Reforming UK Spectrum Policy
In 2001, I was invited by the UK Government to conduct a review of UK Spectrum Management. The report was submitted in March 2002 and this article provides a summary of its conclusions, with some additional commentary.
The goals of the review
The use of radio spectrum has become an integral part of society’s infrastructure. For decades, viewers have taken for granted the reception of clear TV signals, travellers have relied upon assured communications and radio-location for aircraft, and all citizens have benefited from radio connectivity for the public safety services. More recently, the phenomenal growth in personal mobile communications has turned wireless access via mobile phones from a luxury to a necessity for many people.
This value to individuals, businesses and the public sector of access to radio spectrum is becoming increasingly recognised. Radio makes a substantial and increasing contribution to the economy. Recent studies by the Radiocommunications Agency show that even for selected sectors of the economy, the value of radio to the economy as a whole exceeds £20 billion per annum, over two per cent of UK output. Success in managing access to radio spectrum should thus boost the performance of the UK economy.
Looking forward, spectrum is an essential raw material for many of the UK’s most promising industries of the future. Wherever consumers demand mobile and ubiquitous access to communications, wireless products using radio signals will provide the solution. Radio is a uniquely versatile communications medium, essential to connecting up the information society. New products and services typically complement rather than replace existing ones, so adding to the demands on the radio spectrum. Furthermore, the boundaries between new services are blurring, transcending current business models, reducing the predictability of spectrum use, and challenging current regulatory categorisations.
So spectrum management is becoming simultaneously more difficult and more important. But the UK is well placed to respond to this regulatory challenge. The Radiocommunications Agency has a well-deserved reputation as one of the most forward-looking and progressive spectrum managers in the world, having enabled the development of flourishing wireless services in one of the world’s most congested radio environments. In recent years, it has garnered valuable experience of the new market-based tools introduced under the Wireless Telegraphy 1998. My purpose is to look forward to the principles which should guide the Government and Ofcom in managing access to the radio spectrum in the years ahead, in order to derive most value from this national asset for the UK as a whole. To do so, market mechanisms should play a much broader role in allocating and assigning spectrum to its best use, building on regulatory foundations which are essential for any market to work efficiently.
Challenges facing spectrum management
The fundamental building blocks of regulating access to radio spectrum have remained essentially the same during the hundred-year history of radio. Spectrum blocks are allocated, through international agreement, to broadly defined services. National regulatory authorities then assign licences for use of specific frequencies within these allocations within their jurisdictions. The current UK primary legislation for spectrum management, the Wireless Telegraphy Act 1949, is largely based on the Wireless Telegraphy Act 1904.
This regulatory task involves an inherently complex balancing act in a range of dimensions, in each of which there are conflicting considerations:
• Interference. Transmissions interfere unless sufficiently separated in terms of frequency, geography or time. Regulators must strike a balance between reducing the extent of harmful interference, through careful planning, and enabling new and potentially valuable new services to enter the market.
• International co-ordination. The effective use of radio spectrum in the UK will typically require careful co-ordination with neighbouring countries, to mitigate the extent of harmful interference. The Government must weigh up the benefits of co-ordinated and harmonised use of spectrum across Europe against the constraints which this imposes on spectrum management in the UK.
• Investment in equipment. Most radio equipment can operate over only a limited range of frequencies, and so relies on predictable access over time to defined frequency bands. Stability in spectrum to encourage investment in equipment can slow the pace of spectrum re-use. Increasingly, technical specifications are determined internationally to reap economies of scale in production. National regulators need to balance stability and international harmonisation with responsiveness to new technologies.
Developments in technology over the last century have opened up the range of useable radio spectrum, so enabling ever-greater access to new allocations and assignments. While demand from consumers, businesses and public services for wireless communications kept pace with this increased supply over much of the twentieth century, the regulatory regime has proved sufficiently flexible to cope. But with a sharp acceleration in demand in recent years, change in the market place is outpacing the ability of the national and international regulatory regime to respond.
Fundamentally, the spectrum manager is called upon to devise procedures to ration current and future demand for radio spectrum between competing commercial and public service users. To do so centrally would require a detailed knowledge of supply and demand trends, technology developments, and the relative value to society of alternative services. This represents a mammoth central planning task, which is now beyond the scope of any regulatory body, no matter how well staffed and managed. The central regulator is becoming less able to accumulate and assimilate sufficient information to make a correct assignment of spectrum to optimise use overtime.
Instead, spectrum managers will tend, inevitably, to bias decisions in favour of the status quo for a variety of reasons:
• Demand for spectrum. Incumbent users, facing few if any continuous incentives to economise on spectrum use, will tend to ‘over occupy’ spectrum, making wasteful use of it and reducing the amount which can be assigned to new users.
• Interference management. New services could potentially create additional interference to the detriment of incumbent operators. Technical studies can clarify the potential extent of interference, but judgments about results will tend to favour incumbents’ interests.
• Demand for services. New services will be based upon uncertain projections of future demand, against data on actual usage for current operators. The weight of regulatory evidence is likely to be in favour of the latter, particularly where new services will compete with existing ones.
This systemic deficiency of a central planning approach does not detract from the significant steps which the RA has taken in recent years to help meet demands for spectrum from new services. But it does highlight the need to complement the regulatory regime with other approaches to managing access to radio spectrum, in order to enable continued growth of radio-using services in the UK.
Enabling productive and innovative spectrum use
Spectrum is a finite but non-exhaustible resource which is a vital input into an ever widening range of services. The utility of the resource depends crucially on the management of interference from competing users. This has been, and will continue to be, the primary role of the UK’s national spectrum management authority. But the value derived from the economy’s use of radio spectrum also depends on the ability of the system to accommodate shifting demands for spectrum use driven by market changes in technology and consumer preferences. Finally, UK society derives unquantified value from spectrum use by a wide range of public services, from defence to broadcasting, whose reasonable demands for spectrum have to be accommodated within any spectrum allocation regime.
The RA has taken significant steps in recent years to shift its emphasis towards enabling greater economic efficiency in spectrum management. Having assessed the challenges facing spectrum management in the coming years, the review considers that there is an opportunity, and an economic imperative, to move significantly further in this direction. The evidence to date, and prospectively from analysis commissioned by the review, suggests that such a move can be made consistently with maintaining standards of technical efficiency in spectrum use, and with the delivery of a range of public policies which depend upon spectrum as an input.
The fundamental mechanism by which the spectrum management regime could contribute to economic growth is through ensuring that users face continuing incentives towards more productive use of this resource. The review considers that these incentives should be financial and based on the opportunity cost of spectrum use. In this way, spectrum would be costed as any other input into the production process. Price signals about the cost of using spectrum would be disseminated throughout the economy. This information should enable dispersed economic agents to make their own judgments about their use of spectrum and the alternatives open to them to meet their organisational goals.
As with many other input markets, the operation of market mechanisms for spectrum will continue to take place within a framework set by regulation. The intangible nature of radio spectrum and the adverse impacts of unconstrained transmissions on others mean that a considerable degree of regulation will continue to define specific rights to spectrum use. But there is considerable scope:
• to increase the range of spectrum users subject to financial incentives;
• to move such incentives closer to levels at which they reflect the cost to the economy of the spectrum occupied; and
• to increase the flexibility which spectrum users have to respond to these financial incentives.
The application of incentives towards economically efficient spectrum use will vary sector by sector, but can be encompassed in the following figure:
Rapidly changing Environment
Generic allocations, secondary trading of licences, facilitating more rapid ‘refarming’ from one use to another, within a transparent and predictable regulatory framework.
Maximising economic benefits
Auctions and trading of licences where feasible, administratively set spectrum pricing elsewhere.
Protecting social priorities
Make sufficient spectrum available by regulatory rationing for delivery of public services, apply spectrum pricing and positive incentives to share and/or release spectrum into the private sector
The net result of the proposed regime should be to place more information in the hands of spectrum users about the costs of the spectrum they occupy, and more freedom to respond to this information in the choices they make about delivery of their organisational objectives. The aim is to move spectrum as far as possible towards a comprehensive competitive input market,where continuing incentives to economise drive spectrum users towards more innovative and productive use over time.
The benefits of this approach, building on the progress already made in this direction by the RA, will take time to emerge fully. Spectrum use is intimately tied to investment in specific technologies, and major gains in spectrum productivity and innovation are often only possible at step changes in the re-equipment cycle. Lead times between international policy decisions on allocations for new services and the development of commercially viable businesses and technologies can run to decades. Nevertheless, a consistent and comprehensive programme of reforms by the UK should start to bring tangible economic benefits over the next decade. The review sets out an indicative plan for these actions.
International regulatory framework
The international co-ordination of radio spectrum management is an inevitable constraint on the ability of a single country to conduct an autonomous policy for spectrum use within its own jurisdiction. For the UK, as a medium-sized country in a densely populated region, this multilateral approach can bring benefits to consumers and operators. In many areas, the economic value of spectrum in the UK is driven to a great extent by international agreements on technology development and spectrum allocations. Within this framework, though, the review considers that there remain many opportunities for the UK to take a more flexible and market-driven approach to spectrum management, while continuing to benefit from international harmonisation.
To assess whether the international regulatory framework could constrain application of a market-based approach, it is necessary to consider the impact of International Telecommunication Union (ITU), European Community (EC), Electronic Communications Committee (ECC) of the European Conference of Postal and Telecommunications Administrations (CEPT), and bilateral agreements and regulations. Of these, EC regulations and bi-lateral agreements are likely to be the most binding constraints, particularly when considering the scope for enabling market-driven change of use of particular spectrum bands.
If the band in questionis subject to an EC Directive or is judged to be harmonised under the proposed Spectrum Decision, then the new use must be compliant with these regulations. This is an absolute constraint until the band(s) in question is removed from the list of harmonised bands. This seems most likely if the services in question are a commercial failure (e.g. ERMES), or become obsolete (e.g. analogue technology replaced by digital). ECC Decisions become mandatory once signed by administrations (although signing itself is optional).
Bilateral agreements, within the context of ITU regulations which determine which services have primacy in each band, may constrain what actually happens in practice. These are generally framed in terms of the division of frequencies used in border areas and the level of permitted emissions in preferred/non-preferred frequencies across the band and out of band. If the bandwidth of new services differs from that of existing services, then the agreed sharing pattern may not apply and the new use may face harsh emission constraints. This may prevent service deployment in border areas.
A recent study estimates that, in frequencies around 900 MHz andabove, up to 5 per cent of the UK population resides within areas where coordination is likely to be required for most services. The extent to which this would impact on the value of the spectrum would depend on the application and whether additional, unconstrained spectrum were available to support the service. For example, a national broadcaster or fixed wireless access operator could achieve a viable service with less than 100 per cent coverage and would be relatively unaffected by such a constraint, as would a mobile operator which used the spectrum to complement its existing GSM or 3G mobile assignment. The effect of bilateral constraints is likely to affect the UK less than some other European countries which have multiple land borders and/or significant proportions of their populations lying within co-ordination zones.
In summary, theharmonisation of spectrum use under an EC Directive or Decision and bilateral agreements will permit the application of a market-based approach, including secondary trading of spectrum licences, where this does not involve a major change of use. Where the allocated use of a band would change as a result of applying a market approach (e.g. trading, auctions) then the situation is less clear. In some cases, EC Directives could effectively prevent a change of use. Bilateral agreements place a stronger constraint in lower as compared with higher frequency bands. ITU and ECC regulations are a weaker constraint.
So within the current framework of international agreements, the UK has some freedom of action to implement a more market-based approach to spectrum management.
To take advantage of this latitude within international allocations, the range of technically feasible services which can be deployed within specific bands should be widened. This could be achieved through increasing the number of services which are designated co-primary in particular bands, subject to technical studies identifying the extent of service compatibility within and across bands. In the longer term, there may also be scope to widen the definition of these services to encompass a greater range of compatible applications.
This does not preclude harmonization of spectrum use in areas where it can be shown to carry benefits.
A degree of overflow of emissions from one area to another is inevitable. The impact of this ranges from having no effects, through modest inconvenience to individual users to, in the very extreme cases, serious commercial or safety consequences. National regulatory authorities throughout the world have, therefore, regarded it as one of their central duties to ensure both an acceptable interference environment as well as maximising the use of the spectrum. Just as there is an optimal level of environmental pollution, so (conceptually) there is an optimal level of interference which is not necessarily zero.
To pursue the objective of achieving a market-led approach to spectrum management, more freedom and flexibility over the useof licensed spectrum should be devolved to operators. But increased rights over spectrum use would need to be balanced by greater responsibility on the part of operators to participate actively in interference management. This would entail shifting the balance of responsibility for interference management, from the regulators further towards industry. Thus, decisions would be taken at the appropriate level -by those operators who are directly affected. Ways should be found of transferring more responsibility to operators for interference management, in support of wider moves towards using market mechanisms for spectrum management. There would be a continuing need, though, for the central regulator to monitor interference and take enforcement action against breaches of licence terms and illegal spectrum use.
The review considers that a key first step in this process would be the creation of a public on-line database of spectrum assignments. This frequency register should contain a core set of technical and location-based information which would form the basis for operators to carry out the necessary interference co-ordinations associated with any proposed change of use and/or trade within a given band. The RA should also, in conjunction with industry, agree a common understanding of the technical criteria for calculating interference levels.
Market mechanisms for managing spectrum
Creating incentives and opportunities for users to make the most economicallyproductive use of radio spectrum is the primary focus of these recommendations: all spectrum users to the opportunity cost of the spectrum which they occupy. Market-based spectrum
management tools, in conjunction with greater flexibility for spectrum users, are the primary means to this end.
In the vast majority of cases, there are realistic alternatives to the current use of particular frequencies. These alternatives may involve the provision of the same service using more spectrum-efficient technology, or the delivery of a different service using the same or different technology. The transition to alternative uses may only be viable over an extended period, and may involve regulatory action to enable such change. Nevertheless, the existence of these alternatives provides the basis for deriving an opportunity cost of spectrum, based on the full value in the best alternative use to which they could be put.
All classes of users should therefore face financial incentives to economise on the spectrum they occupy. For the majority of frequency bands, where demand exceeds supply, this will entail paying a positive price to obtain access to spectrum. Where trading has been implemented, users will face the opportunity of a positive financial gain from selling access to occupied spectrum.
For some spectrum uses, though, the opportunity cost will be zero. This will occur where use of a particular band in the UK has been exclusively defined through international agreements and incumbents have no scope to change their spectrum use. It will also occur in licence-exempt spectrum where interference is so localised that different spectrum users impose no material constraints on each other’s transmissions.
For other commodity inputs, current market prices generally reflect opportunity costs, because households and firms have the best knowledge of their own costs and preferences and a strong incentive to respond to market signals and put resources to their best possible use. A fully-fledged market in spectrum, should be created through the use of auctions to make primary assignments of spectrum and the introduction of secondary trading. Where this is not feasible, eitherbecause spectrum is reserved for delivery of public services or because the frequency assignments are not suitable for trading, then incentive prices should be set administratively, based upon technical studies to estimate the opportunity cost of spectrum.
The introduction of market-based spectrum management tools is designed tohelp guide spectrum to those who value it highly. But for the UK to benefit from the incentives to innovation and efficiency which auctions, trading and pricing of spectrum are designed to bring, spectrum users need some latitude to respond to market signals. The international allocation process imposes some constraints, as discussed above. But the national regime for assigning licences very often imposes additional constraints which further limit flexibility. These restrictions are typically imposed to achieve other policy objectives which fall outside the remit of spectrum management. They also provide a partial substitute for market-based incentives towards spectrum efficiency.
The review therefore recommends, as a general approach, the reduction of restrictions on spectrum use to the greatest extent possible. This stance should be consistent with the UK’s international harmonisation and co-ordination obligations, and with the maintenance of an effective interference management framework. As market mechanisms are developed further, this should allow the RA to remove licence restrictions (such as requirements for service rollout) which had been designed to mimic the incentives operating in more competitive markets.
Spectrum trading is the most significant step towards a market-based spectrum management regime. It offers great potential benefits to spectrum users, enabling them to enter the wireless market and develop a service by purchasing access to the spectrum they need, when they need it. This in turn should bring benefits to consumers from innovation, greater choice and competition. It should also ease Ofcom’s task, by devolving many complex commercial judgments to the market to resolve, and opening up telecommunications and broadcasting networks to greater competition.
The review strongly advocates the earliest and widest application of spectrum trading possible. Once the necessary liberalising European legislation has been implemented in the UK, Ofcom should move purposefully and progressively towards converting those licences currently used for fully commercial purposes to tradable form.
For trading to bring consumer benefits, then firms must have some freedoms to combine spectrum with other inputsin innovative ways. Ofcom will therefore need to move further than the RA has in defining a generic set of rights and responsibilities for the holder of a spectrum licence. Boundaries of licences will, as ever, need to be carefully defined to help manage interference. But within such boundaries, and subject to any international harmonisation constraints, licensees should be as free as possible to determine the wireless service they provide and the technology they choose to deploy.
Trading should be introduced in a way which minimises transactions costs, consistent with maintaining the integrity of the spectrum management regime. This will entail giving licensees the freedom to divide and partition their licences by frequency and geography for subsequent sale. In these cases, rights and regulatory responsibilities for interference management would be sold together. Spectrum users should also be able to lease access to frequencies to others. In these cases, the original licensee would share access to frequencies while retaining responsibility to the regulator for the conduct of the licence.
As with other markets, trading of spectrum could potentially enable one or more operators to gain and abuse dominance in the spectrum market or in a ‘downstream’ market, which uses spectrum as an input. Government needs to be vigilant against such an outcome, but should deploy the same competition policy tools in spectrum trading as it does for other input markets.
This should be subject to the general competition regime, relying on an ex post analysis of the impact of spectrum trading on competition in defined markets. Where spectrum is an input into a market which is subject to sector-specific regulation, then the objectives of this regulatory regime may be furthered by a more interventionist approach towards spectrum trading, such as ex ante approval of specific trades. These arrangements should be consistent with the UK’s obligations, under the EU Framework Directive, to ensure that competition is not distorted as a result of spectrum trading.
Auctions were first used for the assignment of spectrum licences in the UKin 2000 with the high profile sale for £22.5bn of five licences to use spectrum for Third Generation mobile telephones. They have been used extensively during the 1990s in a number of other countries (notably the USA), for the competitive assignment of commercial wireless licences. Auctions have also been deployed in the UK and elsewhere for the assignment of other scarce resources rationed by regulation (such as commercial broadcasting franchises and mineral extraction rights).
Auctions are well suited to assign spectrum licences to competing users. This should become the default means of assigning licences to exclusive frequency bands. The specific design of individual auctions should be decided on a case by case basis, taking account of competition, marketing and technical analysis.
The RA has been in the vanguard of national regulators in applying pricing to the use of spectrum, with the aim of incentivising more efficient use overtime. This is a valuable complement to the direct market incentives for those licences which have not been assigned by auction, because they have either been reserved for the delivery of public services or assigned under the traditional ‘first come, first served’ basis. Given that there will continue to be large swathes of spectrum reserved for public services, spectrum pricing will need to be maintained for the foreseeable future. As spectrum trading develops over the coming decade, Ofcom should also incorporate price information from marginal transactions in competitive markets into its own administrative pricing procedures.
Administratively set spectrum prices are currently based upon technical assessments of the least cost practicable options for enhancing spectrum efficiency. Prices also vary according to factors such as bandwidth, coverage, degree of sharing, and geographical location. The review agrees with the fundamentals of this approach to deriving spectrum prices.
But current price levels are too low in areas of high spectrum demand to create the incentives towards efficiency. When the Government originally proposed the introduction of spectrum pricing in 1996, it decided that only half the amount of the increases suggested by the preparatory technical study should be implemented. Spectrum prices have now plateaued at this 50 per cent level. The review recommends that, following a re-evaluation of the technical parameters incorporated in the pricing model, the RA should move to full implementation of the prices thus derived. Abstracting from any changes in technology and costs since the original pricing study was undertaken, this move would lead to a near doubling of prices in the sectors and areas of high demand which are currently subject to spectrum pricing.
Commercial spectrum use
Market-based mechanisms should be applied to a range of commercial spectrum uses. Spectrum is not homogenous, and the propagation of signals varies considerably across the frequency range. This has direct implications for the interference management regime, including the degrees of freedom which can be granted to spectrum users. The markets for wireless services are also widely differing, with consequences for the competition regime applying to the auction and trading of spectrum licences. Some specific recommendations are given below: Although broadcasting use of spectrum is increasingly commercial, it is considered separately, given the extensive regulation of free-to-air and publicservice broadcasting which affects spectrum management.
The auction in early 2000 of five licences for Third Generation mobile telecommunications services has set the framework for the future of mobile telephony in the UK. In the meantime, four operators continue to exploit the spectrum licences for Second Generation mobile services, which had been assigned through comparative selection and are which are now subject to spectrum pricing. The majority of spectrum for public mobile telephony is subject to European Union or CEPT decisions harmonising the use of particular frequencies to specific technologies.
Spectrum trading should be introduced as soon as practicable for all public mobile telecoms spectrum. Trading could bring significant benefits, in allowing operators to tailor their spectrum licences to their own needs and enabling new entrants to obtain spectrum for innovative services using compatible technology. Change of use within these bands would initially be constrained somewhat by the European harmonisation decisions. The UK should push for the early retirement of the GSM harmonisation directive, which has now served its purpose, to enable operators more flexibility in the range of technologies and services they deploy. Other harmonisation decisions, which have outlived their usefulness (such as the ERMES public paging directive), should also be retired.
Private mobile radio
Private mobile radio is a complex licence sector, with over 55,000 licensees across the UK including a large number of emergency service and other public safety operators. The RA currently issues a wide variety of licences tailored to the spectrum use defined under the licence. Frequency planning is
primarily managed by the RA, to enable a large number of localised users to share a single national channel. Demand for spectrum in this sector is rationed by the RA, through the application of spectrum prices and the careful assignment of licences. For some bands, spectrum management is devolved to organisations catering for the radio needs of defined groups (such as the utilities).
This central planning approach in a changing market environment inevitably gives rise to inefficiencies and rigidities, with the result that spectrum may be trapped in inefficient uses, exacerbating perceived shortages. There should be a much greater role for the market in future in helping manage access to spectrum currently reserved for private mobile radio.
As a first step towards allowing the market to determine spectrum use, the review recommends that restrictions in licences unrelated to interference management are removed and licences be converted to tradable form. With the publication of a frequency assignment database, these steps should enable a market to develop in spectrum currently allocated to private mobile radio.
In the longer term, a significant amount of the RA’s current frequency planning role should be devolved to commercial spectrum management organisations. Evidence of increased intensity and flexibility of spectrum use in bands managed by such organisations suggests that there could be significant economic gains from extending this approach. The review recommends that Ofcom assign via auction a number of competing national band managers for a range of private mobile radio bands, in parallel with Ofcom’s continued management of the rest of the private mobile radio spectrum. Incumbent licensees within such bands would retain their existing rights to spectrum use, and would become lessees of the commercial band manager. If successful in enabling innovative and intensive use of congested frequencies, this approach could ultimately be extended across the majority of private mobile radio spectrum.
Licence-exempt spectrum use
Many users of spectrum are exempt from individual licensing. In the case of user terminals (such as mobile telephones or televisions), this is because spectrum use is controlled by the licence granted to the system operator. The other broad category of licence-exempt spectrum uses are those where the propagation of radio signals (defined by the permitted power levels and technology standards for the band) is so localised that they do not materially interfere with other spectrum users. In other words, the costs of regulating vialicensing outweigh the potential benefits. These uses are typically confined to bands which are dedicated to licence-exempt use, often those which had originally been designated for industrial, scientific and medical uses.
Use of licence-exempt spectrum is on a ‘non-interference, non-protected basis’. Users of such deregulated spectrum must not cause interference to other authorised spectrum users, nor can they claim protection from interference from such services. With short-range propagation and few devices in any given location, the risk of interference caused by such low power licence-exempt spectrum use has historically been relatively low. At the same time, the absence of regulations covering receivers’ standards has meant that in some cases equipment can be very vulnerable to interference from other services. Technology now offers the prospect of increasing the intensity of spectrum use in these unregulated bands through the use of systems which are automatically self-protecting and ‘polite’. These avoid interference coming into the band and minimise transmitting over other signals within the band.
Licence-exempt spectrum provides an alternative paradigm to regulating for economically efficient spectrum use. Instead of minimising harmful interference through exclusive access to spectrum, the regulator enables multiple re-use of the same spectrum space by limiting the geographical coverage of transmissions. This provides significant flexibility for users, which in turn creates demand for innovative applications of radio technology within these bands. Technological developments are increasingly enabling more valuable broadband applications to be delivered across licence-exempt spectrum.
Public services consume significant swathes of valuable spectrum for the delivery of primarily non-marketed outputs. For example, terrestrial TV broadcasting occupies 40 per cent of the spectrum below 1 GHz, while defence users are allocated nearly 50 per cent of bands in the range 3-10 GHz as well as extensive frequencies elsewhere. It is vital for the productivity of the economy as a whole that such public services face strong and enduring incentives to economise on the spectrum needed to deliver their public service outputs. Without such incentives, there is a growing risk that spectrum hoarding by the public sector will constrain the growth of private enterprise.
There will remain a number of public services for which spectrum is a vital input and for which, in the absence of a fully fledged spectrum market, the current regime of reserving sufficient frequency bands for the delivery of these services should continue through the medium term. In the longer term, as spectrum trading develops, the Government should look to expose more public services’ spectrum use to this market mechanism. In the interim, therefore, the primary means of encouraging spectrum efficiency should be administratively set spectrum pricing, based on the opportunity cost of spectrum occupied. The review recommends that all public services should be subject to this regime, which should provide durable incentives where necessary to economise on spectrum consumption. As noted above, though, for some spectrum, the opportunity cost may be zero as a result of international agreement on the use of the band which gives the UK negligible scope to make alternative use of the spectrum.
As before the application of this approach is described in relation in a number of selected areas.
The Ministry of Defence occupies a privileged position as the largest single user of radio spectrum in the UK. It has de facto management rights over its bands, Historically, it has released a number of valuable bands for civil use (for example, the spectrum used for First Generation mobile telephony). Since 1999 MOD has faced spectrum prices for those bands which it manages where the comparable civil users are charged. It currently pays some £23m per year for the majority of its mobile radio and fixed links bands.
The application of financial incentives to MOD’s spectrum is starting to affect decision-making about rationalising defence needs and release of spectrum for civil use. But major improvementsin spectrum utilisation will only be realised through consistent impact of spectrum pricing on long term decisions about equipment design, procurement and deployment. It should also be recognised that with military requirements for real-time information in ‘battlespace’ and training situations rising, then the MOD’s internal demand for spectrum, even when priced, will often contend with commercial pressures on the spectrum.
Financial incentives on MOD’s spectrum use should be strengthened and widened. Following the recommended revalidation of the opportunity cost calculation of spectrum prices, those mobile and fixed links bands which are currently subject to pricing should be charged at the full opportunity cost level. In addition, MOD’s use of spectrum for ground-based radar in the UK should also be subject to spectrum pricing, in line with the review’s recommendations for the civil aeronautical and maritime sectors.
This could see MOD’s annual spectrum charge rise to over £lOOm. Although this would still be less than one half of one per cent of MOD’s total annual budget, the proposed additional charge on MOD’s programme expenditure should be taken into account public spending reviews.
Where bands are retained for military use but are not fully utilised, MODshould also face positive financial incentives to sharing access to their spectrum with commercial users; it should retain revenues from leasing access to spectrum.
Broadcasters, like other major users of spectrum, must use spectrum efficiently, and there should be effective mechanisms to ensure this. Regulation will continue to play a major role in planning the terrestrial transmission of broadcasting services, given the continuing policy interest in the delivery of public service broadcasting objectives (concerning positive content obligations, free-to-air services, and universal service coverage). In the absence of spectrum pricing across the broadcasting sector, major decisions affecting economically significant spectrum would not properly reflect the opportunity cost of the spectrum asset denied to other users. This is particularly relevant in the approach to digital switchover, which presents a strategic opportunity to improve significantly the spectrum efficiency of broadcasting, and release valuable resources to the rest of the economy.
The Government’s key strategic broadcasting goal is that public service broadcasts should be available to everyone, as now, free at the point of consumption. As alternative delivery platforms (cable and satellite) become more popular, the need for reserved and restricted spectrum for one particular platform (terrestrial transmission) becomes less of a fundamental input for the delivery of public service or commercial broadcasting. Conversely, the convergence of communications services and technologies increases the demand for spectrum which can be used flexibly to deliver a range of wireless broadcast, voice and data services in fixed and mobile environments. So restrictions imposed on spectrum for broadcasting policy reasons become less necessary just as they become more costly in terms of opportunities foregone.
In relation to terrestrial TV transmissions, the current particular circumstances of broadcasting, including the substantial payments already made under the Broadcasting Acts by commercial broadcasters, the level of public service obligations undertaken by the broadcasters, and the forthcoming switchover to digital broadcasting. The review’s recommendations are designed to take account of the various regulatory agreements between Government, broadcasting regulators and individual broadcasters. They also recognise the particular circumstances of public sector broadcasters (the BBC and Channel 4), and those of private sector andinvestor-owned broadcasters. They are also aimed at supporting the Government’s objective of achieving digital switchover in the coming decade.
Spectrum pricing should be applied over the coming decade to all spectrum which is used for broadcasting. The level of prices would be determined by the RA/Ofcom using the methodology outlined by the RA’s original spectrum pricing study11 and would be based on the opportunity cost of spectrum use. Broadcasters should have greater flexibility over the type of transmissions made over spectrum licensed to them, and greater scope to lease spectrum to other users where it is not fully utilised for broadcasting services.
The review considers that commercial independent analogue TV licensees have already paid for their analogue spectrum via their initial bids and ongoing franchise fees for Broadcasting Act licences which allow them to use terrestrial TV spectrum. When these licences are renewed, Ofcom should levy a separate administratively set price for the broadcasters’ analogue TV spectrum. This would be separate from the mechanism used by Ofcom to assign and charge for broadcasting rights-although the existence of a spectrum charge would clearly influence the value of those rights.
Aeronautical and maritime
The Civil Aviation Authority (CAA) and the Maritime and Coastguard Agency (MCA) make extensive use of spectrum reserved for radiolocation, navigation and communications for vessels within UK territory. Marine and aeronautical radars, for example, occupy some 30 per cent of the spectrum in the range 1-3 GHz. Given the global mobility of on-board communications and radar equipment, much spectrum use and associated technology standards in these sectors are subject to extensive and detailed international harmonisation.
The application of administratively set spectrum prices would assist in delivering the best utilisation of spectrum reserved for aeronautical and maritime uses. (Auctions and secondary trading are unlikely to be feasible in these sectors).
In particular, pricing should apply to the use of spectrum by UK ground-based radars, where UK operators subject to pricing have some discretion, over time, to optimise their portfolio of radars and other location devices, in light of the cost of equipment and spectrum. For spectrum reserved for on-board navigation and communications systems, the review considers that the opportunity cost to individual users is, in most cases, effectively zero, since use of this spectrum is mandated internationally, and users are required to adopt specific technologies.
It has been argued above that the time is right for a comprehensive review of spectrum management, with the aim of introducing market principles. Only this will enable allocations and assignments to keep up with the needs of the 21st century. The process will involve further investment in enforcement, as spectrum pricing and trading will impose genuine pressures to make full use of available spectrum.
It would be possible to move immediately to a full market approach, based on trading across all frequencies. What has been proposed above, however, is a dual track approach, under which trading is introduced first into commercial use of spectrum, while public service spectrum continues to be subject to regulatory allocation, but becomes subject more widely to administrative pricings. In time, however, it may become possible to extend the market mechanism to all spectrum use.