Identification of Significant Market Power for the purposes of the Interconnection Directive

II Background

The SMP test

4 The Interconnection Directive (ICD) creates a framework in which all operators who need to interconnect their networks in order to serve end-user customers can do so. In order to be able to offer a telecoms service, an operator needs to ensure that its customers can contact customers of other networks. As competition develops within markets, there is asymmetry, particularly of economic strength, between an incumbent operator with a ubiquitous telecoms network and a new entrant. This is addressed in the Directive by the imposition on operators, who it is determined have SMP, of a set of obligations designed to provide transparency, non-discrimination and cost-orientation.These obligations are intended to prevent them from setting interconnection terms which are unfair or otherwise have an anti-competitive effect.

5 An SMP designation reflects the position of an operator in the sector in which it operates. It is not a commentary on whether or not that operator has acted anti-competitively or whether it is anticipated that it will do so. The designation is intended to ensure that established operators do not unfairly prevent or hinder competitors from becoming established in the market. This ensures that consumers are not disadvantaged by electing to receive service from a new market entrant in preference to one with an existing customer base, where the former may have less commercial power than the latter and the latter can exert proportionately greater control over access to other telecoms users.

6 Rules to control anti-competitive behaviour exist independently of the rules applied to those having SMP. Thus an operator without SMP status might still be found to have a degree of market power in a specific product market and might try to use this power in an anti-competitive way. Oftel will continue to have the ability to act against licensees’ anti-competitive behaviour in particular markets including, but not limited to, abuse of a dominant position in a particular relevant market, regardless of whether or not that operator is deemed to have SMP under the ICD. Likewise, the fact that an operator has SMP does not necessarily mean that they have market power in every product market within the sector.

7 Further, a determination that an operator has SMP for the purposes of the ICD is without prejudice to any decision about SMP under another Directive. It is possible that different operators might be determined to have SMP under the other EC Directives which include an SMP regime. The SMP test is not identical in all material respects in different Directives. Also, the exercise of NRA discretion may give different weight to the factors in the test, according to the objectives of the Directive.

8 The definition of SMP in the Interconnection Directive is set out in Annex A. The trigger for consideration of SMP is set initially in terms of market share. A share of 25% of activity in the relevant sector gives rise to a presumption of SMP. This can be supported or rebutted by taking into account the following factors:

• market power (judged by an operator’s ability to influence market conditions)
• turnover relative to the size of the market
• control of the means of access to end-users
• access to financial resources; and
• experience in providing products and services in the market

Oftel’s view is that the relevant product sectors for the purposes of identifying SMP are the fixed sector, the leased lines sector and the mobile sector as set out in Annex I of the Directive. The relevant geographical market is defined by the limits of the licensed area in which the relevant organisation is authorised to operate.

9 Article 7(2) of the Directive requires the NRA to consider whether mobile operators have SMP on the ‘national market for interconnection’. If they do, they are required to offer cost-orientated interconnect charges.

The SMP determinations

10 The Interconnection Directive and the rules applying to those who have SMP are a part of EC law. It is the Member State’s responsibility to implement EC law correctly. The European Telecommunications Directives create a framework of rules, but leave Member States with discretion as to how they should be applied.

11 It is important to understand the intended relationship between SMP and other tests of market power, such as dominance (particularly given that much of the UK regulatory regime for telecoms is designed around tests aimed at the identification or prevention of abuse of market power). It is clear from the definition set out in the Directive itself that dominance and SMP are not the same (see paragraph 20). They are complementary but different provisions which Member States can use to adapt regulation to market needs.

12 In considering its SMP determinations, Oftel took account of the views of the ONP Committee, which assists the Commission on interconnection matters under Article 15 of the Interconnection Directive. Oftel’s proposals are consistent with the views put forward by the ONP Committee, particularly that the relevant sectors to be considered for the purposes of identifying SMP are the fixed, leased lines and mobile sectors. (See paragraph 2.1 Relevant markets’ of ONPCOM97-41 Issues related to the EP and Council Directive 97/33/EC on Interconnection: Significant Market Power – working document for discussion (December 1997)

III Responses to consultation

BT

13 BT said that the process for determining whether an operator has SMP should depend on an analysis of individual product markets and the positions of different players in them. It took the view that the Interconnection Directive and the SMP test required a full assessment of the existence and extent of market power in the UK, looking at all the individual product markets and did not agree that the Directive required the use of the three broad sectors described in Annex 1. BT commented on the lack of evidence that the markets and the positions of the different players in them had been analysed. It commented on the lack of objective evidence to support Oftel’s conclusion that Cable & Wireless Communications (CWC) had no market power in any of its inland markets. BT disagreed that the relevant geographic market in the case of CWC had to be the national market – arguing that although CWC are nationally licensed, the geographic limits of their market consist of several individual product markets. It argued that if the ability to raise prices above competitive level was the test, and if BT was said to have that power, then CWC and other cable companies also have it in relation at least to charges for terminating calls. It also commented that no analysis had been published to support Oftel’s view that BT had such power in the fixed and leased lines markets in the UK and said that in the competitive market which exists across all sectors it did not believe it to be the case. BT did not agree that the analogue and digital mobile markets should be treated as one market, or that market share should be measured across all leased circuits, analogue and digital, national and international. BT suggested that other ways of analysing market power, such as on the basis of net growth, would provide a more accurate view of the current position in the markets than looking only at a total share. BT did not agree that there were no other “well established” operators, that newer market entrants had access only to limited financial resources, or that companies such as CWC could be considered “inexperienced”.

NTL

14 NTL also sought to align investigations of SMP with the methodological approach used to test for dominance in competition cases, and showed particular concern about regulatory uncertainty if the SMP test were allowed to differ from the usual approach to testing for dominance. Their response considered that the prime focus of implementation should be the need to decide whether market power exists, and if so how to contain it.

Kingston

15 Kingston supported Oftel’s interpretation of the objectives of the Directive but questioned the appropriateness or desirability of designating Kingston as an SMP operator. In particular, Kingston endorsed the distinction between the considerations of an SMP determination and the economic analysis undertaken in a competition or fair trading investigation. However, because they do not themselves provide the facilities for long distance or international calls, they felt that their market share of the Hull area by value was much less than 100%. In addition they disagreed with some of Oftel’s conclusions on the extent to which they satisfied some of the other criteria in the SMP test, particularly in relation to market power. Whilst recognising the effectiveness of the SMP framework in Member States where competition needs to develop quickly, they questioned its efficacy or suitability in the already competitive UK markets. Kingston believed that they were not treated in the same way as other regional operators with whom they should be compared, but rather that Kingston’s position in the Hull area market was being treated as equivalent to BT’s position in the UK as a whole. Kingston wanted the SMP analysis to reflect national, as opposed to local, market conditions. On this basis, they believed they could not be said to have SMP.

Cellnet

16 Cellnet believe that none of the 4 UK mobile network operators should be deemed to have SMP, due to the extent of competition in the UK mobile market. They sought clarification of why market share was measured in revenues and not another basis, such as traffic levels or net connections – which give demonstrably quite different results. They argued strongly that they do not have market power, given the need to respond to the commercial initiatives of their competitors in the mobile market. They suggested that the factors listed in Article 4(3) of the Directive were not exhaustive and that other matters could also be taken into account. Cellnet suggested that the analogue and digital mobile markets should be treated as separate markets for the purpose of analysis of market power, since the Directive does not preclude, and, in Cellnet’s view, positively allows for, the three product markets referred to being divided into sub-markets. Cellnet therefore took the view that it was appropriate that Oftel’s earlier decision that the analogue and digital mobile markets should be treated as a single market, was reviewed.

Vodafone

17 Vodafone, on the other hand, took the view that all 4 mobile operators should be regarded as being operators with SMP. This was on the grounds that every operator with end-user customers has bottleneck control over call termination and as such the NRA must have powers to control their behaviour. Vodafone were concerned that failure to designate an operator with bottleneck control as having SMP could provide opportunities for that operator to discriminate in providing interconnect and setting charging levels. They argued that call termination, not retail services, should be viewed as the relevant market. Such an approach would give due weight to an operator’s ability to control access to customers, one of the criteria in the SMP test. Vodafone took the view that undue reliance on the 25% threshold as a test for SMP would not give a sufficiently complete and balanced view of the test and the supporting conditions.

Cable & Wireless Communications

18 CWC argued that the relevant geographic market should depend on the geographic coverage of the combined licences of an organisation, looked at as a single legal entity. Treating each individual licensee as an ‘organisation’ for the purposes of identifying SMP could, they said, give rise to confusion, particularly where operators hold multiple, overlapping licences. CWC questioned the future suitability of treating the fixed and mobile markets as separate – convergence between fixed and mobile markets could lead to a distorted picture of market power should this distinction be strictly followed. They also commented that SMP as a tool may be suitable for regulating former incumbents in newly liberalised environments, but is likely to be less appropriate for the regulation of operators with market power in dynamic competitive markets. They pointed out the importance of the availability of alternative, more flexible, competition based controls for regulating individual markets.

Telewest

19 Telewest supported geographic definition on a licence by licence basis but supported the use of revenues, not number of exchange lines for measuring regional market share. On this basis, they disagreed that CWC’s market share should be assessed on a national basis.

Other

20 The Communication Workers Union sought the collection and publication of regional information on a systematic basis so that interested parties could judge for themselves the extent of developing regional competition. Reuters argued that Cellnet and Vodafone had sufficient shared power in the national interconnection market to justify the imposition upon them of cost-orientated interconnect charges. The Consumers Association encouraged flexibility in the consideration of the relevant geographic market, to take account of the uneven development of network competition in the UK, and to consider additional factors when assessing market power, such as the level of sunk costs, the degree of excess capacity available to operators and any evidence of parallel pricing by alternative operators . They also encouraged Oftel to use its discretion to determine that operators with less than 25% market share are SMP operators.

IV Oftel position on the main issues which arose in consultation

SMP and dominance

21 Oftel sought to explain the relationship between SMP and dominance in the consultative document. Dominance is a familiar economic concept and the procedure for identifying it in telecoms markets has been widely discussed with the industry. Dominance has been defined by the European Court as an operator’s ability to act independently of competitors and customers. It depends on the position of an operator in an economic product market. Many respondents sought to place SMP in the same framework. However, this is not the approach of the ICD. SMP under the ICD is not related to specific product markets as defined under EC competition law, but is considered within the three broad sectors – fixed, leased lines and mobile – specified by the Directive. It is true that both are tests which analyse an operator’s position in markets in which it operates and which is intended to deal with the ability of firms with market power to distort the fairness of market conditions. But even at such a general level, they are not the same – dominance for EU purposes is generally considered when an operator has a market share above 40%, yet the presumption of the existence of SMP arises where a market share is around 25%.

Using the SMP framework and controlling anti-competitive behaviour

22 The SMP test is not a basis for regulating anti-competitive behaviour. It is a framework for the application of specific rules with which an operator determined to have SMP must comply, in order to ensure fair terms for interconnection for other operators and their customers. Oftel will only use the SMP test to identify operators on whom the SMP obligations in the ICD should be imposed. Fair trading rules exist for the control of anti-competitive behaviour. The two systems – a regulatory framework for interconnection combined with controls on anti-competitive behaviour – are complementary and serve different functions. The determinations will not affect Oftel’s general approach to regulating anti-competitive behaviour under licence conditions.

23 SMP, in the context of the ICD, is intended to take into account the interdependence of operators in a networked industry, and gives recognition to the importance of the position of an established operator in the essential network function of interconnection. It ensures fair conditions for other operators who are inevitably heavily dependent on getting fair interconnection from that operator if they are to develop and run their businesses. On the other hand, sanctions for anti-competitive behaviour can still be applied to other operators not designated as having SMP, in specific economically defined product markets, where that operator has market power or is dominant. The relevant economic market in such cases may be relatively small in comparison with the whole fixed, leased lines or mobile sector, or not particularly significant in the overall scheme of interconnection between networks.

Defining an organisation

24 The statutory instrument implementing the ICD proceeds on the basis that the term ‘organisation having SMP’ means an individual licensee (as opposed to a wider definition of ‘organisation’ to include all subsidiaries and other group companies, equivalent to an ‘undertaking’ in competition law). This approach is in line with the way the UK has implemented EC telecoms legislation to date.

Defining the relevant market for the purposes of the ICD

25 The Directive does not set out how the markets in which operators have SMP should be defined except where Annex 1 of the Directive describes the networks and services considered to be of major importance at EU level. It is Oftel’s view that the three ‘sectors’ set out in Annex 1 should be treated as the relevant markets, each taken individually. This approach has been endorsed by the Commission. It allows the Director General to continue to exercise his general competition powers as before, by looking at specific product markets within each of the three sectors and analysing the level of market power in each of the 3 sectors. (The criteria which Oftel use for defining markets for the purpose of enforcing competition rules have been set out elsewhere – see Oftel’s statement, Effective Competition Review – market definition and competition analysis published in February 1998).

26 Moreover, the obligations of the SMP framework are likely to be most effective on a broad sectoral level. If SMP were equated with dominance, for example, and narrowed down to specific product markets, SMP obligations could probably only be imposed on an operator’s activities in that specific market where SMP was determined to exist. But if accounting separation were imposed on a single product market, it is not obvious how it would effectively serve its purpose, such as allowing identification of cross subsidy or other forms of unfair pricing behaviour between that and related or separate markets.

Relationship between market share and market power

27 Several respondents questioned the use of the 25% market share test as the trigger for considering the existence of SMP, in favour of a more sophisticated economic analysis based on the existence of dominance in a relevant market. Given that SMP is not the same as dominance, Oftel used a 25% market share trigger because that is what is required by the Directive. Having a market share above that level does not of itself mean that an operator has SMP, but raises a presumption to be considered in the light of other factors, particularly those set out in the test.

The measure used to calculate market share

28 Some respondents suggested alternative measures for the calculation of market share for applying the SMP test or questioned Oftel’s choice of retail revenues as the most appropriate measure. The ICD does not specify what measure to use for the calculation of market share. The key requirement for a measure is that it reflects the relative market positions of all the players in the market. The choice of measure also took into account the need for transparency and the availability of information. While Oftel collects a great deal of information about the telecommunications market, the value of the information collected needs to outweigh the regulatory burden on the industry of providing it. Retail revenues was chosen as the measure because it is transparent, is broadly based and is not distorted by choices made by individual operators on the packaging of their tariffs. Importantly, the potentially distorting effects of bundling of services are avoided by looking at total revenues.

29 Some respondents suggested billed minutes as a more appropriate measure. In industries where the product is relatively homogeneous, volumes do not differ significantly from value measures as an appropriate measure for the calculation of market share. However, telecoms services to end-users are far from homogeneous, there is a whole range of different prices for different types of call and a variety of tariff packages, including ones which bundle call minutes. This means that the use of billed minutes on its own could be misleading. For instance, some operators have tariff packages which offer certain types of call at no additional cost to the subscriber. Instead, part of the monthly or quarterly payment covers the cost of these bundled calls (eg ‘free’ weekend local calls or ‘free’ local calls between subscribers to the same cable company). These tariff packages often generate significant volumes of traffic, but might not generate revenue and thus would not necessarily be a reflection of the operator’s position in the market. In view of these distortions it was felt that call minutes was not an appropriate measure at least for the purposes of determining SMP for the ICD.

30 Other respondents suggested that some measure relating to call termination would be the most appropriate as it is directly related to the intention of the ICD – ie regulating interconnection. However, the SMP test is a test of an operator’s position in the appropriate sector as a whole. Call termination is a specific interconnection market, in which arguably every operator with directly connected customers has market power in relation to termination of calls to its own customers. This could produce the result that every operator with end customers could be deemed to have SMP, which is clearly not intended by the ICD.

Measuring the market shares of regional operators

31 A number of respondents questioned the use of a different indicator, the number of exchange lines, for regional licensees. Oftel accepts that a market share calculation based on exchange lines is not directly comparable to the retail revenue measure used at a national level. For reasons of consistency it would obviously be preferable to use the same measure to calculate market share for all operators, including those operating on a regional basis. Currently Oftel does not, however, collect revenue figures for regional geographic areas. Past investigations of the practicality of collecting this type of information have thrown up a number of problems. These include assignment of revenues to individual premises for multi-site customers and the difficulty of collecting information from the full range of operators on a consistent set of relevant geographies. In Oftel’s view the difficulty and cost to the industry of compiling this information would be disproportionate to the benefits of having it (but Oftel would be interested in the views of operators who would be called on to provide it).

V Next steps

32 The SMP test and its interpretation, continue to be discussed by the ONP committee. Oftel will continue to contribute to these discussions, taking account of views expressed by operators and users.

ANNEX

ARTICLE 4(3)

An organization shall be presumed to have significant market power when it has a share of more than 25% of a particular telecommunications market in the geographical area in a Member State which it is authorized to operate.

National regulatory authorities may nevertheless determine that an organization with a market share of less than 25% in the relevant market has significant market power. They may also determine that an organization with a market share of more than 25% in the relevant market does not have significant market power. In either case, the determination shall take into account the organization’s ability to influence market conditions, its turnover relative to the size of the market, its control of the means of access to end-users, its access to financial resources and its experience in providing products and services in the market.

ICD ANNEX I

SPECIFIC PUBLIC TELECOMMUNICATIONS NETWORKS AND PUBLICLY AVAILABLE TELECOMMUNICATIONS SERVICES (referred to in Article 3 (2))

The following public telecommunications networks and publicly available telecommunications services are considered of major importance at European level.

Organisations providing the public telecommunications networks and/or publicly available services identified below which have significant market power are subject to specific obligations with regard to interconnection and access, as specified in Articles 4(2), 6 and 7.

Part 1

The fixed public telephone network.

The fixed public telephone network means the public switched telecommunications network which supports the transfer between network termination points at fixed locations of speech and 3, 1 kHz bandwidth audio information, to support inter alia:

• voice telephony;

• facsimile group III communications, in accordance with ITU-T Recommendations in the “T-Series”;

• voice band data transmission vis modems at a rate of at least 2 400 bit/s, in accordance with ITU-T Recommendations in the “V-series”.

Access to the end-user’s network termination point is via a number or numbers in the national numbering plan.

The fixed public telephone service according to Directive 95/62/ec of the European Parliament and of the Council of 13 December 1995 on the application of open network provision (ONP) to voice telephony (1).

The fixed public telephone services means the provision to end-users at fixed locations of a service for the originating and receiving of national and international calls, and may include access to emergency (112) services, the provision of operator assistance, directory services, provision of public pay phones, provision of service under special terms and/or provision of special facilities for customers with disabilities or with special social needs.

Access to the end-user is via a number or numbers in the national numbering plan.

Part 2

The Leased lines service

Leased lines means the telecommunications facilities which provide for transparent transmission capacity between network termination points, and which do not include on-demand switching (switching functions which the user can control as part of the leased line provision). They may include systems which allow flexible use of the leased line bandwidth, including certain routing and management capabilities.

Part 3

Public mobile telephone networks

A public mobile telephony network is a public telephone network where the network termination points are not at fixed locations.

Public mobile telephone services

A public mobile telephone service is a telephony service whose provision consist, wholly or partly, in the establishment of radiocommunication to one mobile user, and makes use wholly or partly of a public mobile telephone network.