public's airwaves." The transfer of these profits to a private individual or firm seemed especially ironic in view of the federal government's chronic budget deficits. Frustration with the lottery experience and the potential for added revenue sources prompted serious consideration and eventually the adoption of legislation authorizing the FCC to auction licenses in some services (see section V).

E. Pioneer's Preference

One of the problems with the traditional administrative method of allocating and assigning spectrum is that it can discourage innovation and the provision of new services. Under the traditional system in the United States, an inventor or entrepreneur envisioning an entirely new service (for example, one that does not conveniently fit into any existing service category) must submit to a two-step process.15 First, the inventor or entrepreneur is required to petition the FCC to allocate spectrum for the service envisioned. To persuade the FCC to allocate the spectrum, the petitioner has to reveal the nature of the proposed service and the technology to be employed. Second, if the spectrum is reallocated as requested, the petitioner may have to compete in a comparative hearing or, worse, in a lottery with other applicants who have done little more than copy the idea.16 As a practical matter, a venture capitalist or other investor would hesitate to support the inventor or entrepreneur when there is no reasonable assurance that a license will be awarded at the end of the process. In other words, the traditional system imposes a significant regulatory risk that effectively discourages innovation.

In 1990, the FCC set out to remedy this situation by unanimously deciding to grant preferential treatment in its licensing processes for parties requesting spectrum allocations in conjunction with the development of new communications services and technologies. According to the FCC:17

This preference will help ensure that innovators have an opportunity to participate either in new services that they take the lead in developing or in existing services to which they wish to apply new technologies. The Commission believes that a new "pioneer's preference" will foster the development of new services and improve existing services by reducing for innovators the delays and risks associated with the Commission's allocation and licensing procedures.

At the outset, the chairman of the FCC touted the adoption of the pioneer's preference, calling it a valuable part of the "national communications policy blueprint."18 The FCC subsequently awarded two licenses based upon pioneer's preferences, the first for a "little" low-earth-orbit satellite (LEO) system, and the second for an advanced nationwide paging system in the 900 MHz band. The agency also made tentative preference awards for 3 companies (out of 97 that applied) in its current proceeding to reallocate a large amount of spectrum for personal communications services (PCS).

In spite of its initial enthusiasm, the FCC recently hinted at the possibility of a total reversal of its policy of awarding pioneer's preference. The reasons for the proposed reversal of policy are not entirely clear,19 but in launching a new rulemaking proceeding on the subject of the pioneer's preference, the agency indicated that such a preference might not be necessary if an inventor or entrepreneur could simply acquire the needed spectrum in an auction. Beyond this, one can only speculate, but it is clear that choosing the true pioneers from among a large number of applicants is very difficult. Although the process was not designed as a comparative hearing, in reality, it poses many of the same difficulties. Moreover, in anticipation of an auction environment, it is not clear exactly how a preference should be awarded (for example, through an outright grant or a "discount" in the auction price) and, of course, any such preference would reduce the revenues flowing to the government from the auction.

Perhaps it is too early to judge the merit of the pioneer's preference policy in spectrum management, especially in view of the agency's newly acquired auction authority. The author's personal view of the policy is very favorable; but it is also probably true that, like comparative hearings, it tends to look better in theory than in practice.