This report has been prepared for the Radiocommunications Agency (RA) by The Smith Group and National Economic Research Associates (NERA). It details the results of a study to examine the issues surrounding the application of economic value-based pricing to a number of sectors and licence classes under the provisions of the Wireless Telegraphy Act 1998. The sectors and licence classes considered are:

Amateur and Citizens’ Band (CB);
Aeronautical;
Maritime;
Fixed Satellite Services;
Fixed point-to-multipoint services;
Fixed Wireless Access (FWA);
Test and Development licences (T&D);
Programme Making and Special Events (PMSE).

The report presents the results of the study regarding the feasibility of introducing economic value-based pricing in each of the areas, and where appropriate calculates marginal values on which economic value-based prices should be based.

The objective of the Wireless Telegraphy Act 1998 is to base charges for wireless telegraphy licences on spectrum management considerations. That is to say, fees to install or use radio equipment should reflect more closely the value of spectrum instead of being tied to the fully allocated costs of spectrum management, as was previously the case. In previous reports, the new fees have been described as administrative prices. This is perhaps a little confusing, since in practice cost recovery based prices are also administratively set. To distinguish the two approaches more clearly, this report uses the term economic value-based pricing to describe fees set on spectrum management considerations.

Economic value-based prices are set at a level so as to influence the choices made by spectrum users, such that their use of, and requests for spectrum reflects the value they place on it. Under such a regime, users (and potential users) have incentives to release unused or under-used spectrum, to consider alternative services or less congested frequency bands, and to implement more spectrally-efficient technologies.

The primary objective of economic value-based pricing is to improve the efficiency with which a scarce resource (the radio spectrum) is managed, not to raise money. The general principle we adopt is that such a move is only likely to result in benefits if it induces users to change their spectrum use. Although such changes are likely to be slow to occur, because of the significant investment in existing technologies and services, this is not an argument against the application of economic value-based pricing. The application of the existing administrative rules to clear spectrum or to introduce new more efficient technologies also takes many years to implement for exactly the same reasons.

Under the Wireless Telegraphy Act 1998, spectrum licence fees are to be based on spectrum management considerations for all sectors. The extent to which it is appropriate to apply economic value-based pricing may, however, vary due to a range of constraints which limit users’ ability to change their spectrum use in response to the new charges. The principal constraint is found to be where assignments are co-ordinated for international use, such that alternative uses of spectrum are limited or where changes to equipment used needs to be agreed internationally.

Table 1 summarises the suitability of the candidate sectors considered in this report for economic value-based pricing.

The prices recommended in this report have been calculated with reference to mobile and fixed link Spectrum Tariff Units (STUs). Following consultation on the first round of spectrum pricing, the RA has adopted the STU as a reference measure against which spectrum charges for different services are established. The use of the STU allows easier comparison of spectrum prices across different services and provides for greater transparency in licensing procedure and process.

Two widely accepted STU values have been established for mobile services and for fixed links, and have been applied to fees for services covered under the first two rounds of spectrum pricing. Since these services include the major and highest value spectrum uses, it is likely that they will represent the most likely alternative use and users for the services considered in this report. We therefore consider it valid to use these STUs as the basis for calculating target prices for these services.

The mobile and fixed link STUs are based on the estimated value of 1 MHz of spectrum over 1km² across the UK. They were originally set at 50% of the calculated marginal value of the spectrum concerned. There are strong grounds for this cautious approach. In particular, while the marginal values on which they are based give a reasonable estimate of the prices that should be applied to ration spectrum in a static world, in reality the world is far from static and may give rise to reasons for both higher or lower values over time. As demand grows, marginal values might be expected to increase, although technological developments could help to reduce the costs of alternatives. Also, as users respond to prices, more spectrum should become available, putting downwards pressure on prices. The RA will therefore need to monitor subsequent market developments and modify prices accordingly.

Terrestrial Flight Telephone System (TFTS)

CSR (UK) is currently priced at a flat fee of £180 pounds per channel per site. The main alternative use for the CSR (UK) band is PBR. We therefore recommend that spectrum fees for CSR (UK) should be based on PBR fees, which are already set subject to economic value-based pricing.

Based on 1999 PBR fees, this implies the following pricing regime for CSR (UK):

Minimum configuration CSR (UK); for each 12.5 kHz channel per site:

Maximum configuration CSR (UK); for each 2x12.5kHz channel per site:

In order to provide an incentive for users to keep the transmitter power to the minimum required, a sliding scale of charges could be adopted between the maximum and minimum configurations.

We have calculated economic value-based prices for satellite Earth station uplinks by reference to the STU for fixed links, with which much of the spectrum is shared. Current licence fees and calculated prices are given in the table below.

There are grounds for price increases for classes II and III and price reductions for class I earth stations. In addition, there is a case for introducing different prices for congested and non-congested areas in a manner similar to that adopted for fixed links (since fixed links are the main alternative use). We suggest that the RA consult with industry about how best to introduce differential fees for different area types.

We have calculated an economic value-based price for scanning telemetry by reference to the STU for mobile spectrum, since there is already a high degree of sharing with PBR and public networks. The calculated price is £433 per base station per 2x12.5 kHz channel, compared with £40 per station (base station and outstations) at present.

Since the main alternative use of the spectrum is PBR, an obvious approach is to align prices with wide area PBR, with congestion defined in the same manner. This implies the following structure of fees:

However a possible problem with this approach is that it may reduce the fees paid for what is already a congested part of the spectrum, with a risk that the congestion is exacerbated. The current fee of £40 per station is not directly comparable with the above, but based on a mobile STU of £1.65/MHz/km², the 2 MHz of national spectrum allocated to scanning telemetry is worth £800,000. The 11,000 scanning telemetry stations currently in operation therefore generate fees of about £440,000 which indicates that the fees are already close (~55%) to the reference value of the spectrum.

Nevertheless, there may be some merit in adopting a pricing structure more closely matched to PBR, in particular a structure based more closely on bandwidth and area sterilised. An alternative simpler approach to the above might also be considered, in which a flat fee of £433 per base station is charged, irrespective of the number of outstations.

It may be noted that, as the operators already pay a significant fraction of the reference value of the spectrum and still suffer congestion, there might be a case for allocating more spectrum to them to reduce the amount of congestion.

As the operators form a well-coordinated community who co-operate extensively already, any change in the pricing structure should be made in consultation with the operators. Since the band is going to be re-planned in 2004 this might provide an opportunity to implement any changes to the current regime, giving the RA enough time to consult with the industry and to establish a pricing structure based more closely on the area sterilised and bandwidth used.

We have calculated prices for FWA by reference to the fixed link STU. Current licence fees and calculated prices are shown in the table below.

Table 5: Current and reference prices for FWA licences

In order to derive a price for regional licences the area covered needs to be considered as a proportion of the national value. It would also be appropriate to apply a weighting factor to take into account the population within the area covered by the licence.

A move towards economic value-based prices would imply a significant increase over current fee levels. If economic value-based pricing is to be introduced, then as before we suggest that increases are phased in over a three year period.

Prices for the 2 GHz band should be reviewed following the outcome of the UMTS auction, which will establish a market price for UMTS spectrum. The RA may also wish to consider the use of modifiers to adjust fees where the spectrum is shared with other users.

This introduces a number of administrative complications to implementing a revised pricing regime. We suggest the RA consults further on this issue following publication of its study on the demand for PMSE spectrum. Consideration should be given to reducing the number of licence classes issued, and the impact of revised fees on different user types.

Amateur and Citizens’ Band (CB);
Aeronautical;
Maritime;
Fixed Satellite Services;
Fixed point-to-multipoint services;
Fixed Wireless Access (FWA);
Test and Development licences (T&D);
Programme Making and Special Events (PMSE)

This report presents the results of the study regarding the feasibility of introducing economic value-based pricing in each of the areas, and where appropriate calculates marginal values on which economic vale based prices should be based.

The information in this report is drawn from available literature, interviews with relevant RA staff and, where appropriate, interviews with personnel from other external organisations.

Under the Wireless Telegraphy Act 1998, spectrum licence fees are to be based on spectrum management considerations for all sectors. The extent to which it is appropriate to apply economic value-based pricing may, however, vary due to a range of constraints which limit users’ ability to change their spectrum use in response to the new charges. The principal constraint is found to be where assignments are co-ordinated for international use, such that alternative uses of spectrum are limited or where changes to equipment used needs to be agreed internationally.

Table 2-1 summarises the suitability of the candidate sectors or licence classes for economic value-based pricing.

using more spectrally-efficient equipment;
adopting different, possibly non-radio based, alternatives;
moving to different (non-congested) bands;
sharing capacity with other users/applications?

In answering this question it is necessary to take account of the technical and commercial feasibility of possible alternatives.

Are there other uses? Even if users have no alternative but to stop using the spectrum altogether, it is necessary to consider whether this would release spectrum for another application which could realistically use it, taking account of technical considerations and international co-ordination and allocation requirements. It is also possible that economic value-based pricing might lead users to take actions to press internationally for more efficient use of spectrum.

Are there major policy/political impediments to changing prices? We would suggest that, ideally, economic value-based prices should apply to all users. Exceptions should be reserved for special situations, for example where there are reasons for change (eg the proposed UHF band reversal), and cases where the user genuinely provides social benefits which exceed their private benefits from spectrum use. Discounts might be argued on pioneer's preference grounds, for example, but again the grounds should be tightly specified to fit the economic or social justification for the discount.

Is it feasible to administer spectrum pricing? A final test is the likely practicality of collecting licence fees and enforcing spectrum disciplines. If the introduction of higher spectrum licence fees is likely, for example, to result in an increased number of users attempting to evade payment and/or operating on an unlicensed basis and if the costs of enforcement are likely to be prohibitive, then there may be little merit in applying revised charges.

By testing each licence class against these criteria, it is possible to arrive at a conclusion about its suitability for economic value-based pricing.

2.3 Amateur and CB

This section provides a summary of the current situation in the area of amateur and CB radio. This area is taken to include:

Amateur Service;
Amateur Satellite Service;
Citizens’ Band (CB) Radio.

The amount of spectrum allocated to these services is summarised in the table below.

Table -2: Summary of spectrum allocated to Amateur and CB Services (Source: RA)

Class B Novice Licence: requires completion of a practical training course and a multiple-choice exam. Allows the user to access a restricted number of frequencies above 30 MHz in the VHF, UHF and SHF bands.

Class A Novice Licence: requires completion of a practical training course, a multiple-choice exam and a Morse code test (5 words per minute). It allows the user to access a restricted number of frequencies in the HF, VHF, UHF and SHF band, including some below 30 MHz.

Class B Full Licence: requires a more detailed multiple-choice exam. This allows users access to all amateur frequencies above 30 MHz in the HF, VHF, UHF and SHF bands.

Class A Full Licence: requires a more detailed multiple-choice exam and Morse code test of 12 words per minute. This give the user access to all the frequencies allocated to the amateur services, including those below 30 MHz.

Class A/B Full Licence: technical requirements as Class A, but only requires a 5 words per minute Morse code test.

CB radio licence: this licence authorises any individual to use CB equipment and frequencies. Each licence costs £15. During the Autumn of 1999, the RA is going to introduce a free CB licence for the under 21’s. CB licences are issued on application, with no tests involved.

Table 2-3 summarises the types of licence together with the frequencies allocated to them, their price, and the estimated number of licences in January 1999.

Table -3: Amateur and CB Licences (Source: RA)

At the moment radio amateurs have to pass a Morse code test to obtain access to the HF band. This is an international requirement. However, negotiations are underway internationally to do away with the compulsory need to demonstrate a knowledge of Morse code as a prerequisite to using the HF bands. The RA and RSGB are working towards the introduction of an incentive based licensing structure and will implement new exams once the updated requirements have been agreed. The general structure of the licences will nonetheless remain.

This section considers the possible methods open to users for relieving congestion of the spectrum, and the scope for alternative uses or users of the spectrum taking into account technical factors and any international co-ordination issues that may apply.

What alternatives do users have?

There appears little scope for CB users to modify their spectrum use. Existing equipment is normally designed to operate only on the recognised UK and CEPT CB channels. Therefore, users could only migrate to other bands, assuming spectrum became available, by purchasing new equipment.

There might be some scope for amateurs to move to different bands if suitable, less congested, frequencies could be found. In contrast to CB, enforcement issues are unlikely to be a problem.

Possible alternative uses/users of the spectrum

For spectrum used by Radio Amateurs, there are possibilities that part of the allocation could be used by other groups. As was mentioned above, there are segments of the spectrum currently allocated to this service which could be used by fixed and mobile services. Furthermore, it would seem that there is some precedent for changing the bands available to amateur users, and this has not presented a serious problem in the past.

As well as international allocation agreements, which could affect the possibility of re-assignment of some bands currently used by Radio Amateurs, there are other policy constraints operating in this area. The RA’s priority in allocating spectrum to amateur radio is to encourage its wider use for educational, leisure and amateur research purposes. It is felt that the service represents a valuable source of ‘engineers of the future’ and should not be undervalued as such.

For CB, the cost of policing this service outweighs its revenue. This also means that, were the bands currently used by this service to be re-allocated to another service, the large number of users would make the task of freeing the CB bands very difficult. There is, moreover, little scope for persuading current users to buy more spectrally-efficient equipment, especially whilst such equipment is still relatively expensive and many CB users are unwilling even to buy a licence. Thus, in the case of CB, there is little scope for using spectrum pricing to encourage the use of less spectrum in this area.

There are major practical and policy obstacles to the introduction of economic value-based pricing for either CB or amateur use. While congestion has not been technically verified by the RA, there is significant demand for the small amount of spectrum available from existing users and there is at least a theoretical possibility of alternative uses of the spectrum. In practice, however, unless a practical solution is found to enforce licence conditions and protect legitimate users from those that abuse the service, the introduction of such charging would appear self-defeating.

Spectrum pricing could play a role in relieving congestion in the amateur bands, and parts of the spectrum could be used by fixed and mobile services. There are, however, perceived to be wider social benefits from the encouragement of amateur use suggesting that there is a case for preferential treatment relative to other users. It is the RA’s policy to encourage the use of amateur radio in the UK, including the use of voluntary band plans.

2.4 Aeronautical

This section provides a summary of the current situation in the area of aeronautical radio. This area is taken to include:

Aeronautical mobile communication;
Radio navigation;
Radio location.

The amount of spectrum allocated to these services is summarised in the table below.

Table -4: Summary of spectrum allocated to aeronautical services (Source: RA)

Table 2-5 contains estimates for the user population of each of the various licences listed above. Each licence is valid for a year.

Table -5: Aeronautical Licences (Source: RA)

2.4.3 Planned Changes

Recently, the European Aviation Authorities have agreed to change, above Flight Level 245, the bandwidth of channels used in the frequency band between 118 MHz-137 MHz from 25 kHz to 8.33 kHz, thus effectively tripling the number of available channels. This change is due to be implemented in most European countries, excluding Britain amongst others, on 7 October 1999. The required change in radio equipment in all AGS and aircraft together with necessary training of staff is underway and should be completed on time. The CAA is keen to see these changes introduced in the UK as soon as possible. Furthermore, they would like to introduce 8.33 kHz channels at all flight levels due to the increasing demand.

This section considers whether there is excess demand for the spectrum from actual or potential users of the spectrum (the scope for possible alternative uses of the spectrum is considered in the next section). In this sense, the concept of congestion used is intended to include potential or latent as well as current congestion.

Frequencies allocated to aeronautical mobile (HF and VHF communication) and used for air-to-ground communication and vice versa are heavily used and hence can become congested at times. This is exacerbated by the fact that these frequencies are used world-wide and demand is growing constantly. The spectrum is limited mainly because it cannot be shared easily. This is due to safety issues requiring that channels be protected and can only be shared if the operators are geographically separated and therefore do not interfere with each other's operation.

Due to the fact that two organisations are involved, a lack of information sharing can sometimes lead to inefficiency in the licensing process. One example would be an airfield operator going out of business, which would free the channels allocated to them for use by other users. However, the CAA will not be aware of this change unless the RA inform them because the CAA are not involved in the process of renewing licences. Therefore, they are not automatically aware that they are able to utilise these channels. Notwithstanding this, there have been no significant assignment problems to date. There are no refusals of applications for ground station or navaid licences on these grounds. Furthermore, there is no increase in applications foreseen which would change this situation.

The situation for navaids is less clear than for mobile communications. There is some congestion for particular systems such as the Instrument Landing System (ILS), where there is only a single frequency band (40 channels) available for all systems world-wide, and it is likely that the pressure on other systems such as Mode S radar will increase as the demand for data communications grows. Some primary radar frequencies are shared with other users, particularly the MoD and broadcasting assignments.

What alternatives do users have?

There is some scope for users to adopt more spectrally efficient equipment. The scope for change in the short term is, however, limited due to the need to agree changes internationally (see section 2.4.6).

Possible alternative uses/users of the spectrum

Much of the demand for spectrum in this area comes from other aeronautical users. Alternative uses of the spectrum are limited due to the international nature of the allocations.

The scope for meeting users' demands for spectrum through the use of more spectrally efficient equipment is limited by the need to agree changes to on-board equipment internationally. This is invariably a slow process, with due notice of 7 years required before any changes can be imposed.

Changing any of the internationally allocated frequencies cannot be considered as a feasible option in the spectrum pricing calculation. The emphasis should thus lie on making more efficient use of the allocated spectrum.

There is growing demand for aeronautical spectrum, however, the scope for administrative pricing to modify users' behaviour is limited by the international nature of the allocations and particularly the need to change on-board equipment. Any move towards the use of more spectrally efficient equipment will require international agreement, and is probably best sought through lobbying rather than any unilateral pricing initiative by the UK authorities. It does appear, however, that some efficiency gains might be achieved through better co-ordination between the RA and the CAA.

This section provides a summary of the current situation in the area of maritime radio. This area is taken to include:

Maritime mobile radio;
Radiodetermination.

Table -6: Summary of spectrum allocated to maritime services (Source: RA)

Table -7: Maritime licences (Source: RA)

This section considers whether there is excess demand for the spectrum from actual or potential users (the scope for possible alternative uses is considered in the next section). In this sense, the concept of congestion used is intended to include potential or latent as well as current congestion.

The use of maritime radio frequencies assigned to CSR (International) is very high especially in the Thames Estuary area and along the Solent, as many licensees share the same channel and hence encounter interference. The situation is similar for the licence CSR (Marina) channels. The channels assigned to the licensee are not exclusive and are subject to heavy traffic. Furthermore, interference from port operators on the continent is also experienced. CSR (UK) spectrum also experiences significant congestion.

In general, navigational aids have not suffered from any interference or congestion. The only exception is for DGPS VHF and DGPS MF/Low HF where users have been refused licences due to congestion and possible interference.

What alternatives do users face?

Over the longer term, it is likely that the introduction of newer, more spectrum efficient equipment including narrowband systems, will provide a means for relieving congestion and possibly freeing spectrum for other uses. The scope for achieving this through the unilateral introduction of higher licence fees in the UK is, however, limited by the need to agree technologies and to co-ordinate allocations internationally. A possible exception arises in the case of CSR (UK) since the RA has direct control over the assignment of these frequencies, and they are used exclusively through UK coastal stations. The feasibility of introducing 12.5 kHz or 8.33kHz channel spacing should be considered.

Possible alternative uses/users of the spectrum

The major alternative use for the maritime spectrum is for land mobile purposes (the maritime band sits alongside the land mobile allocation).

A further problem is the lack of enforcement. There are about 120,000 registered vessels in the UK, but only 60,000 issued licences. However, it must be remembered that the use of radio on pleasure vessels is voluntary and users of pleasure vessels who have a radio are not always aware of the need for a licence to operate it.

Furthermore, even though the administration of the ship radio licences is contracted out (£6 of the £22 licence fee is retained by the contractor who deals with vessel call signs and notification of international authorities), the RA still deals with a large number of queries and other non-routine administrative tasks.

Ostensibly, maritime appears to be a prime candidate for economic pricing since there is considerable congestion and high demand. Whether in the long-run the more efficient solution is to move to more spectrally efficient equipment or to allocate more spectrum to the service is a question which would be easier to answer and achieve if the spectrum were properly valued. In practice, these are issues which lie outside the control of the UK since technologies and allocations need to be agreed internationally. Higher licence fees in the UK, might simply lead to users registering their vessels elsewhere and could further add to the current enforcement problems. It therefore appears that there is little benefit to be derived from the UK unilaterally introducing administrative pricing for most of the maritime band.

A possible exception is the case of CSR (UK), which is under direct UK control and for which spectrum pricing might play a role in relieving congestion if a practical solution to the problem of enforcement could be found. Since most of the spectrum sits alongside the land mobile allocation, there appears a strong case for aligning CSR (UK) prices with that of PBR.

This section provides a summary of the current situation in the area of fixed satellite links. The area is taken to include:

Permanent Earth Stations;
Transportable Earth Stations;
Very Small Aperture Terminals (VSATs);
Miscellaneous Earth Stations.

The amount of spectrum allocated to these services is summarised in the table below.

Table -8: Summary of spectrum allocated to fixed satellite services

Table -9: Fixed Satellite Links licences

Another area of growth is services operating in the 20/30 GHz band. The ITU has received notifications for satellite networks using both geostationary and non-geostationary orbits for services to small aperture interactive terminals. ETSI is currently formulating standards for these terminals with the possibility of interactive services being operational at the beginning of 2000. The successful implementation of these systems could lead to the use of hundreds of thousands of terminals throughout Europe and the rest of the world. This would be likely to result in sharing difficulties in the shared bands if such high density FSS terminals were to operate using the same frequencies as high-density fixed services. These sharing issues are the subject of ongoing work in CEPT and the ITU.

This section considers whether there is excess demand for the spectrum from actual or potential users of the spectrum (the scope for possible alternative uses of the spectrum is considered in the next section). In this sense, the concept of congestion used is intended to include potential or latent as well as current congestion.

At an international level the ITU provides co-ordination procedures to identify suitable spectrum for particular satellites in particular countries. The aim of the co-ordination procedure is to ensure that once the frequencies have been identified, an operator does not suffer any interference from other satellite operators. Due to the ever increasing development of FSS notifications there is a high demand for orbital locations and spectrum, particularly in the lower FSS bands. This has resulted in an over filing of applications for orbital resources within the ITU, so-called ‘paper satellites’, and co-ordination difficulties can sometimes mean that operators have to operate at frequencies that are different to those for which they originally applied.

The UK has led the de-regulation of satellite user terminals. Consequently, there is quite a high number of satellite earth stations in this country. There is so far no congestion in the sense that operators have been refused a licence because not enough spectrum was available.

There is, however, a high degree of sharing between FSS and terrestrial fixed links. Co-ordination areas for each Earth station can be quite large (up to a radius of 350km). This area is not sterilised, but requires careful co-ordination with other users, eg terrestrial fixed links, to minimise interference and provide suitable assignments. Interference patterns depend on the location of the satellite and the characteristics and location of the Earth station; for example, larger dishes create less interference due to higher gain, but they can also be more sensitive to interference. Satellite operators can reduce the interference from terrestrial fixed links by providing sufficient shielding for their equipment. This, however, adds to the overall cost of the earth station. Some operators using receive-only equipment (which does not require