European Affairs

The demand for radio frequencies is simultaneously growing and changing fast, especially in bands suitable for mobile communications. Consumers increasingly demand freedom from the physical constraints of wires.

At the same time, technical and market developments - such as the convergence of broadcasting, information technology, and telecommunications - are accelerating in unpredictable ways. New types of service are constantly demanding access to spectrum.

Conventional forecasting techniques are not well suited to this pace of change or degree of uncertainty, so the UK Radiocommunications Agency is using a technique known as "scenario planning." A report produced for the Agency (available on envisages four possible scenarios over the next ten years:

  • Internet convergence: Services converge around the open Internet forming part of the fabric of everyday life for consumers.
  • Digital islands: Consumers, deterred by the sheer scale of choice on the open Internet, and worried about fraud, privacy, and pornography, turn to the safety of 'walled gardens.' These are proprietary closed community networks offering security, convenience, and simplicity. They may be based on leading portals2 or interactive digital television, and may use a variety of standards and technologies.
  • Total mobility: Everything is personalized and untethered; lifestyle and working habits drive the use of myriad mobile, converged applications.
  • Broadband revolution: Consumers' voracious demand for bandwidth outstrips the capacity of mobile communications and can only be met by wireline networks. Tremendous increases in speed and capacity totally revolutionize communications, information processing, and entertainment, e.g. with applications like video on demand and networked game playing. Wireless networks cannot match broadband optical fiber networks and are used only for true mobility, temporary and supporting roles.

These scenarios provide an analytical framework for considering and formulating future spectrum strategy and the implications of convergence for spectrum demand. They are not predictions, nor are they options from which the spectrum manager needs to select. They are not mutually exclusive, and are likely to interact in a complex fashion over the next ten years.

A key conclusion from scenario planning is that spectrum should be managed in a coherent, §exible, and responsive manner. I believe market-based spectrum management tools are well adapted to do this.

They will lead toward the time when technologies like software-defined radio3 and ultra-wideband, now in their infancy, have developed and require further refinement of spectrum management techniques.

In addition, scenario planning has confirmed my view that there is a need to re-think definitions of conventional services, some of which are based on demarcations between services such as broadcasting and telecommunications that are, in reality, becoming increasingly blurred.

Since the development of radio at the dawn of the 20th century, spectrum managers have generally relied on regulation as their sole tool for allocating spectrum to different services and then assigning it to individual users.

This worked well while spectrum was plentiful, but as demand has grown and change has accelerated, the UK Radiocommunications Agency, like other spectrum managers, has found it increasingly difficult to respond well enough and quickly enough. This is for two reasons. First, a central regulator inevitably has less information than those directly involved in the market. Second, legal requirements mean regulatory procedures tend to be time-consuming and in§exible.

As circumstances change, so does the optimal distribution of spectrum, sometimes rapidly. New services are at real risk of being delayed or denied if regulation is the only spectrum management tool. Another harmful consequence can be that users have little incentive to give up unused or under-used spectrum or to invest in more spectrum-efficient technology or services. This creates a self-perpetuating circle of shortage and hoarding.

The UK is making increasing use of market-based spectrum management tools. These allow market participants to make decisions on the basis of their own private information and preferences, offer greater §exibility and dynamic responsiveness to change, and are therefore more likely to maximize the overall economic benefits derived from the use of the spectrum.

UK law allows two forms of spectrum pricing as spectrum management tools: auctions, in which fees are set directly by the market; and "administrative incentive pricing," in which fees are set by regulation on the basis of spectrum management criteria.

In accordance with the European Union Licensing Directive, as well as its proposed replacement, it is a cardinal principle in the UK that spectrum pricing should be used to achieve spectrum management objectives, not to maximize license revenue.

I believe it is important, however, to emphasize two key points. First, market forces are used in the UK as a complement to regulation. For example, market forces may not take full account of external factors, such as interference, or of the spectrum needs of public services for which there is no commercial market. Regulation will, therefore, continue to play an important role.

Second, solutions adopted in the UK will not necessarily be applicable elsewhere. National circumstances, such as the balance between spectrum demand, availability in particular frequency bands, and the pace of market change, will differ from country to country and will affect the selection of spectrum management policies. Auctions provide a mechanism for both pricing and awarding licenses. When contrasted with the alternative of comparative selection, in which the regulator decides, they offer the following important advantages:

Economic efficiency: A well-designed auction with well-informed bidders ensures that licenses are awarded to operators who value them most and can generate the greatest economic benefit. This can be achieved by an auction design that requires bidders, through their bids, to reveal information on their valuations of the licenses on offer.

This information would not be revealed in a comparative selection. Well-informed bidders can be expected to bid up to their valuations and to avoid bidding more than they believe a license is worth.

  • Fairness: Comparative selection is more subjective and less transparent.
  • Greater opportunities for new market entrants: Comparative selection tends to favor incumbents with established track records.
  • Speed and certainty in the selection process.

It is very important to select the most appropriate of the different auction designs to meet an auction's objectives. Auctions, however, are not suitable in all circumstances. For example, they would be impracticable for high volume-low value licenses such as those used by taxis or individual fixed links because the cost of running auctions and of participation by bidders would be relatively high.

Licenses for "Third Generation" mobile telecommunications spectrum were the first to be auctioned in the UK. Five licenses were sold, with the largest license reserved for a new entrant to promote competition and innovation. The auction started in March 2000, and lasted eight weeks. It resulted in total proceeds of £22.5 billion for the five licenses on offer.

In November 2000, the UK's second spectrum auction successfully assigned licenses covering 60 percent of the population for Broadband Fixed Wireless Access at 28 GHz. The license winners will provide competition to fiber, cable links, DSL phone lines, and satellites that can also offer access to broadband services.

In this auction, 26 licenses were left unsold, an outcome seen by some commentators as a failure. I take a different view. I want the spectrum to be used for the economic benefit of the UK, so I would rather retain control of spectrum that is not required by the market at a particular time.

We are currently considering how best to re-offer these licenses to entrepreneurs.

With well-designed auctions, the winners will be those who can derive the most benefit from the licenses. This, plus competition, should strongly encourage winners to provide services quickly to recoup their outlay, to the advantage of consumers.

When administrative pricing is used, license fees are set on the basis of criteria which include: the balance between spectrum availability and current and expected future demand; and the desirability of promoting efficient spectrum use and management, economic benefits, development of innovative services, and competition.

This method has been progressively implemented in the UK, making it affordable to license holders, including small businesses. In fact, tens of thousands of smaller UK business users are paying no more than before, or have benefited from fee reductions. Some users with exclusive national channels, or in congested parts of the country, are paying higher fees. These increases, however, have been no greater than necessary for spectrum management purposes.

The public sector is a major user of spectrum, and should have incentives to use it more efficiently. Under spectrum pricing, the public sector, including the armed forces and emergency services, is charged for spectrum on a comparable basis to the private sector through the application of administrative incentive pricing principles.

The creation of a spectrum market would be a logical development of spectrum pricing. This would enable users to trade spectrum instead of only being able to obtain it from the government. It would have the important advantage of providing an increased commercial incentive to relinquish surplus spectrum, and an additional mechanism to help new services gain access to spectrum more quickly.

The UK Government White Paper "A New Future for Communications," published in December 2000, says that the UK will introduce spectrum trading, depending on amendments to European law. The European Union Licensing Directive currently precludes spectrum trading, but its proposed replacement would allow it.

Every country is different, of course. As demand for spectrum grows, however, I believe that other countries in Europe and beyond will increasingly need to choose tools from the full set, including those that are market-based, if they are to meet their spectrum management objectives in a §exible and positive manner.


This article was published in European Affairs: Volume number II, Issue number II in the Spring of 2001.