Commission Decision of 12 October 2006 the United Kingdom's application of the tax on non-domestic property to telecommunications infrastructure in the United Kingdom (No C 4/2005 (ex NN 57/2004, ex CP 26/2004) (notified under document number C(2006) 4378) (Only the English text is authentic) (Text with EEA relevance) (2006/951/EC)

Commission Decision

of 12 October 2006

the United Kingdom's application of the tax on non-domestic property to telecommunications infrastructure in the United Kingdom (No C 4/2005 (ex NN 57/2004, ex CP 26/2004)

(notified under document number C(2006) 4378)

(Only the English text is authentic)

(Text with EEA relevance)

(2006/951/EC)

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, and in particular Article 88(2) thereof,

Having regard to the Agreement on the European Economic Area, and in particular Article 62(1)(a) thereof,

Having called on interested parties to submit their comments pursuant to the provisions cited above(1) and having regard to their comments,

Whereas:

I. PROCEDURE

(1) On 17 February 2004, the Commission registered a complaint from Vtesse Networks Ltd (‘Vtesse’), a UK telecommunications operator, concerning alleged preferential tax treatment in favour of B T plc (‘BT’), the incumbent telecommunications operator in the United Kingdom.

(2) By letter dated 19 January 2005, the Commission informed the UK authorities that it had decided to initiate the formal investigation procedure laid down in Article 88(2) of the EC Treaty in respect of the measure at issue (‘decision to initiate the procedure’). The Commission decision to initiate the procedure was published in the Official Journal of the European Union on 12 March 2005(2). The Commission called on interested parties to submit their comments on the measure.

(3) Following the extension of the deadline, the UK authorities responded by letter of 4 April 2005 to the request for comments in the decision to initiate the procedure.

(4) The Commission received comments from the following interested parties:

• AboveNet Communications UK Ltd. (‘AboveNet’), received by letter dated 29 April 2005

• Altnet Task Force (‘Altnet’), received by letter dated 2 May 2005

• Broadband Stakeholders Group, received by letter dated 13 May 2005

• BT, received by letter dated 18 May 2005

• Cable and Wireless Group plc. (‘Cable and Wireless’), received by letter dated 29 June 2005

• Communications Management Association, received by letter dated 30 March 2005

• Easynet Group Plc. (‘Easynet’), received by letter dated 3 May 2005

• Gamma Telecom Ltd., received by letter dated 1 December 2005

• Hutchinson Network Services UK Ltd. (‘Geo’), received by letter dated 6 May 2005

• Global Crossing Ltd., received by letter dated 6 May 2005

• GVA Grimley, received by letter dated 27 April 2005

• Kingston Communications Plc, received by letter dated 29 April 2005

• NTL Group Ltd., received by letter dated 29 April 2005

• Telewest Broadband Ltd., received by letter dated 28 April 2005

• Thus Plc., received by letter dated 4 May 2005

• UK Competitive Telecommunications Association (‘UKCTA’), received by letter dated 25 April 2005

• Vanco Ltd., received by letter dated 19 April 2005

• Viatel (UK) Ltd., received by letter dated 6 July 2005

• Vtesse, which provided comments and further information by letters or e-mail dated 28 January 2005, 4 February 2005, 16 February 2005, 22 February 2005, 15 March 2005, 29 March 2005, 28 April 2005, 28 July 2005, 17 August 2005, 2 September 2005, 4 November 2005, 23 November 2005, 30 November 2005, 1 December 2005, 11 January 2006, 19 January 2006, 16 March 2006, 20 March 2006 and 8 July 2006

• Anonymous, received by letter dated 1 February 2005

(5) The Commission forwarded these comments to the UK authorities by letters dated 20 June 2005, 4 August 2005 and 9 August 2005, in order to give them the opportunity to react. They provided their comments by letters dated 10 October 2005 and 22 November 2005.

(6) By letter dated 26 July 2005, the Commission asked OFCOM, the UK independent authority regulating the telecommunications sector, to submit its comments on the decision to open the procedure. OFCOM replied by letter dated 21 September 2005. OFCOM provided further information on 7 December 2005, 6 January 2006 and 21 March 2006.

(7) On 14 July 2005 and 4 October 2005, two meetings were held with representatives of BT, which provided further information on 23 November 2005 and on 23 March 2006.

(8) On 16 March 2005, 9 September 2005 and 4 July 2006, three meetings with the UK authorities took place.

(9) By letter dated 26 September 2005, the Commission requested further information from the UK Authorities, to which they replied by letter dated 18 November 2005. They provided further information on 23 November 2005, 23 January 2006, 21 February 2006, 13 March 2006, 21 March 2006, 27 March 2006, 20 April 2006, 2 May 2006, 4 July 2006 and 20 July 2006.

(10) On 23 November 2005, a meeting took place with Vtesse, the complainant.

II. DETAILED DESCRIPTION OF THE MEASURE

A. Description of the measure

Rating system

(11) Business rates are a UK property tax levied on non-domestic property. Its aim is to contribute towards the costs of services provided by Local Authorities. The primary legislation on non-domestic rating in England and Wales is the Local Government Finance Act 1988 (‘LGFA 1988’). The implementation of the system is further regulated by statutory instruments and case-law.

(12) The property unit that is valued for rating purposes is called the ‘rateable hereditament’ and includes land, buildings and rateable plant and machinery. Liability for the rates falls on the person in occupation of the rateable hereditament, i.e. the person who is in ‘paramount control’ of the assets, defined by case-law as the actual, exclusive and non-transient occupation of the hereditament, which brings commercial benefits to the occupier.

(13) The tax rate, which applies equally to all business properties is determined annually by the Secretary of State and applies equally to all operators. It was set at 45,6 % for the fiscal year 2004/5 in England and 45,2 % in Wales. The base of this tax, the so-called rateable value, is the hypothetical annual rent at which the hereditament might reasonably be expected to be let from year to year in an open market transaction at a certain valuation date. The full statutory definition is contained in Schedule 6 to the LGFA 1988.(3)

(14) The valuation of this hypothetical annual rent is carried out by the Valuation Office Agency (‘VOA’), an executive agency of the Inland Revenue, part of central government. Revaluations are carried out every five years. For the period 1995-2000, the rateable value is the hypothetical annual rent of the property if available on the market on 1 April 1993, which is called the ‘antecedent valuation date’. For the period 2000-2005, the antecedent valuation date was 1 April 1998 and for the on-going period 2005-2010, the date was 1 April 2003. The lapse of time between the antecedent valuation date and the date when the valuation enters into force is meant to allow the VOA to carry out the valuation based on actual evidence at the antecedent valuation date.

Valuation

(15) The VOA has a choice of four methods for calculating the ‘rateable value’, i.e. the hypothetical annual rent, of the hereditament, with a hierarchy between the different methods according to the decision to initiate the procedure:

(a)

If there is direct and actual evidence of the property's rent, the rental method is to be applied.

(b)

If direct rental evidence is not available, a comparison with the rental evidence of other comparable properties can be made. This is the Tone of the list method. For matters of convenience, both the rental method and tone of the list method, which both rely on market rental evidence, will be referred to in this decision as ‘rental method’.

(c)

The Receipt and Expenditure (‘R&E’) method is to be applied for properties that are seldom let or are difficult to replicate, typically utilities. It seeks to replicate the thought process that a putative tenant might go through when assessing the profitability of a commercial venture involving the renting of a property. He would estimate his future receipts and deduct his future costs. From that, he would deduct the return he requires, the rest representing the maximum rent he would be prepared to pay.

(d)

The Contractor's basis method is based on the replacement costs of a business property: a putative tenant would be unwilling to pay more as an annual rent for a hereditament than it would cost him in the way of annual interest on the capital sum necessary to build a similar hereditament. This method is normally used for hereditaments that are not let or do not generate profits (e.g. hospitals).

Rating of telecommunications infrastructure

(16) Business rates are levied on telecommunications infrastructure as on other non-domestic properties. The Valuation for Rating (Plant and Machinery) (England) Regulations 2000(4) lay down that plant and machinery that are rateable for telecommunications infrastructure are ‘cables, fibres, wires, conductors or any system of such items …, used or intended to be used in connection with the transmission of communications signals’.

(17) Telecommunication property has been subject to business rates since 1855. Following the privatisation of BT in 1984, this tax was also levied on the latter's infrastructure, as well as on Mercury Communications, BT's only competitor during the duopoly situation of the 1980's. Initially, for the 1990-1995 period, the tax amount was determined by applying a prescribed ‘standard formula’. When the telecommunications industry was gradually opened to competition, starting in 1992, the tax was levied on all telecommunications infrastructures.

Application of the property tax to BT

(18) For the period 1995-2000, the VOA applied the R&E method to BT. This valuation was adjusted following a negotiated settlement with BT which had constructed its own valuation on the Contractor's basis method and appealed against the original valuation to the Central London Valuation Tribunal and then to the Lands Tribunal. BT and the VOA reached a settlement in 2000. The outcome was an agreement to set the rateable value of BT's network assets at GBP 445 million in England and GBP 25 million in Wales for the fiscal year 1995/1996. With a multiplier of 43,2 %, BT's tax liability amounted to GBP 203 million for that fiscal year, which represented 2 % of BT's relevant annual revenues.

(19) For the 2000-2005 period, the VOA settled BT's asset valuation at GBP 467 million in England and GBP 26 million in Wales. The valuation was carried out on the basis of the R&E method and was based on the principles and values agreed at the same time the 1995 list was settled. For the 2005-2010 period, the VOA applied again the R&E method to BT's network.

(20) In principle, a review of the valuation takes place every five years. However, in its decision to initiate the procedure, the Commission noted that subsequent downward revisions were made to reflect the loss of market share of the company in the UK fixed telecommunications market resulting from the physical roll-out of competing telecommunications networks. The decision to initiate the procedure noted that no similar upward adjustment was apparently made to reflect possible extensions or upgrades of the network.

Application of the property tax to Kingston

(21) Kingston Communications plc (‘Kingston’) is the incumbent telecommunications operator and the owner of the only local access network in the region of Hull. During the privatisation of most publicly-owned telecommunications networks in the United Kingdom in the late 1980s, the company remained the property of the municipality, until the share capital was opened to the public in 1999. Hull City Council remains the only substantial single shareholder. Kingston was the only other telecommunications operator valued under the R&E method.

(22) However, Kingston also owns a subsidiary, Torch Communications Ltd, which constructed a fibre backbone network outside Kingston's local access network area (i.e. Hull conurbation) and operates that backbone network as a separate subsidiary. As a result, the Torch backbone network can be separately valued on the basis of the rental method.

Application of the property tax to Vtesse

(23) Vtesse is the complainant. It is a ‘fibre to the business’ service provider offering high capacity retail leased lines, mainly supplied to large corporate users. It directly competes with BT's leased line offers. In order to connect its customers, Vtesse leases ‘dark’ fibre from other backbone operators and complements this with its own built infrastructure. The property rented from other parties is included in the rateable hereditament of Vtesse. The VOA has decided to rate Vtesse based on the ‘tone of the list’ method, which is then applied each time new fibre is rented and lit.

(24) Based on benchmarks of actual rents paid, the VOA has set the annual rate under the rental method at GBP 1 200 per km of pair of optic fibre in the London metropolitan area and GBP 1 000 per km in the other parts of the United Kingdom for the 2000-2005 period. In addition, in 2001, a so called ‘oversupply allowance’ was granted by the VOA to fibre providers assessed under the rental method, equalling a 15 % discount on the rateable value effective as of 1 April 2001 and 25 % as of 1 April 2002. This allowance was meant to reflect the overinvestment in fibres during the telecommunications boom of the late 1990s and the subsequent drastic reduction in value of these assets due to physical oversupply. There is a further discount of 10 % for networks over 3 000 km. Following these discounts, the rental value of a fibre pair fell to GBP 900 per km in the London area and GBP 750 per km in the other parts of the country in the years 2002-2004. For the 2005-2010 period, the rental value of a fibre pair was set at GBP 600 per km in the London area and GBP 500 per km in the rest of the United Kingdom.(5) According to Vtesse, its tax liability was around 7 % of its recurring revenues in 2003/04.

(25) As to revisions, the tone of the list rate is applied to each fibre which is rented and lit. Vtesse has the obligation to regularly inform the VOA of extensions to its network so that the valuation of its hereditament can be adjusted. As a consequence, whenever it procures fibre to provide services to its customers, Vtesse claimed that it is liable for rates that may equal up to 20-30 % of the revenues of a new contract.

(26) As to the other telecommunications operators in the market, they are valued under the rental method, with the exception of cable TV operators, which are valued on the basis of a method derived from the contractor's basis method.(6)

B. Reasons for initiating the formal investigation procedure

(27) In its decision to initiate the procedure, the Commission considered that the application of the property tax to BT and Kingston could confer an advantage on these two firms. They seem to benefit from a reduction of their tax base as compared with the other competitors. This was reflected by the fact that BT paid in the range of around 2 % of relevant revenues in tax, while Vtesse and other competitors might have to pay in the range of around 7 %, and up to 20-30 % on an incremental basis.

(28) The decision to initiate the procedure expressed the view that this apparently discriminatory taxation could be the result of the application of a particular asset valuation method to BT and Kingston, while the other telecommunications operators are valued under the rental method, again with the exception of cable TV operators, which are valued on the basis of a method derived from the contractor's basis method. The non-uniformity of the system and the discretionary powers which are given to the VOA to apply the general provisions to specific operators and which were used in the context of the settlement negotiated with BT, may have resulted in a specific advantage to BT and Kingston.

(29) The decision to initiate the procedure questioned the view that the rental method could not be applied to BT's and Kingston's hereditaments. It suggested that the rent received by BT when renting out its own infrastructure could serve as guidance for the value of the network. Moreover, BT's retail business units effectively rent infrastructure and buy network services from BT Wholesale Ltd. (‘BT Wholesale’), the business unit that manages BT's infrastructure: the internal transfer prices between BT's retail business and BT Wholesale could provide a basis for estimating the rental value. Moreover, liberalisation and network element unbundling have resulted in a variety of wholesale products which could serve as benchmarks for valuing the rental value of the various infrastructure elements of BT's and Kingston's networks. As to the argument that there is no network nearly as large or diverse as BT's whose rental value could serve as benchmark, the decision to initiate the procedure suggested that the taxation of cable TV companies, which are also active in the telecommunication market, could serve as a benchmark.

(30) As regards the application of the R&E method itself, the Commission noted that this method estimates the rental value of the assets concerned on the basis of the revenues derived from the use of these assets. The result would therefore depend on the level of profitability with which these assets are used, which could lead to results that penalise operators whose assets are valued under a different method. The Commission questioned whether the necessary adjustments were made when applying the R&E method to BT and Kingston in order to assume that their assets are used at capacity and profitability. The Commission also enquired how the universal service obligations are taken into account in the valuation of BT.

(31) The decision to initiate the procedure further questioned the method used for revisions of the rateable value. It noted that BT seems to enjoy a downward revision mechanism, while there does not seem to be a similar systematic review of the market conditions faced by competitors. The apparent failure to take into account the increase in the value of BT's infrastructure between two five-year review periods while charging competitors incrementally as the value of their network increases appears to confer an advantage on BT.

(32) Finally, the decision to initiate the procedure underlined the fact that the VOA apparently enjoyed broad discretionary powers, which enable it to negotiate settlements. It underlined that the settlement reached with BT should respect the principle of non-discrimination, and should not unduly discriminate among operators.

III. COMMENTS FROM INTERESTED PARTIES

A. Comments from interested parties other than BT and Kingston

(33) The comments submitted by interested parties other than BT and Kingston generally support the view that the application of the property tax to BT and Kingston may give them an advantage.

(34) Vtesse first pointed out that, after the decision taken in March 1998 by the Valuation Tribunal which confirmed that the R&E method had to be applied to BT's network and that its rateable value was GBP 553 000 000 for England and Wales, BT appealed the decision before the Lands Tribunal. Before the appeal was heard, BT and the VOA settled the valuation at GBP 470 000 000, which represents a 15 % reduction compared to the initial valuation. Vtesse suggests that BT and the VOA settled a final amount and that the VOA adjusted the numbers to match the agreed result.

(35) Vtesse put forward new elements that tend to show that the hereditament of BT is under-valued under the R&E method. Thus, supposing that the rateable value of BT's whole network(7) was attributed to its optic fibre network (7 300 000 km in 2005), the annual rateable value per km of fibre would be GBP 74. According to Vtesse, this figure is to be compared to the annual rateable values of GBP 1 000 and GBP 1 200 per km of optic fibre pair for telecommunication operators that are assessed under the rental method. Similarly, setting the value of BT's optic fibre to zero, and given that BT has 29 million final users connections, the annual rent per line would be GBP 18,57. This figure is to be compared with the annual GBP 122 paid by telecommunication operators for access to BT's unbundled local loops(8).

(36) Vtesse also notes that, in 2001, BT transferred the leasehold and freehold of most of its property portfolio to ‘Telereal’, a joint venture between Land Securities and the Williams Pears Group for GBP 2,4 billion. The deal involved 5,5 million sq m in total. Following the deal, BT paid GBP 190 million in annual rents for the freehold plus GBP 90 million for the leasehold in 2001. On the basis of this transaction, Vtesse concludes that BT pays GBP 35 per sq m in rents for its buildings, while, according to Vtesse, other telecommunication operators are rated on average at GBP 115 per sq m on their commercial property. Vtesse asserts that the actual rent paid in 2001 by BT for its buildings to Telereal, i.e. GBP 280 million, represented more than 50 % of its rateable value that year, despite the fact that most of this rateable value should be mainly attributable to its network and not to its buildings. Vtesse seems to imply that this is evidence that BT's network is undervalued under the R&E method.

(37) In Vtesse's submission, the discrepancy between the rateable value of the different components of BT's network and those of the networks of the other telecommunication operators proves that the R&E method has resulted in a lower valuation of BT's hereditament than would have been achieved under the rental method.

(38) Vtesse provides further elements to compare the respective burden of business rates on BT and on other telecommunications operators. Vtesse advocates the rates/receipts ratio, which allegedly amounted to 13,46 % in its case in 2004. It claims that this ratio must be adjusted in order to allow direct comparison with BT. According to Vtesse, since most operators pay significant fees for the carriage and termination of traffic and since those connection fees do not relate to their rateable assets, the fees should be deducted from their receipts when calculating their rates/receipts ratio. This view is supported by Gamma Telecom which, for the same reasons, argues that the most appropriate ratio for comparing the respective burden of business rates should be the rates/added value ratio.

(39) Furthermore, Vtesse compared the prices of BT's Wholesale Extension Services (‘WES’) with the rates that an operator valued under the rental method would have to pay if it were lighting optic fibres for the purpose of providing the same service. Depending on the capacity and distances of the WES, Vtesse claims that such an operator could have to pay rates worth up to 87 % of the price charged by BT. Vtesse considers that this is further evidence of the fact that BT's optic network is undervalued: if its network were valued in the same way, BT should charge more for its WES in order to cover the real costs of the business rates.

(40) Vtesse also makes the claim that the rates/receipt ratio is particularly high for new entrants such as Vtesse. More established firms can load up fibres with as much traffic as possible in order to amortise the rental and tax costs over as many customers as possible. BT is not subject to this constraint because allegedly it is not taxed marginally. That means that, for established firms, the rates/receipts ratio will decrease in the long run.

(41) On the question of marginal taxation, Vtesse emphasises that the disadvantage that it suffers stems from the fact that BT is rated as an indivisible network, while Vtesse is rated on each individual metre of optic fibre cable that it lights. It considers that the unit of analysis should be the individual contract in the market, on which Vtesse and BT compete. Vtesse gives more precise figures on the impact of these differences in marginal taxation of contracts: taking the example of the bid launched by Kent MAN for the provision of backbone and access telecommunication links, Vtesse claims that its rates liability on that bid represented 16 % of its revenue. On the same bid, Vtesse claims that BT was not taxed marginally, or was taxed, at most, at the average of 2 % of revenues, which BT allegedly pays in rates. That difference would explain why Vtesse lost the bid to a consortium involving BT.

(42) Vtesse also notes that BT has recently rented some 2 000 route kilometres of optic fibres on the network of Geo, one of Vtesse's suppliers, probably on the same commercial terms as Vtesse. In Vtesse's view, BT should be considered in occupation of these fibres. Vtesse questions whether this extension of BT's network was taken into account in its valuation, and suggests that BT should have been assessed on similar terms as Vtesse when it rents fibres from the same operator.

(43) The Altnet Task Force, which represents a number of alternative fixed-line telecommunication operators, submits that the R&E method should only be applied to regulated monopolies with little comparable rental evidence. It considers that the VOA had ample evidence of rental value to evaluate BT on a rental method, such as BT's wholesale line rental tariffs or its published prices for Local Loop Unbundling products which could be used to evaluate the rental value of its copper cables. It also considers that the rental method has an incremental effect on the operators' tax liabilities. No such effect occurs under the R&E method, since the value of the infrastructure is established in advance and therefore BT's decision to light dark fibres is fiscally neutral. According to Altnet, the effects of these distortions were made worse by the fact that the settlement reached by BT and the VOA was perpetuated for two valuation periods, which gave a substantial advantage to the incumbent, in terms of increased certainty, while Altnet firms are facing costly and outstanding appeals and negotiations. Altnet claims that BT faces 74 % of the total tax burden, in relation to telecommunications networks, while its market share is at least 80 %.

(44) Global Crossing confirms that incremental taxation (i.e. on newly lit fibres) favours BT since, when Global Crossing illuminates a new fibre, the incremental taxation liability is incurred on the entire length of the circuit from its customer's site to its core network nodes. Global Crossing claims that it has lost business to BT because of the unfavourable taxation treatment.

(45) Thus plc and Viatel (UK) Ltd. also claim that they face incremental tax liability as a result of connecting new customers' sites to their networks, while BT, their main competitor, does not incur such an incremental tax liability. Above Net makes the same claim, indicating that it lost business to BT, possibly because of BT's alleged lower taxation

(46) UKCTA, a trade association representing the interests of fixed-line communications operators competing against BT for voice and data services, expressed the view that the property taxation system is severely lacking in transparency, since it is characterised by confidential bilateral deals agreed between network operators and the tax authorities. UKCTA considers that the settlement reached by BT has provided it with ten years of certainty and stability, while other operators did not benefit from a similar level of predictability. It notes that the property tax system has entrenched BT's dominant position, since taxation has been more affordable to BT and has therefore impaired new entrants' ability to compete fairly, and since BT, unlike other competitors, does not face direct increases in business rates liability for lighting more fibres. However, following a request from the Commission, UKCTA indicated that it was impossible to indicate what Altnet operators might have had to pay in rates, had they been assessed on the same basis as BT.

(47) GVA Grimley, consultants in the field of business rates for telecommunications companies, criticise the application to BT of an adjustment mechanism which reflects change in its market shares and whose application is opaque. GVA Grimley claims that for all other hereditaments valued under the R&E method, normal practice is for there to be a five yearly revaluation without a market share mechanism. Furthermore, they question the way this market share adjustment mechanism is applied to BT, noting that it is difficult to see why BT's rateable value should have been declining from 1995 to 1999, in a context of improving turnover and profit. They further claimed that the VOA declared that a pair of unbundled copper wires had a rateable value of circa GBP 50, while BT's local loops are valued at around GBP 16 (dividing BT's total rateable value by the number of local loops). The discrepancy would indicate that BT's network is undervalued.

(48) Some of the respondents already mentioned and others, such as Telewest Broadband, Cable and Wireless, or Easynet, raise other issues that concern business rates, but do not appear to be directly related to the facts described in the decision to initiate the procedure. For instance, some of those firms are concerned about the possible application of the property tax treatment in the context of local loop unbundling. They also complain about the fact that the business rates are set at too high a level(9), which represents an impediment to investment, and that a period of five years is incompatible with the structure and pace of technological change in the telecommunications industry(10). One firm also claims that some competitors are not rated(11).

B. Comments from BT and Kingston

Comments from BT

(49) BT first explained that, contrary to what is claimed in the opening decision, there is no accepted hierarchy between the valuation methods, and no principle that the rental method, if rental evidence is available, is the most appropriate method. The method which the VOA uses is dictated by the nature of the property being assessed and by available evidence. Thus, there is not discretion for the VOA as to the choice of method.

(50) Although BT is valued under a different method, there is no discriminatory treatment between BT and other ratepayers: the statutory test is the same, and all methods aim at assessing the hypothetical rent at which it is estimated that the business property in question might reasonably be expected to be let from year to year.

(51) BT claims that there is no evidence that the rateable value assessed for BT's hereditament has given BT an advantage in the form of a reduced tax base compared to competitors. The only evidence the Commission has provided to support this allegation is a comparison of rateable values with respect to turnovers.

(52) BT further submits that the VOA correctly concluded that the R&E method was the only appropriate method to value BT's network, mainly because of the lack of rental evidence to estimate the rateable value of BT's hereditament as a whole. Neither local loop(12) nor wholesale rental evidence(13) was available at the time of the 1995 and 2000 valuations.

(53) BT denies that there was any bias in the appeal system or in its application to BT. Everybody has a right to appeal and a possibility to settle. The valuation of BT's network was the object of an appeal in 1995. Its settlement before the Valuation Court was not unusual, and does not imply discretion on the part of the VOA: the possibility of reaching a settlement does not relieve the VOA of its statutory duty to determine the annual rental value of the hereditament under the hypothetical tenancy. According to BT, it is also wrong to say that the 1995 valuation also served as a basis for determining the 2000 values and conferred ten years of certainty on BT: the settlement concerning the 1995 valuation only took place in 2000 because of delays in the legal proceedings. Thus, BT was subject to five years of uncertainty as regards the 1995 valuation. The 2000 list was settled at the same time.

(54) BT explained that in the course of the five-year period, revisions of the valuation can take place in cases of ‘material changes in circumstances’ (‘MCCs’). A market share adjustment mechanism was seen as an efficient means of reflecting, on an annual basis, the net effect of these MCCs(14).

(55) Therefore, BT explained that the Commission's decision to initiate the procedure was wrong to claim that any increase in the value of BT's infrastructure is not taken into account: in fact, BT has been subject to upward as well as downward revisions, depending on whether factors increasing BT's value outweighed decreasing factors(15).

(56) According to BT, the revision system applicable to Vtesse appears to be based on the same principles, albeit on a different manner: newly occupied fibres become an additional part of its hereditament and are valued as part of it. In the case of Vtesse, market share is not a suitable means for assessing changes in rateable value since any changes are likely to be imperceptible due to Vtesse' small market shares.

(57) Finally, if the Commission were to decide that BT did receive State aid, BT submits that the Commission could not order recovery because of BT's legitimate expectation that any such aid was lawful, or, alternatively, that it would constitute existing aid.

Comments from Kingston

(58) Kingston claims that the Commission has not established the existence of an advantage, either by showing that BT and Kingston have benefited from a selective reduction in the burden of tax, or by showing that other operators face an exceptional increase. It notes that in the case of Kingston, the Commission has no idea how it has been rated.

(59) Kingston considers that the only evidence that an advantage may have been granted to BT is the difference in the rates/revenue ratio between BT and Vtesse. However, this ratio is irrelevant in the case of property tax, which is not a tax on revenues. For such ratio to be relevant, Vtesse and Kingtson or BT should be directly comparable firms, which is not the case given the differences in size, turnover and economic activities. Finally, it estimates Vtesse's ratio at 0,39 %, while its own rates/revenue ratio is allegedly at 4,5 %(16).

(60) It denies that the existence of several methods bestows a selective advantage on Kingston. The VOA has, in its view, some discretion in the valuation process, but this is not unfettered discretion: the nature of the hereditament and the nature of the evidence determine its choice, which is further constrained by a considerable body of case-law. It is not possible to demonstrate that the R&E method favours Kingston or BT, since the rental method is not applicable to them, while the R&E method is not applicable to Vtesse, since it is not profitable.

(61) It further submits that there is no demonstration of a distortion of competition, with no reference to the other operators that compete with Kingston. As a subsidiary legal submission, it suggests that any potential aid granted through the application of the UK rating system would be existing aid.

IV. COMMENTS FROM THE UNITED KINGDOM

(62) The UK authorities submit that the rates system is a uniform statutory system, applicable to all non-domestic hereditaments in England and Wales, which aims at establishing a common yardstick, by which to assess the value of the premises. This yardstick is the rent at which it is estimated that the hereditament might be expected to be let from year to year, under certain statutory assumptions concerning the state of repair of the hereditament and the terms of the hypothetical tenancy.

(63) This value is established by the VOA, which employs professionally qualified chartered surveyors, and is required by law to exercise its expertise independently. Its decisions are subject to the independent scrutiny of the Valuation Tribunal, and upwards to the Lands Tribunal and the Court of Appeal.

(64) The UK authorities further explain that, in order to value this hypothetical rent, the VOA can use several methods. These valuation methods have the same objective, namely the ascertainment of the hypothetical tenancy and should all, properly applied, yield the same result. The existence of several methods is meant to cater for the wide range of circumstances to which rates are applied. The selection of the appropriate valuation method is not a discretionary decision: it will depend on the nature of the hereditament and the nature of the available evidence. Thus, for instance, if there is no profit motive (hospitals, schools) for occupation, the R&E method is not applicable and the VOA is likely to judge that the most appropriate method is the contractor's basis.

4.1. Choice of the method applied to BT and Kingston

(65) The UK authorities explain that the R&E method was the only one that was available to assess BT's and Kingston's hereditament. They claim that the VOA's team of independent experts(17) that worked on assessing their hereditament concluded for all valuation periods, including 2005-2010, that there was not enough evidence to assess the rateable value of BT's and Kingston's networks on the basis of the rental method. This finding was accepted by the Valuation Tribunal in the case at hand(18). They further claim that the R&E method is the only one that could reflect the unique advantage of ubiquity which BT enjoys.

(66) The UK authorities claim that the rental method was not applicable to BT and Kingston. On this point, they first emphasise that the question of the availability of rental evidence has to be assessed at the time when the valuation was conducted. Thus, for instance, the possibility of applying the rental method for the period 1995-2000 can be considered only on the basis of information that was available at the ‘antecedent valuation date’, i.e. 1 April 1993. Therefore, the UK authorities claim that third parties and the Commission are wrong to assume that market evidence that is now available is relevant to the question as to whether the 1995 and 2000 lists could have been compiled on the basis of the rental method.

(67) Therefore, with respect to the 1995 lists, only information available in 1993 could be taken into account. In its decision to initiate the procedure, the Commission suggested two possible sources of rental evidence: either the rent received by BT when renting out its own infrastructure or transfer prices between BT's retail business and BT's wholesale business. The UK authorities claim that neither of those approaches was available to the VOA in 1993. In 1993, BT was not renting out any significant part of its infrastructure and it was not structured in a way that separated its retail and wholesale businesses.

(68) Concerning the 2000 list, based on information available in 1998, the UK authorities admit that by that date, there was limited rental evidence from fibre optic networks which started to be provided to operators under so-called Indefeasible Rights of Use (IRU) contracts. However, that evidence was limited and applied only to fibre optic networks used for carriage of high volume services, and was therefore relevant to an extremely limited proportion of BT's hereditament. They also admit that there were separate retail and wholesale divisions in 1998, but the access business, which is an essential part of BT's activities, was operated on an ‘end to end’ basis with no transfers to the other divisions. Therefore, transfer payments could not be used to value the whole of BT's network.

(69) As to the 2005 list, the same remarks as above apply. The UK authorities also deny that, contrary to the suggestion made by Altnet, tariffs for local loop unbundling provide any rental evidence on the basis of which BT's hereditament could have been assessed. For instance, the Wholesale Line Rental (WLR) tariff was available from August 2002. It is a product which BT is obliged to provide to other telecommunications operators, and which enables them to offer both line rental and calls to customers over BT's local network. However, the payment made by the telecommunications provider to BT is not equivalent to a rent for the local network. It is a payment for a service, since BT has to provide, inter alia, maintenance of the line, exchange accommodation, billing and R&D, fault repair, pair gain equipment, and power. Therefore, the UK authorities claim that, contrary to Vtesse and Altnet's assumption, it cannot be equated with a rent per line for the local loops. They come to the same conclusion with respect to the Calls and Access tariff, which was a precursor to the WLR product, was available from 1998 and covered the same services as the WLR. Furthermore, BT's wholesale line rental tariff and local loop unbundling fee are regulated by OFCOM. They are not open market rents and as such little weight can be given to these tariffs as evidence of rental value(19). Therefore, the VOA claims to have no rental evidence on the essential parts of the networks of BT and Kingston, i.e. their local access networks.

(70) Furthermore, the UK authorities explained that rental evidence used when applying the rental method to telecommunications operators other than BT and Kingston cannot be used to value the hereditament of these two firms. BT and Kingston, on the one hand, and other telecommunication operators, on the other hand, are not remotely comparable. The networks of BT and Kingston are primarily local access networks, serving millions of individuals in the case of BT, while the networks of most other telecommunication operators are primarily core networks serving a small number of high value, high volume customers. The value of a hereditament is influenced by a number of factors including its function, its physical layout, its extent and the regulatory environment under which it operates. Thus, given the differences in the hereditaments and the lack of comparability, rental evidence from telecommunications operators, for instance on their optic fibre network, cannot be applied directly to BT/Kingston. For the same reasons, the UK authorities claim that, contrary to the suggestion made by the Commission in its decision to initiate the procedure, rental evidence from Cable TV companies cannot be applied to BT or Kingston.

(71) As to the suggestion that parts of BT's hereditament could have been valued by reference to rental evidence derived from other networks and then on an R&E method for the remainder, the UK authorities claim that such an approach was not feasible. Such an approach would pre-suppose that there are distinct physical parts of the BT network for each of the different businesses in which BT is engaged. However, the UK authorities note that according to British case-law on the application of business rates(20), contiguous property in the same occupation, such as BT's and Kingston's networks, must be valued as a single hereditament. Disaggregating property in single occupation would introduce valuation distortions, since the sum of the value of the parts of a property rarely equates to the value of the whole.

(72) More specifically, to seek to disaggregate BT's core trunk route, high volume, fibre optic network, which competes with Vtesse's business, and to assess its rateable value by reference to rents paid by competing telecommunications providers is not possible. That is because that same infrastructure is also used by low volume telephony services provided to BT's 29 million domestic customers. BT's or Kingston's networks are physically incapable of being divided into self-contained units capable of being separately let or occupied, by reason of their physical and functional integration.

(73) In any case, even if it were possible to value separate parts of BT's and Kingston's hereditaments and value them under a different method than the R&E method, the problem would remain of how to value the remainder, since the R&E method can be applied only in respect of the whole of those undertakings' networks.

4.2. The application of the R&E method to BT and Kingston.

(74) The UK authorities provided detailed explanations on the way the R&E method is applied to BT and Kingston.

Valuation of BT

(75) The valuations are carried out by independent experts in accordance with the general rules on the application of the R & E method(21). Although the modelling and the resulting calculations are complex, the following paragraphs provide a brief overview on how the valuation was done for the 1995-2000 period.

(76) The R & E valuation takes the expected receipts, which are relevant to the rateable occupation of BT's hereditament, and deducts the relevant expenses. Specialist consultants (National Economic Research Associates — NERA) were commissioned by the VOA to prepare a model to provide an estimate of future revenues and operating costs over a five-year period arising from occupation of the BT Central List hereditaments in England & Wales, together with the remainder of the network in Scotland and Northern Ireland. A summary description of the NERA model and underlying spreadsheets for the 1995, 2000 and 2005 valuations were submitted by the UK authorities.

(77) The objective of the NERA model was to predict the revenues and operating costs likely to arise from the BT network throughout the United Kingdom, in the context of the rating hypothesis over a five year period. That hypothesis envisages a letting of the rateable portion of the BT network at 1 April 1993, but having regard to its physical state and that of competing networks and the remainder of the physical environment as at 1 April 1995. The NERA model proceeded from the basis of actual revenues and operating costs, arising in the year preceding the antecedent valuation date, as these would be known to the bidders for the hypothetical tenancy. Observations of a long series of past statistics were used to derive elasticities which were used to forecast movements in the total telecoms market and BT's share thereof over a five-year period.

(78) In the case of BT, the VOA decided to use forecast costs and revenues over a five-year period rather than the usual one-year period. The UK authorities explained that in certain circumstances, the application of the R&E method may involve considering levels of profitability over a number of future years if performance over a single year is not likely to be typical of the future. In the case of BT, the VOA anticipated that, viewed from the antecedent valuation date (‘AVD’), the future pattern of profitability would be affected by the prospect of increased competition, regulatory price control, a significant growth in competitors' networks and changes in the types of services offered. For these reasons, an approach based on anticipated profitability over a five year period was necessary to value the rent that a hypothetical tenant, looking at the AVD, would be ready to pay for BT's network over that period(22).

(79) Having anticipated receipts and operating costs over five years, the next stage in the R&E valuation is a deduction of the hypothetical tenant's estimated capital expenditure over the five-year forecast period on items which do not form part of the hereditament, but which are necessary in order to earn the forecast receipts. These items fall into two categories: (a) non-rateable plant & machinery required to be procured during the five year period, whether as replacement for existing equipment or otherwise; and (b) additions to the network not in existence at 1 April 1995, but which are forecast to be made over the five-year period.

(80) The next step is the predicted change in working capital over each of the five years of the forecast period. The annual change is deducted from receipts net of expenses and the capital expenditure described above. This produces a divisible balance representing the total of rent and the tenant's required cash flow. The next step is the estimation of the tenant's required cash flow.

(81) This is the tenant's required return on his capital at the commencement of the year, adjusted for change in that capital over the course of the year. The tenant's capital at the commencement of the hypothetical tenancy consists of working capital and tangible assets, namely non-rateable plant and machinery. The change in the tenant's total capital is estimated for each year. Changes in tangible asset values will occur each year as a result of changes in replacement asset costs, depreciation of existing assets, and acquisition of new or replacement assets, and these changes have been forecast over a five-year period.

(82) The tenant's required return was taken to be 16,2 % in nominal terms(23). This is an estimate of BT's weighted average cost of capital (‘WACC’) in nominal pre-tax terms at 1 April 1993 — the relevant antecedent valuation date. In this way the tenant's required cash flow at the year end is computed. Since receipts, expenses and capital expenditure are incurred throughout the year, just as rent is payable during the year, the year-end required cash flow is expressed in mid year terms by applying a present value factor for 6 months at 16,2 %.

(83) The tenant's required cash flow is deducted from the divisible balance, leaving a residual cash flow for each year, since that cash flow arises from transactions taking place throughout the year. The present value of each year's cash flow (at 16,2 %) is then summed up and divided by the sum of the present value factors (equalling 5 years purchase at 16,2 %).

(84) The result, GBP 700 636 855, represents the estimated rental bid which the hypothetical tenant could afford to make in order to retain a return of 16,2 % on the capital investment which he would of necessity make in order to generate the receipts upon which his bid would be predicated. The penultimate stage of the valuation involves the apportionment of the UK ‘cumulo’ value of over GBP 700 million to England and Wales. In the absence of any direct evidence of telecommunications traffic in each of the countries of the United Kingdom, it has been assumed that the apportionment most likely to be accurate should have regard to the distribution of economic activity between the constituent parts of the United Kingdom. On the basis of this evidence, 84,8 % of the UK ‘cumulo’ was apportioned to England and 4,2 % to Wales.

(85) The resulting figures include a number of properties which are owned and occupied by BT but not included in the Central List hereditaments. These properties were capable of being valued directly on the basis of the rental method and separately assessed at GBP 149 167 870 in England and GBP 3 822 485 in Wales. The final stage of the valuation involves the subtraction of those separately assessed rateable values in order to arrive at the final rateable value for the Central list hereditaments, which was GBP 445 million in England and GBP 25 million in Wales.

(86) The same method was applied for the following two valuation periods, using as a basis BT's actual accounts for the year 1997/98 for the 2000-2005 period, and for the year 2002/03 for the 2005-2010 period. The VOA used a tenant's required return of capital investment of 14,25 % for the 2000-2005 period and a required return of 12,25 % for the 2005-2010 period, this decrease reflecting the evolution of inflation.

(87) The application of the R&E method to BT has resulted in the following rateable values since 1995:

List	England	Wales
1995-2000	GBP 445 million	GBP 25 million
2000-2005	GBP 467 million	GBP 26 million
2005-2010	GBP 530 million	GBP 24 million

(88) On the specific question of the 1995-2000 valuation, the UK Authorities acknowledge that the application of the R&E method to BT for that period was the object of a settlement, following more than five years of litigation and subsequent negotiations. The VOA is entitled to settle, which is common practice in the application of the property tax(24). It cannot reach a settlement on a basis that does not represent the correct rateable value of the hereditament concerned. The UK Authorities emphasise that the settlement concluded with BT confirmed that its hereditament was to be valued on the basis of the R&E method rather than the contractor's basis method, which was favoured by BT, and that it resulted in a rateable value that was significantly higher than the value BT contended for.

(89) Concerning the allegation made by Vtesse that, after the decision of the Valuation Tribunal(25), the VOA and BT settled on a rateable value of BT's network that was 15 % lower than the initial valuation made by the VOA, the UK authorities explained that the difference can be accounted for by the fact that the valuation carried out after the Valuation Tribunal's decision was the product of a more thorough analysis. When preparing the first valuation, which was presented before the Valuation Tribunal, the VOA and its economic consultant NERA were not able to obtain full details of BT's historic revenues, prices, volumes and costs since, in particular they did not have the tools to force disclosure of this information by BT. BT appealed to the Lands Tribunal against the decision. The Lands Tribunal hearing is a de-novo hearing(26). Furthermore, at that stage of the procedure, the VOA could issue information injunctions and obtain disclosure of relevant information by BT. In particular, for the 2000 valuation, the VOA and NERA were able to obtain significant data which were not available to them in 1997, such as a full breakdown of BT's 1992/3 revenues and costs, BT's five-year plan produced before the AVD, and historic data on line and call volumes and prices and leased line volumes and prices. This new information affected all the main parts of the valuation, including forecast of revenues and operating costs, future capital expenditure and the hypothetical tenant's required return.

(90) Some of the changes tended to increase the rateable value. This was the case of the following factors:

• The VOA finally used a tenant's required return of 12,27 % in real terms, instead of 12,5 %, as initially adopted by the Valuation Tribunal.

• Capital expenditure on the tenant's non-rateable assets were significantly revised downward, thanks in particular to the information provided by BT's 1993 business plan, which was not available in 1997.

• The cost per redundancy was reduced from GBP 50 000 to GBP 33 900 in 1992/3.

(91) The factor with the main downward effect on rateable value was the change in forecast of future receipts and costs. The revised version predicted higher profits in the first year but lower in the subsequent years, compared with the forecasts used in the 1997 valuation. More precisely, these changes can be attributed to the following elements:

• The X factor in price cap(27) in 1997/8 was changed from -4,5 % to -7,5 % since the move to -4,5 % which took place in 1997 and was used in the 1997 valuation could not have been reasonably anticipated at the AVD.

• The forecast for the annual reduction in access deficit contribution per minute(28) was increased from - 3 % to - 12 %, since the 1997 forecast did not take sufficiently into account the reduction in the size of the access deficit as exchange line prices increased faster than costs.

• In the 1997 valuation, the VOA and NERA assumed that inland leased line prices would be rising with inflation. This assumption was not supported by the data on historic leased line prices which was subsequently released by BT, and these prices were assumed to remain constant in the context of the 2000 valuation.

• In the 2000 valuation, an allowance was made for additional outsourcing costs.

(92) The combined effect of these factors meant that the final valuation was lower than that which had initially been calculated by the VOA in 1997.

(93) Finally, on the point raised by the Commission in its decision to open the procedure as to whether the R&E method correctly takes account of the possible inefficiencies and lower profitability of the firms to which the R&E method is applied, the UK authorities reply that valuation proceeds on the assumption of an average level of efficiency. Where there is evidence of less than average efficiency, the valuer using the R&E method must make adjustments to the estimated fair receipts and expenses. The UK authorities claim that there is evidence that BT has an average level of efficiency(29) and no adjustment was deemed necessary to make a correct valuation of their hereditament. On the question of the universal service obligations (‘USO’) and the way in which they are taken into account in the valuation of BT, the UK authorities explained that no specific allowance is made for such USO. The costs of the USO are included in the overall predicted capital expenditure for growing the network and in the overall operating expenses. The increased revenues resulting from the service of customers under the USO and from the brand enhancement resulting from the USO, if it exists, are included in the anticipated receipts.

Valuation of Kingston

(94) A similar but much simpler staged approach to the R&E method was applied to Kingston. No model predicting the turnover likely to be derived from occupation of the hereditament over a five-year period was prepared. The reason for this is that there are no networks serving residential customers competing with Kingston in the Hull conurbation area, no cable TV and little competition for commercial lines. Due to the more stable situation in Kingston's Hull market, it was decided that a five-year projection was not necessary.

(95) Instead, for the 1995-2000 period, regard was had to actual turnover and operating costs for the 3 years preceding the AVD and based on this, it has been assumed that the hypothetical tenant might reasonably anticipate a net income of around GBP 9,5 million per year before rent. The tenant's required return has been taken to be 12,5 % on the value of the non-rateable assets (assumed to be GBP 64 million), giving a return of GBP 8 million to be deducted from net income. The residue of GBP 1,5 million is available for rent and was set as Kingston's rateable value for that period. The same method, using the figures available in 1998 and 2003 gave a rateable value of GBP 5,1 million for the 2000-2005 period and GBP 7 million for the 2005-2010 period respectively.

4.3. Revision mechanism

(96) The UK authorities also provided explanations concerning the revision mechanism applied to BT. They explained that under Schedule 6 to the LGFA 1988 BT, like any other rate payer, is entitled to have its rateable value altered if a material change in circumstances (‘MCC’) occurs during its five year valuation period. The MCC is taken into account as if taking place at the antecedent valuation date. In the case of BT, MCCs take place every day, either because of the extension of BT's network (upward effect) or because of the roll-out of competing networks (downward effects)(30). Theoretically, a valuation using the R&E method should take place every year, taking into account how these MCCs may have changed BT's five-year plan at the time of the antecedent valuation date. In preference to this complex method, the UK authorities and BT agreed on a proxy method of adjustment, based on the evolution of BT's market share in comparison with its market share in the year preceding the commencement of the rating list, and using data published by OFCOM.

(97) This market share was calculated by dividing the total of all categories of BT's relevant revenues by the total of all corresponding categories of UK market revenues for fixed telephony. The relevant categories are: (a) local calls; (b) national calls, (c) international calls, (d) calls to mobiles; (e) other calls, (f) exchange line connections, (g) exchange line rentals, (h) inland private leased rentals (this category includes the market on which BT and Vtesse compete), (i) inland private leased connections, (j) international private leased circuits, (k) interconnect and (l) telex (in 1995 list only) .

(98) Taking the example of the 2000-2005 list, BT's market share thus calculated was 68,47 % in 2000. The rateable value of its hereditament was GBP 467 million in England and GBP 26 million in Wales in 2000. For the subsequent year, BT's market share had fallen to 65,03 %. The rateable value for 2001 for therefore set at GBP 467 million ×65,03 %/68,47 % = GBP 443 million for England and GBP 26 million × 65,03 %/68,47 % = GBP 24,7 million for Wales. The following year, OFCOM data suggested a small reduction of its market share, which however led to no reduction of BT's rateable value. In the years 2003/4 and 2004/5, BT's market share increased to 67,6 % and 68,0 % respectively. Compared to the benchmark market share in 2000(31), it led to rateable values of GBP 447,5 million and GBP 450,6 million respectively(32).

(99) This method gave the following results for the 1995-2000 and 2000-2005 periods:

BT	List	England	Wales
01.04.1995	1995	GBP 445 million	GBP 25 million
01.04.1996	1995	GBP 412 million	GBP 23 million
01.04.1997	1995	GBP 392 million	GBP 22 million
01.04.1998	1995	GBP 366 million	GBP 20 million
01.04.1999	1995	GBP 347 million	GBP 19 million
01.04.2000	2000	GBP 467 million	GBP 26 million
01.04.2001	2000	GBP 443,5 million	GBP 24,7 million
01.04.2002	2000	GBP 443,5 million	GBP 24,7 million
01.04.2003	2000	GBP 447,5 million	GBP 24,9 million
01.04.2004	2000	GBP 450,6 million	GBP 25,1 million

(100) For the 2005-2010 period, the UK authorities indicated that, due to the lack of available data from OFCOM, the revision mechanism applied for the first two periods will be discontinued(33). In place of the former system, the UK authorities intend to apply a full R&E method annually, which will reflect all physical changes, including new lit fibres, but in the context of the economic conditions existing at the antecedent valuation date(34). This system is still being finalised.

(101) In the case of Kingston, no revision mechanism was applied. According to the UK authorities, no adjustment was undertaken for the 1995 and 2000 rating lists on the basis that no competitors built networks in the Kingston upon Hull conurbation area and that Kingston already had a network coverage of almost 100 % in the area, so it did not extend its network.

4.4. Existence of an advantage

(102) The UK Authorities first emphasise that all valuation methods should arrive at the same result: the hypothetical rent that would be paid for the hereditament. Therefore, the fact that a different method of valuation was applied to BT and Kingston, on the one hand, and to the other telecommunications operators, on the other, is not an indication that an advantage could have been granted to the former.

(103) They are of the opinion that there are two means by which the existence of a selective advantage could be demonstrated:

a)

Either by showing that an R&E valuation of telecommunications operators other than BT and Kingston would have yielded a lower rateable value than that derived from the rental method, so as to demonstrate that the R&E method is flawed or systematically understates value;

b)

Or by showing that, using correct and relevant comparisons, other telecommunications operators had a disproportionately high rateable value when compared with BT and Kingston.

(104) In order to establish point (a), it would be necessary to carry out an R&E valuation of the other telecommunications operators in order to see whether comparable results were obtained under the R&E method and under the rental method. None of the telecommunications operators has undertaken such an exercise and the UK authorities assume that if indeed the valuation of their network had been lower under the R&E method, members of Altnet and Vtesse would have raised this point(35). As to the VOA, it does not have these telecommunications operators' business plans. Therefore, it is not in a position to assess their anticipated receipts and expenditure deriving from the occupation of their network, and determine the rent they would be ready to pay for it under the R&E method.

(105) Concerning point (b), the UK authorities note that Vtesse and the Commission suggested comparing the percentage of receipts paid by Vtesse in rates compared with the percentage of receipts paid by BT and Kingston. The UK authorities first submit that this ratio is irrelevant. They argue that rates are a tax on the value of property, not a tax on receipts, profits or capital gains. The value of a property is not a function of the turnover of the particular occupier: two businesses may occupy identical properties, yet their turnover could be very different depending on the nature and success of their business. Even tenants in the same area of business may have dramatically different turnovers depending on their efficiency. Consequently, revenue is no indication as to the rent which would be arrived at in the ‘haggling of the market’ and placing any weight on an alleged disparity between the percentage of receipts spent on rates is wrong.

(106) In any case, the UK authorities submit that the correct rates/receipts(36) ratio for BT and Kingston in the UK were the following :

	1995/96	2000/01	2005/06
BT	3,0 %	2,7 %	3,6 %
Kingston	1,0 %	4,0 %	3,4 %

(107) The UK authorities explained that the difference between the ratio that they obtain for BT and the ratio calculated by Vtesse for BT (2 %) can be explained by the fact that Vtesse compared BT's receipts in the UK with its network rateable value in England and Wales. Vtesse failed to take into account separately assessed properties and the rateable value of BT's hereditament in Scotland and Northern Ireland. The difference between the ratios provided by Vtesse and the UK authorities can be further explained by the fact that the UK authorities calculated the rates/receipts ratio by using revised revenues for BT and Kingston which were adjusted to take account of non-relevant revenues and were normalised to reflect projected future trends.

(108) In any case, even if the percentage-of -receipts approach was to have any validity, the UK authorities claim that the figures provided by Vtesse and used by the Commission in the decision to initiate the procedure, do not show the existence of an advantage to BT. On the contrary, they note that, when all the Altnet telecommunication operators as a whole are compared with BT, the percentage of receipts shows that BT does not pay less than the Altnet operators. In the final year of the 2000 list, BT had about 70 % of the total fixed telecoms rateable value and about 71,6 % of call and access revenues and a smaller percentage of other fixed market revenues. According to the UK authorities, these figures show that there is no systemic problem with the valuation in the 1995 and 2000 lists.

(109) On the question of marginal taxation, the UK authorities deny that the different way of rating additional fibres confers any advantage on BT or Kingston. The trigger for rateability of telecommunication operators' fibres under the rental method is the lighting of the fibres. This is not the way new fibres are treated in the case of BT/Kingston under the R&E method.

(110) The R&E method anticipates the receipts and expenditures over the next five years. The fibres necessary for BT's business throughout the five-year life of the list are taken into account in the assessment of the rateable value because the receipts from these fibres are included in the R&E valuation. Business rates are therefore payable in respect of optic fibres from the commencement date of the rating List, not merely from the day when they are lit. Physical additions to BT's hereditament are further taken into account by the annual revision mechanism which is based on the evolution of BT's market shares described in recitals 96 to 100.

(111) The UK Authorities acknowledge that the valuation methods, of necessity, require a different approach to the treatment of the lighting of additional fibres, but underline that the end result of the valuation exercise in both cases is that the value of the fibres to the occupier is reflected in the rateable value.

V. COMMENTS FROM OFCOM

(112) OFCOM is the independent regulator and competition authority for the UK communications industries. The comments from the UK authorities were provided by the UK central government and do not represent the views of OFCOM. As a result, the Commission asked OFCOM to provide its comments on the decision to initiate the procedure by letter dated 26 July 2005. They were provided by OFCOM by letter dated 21 September 2005.

(113) OFCOM considered the extent to which differences in methodology and practice may have effects on the conditions of competition. Its general conclusion was that the non-domestic rating system is currently not having an adverse effect on the conditions of competition in the telecommunications market.

(114) Concerning the rates/revenue ratio, OFCOM is of the view that this ratio is not pertinent for assessing the fairness of the non-domestic rates system. It points out that rates are related to the expected market rents, which are likely to reflect the efficient use of the assets. If an asset is used inefficiently by a firm, rates may be a significant proportion of its revenues. For example, the rates/revenues ratio for a fibre network may depend on the amount of data for which it is used. Differences in the rates/revenue ratio of different companies may thus simply reflect differences in the value of the services they provide to customers through similar rateable assets.

(115) Furthermore, OFCOM calculated the rates/revenue ratio for a certain number of telecommunications operators in the United Kingdom:

	2005 rate liability (GBP million)	2005 revenues(GBP million)	Rates/revenue( %)
BT	238,31	18 623	1,28
NTL	24,26	1 930	1,26
Telewest	15,49	1 300	1,19
C&W	16,37	1 602	1,02
Kingston	3,35	364	0,92
Energis	3,27	720	0,45
Global Crossing	3,13	270	1,16
Easynet	1,77	78	2,28
Thus	2,41	341	0,71
Your Communication	1,09	186	0,59
Fibernet	1,08	47	2,32
Gamma Telecom	0,46	72	0,63

(116) OFCOM also calculated Vtesse's ratio for 2003, the only year for which it had available figures. This ratio was 2,1 %. These figures show that the rates/revenue ratio differs significantly and with no clear pattern to volumes of revenues. OFCOM concluded that not only is this ratio a flawed methodology for assessing the fairness of the non-domestic rating system, but the allegation that BT generally pays a smaller amount in rates compared to its revenues than its competitors appears to be false.

(117) OFCOM underlines the difficulty of making appropriate comparisons between very different telecommunications operators (e.g. the average rateable value of a route kilometre of a fibre network can be very different according to the characteristics of the network to which it belongs and to the number of lit fibres on that route). It nevertheless suggested some alternative ways of comparing BT's taxation with that of its competitors. For instance, it calculated the ratio rateable value/connection for BT and the three companies which are most similar to it regarding the local access network, i.e. Kingston, NTL and Telewest. All four are primarily local access providers. BT and Kingston are valued on the basis of the R&E basis and NTL and Telewest are valued on a method derived from the contractor's basis method. The results of OFCOM's calculations are shown in the table below. In the view of OFCOM, these rateable values per connections do not vary much from one operator to the other.

	Rateable value/connection
BT	GBP 21
NTL	GBP 22
Telewest	GBP 21
Kingston Communication	GBP 26

(118) Therefore, in general, OFCOM concludes that there is little evidence to support an argument that BT pays significantly less than fibre backbone operators on any measure.

(119) OFCOM also analysed whether the rating system introduces undue asymmetry in terms of marginal rate liability. On this point, OFCOM notes that the R&E method uses projected figures for revenues and costs and therefore conceptually includes any network build required to meet these business plans. The R&E basis is not limited to the state of BT's network at the start of the five-year period but reflects the necessary changes that will be made to it over the following five years, with annual reassessments to capture and rate material changes to BT's hereditament. In the view of OFCOM, although the incremental charge is more apparent to smaller operators, as each fibre lit leads to a specific additional rates charge, this does not mean that BT's rateable value does not include amounts for similar changes. OFCOM concluded that it has no evidence to suggest that, unlike telecommunication operators assessed under the rental method, BT pays no additional rates when it lights a fibre. The system would appear to provide a materially symmetrical outcome in the long term in each case, albeit through different routes and with low levels of transparency.

VI. LEGAL ASSESSMENT

State aid assessment pursuant to Article 87(1) EC

(120) According to Article 87(1) EC, ‘any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the common market.’ Therefore, a measure constitutes State aid within the meaning of Article 87(1) EC if it fulfils the following four criteria:

• It confers a selective advantage on the recipient(s)

• It is granted by the State through State resources

• It distorts or threatens to distort competition

• It is likely to affect trade between Member States.

(121) In order to determine whether the application of the property tax to BT and Kingston has resulted in an economic advantage to these two firms, the Commission's analysis must be two-fold:

• It must first determine whether the rules concerning the property tax have been correctly applied to BT and Kingston and in particular whether the application of the R&E method was justified and properly applied.

• It must also consider whether the application of the R&E method to BT and Kingston confers an advantage on those firms in comparison with their competitors that are valued under the rental method.

6.1. Correct application of the rules on property tax to BT and Kingston

6.1.1. Choice of the method applied to BT and Kingston

(122) The first point considered is whether it was appropriate for the UK authorities to apply the R&E method to BT and Kingston instead of the rental method, which appears to be the generally applied method in the telecommunications sector. The question is whether there was enough rental evidence to assess BT's and Kingston's networks taking into account the rental evidence for the 1995-2000, 2000-2005 and 2005-2010 periods.

(123) In its decision to initiate the procedure, the Commission suggested that the internal transfer prices between BT's retail businesses and BT wholesale as well as the rent received by BT when renting out its own infrastructure could provide a basis for applying the rental method. Among the third parties that commented on the decision, the Altnet Task Force suggested specific sources of rental evidence, such as BT's wholesale line rental tariffs or its published prices for unbundled local loops, which could be used to determine the rental value of its copper cables.

(124) On these points, BT and the UK authorities have explained that the rental method could not be applied to assess BT's and Kingston's hereditaments. First, the evidence which the Commission or Altnet suggested using was not available for the first two valuation periods (i.e. 1995-2000 and 2000-2005). That is the case of the LLU rental tariffs, which were applied from 31 May 2001 or the set of regulatory accounts that reflected transfer charges for BT's access services. Furthermore, even when that information was available, for instance for the latest valuation period of 2005-2010, it was not relevant and could not be used for a valuation of BT's and Kingston's very specific networks. For instance, the UK authorities have explained that the Wholesale Line Rental Tariff and the LLU Tariff are a payment for a service and, under existing case-law on the property tax, cannot be considered rental evidence to value the relevant part of BT's network. Furthermore the tariffs are regulated by OFCOM and as such are not ‘open market rents’. As such, little weight can be given to those fees as evidence of rateable value. For these reasons, there is no actual market-based evidence concerning the access part of BT's or Kingston's networks, which is a unique characteristic of their networks.

(125) It also appears that even when rental evidence derived from other telecommunication operators exists, for instance on optic fibres, the differences between the use of optic fibres by those operators and the use that is made of them by BT and Kingston means that this rental evidence is not fully relevant to value the latter undertakings' hereditaments. The fibre rent quoted by Vtesse is based on market evidence for fibre trunk networks serving a small number of high value, high volume customers. BT's and Kingston's network are mainly local access networks serving millions of individuals in the case of BT. The role of their optic fibre network is mainly that of a backbone which interconnects their local loops. Their networks are for this reason unique and cannot be compared with any other in the United Kingdom, including large networks such as that of Cable & Wireless. It is true that parts of BT's network may be used by BT to provide services that compete with Vtesse, but they remain physically and functionally incapable of being disaggregated from BT's whole network. Therefore, on the whole, the two types of network do not fulfil the same function, and are unlikely to have the same rental value. This view is confirmed by OFCOM in particular, which pointed out that the average value of a route kilometre of a fibre network can be very different according to the characteristics of the network to which it belongs and that BT's core network topology is likely to be very different to that of a backbone operator catering for the corporate market.

(126) As to the possibility of applying the rental method to parts of BT's or Kingston's hereditaments for which rental evidence may be available and the R&E method to the rest, that has also been ruled out by the VOA. The reason is that the R&E method can only be applied to the whole hereditament and it is not possible to separate parts of these firms' hereditaments, such as their optic fibre networks, since the hereditament has to be assessed as a whole under British case-law, since these parts are fully integrated with the rest of the network(37).

(127) The independent experts of the VOA therefore concluded that the rental method cannot be applied to BT and Kingston. That conclusion that the R&E method was the correct one, was confirmed by the Valuation Tribunal(38). The Commission considers that there is no probative evidence showing the contrary. Consequently, the Commission concludes that there was not enough rental evidence to assess BT's and Kingston's networks in the periods under consideration and that the application of the R&E method to assess the rateable value of their networks was therefore justified.

6.1.2. Correct application of the R&E method to BT and Kingston

(128) The next question is whether the R&E method has been correctly applied to BT and Kingston. On the basis of the extended explanations provided by the UK authorities and described at recitals 74 to 93, it appears to the Commission that the method was applied in accordance with the general rules on the R&E method, as laid down in the RICS Guidance Note on the R&E method of valuation for non-domestic(39). The use of anticipated revenues and costs over a five-year period in the case of BT is not explicitly contemplated in the RICS Guidance Note. However, the guidance note does refer to the estimated future profitability when considering the hypothetical tenant's bid in the light of cyclical economic changes which are predictable as at the AVD. Furthermore, a method based on anticipated receipts and costs over a certain number of years has also been used in the assessment of the rateable values of hereditaments of gas and electricity operators, and is therefore not specific to BT. In the case of BT, given the prospect of increased competition and significant growth in competitors' networks which characterise the telecommunications sector, the use of forecast revenues and costs over five years is a reasonable methodology for assessing the hypothetical rent that a tenant with a reasonable prospect of continued occupation would be ready to pay for BT's network. Conversely, the stability of the market on which Kingston is present, which is characterised by limited competition, justified that a five year projection was not deemed to be necessary.

(129) On the specific question of the settlement concerning the 1995-2000 period, the UK authorities have provided detailed explanations as to the reasons why the final rateable value (GBP 445 million in England, GBP 25 million in Wales) was lower than the rateable value initially set by the Valuation Tribunal in its decision of 19 March 1998 (GBP 523 million for England and GBP 30 million in Wales). These explanations, which are summarised in recitals 88 to 92, show that the final lower rateable value of BT's network was not negotiated between the VOA and BT, as claimed by Vtesse, but was arrived at through the use of revised and more precise data, which were disclosed by BT following the decision of the Valuation Tribunal and which were not available at the time of the 1997 valuation, and through the use of more precise methods of valuation following the Valuation Tribunal's decision.

(130) In relation to a point raised in the decision to initiate the procedure, it is also to be noted that, when applying the R&E method, the UK authorities do consider the level of profitability of BT. As already explained in recital 93, valuation under the R&E method must proceed on the assumption of an average level of efficiency, and the UK authorities have established that BT fulfils this criterion. The Commission also notes that the Universal Services Obligations are correctly taken into account since both the costs and revenues deriving from such obligations are included in the overall valuation.

(131) Concerning the application of a revision mechanism based on the evolution of market shares, described at recitals 96 to 99, and discontinued in 2005, the information received indicates that this was an ad-hoc adjustment mechanism. It is the Commission's understanding that, under Schedule 6 to the LGFA, the VOA is under a statutory obligation to alter the rateable value of an undertaking if a material change in circumstances (MCC) takes place and to take this MCC into account as if taking place at the Antecedent Valuation Date. Given the size of BT's network, and the conditions of the telecommunications market, with the increasing development of competing networks, it is accepted that MCCs take place on a daily basis and are therefore difficult to take into account, especially in the context of the application of the R&E method. This market share based mechanism is supposed to reflect physical extensions of BT's network (which should result in an increase of its market shares) and the roll-out of its competitors' networks (which should result in a decrease of BT's market shares). Therefore, this revision mechanism appears to be a reasonable way of taking MCCs into account and can be justified by the characteristics of BT's network and the telecommunications markets.

(132) The result of the application of this revision mechanism has led to fluctuations in the evolution of BT's rateable values. For instance, in 1999, in the last year of the 1995-2000 period, BT's rateable value in England and Wales was GBP 366 million. In 2000, in the first year of the 2000-2005 period, its rateable value was set at GBP 493 million. These fluctuations can be explained by the characteristics of the property tax and by the economic situation of the telecommunications sector in the United Kingdom(40).

(133) It must first be recalled that, under statutory requirements, for the 1995 valuation, BT's network was valued as it physically existed as at 1 April 1995, but on the assumption that it existed in that state in the economic conditions of 1 April 1993. Material changes in circumstances were also treated as having occurred at 1 April 1993. For the 2000 valuation, the same exercise was carried out, but having regard to the assumption that the hereditament was let in the economic context of 1 April 1998. This is also true for Kingston.

(134) That means that the 1999 rateable value is the value of the hereditament as it physically existed in 1999, but in the economic context of 1993, i.e. at 1993 price levels and in a telecommunication market of the size it was in 1993. However, the following year's rateable value, i.e. in 2000, was calculated by reference to the economic conditions, price levels and size of the telecommunications market as they were in 1998. Similarly, the 2004 rateable value is set by reference to the economic conditions of the 1998 market, while the 2005 valuation is set by reference to the 2003 market conditions(41).

(135) The UK authorities have pointed out that the economic conditions in the telecommunications sector were substantially different in 1993, 1998 and 2003. First, the telecommunications market grew significantly between 1993 and 1998, and between 1998 and 2003. Thus BT's physical network in 2000 was considered in a market that was bigger than the market in which BT's physical network was considered in 1999. This necessarily meant that its 2000 rateable value was likely to be significantly higher than in 1999. Furthermore, in 1993, the telecommunications sector was in the process of opening up to unprecedented competition and the effect on BT's trading position was likely to be profound. That is confirmed by the fact that BT's market shares fell from 87 % to below 70 % over that period. This anticipation had a downward effect on BT's rateable value for each of the years of that period. Conversely, in 1998, less competition could be anticipated, which is again confirmed by BT's relatively stable market shares of around 70 % over the 2000-2005 period. This anticipation was likely to have an upward effect on BT's valuation over that period.

(136) Furthermore, the most valuable parts of BT's and Kingston's networks consist of their local access networks, the value of which has significantly increased with the exploitation of broadband DSL technology. This development, which could not be fully anticipated in 1993, created an upward effect on the valuations of BT's and Kingston's networks and can explain why their valuations were significantly increased in 2000 and 2005.

(137) The Commission also notes that the fact that only physical changes can be taken into account in the economic context of the AVD can explain why BT's rateable value fell over the life of the 1995-2000 and 2000-2005 rating lists despite the fact that its receipts or profits were stable or improving. Indeed, such elements as the growth of the market, the increase in price levels, technological changes or changes in the pattern of demand, which are factors that can have a positive impact on BT's results, cannot be taken into account by the VOA when carrying out its yearly revisions of the rating list.

(138) The Commission finally takes note that a new annual revision system which will be applied to the 2005-2010 period is no longer based on the evolution of market shares, but will consist of the application of a full annual R&E valuation. The UK authorities have indicated that this system will fully reflect all physical changes, including newly lit fibres. However, since this revision system has not been finalised yet, the Commission has not had the opportunity to analyse it and its final conclusion in the Articles of this Decision will therefore be limited to the 1995-2005 period.

(139) The Commission concludes that the UK authorities applied the appropriate valuation method to BT and Kingston and that the general principles concerning the application of the R&E method were respected by the UK authorities when they applied that method to BT and Kingston.

6.2. Does the R&E method confer an advantage with respect to the rental method?

6.2.1. Does the R&E method result in an overall undervaluation of BT's and Kingston's networks?

(140) The next point to be considered is whether the application of the R&E method to BT and Kingston confers an advantage on those firms in comparison with their competitors that are valued under the rental method. The question is whether this different treatment is likely to lead to a different result than that which would arise if the valuation of those firms were carried out using the same methodology. In line with the comments of the UK authorities, the Commission is of the view that the existence of an advantage to BT and Kingston could be demonstrated in the following ways :

(a)

by showing that the application of the rental method to BT and Kingston would be possible and result in a higher rateable value than the R&E method, or

(b)

by showing the application of the R&E method to the other telecommunication operators would be possible and would systematically result in a lower rateable value than that derived from the rental method, or

(c)

by showing that, using other correct and relevant comparisons, telecommunication operators that are assessed under the rental method are taxed more heavily as compared with BT and Kingston.

(141) Demonstrating the existence of a possible advantage to BT and Kingston following the reasoning under point (a) supposes that the rental method is applicable to BT and Kingston. For the reasons explained in recitals 124 to 127, there is not enough rental evidence to assess the rateable value of their hereditaments. Consequently, the Commission concludes that it is not possible to establish that BT's and Kingston's rateable values would have been higher if the rental method had been applied to them.

(142) Demonstrating the existence of an advantage to BT and Kingston following the reasoning under point b) would involve applying the R&E method to a representative sample of telecommunication operators which have been valued under the rental method. The Commission is not in a position to carry out these valuations itself. The VOA has also explained that, without the business plans of these firms, it cannot carry out these valuations under the R&E method. It is also noted that none of the third parties has produced evidence that the application of the R&E method would result in lower valuations of their hereditaments: in its submission, UKCTA indicated that it could not calculate what its members would have paid in rates under the R&E method.

(143) On the basis of those considerations, the Commission must therefore conclude that the method in point b) cannot be used to demonstrate the existence of a possible advantage to BT and Kingston.

(144) The method suggested in point c) would be to use relevant comparisons to establish that BT and Kingston are under-taxed under the R&E method. This method was suggested by the Commission in the decision to initiate the procedure: according to the information that was then available, BT was paying 2 % of its receipts in property tax, while Vtesse was paying two or three times as much.

(145) As a preliminary remark on the use and application of the rates/receipts ratio, the Commission notes that, according to Vtesse's comments, the revenue of telecommunications operators should be adjusted in order to allow a direct comparison of BT's rates/receipts ratio with that of other telecommunication operators. Vtesse claims that most operators pay significant fees to BT for the carriage and termination of traffic and that these fees do not relate to the rateable assets of these operators and therefore should not be used in any comparison of the rates burden.

(146) That argument cannot be accepted. The ratios to be compared should relate rates either to gross receipts or to net receipts. The ‘halfway’ approach suggested by Vtesse which consists of deducting some specific costs, such as interconnection fees, and not others does not seem accurate. First, it is not correct to assert, as Vtesse does, that the fees for the carriage and termination of traffic do not relate to the rateable assets of the operator under consideration: those fees are on the contrary necessary for the origination and termination of some of the traffic carried over their networks and therefore they directly relate to the operation of those networks. Secondly, those fees have a direct impact on the hypothetical rent that a putative tenant would be ready to pay for the network under consideration, since they are part of the costs that are borne for the use of that network. Finally, those interconnection fees are also a significant part of BT's income and as a result, they have a clear upward effect on BT's rateable value. Since the fees affect the hypothetical rent and rateable value of all telecommunications operators concerned, there is no justification for deducting them from the revenues generated by the telecommunications operators when carrying out a proper comparison between their rates/revenues ratios.

(147) On the question of the relevance of the rates/receipts ratio, the Commission notes that, in the view of the VOA and OFCOM, the rates/receipts ratio is not pertinent for assessing the burden of rates on the different operators. In the light of their comments and objections, the Commission can agree that this ratio is not a correct tool for assessing the fairness of the rates system, which is a tax on the value of a property, and not a tax on receipts or added value(42). As argued by OFCOM in particular, it appears that this ratio will vary according to the value of the services the operators provide to customers through similar rateable assets and according to how efficiently they use their assets. The Commission can therefore conclude that this ratio is at best indicative and that no firm conclusion can be drawn from it.

(148) Moreover, even if this ratio is to be taken into consideration, it does not allow any useful conclusion to be drawn. According to figures provided by OFCOM for twelve telecommunication operators, the rates/receipts ratio ranges from 0,59 % for Your Communication to 2,32 % for Fibernet, with BT standing at 1,28 % and Kingston standing at 0,92 %. It appears that neither BT nor Kingston pays a significantly smaller share of their revenues in business rates than most of the other telecommunications operators. On this question, it is further to be noted that there is some disagreement on the calculation of Vtesse's rates/receipts ratio. For 2004, Vtesse claims that this ratio was 13,46 %. This figure was based on a rateable value of GBP 429 000 (which implied rates of GBP 195 624) and ‘recurring revenues’ of GBP 1 453 146. However, according to the UK authorities, Vtesse' rateable value for 2004 was subsequently reduced to GBP 145 000 (implying rates of GBP 66 120), since it appeared that it was not actually in occupation of a hereditament in Hounslow, valued at GBP 284 000. Furthermore, it appears from Vtesse's report and financial statements for the year ended 31 December 2004 that its total turnover in 2004 was GBP 4 744 228. On the basis of its total turnover, which is the figure used by OFCOM for comparison in the case of the other telecommunications operators, Vtesse's rates/turnover ratio in 2004 falls to 1,4 %, or 2,2 % if one includes the rates paid by Vtesse on its offices, which are valued separately. These figures are more in line with the conclusions of OFCOM, which found that Vtesse had paid 2,1 % of its revenues in rates in 2003. On the basis of these figures, there is no clear difference in the tax burden borne by BT and by Vtesse.

(149) Other ways of comparing the tax burden imposed on the different telecommunications operators have been suggested by OFCOM and the UK authorities. OFCOM suggested that the rateable value/connection ratio for BT, Kingston, NTL and Telewest should be compared. The ratio is very similar for these four companies. In fact, Kingston is valued at a slightly higher level than the other operators, which may be explained by the fact that, being significantly smaller than the others, it may not have the same economies of scale. OFCOM's suggested comparison is all the more relevant given the similarities between these four companies, as locally provided access network services are an important part of their overall business. It is true, however, that NTL and Telewest are valued on the basis of method derived from the contractor's basis method. Therefore, such a ratio is not directly pertinent for assessing the alleged discrepancy between the valuations carried out under the rental method and under the R&E method. Nevertheless, it shows that using different methodologies and comparing BT and Kingston with firms more directly comparable than Vtesse leads to similar results, which tends to corroborate the VOA's assertion that, in principle, all valuation methods should lead to the same result.

(150) As to the UK authorities, they indicated that, in the final year of the 2000 list, BT represented about 70 % of the total rateable value in the fixed-line telecommunications sector, and about 71,6 % of fixed-line call and access revenues according to OFCOM figures(43). The Commission can agree that a comparison of BT with all the Altnet operators taken together may provide more pertinent conclusions than a comparison of BT individually with small operators such as Vtesse. On this basis, it appears that BT pays a proportion of the total rateable value in the fixed-line telecommunications sector that is very similar to its share in the fixed-line telecommunications market. It must be recalled, however, that this is again a comparison between rates and revenues, the relevance of which is limited and at best only indicative.

(151) It must be noted that, instead of BT's market share of the total call and access revenues, Vtesse recommends that BT's market share of access revenues be taken into account. On that market, BT commands a share of over 80 %. In the view of Vtesse, this is further evidence that BT's hereditament is undervalued since BT only represents 70 % of the total rateable value in the fixed line telecommunication sector. The Commission cannot agree with this argument. First, BT's share in total access and call revenues seems a more appropriate benchmark for the purpose of assessing the burden of business rates: call revenues represent a larger share of BT's total revenues than access revenues(44). Therefore, call revenues have a very significant impact on the hypothetical rent and rateable value of BT's network and should be taken into account in any comparison between revenues and rateable values. Secondly, even if BT's share of access revenues only is to be used, the difference between that share (80 %) and BT's share in the total rateable value in the fixed-line telecommunications sector (70 %) at a specific moment in time is not such as to raise questions on the level of BT's rateable value, especially in view of the limited value of any comparison between rates and revenues.

(152) Vtesse suggests another method for proving the existence of an advantage to BT. It consists of pointing to the alleged discrepancy between the rateable values per unit of the different components (e.g. optic fibres, connections, buildings) of BT's network and those applied to other telecommunications operators. On that basis, Vtesse concludes that the R&E method has resulted in a lower valuation of BT's hereditament than would have been achieved under the rental method. Vtesse points out, for instance, that BT is rated at GBP 34 per square metre for buildings compared with GBP 115 per square metre in the case of other telecommunication operators, at GBP 74 per kilometre of optic fibre compared with the GBP 1 000/1 200 per optic fibre pair in the case of other operators, and GBP 19 per connection per year, compared with the annual fee of GBP 122 actually paid by operators for unbundled local loops.

(153) That approach is not satisfactory for a number of reasons. First of all, the UK authorities have convincingly argued that those figures, if they are correct, are not directly comparable. For reasons already stated in recital 125, BT's and Kingston's optic fibre networks are of a different nature from, and not comparable with, those of other operators. It is therefore not possible to apply rental evidence derived from other operators' optic fibre networks to BT or Kingston. In any case, the calculations and comparisons made by Vtesse are flawed.

(154) First, for an appropriate comparison, it would be necessary to take only BT's lit fibres into account since only lit fibres are rated. More importantly, Vtesse compared BT's average rateable value of km of optic fibre with the rental value of a fibre pair per route. In fact, the rental value of an optic fibre as applied by the VOA will vary according to the number of optic fibres per route. The more fibres per route, the lower their rental value will be(45). Thus, taking the lit fibre rent tone applied as of 2005, it appears that the rental value of two optic fibre (i.e. a single fibre pair) per route km for networks above 3 000 km(46) will be GBP 450(47). The rental value of three optic fibres per route km will be GBP 540, i.e. GBP 180 per optic fibre(48). At the lower end of the table, the rental value of 48 optic fibres per route km will be GBP 1 728, i.e. GBP 36 per optic fibre(49). The VOA claims that these values are market based(50).

(155) It is clear that BT's core network has, on average, dozens of lit fibres per route. It means that the rateable value of BT per km of optic fibre should be compared with the lower end of the rating list used by the VOA. On that basis, it appears that, contrary to what is claimed by Vtesse, a comparison between BT's rateable value per km of optic fibre and the rental values used by the VOA does not show any significant discrepancy or indicate an undervaluation of BT's network.

(156) Concerning the comparison between the LLU tariff and BT's rateable value per connection, it is noted, first, that, according to the UK authorities, the LLU tariff is not the same as the rent for the rateable asset, since a significant proportion of the tariff relates to system and maintenance costs, billing, and costs of using BT's non-rateable plant and machinery. It is therefore not directly comparable with the legal definition of the rateable value. Secondly, the UK authorities have indicated that some local loops will in fact have a high value, some a low value and some may even have a negative value. On the other hand, according to OFCOM, the provision of LLU services is aimed at stimulating competition in broadband services(51). Local Loop Unbundling is therefore also affected by the demand for broadband services and there is no demand for broadband services on a large number of BT's local loops, which are only required for telephony. Therefore, the LLU tariff does not reflect the average value of all BT's connections, and no conclusion can be drawn from a comparison of BT's rateable value per connection and the LLU tariff. As shown by OFCOM in its comments on the decision to initiate the procedure, this ratio is almost identical for BT, NTL and Telewest, despite the fact that a different valuation method was applied to the last two companies. For the reasons explained above, the ratio for NTL and Telewest appears to be more appropriate a benchmark than the LLU tariffs quoted by Vtesse, and the comparison between these ratios does not show any discrepancy between the different operators. It is further to be noted that, under its assessment, the VOA is considering applying a rateable value in the range of GBP 15 — GBP 20 per unbundled local loop(52), which again is in line with the valuation of BT's rateable value per connection.

(157) As to the alleged difference in rateable value per square metre for buildings, the Commission first notes that the correct figure derived from the Telereal transaction(53) is not GBP 35 per square metre as alleged by Vtesse but, applying a calculation similar to that of Vtesse, GBP 51 per square metre. Furthermore, this figure is compared by Vtesse with an alleged GBP 115 per square metre for other telecommunications operators. The basis and comparability of this figure is not clear and it does not reflect the variations that are likely to exist as a result, in particular, of the respective geographical locations of the telecommunication operators. The UK authorities have pointed out that BT's property portfolio covers the whole of the United Kingdom, including its most deprived and isolated areas, which has a downward effect on the average rental value of its properties.

(158) More importantly, it should be recalled that BT's general purpose buildings, such as offices, or warehouses, which are not related or linked to the network, are valued locally and separately, on the basis of the rental method. In 2000, their rateable value was GBP180 million. Operational and specialised buildings, such as telephone exchanges are considered to form part of BT's network and are centrally valued by the VOA using the R&E method. The rent paid by BT to Telereal following the transaction was GBP 280 million. The total rateable value of BT's buildings includes both the rateable value of BT's separately assessed buildings, which was approximately GBP 180 million in 2000, and the rateable value of the buildings included in the overall valuation of its network, which it is not possible to specifically identify but which represents a significant share of the GBP 560 million rateable value of BT's network in the United Kingdom. These figures do not show that there is a significant discrepancy between the actual market value of BT's buildings, reflected by the Telereal transaction(54), and their rateable value as calculated by the VOA(55). Therefore, these figures do not provide any evidence that this rateable value is undervalued.

(159) Vtesse put forward a secondary argument on the basis of the Telereal argument. Vtesse suggests that the rent paid by BT to Telereal for its buildings represents roughly 50 % of its rateable value in 2001, despite the fact that most of the rateable value should be mainly attributable to its network and not its buildings. From that, Vtesse seems to derive the conclusion that BT's network is obviously undervalued. This argument, however, is based on incorrect premises, since part of the property included in the Telereal deal are those that were separately valued at GBP 180 million. BT's overall rateable value for the whole of United Kingdom at the time of the transaction was therefore equal to GBP 560 million for the network and linked properties plus GBP 180 million for the separately valued buildings, which gives GBP 740 million. The rent of GBP 280 million paid by BT to Telereal amounts to 38 % of its overall rateable value. No useful conclusion can be drawn on the basis of this figure.

(160) In any case, a methodology that consists in separating parts of BT's hereditament and valuing them in isolation seems questionable. As confirmed by British case-law, the VOA must value the hereditaments of telecommunication operators as a whole. This rule is applied without discrimination to all operators. It is an inherent feature in any type of property valuation that the market rent of a whole property is usually different from the sum of the market rents of parts of that property: it is often lower(56), it may also be higher in certain circumstances. It is therefore likely that the market rent of BT's property portfolio in isolation will be different from its value when included in BT's hereditament.

(161) There is thus no clear evidence that the R&E method would generally result in a lower rateable value than the rental method, or that the VOA, when applying the R&E method to BT and Kingston, erred and undervalued the rateable value of their hereditaments.

6.2.2. Does the R&E method confer an advantage at the margin?

(162) It should be noted as a preliminary point that the question of marginal taxation, i.e. taxation on newly lit fibres, is relevant only as regards the rating assessment of BT but not as regards Kingston. As stated in recital 22, Kingston owns a subsidiary, Torch Communications Ltd, which operates a fibre backbone network outside Kingston's local access network and which is valued on the basis of the rental method. As a result, Torch Communications Ltd is fully taxed on newly lit fibre. The R&E method is applied only to that part of Kingston's network which is essentially a local access network. As already explained(57), its hereditament is not subject to a revision mechanism. That reflects the fact that, according to the UK authorities, no competitor has built networks in the Kingston upon Hull conurbation area and Kingston already had network coverage of almost 100 % in the area, so it did not extend its network. There was therefore no claim of material changes of circumstances which could have significantly altered the rateable value of Kingston's network. The Commission accepts this argument and notes that no third party has claimed that Kingston was benefiting from any advantage at the marginal level.

(163) In the case of BT, however, the fact that there is no evidence of the existence of an advantage at the level of the overall valuation of its hereditament does not necessarily exclude the possibility that marginal taxation, i.e. on the rating of newly lit fibres, does grant an advantage to this firm. In fact, the comments provided by several third parties indicate that the different way of rating BT and the other telecommunication operators when they light a new fibre may result in distortions of competition when these firms compete for the same contract. This problem raises the more general question of the revisions of rateable values that are conducted annually by the VOA.

(164) As explained above, when a telecoms operator to which the rental method is applied connects and lights a new fibre, its rateable value is increased by the rateable value of this new fibre. The opposite takes place when it switches off and disconnects a fibre: the fibres' rateable value is then deducted. Telecommunication operators that are assessed under the R&E method are not subject to this type of precise adjustment. In the case of BT, the annual revisions are based on a market adjustment mechanism: changes in BT's market shares are supposed to reflect changes in the value of its network. The question is therefore whether the different way of taking into account changes in the networks, such as the lighting of new fibres, is likely to confer an advantage on BT.

(165) It is the understanding of the Commission that, on the basis of the business plans provided by BT, the VOA assesses the hypothetical rent that can be expected from BT's hereditament on the basis of the anticipated receipts and expenditure deriving from the occupation of that hereditament. The valuation will therefore include any extension of the network, including the lighting of new optic fibres, which will be required to meet the anticipated demand. The lighting of new optic fibres, whether leased or actually owned by BT(58), is therefore anticipated and taken into account by the VOA. In fact, because of this valuation method, and unlike the telecommunications operators that are assessed under the rental method, BT is rated on dark fibres as well, on the assumption that some of these dark fibres will be lit in the course of the following five year valuation period. Finally, the annual reassessment based on the evaluation of BT's market shares is meant to capture the material changes to BT's network(59).

(166) Therefore, it is not correct to say that BT does not face any incremental tax liability, as suggested by the complainant or third parties such as Altnet. Certainly, with a network hereditament in England currently with a rateable value of GBP 530 000 000 which is valued as a whole, a small incremental change such as the additional lighting of a fibre is less likely to have a discernible impact on BT than on a small operator such as Vtesse. Furthermore, increases in value of BT's network are offset by factors resulting in a decrease of that value, such as the rolling out of competing capacity. This makes the impact of the lighting of new fibres by BT even less visible. That does not mean however, that these newly lit fibres are not rated.

(167) It is also alleged by Vtesse that, in a typical bid such as the one for the provision of backbone and access telecommunication links to Kent MAN(60), its rate liability represented 16 % of expected recurring revenues. On the same bid, it claims that BT was, at most, rated at 2 %, which is BT's alleged average rates/receipts ratio. Vtesse concludes that BT is much less taxed on a comparable and in this case identical transaction and therefore that it receives a clear advantage.

(168) That approach cannot be retained. First of all, it is not correct to assume that, since BT has an average rates/receipts ratio of 2 %(61), it was taxed at 2 % on that bid. That ratio is calculated on the basis of BT's overall rateable value, to which the value of the fibres lit by BT following the Kent MAN contract were added, either in anticipation or on the basis of the market share based revision mechanism. In order to make an appropriate and correct comparison with Vtesse's tax liability on that bid, it would be necessary to determine by how much the contract concerned increased the rateable value of BT's network, either in anticipation or via the annual revision mechanism. The anticipated valuation makes an overall assessment from which it is not possible to identify individual contracts. Therefore, it is not possible to determine by how much the Kent Man contract has increased the rateable value of BT's network. In any case, it would not lead to definitive conclusions concerning the possible existence of an advantage and therefore the existence of State aid. Vtesse's rate liability on this bid cannot be seen as a benchmark on the basis of which it is possible to assess the existence of an advantage to BT within the meaning of Article 87(1) EC. The tax liability is in fact likely to change depending on the technical solution chosen, the cost structure of the firms presenting a bid and the length of the optic fibre network operators will use to provide the services and, most importantly, will depend on the extent to which they will have to light new fibres(62). Thus, under the rental method, a telecommunications operator who provided the service by using an already lit fibre would not face any increase in rates(63).

(169) The same reasoning can be applied to the argument Vtesse based on a comparison between the prices of BT's Wholesale Extension Services (‘WES’) and the rates that an operator valued under the rental method would have to pay if it were lighting optic fibres for the purpose of providing the same service. Depending on the capacity and distances of the WES, Vtesse claims that such an operator could have to pay rates worth up to 87 % of the price charged by BT and sees that as further evidence of the fact that BT's optic network is undervalued or not properly taxed at the margin. However, the argument is based on the assumption that the hypothetical operator would have a single pair of lit fibre on the same route as the one chosen by BT. However, if this operator uses an already lit fibre in order to provide its WES, the extra costs in rates under the rental method will be nil. Furthermore, if he lights new fibres on a route on which he already has several lit optic fibres, than the marginal costs of the operation will be much lower since the more optic fibres per route, the lower the rates per fibre. Again, no conclusion can be drawn from this argument.

(170) In the light of these facts, the allegation that BT is not rated at the margin is not correct. Even if, as suggested by Vtesse, the unit of analysis is the taxation of contracts on which Vtesse and BT compete, it appears that there is no obvious advantage to BT: it is taxed on newly lit fibres, albeit differently and less transparently. The Commission also notes that, because the R&E method is based on anticipated revenues, BT is rated on dark fibres on the assumption that these fibres will have to be lit to meet the anticipated demand. Therefore, unlike other telecommunications operators which are rated only once they light a fibre, BT is also rated on dark fibres.

(171) The Commission can therefore concur with OFCOM when it says that ‘the system [of revisions] would appear to provide a materially symmetrical outcome in the long-term in each case, albeit through different routes and with low levels of transparency.’

(172) The absence of any obvious competition issue is further confirmed when the market shares on the market for the provision of high bandwidth services to corporate customers are considered, on which BT is in direct competition with Vtesse(64). According to figures provided by BT, BT's share on that market is 12 %, and it is only the fourth largest player. According to figures provided by OFCOM, BT has [10-15] %(65) of the market in the provision of traditional interface leased lines of 622 Mbit/s and above(66). Despite being the incumbent in the telecommunications sector, it therefore appears that BT had a relatively low market share in this specific market. It could be assumed that, if BT was indeed enjoying such a substantial advantage in the form of an exemption from marginal taxation on new fibres, its market share would be significantly higher. Therefore, data available to the Commission do not confirm the allegation that BT benefits from a competitive advantage which would make it very difficult for other telecommunications operators to compete with it on this specific market(67).

(173) As a final point on the question of revisions, the Commission also raised the concern in its decision to initiate the procedure that ‘while BT does enjoy a downward revision mechanism, there does not seem to be a similar systematic review of market conditions faced by competitors’. In view of the explanations provided by the UK authorities and BT, the Commission is satisfied that this concern is unfounded. BT is not subject to a downward revision mechanism only: as already explained, the revision mechanism applied to BT is based on the evolution of its market shares. It is true that for a certain number of years, its market shares, and therefore its rateable value, decreased, which reflected a loss in value due to increased competition. In 2003 and 2004, however, its rateable value increased by GBP 4 million and GBP 7,1 million respectively. As to competitors, they can, and have indeed, benefited from downward revisions: for instance, they received a 15 % discount on the rateable value of their optic fibre networks as of 1 April 2001 and a 25 % discount as of 1 April 2002 in order to reflect oversupply in that sector. As a result, the rateable value of a fibre pair fell from GBP 1 000 per km outside the London area in 2000 to GBP 850 per km in 2001, and GBP 750 between 2002 and 2004. Under the 2005-2010 list, the rateable value of km of fibre pair has further fallen to GBP 500. It means that an operator whose optic fibre network remained stable over that period would have enjoyed a 50 % reduction of its rateable value.

CONCLUSION

(174) In conclusion, it should be recalled that business rates are a tax on the value of the property concerned. They are not a tax on profits or revenues. They are normally applied on all non-domestic properties, and consequently are applied to all telecommunications networks. According to British case-law, all telecommunications networks are valued as a whole. There are several methods for valuing such property. When all methods can be applied, they should result in the same valuation. The use of a specific valuation method depends on the circumstances of the case.

(175) It now appears that the VOA has applied to BT and Kingston the general rules concerning business rates as laid down in the legislation and case-law. It is clear that the valuation of BT's and Kingston's hereditaments as well as the revisions of these rateable values, are carried out on the basis of a different method than in the case of their competitors. However, the Commission can conclude that there is no evidence that the use of this different method is not justified by the objective differences between those firms and their competitors and by the extent of the evidence available to the VOA.

(176) There is no evidence that the application of a different valuation method to BT and to Kingston has resulted in an advantage to these firms in comparison with their competitors. Since there is no evidence of an advantage, the Commission can conclude that the non-domestic rates system has not provided State aid to BT and/or Kingston within the meaning of Article 87(1) EC during the period considered by the Commission i.e. 1995-2005.

HAS ADOPTED THIS DECISION:

(1)

OJ C 62,12.3.2005, p. 10.

(2)

See note 1.

(3)

Available at http://www.opsi.gov.uk/acts/acts1988/Ukpga_19880041_en_1.htm

(4)

S.I. 2000/540.

(5)

The 2005-2010 rental value of optic fibre set by the VOA is the object of an appeal, and could be lowered if the appeal is successful.

(6)

They are actually valued on a rateable value per home connected. This method is based on a contractor's valuation involving decapitalised costs with end allowances to reflect penetration rates.

(7)

GBP 554 100 000 in 2005 for England and Wales.

(8)

The local loop is the last mile connection linking the final customers to the local telephone exchange. BT has to provide access to these local loops to its competitors, at regulated conditions.

(9)

Comments from Cable and Wireless, Telewest and Viatel.

(10)

Comments from Telewest.

(11)

Comments from Easynet.

(12)

The first local loop unbundling rental tariff applied from 31 May 2001, considerably after the relevant antecedent valuation dates of 1 April 1993 and 1 April 1998.

(13)

The first set of regulatory accounts that reflected transfer charges for BT's access services were produced for the year 2001/02.

(14)

The details of this market-share based mechanism are explained in the description of the comments from the UK Authorities.

(15)

Thus, in 2003, BT's rateable value in England was increased from GBP 443,5 million to GBP 447,5 million. In 2004, it was increased in England to GBP 450,6 million.

(16)

Kingston' hereditament in the Hull conurbation area which was assessed under the R&E method was valued at GBP 5,1 million in the 2000 list. Kingston claims that the turnover derived from that area was approximately GBP 50 million. Applying the appropriate multiplier of 44 %, this would result in a revenue/rate ratio of about 4,5 %.

(17)

These included specialist consultants from National Economic Research Associates (NERA), who were commissioned by the VOA to prepare a model providing an estimate of BT's future revenues and operating costs.

(18)

British Telecommunications plc v Central Valuation Officer, 19 March 1998.

(19)

The UK authorities quote Poplar Union Assessment Committee v Roberts [1922] 2 AC 93

(20)

See for instance Brewin (VO) v Railway Executive (1952) 45 R&IT 553

(21)

The Receipts and Expenditure Method of Valuation for Non-Domestic Rating, A Guidance Note — July 1997. Published by the Royal Institution of Chartered Surveyors (RICS).

(22)

It must be borne in mind that, although the hypothetical tenancy must be from year to year, it must not be assumed that it will last only for a year. On the contrary, it must be assumed that there is a reasonable prospect of continuance. See R v. South Staffordshire Waterworks Co (1865) 16 QBD 359 at 370.

(23)

12,27 % in real terms.

(24)

For instance, for the valuations of the 2000-2005 rating lists, there were more than 1,17 million appeals, of which about 845 000 were withdrawn or have been settled.

(25)

See note 18.

(26)

The appeal to the Lands Tribunal involves a complete de novo hearing over a period of 13 weeks before a panel of two surveyors members and a QC. In contrast, the Valuation Tribunal featured a lay panel and took a mere four weeks.

(27)

BT's core retail telecommunications services were subject to regulatory price control. In a given year, the price change for BT's regulated retail services was not allowed to exceed RPI-X, where RPI is the annual percentage change in the whole economy retail price index. The X factor was set by OFCOM. In 1992, OFCOM had determined that the value of the X factor would be 7,5 % in each year from 1993/4 to 1996/7.

(28)

BT's access deficit is the loss it made in providing exchange lines.

(29)

The UK authorities had regard to evidence from studies indicating that BT is efficient. For example, when compared with US local exchange carriers, which are generally regarded as being reasonably efficient and for whom detailed cost and volume data is available, BT is at or close to the upper decile in terms of efficiency.

(30)

The matters that can be considered as material changes of circumstances are described in paragraph 2(7) of Schedule 6. These matters include in particular ‘matters affecting the physical state or physical enjoyment of the hereditament’, and ‘matters affecting the physical state of the locality in which the hereditament is situated’.

(31)

Using a revised market share for BT in 2000, following fresh data provided by OFCOM. This market share was 70,5 %.

(32)

[…]. The VOA is contending that the loss of traffic to mobile telephony is not a material change within the 2000 rating list.

(33)

Since BT's position in the market has changed, OFCOM no longer has the same competence as before to gather relevant market data.

(34)

It means that the forecast will be that which would have been anticipated looking forward from 1 April 2003, but on the assumption that the physical environment (including the physical extent of the network of both BT and other operators) was in the same state as it actually was at the commencement of the year in which the revision is made.

(35)

The UK authorities note that, in its submission, Altnet suggested that an R&E valuation for their networks for the 2000 list would have yielded a nil valuation. The UK authorities responded that this is inconsistent with the rents actually paid and with the business plans upon which those actual rents must have been justified.

(36)

The level of receipts used in these valuations is a figure adjusted to the United Kingdom, which takes out non-relevant revenues.

(37)

See the explanations provided by the UK authorities, described at recitals 70 to 72.,

(38)

British Telecommunications PLC v Central Valuation Officer, 19 March 1998.

(39)

See note 21.

(40)

The rateable values of the telecommunications operators valued under the rental method do not show the same fluctuations. It must, however, be stated that any comparison is likely to be misleading since their networks are rapidly expanding ones, while BT's and Kingston's networks are relatively stable. In any case, those telecommunications operators also enjoyed significant decreases in their rateable values, as explained in recital 173.

(41)

This explains why, contrary to the argument brought forward by BT, the VOA does not accept that BT's rateable value in 2003 and 2004 should be affected by the growing popularity of mobile telephony. The VOA considers that this is an economic change which took place in the life of the 2000-2005 list, and cannot be seen as a material change in circumstance to be taken into account as if taking place in the economic environment of 1998. Similarly, the VOA considers that it would not be correct to have regard to the opposite, upward effect of subsequent extra demand from internet access not predicted in 1998. Both economic phenomena could be taken into account only in the context of the elaboration of the 2005-2010 list.

(42)

This point is sufficient to rebut the suggestion, made by Gamma Telecom in its comments, that the respective burden of the business rates should be assessed on the basis of a comparison between the rates/added value ratio of BT and its competitors.

(43)

Available at http://www.ofcom.org.uk/research/cm/nov05/

(44)

In 2004, call revenues represented 59,5 % of BT's call and access revenues.

(45)

It reflects the general principle that the rent per square metre of a large building rented as a whole will be lower than in the case of a small building. OFCOM confirmed in its comments that, on the basis of the available information, rental prices for fibres are lower where more fibres are rented.

(46)

As explained above, networks above 3 000 km benefit from a 10 % reduction, which is based on market evidence. This is the proper benchmark to make a comparison with BT's network given its size.

(47)

GBP 500 in the London area.

(48)

GBP 720 and GBP 240 respectively in the London area.

(49)

GBP 2 256 and GBP 47 respectively in the London area.

(50)

It is the Commission's understanding that the VOA has no direct evidence above 12 fibres and therefore, the top of the scale is based on decapitalised costs, i.e. on the assumption that an operator would set its maximum rental bid a network with 13 fibres and above at no more than the decapitalised costs of building the network itself.

(51)

OFCOM consultation document, ‘Local loop unbundling: setting the fully unbundled rental charge ceiling and minor amendment to SMP conditions’, 7 September 2005.

(52)

In their submissions, GVA Grimley indicated that the VOA was of the view that a pair of unbundled copper wires was worth circa GBP 50 rateable value. In fact, at the time, and at a very preliminary stage in the analysis, the VOA suggested that this rateable value could be in the range of GBP 12 to GBP 50.

(53)

See recital 36.

(54)

The Telereal transaction covered almost the entirety of BT's portfolio: only 220 out of 6 700 properties remained in BT's ownership.

(55)

Account should also be taken of the fact that the actual rental value of a property should not necessarily be equated with its hypothetical rateable value.

(56)

Market evidence usually reflects the fact that the larger the property, the smaller the unit price. This principle is taken into account by the VOA, since, on basis of the market evidence, it grants a 10 % discount to networks over 3 000 fibre kilometres in length.

(57)

See recital 98.

(58)

Vtesse raised the point that BT has allegedly rented some 2 000 km of optic fibres from GEO, one of Vtesse's suppliers and inquired whether this extension of its network had been properly assessed. The VOA has obtained confirmation from GEO that it had rented 1 717 route km of fibres to BT. It is not clear at this stage whether the fibres leased by BT have been lit, if they are replacement fibres for obsolete BT fibres or if they are new route. Concerning the way these fibres will be taken into account in the review of BT's rateable value, the VOA explained that the existing R&E valuation is based on a model which anticipates future income streams and it will, as part of the review, take into account the income stream expected from the fibres which BT has leased from GEO. This review will be carried out on the basis on the new revision mechanism, which is still being finalised.

(59)

The UK authorities have confirmed that the revenues from broadband services were taken into account in the annual revision method applied to BT and based on market shares.

(60)

See recital 41.

(61)

As already explained the precise ratio for BT varies according to the assessments and varies from 1,28 % (OFCOM) and 2 % (Vtesse) and 3 % (UK authorities). The discrepancy between the figures provided by Vtesse and the UK authorities has been explained at recital 107. The lower figure obtained by OFCOM may be explained by the fact that it used the total turnover of the BT group, including the turnover of its activities outside the UK.

(62)

It is technically feasible to use already lit fibres to supply high bandwidth services to corporate customers. In that case, under the rental method of valuation, there will be no increase in the rateable value of a telecommunications operator's network.

(63)

The operator may have to pay rates only on the spurs linking the fibre to the customer's premises.

(64)

Vtesse describes itself as a carrier providing innovative very high capacity services (up to 66 Gigabits per second).

(65)

Confidential information.

(66)

For the lower bandwidth segment, it is also useful to point out that OFCOM considers that BT does not have significant market power in the retail market of leased lines of 8 Mbit/s and above. See http://www.ofcom.org.uk/consult/condocs/llmr/

(67)

Vtesse actually won a large contract against BT, in 2004, for the provision of high bandwidth services to Lloyds TSB.