The Scope of the Commission Three-Criteria Test: A System Fault?

01 November 2006

Peter Eijsvoogel and Bas Braeken, Allen & Overy LLP

The recommendation constitutes the point of departure for the market analyses that the Framework Directive requires the national regulatory authorities (NRAs) to carry out. These analyses must be performed on the basis of general competition law principles and must – obviously – be based on the particular circumstances of the market concerned. We have found, however, that in its review of the NRAs’ draft decisions the commission adheres very closely to the recommendation. Thus far, the commission has granted limited scope for NRAs to deviate from the recommendation, which in some cases has proven to be nothing less than a straitjacket for national regulators. Also, with the notable exception of access and call origination on mobile networks, in certain instances NRAs that wanted to deregulate markets mentioned in the recommendations have found the commission in their path.

The legal basis for the recommendation is provided for in article 15 of the Framework Directive, which stipulates that the recommendation identifies those product and service markets within the electronic communications sector, “the characteristics of which may be such as to justify the imposition of regulatory obligations.” In paragraph 9 of the preamble to the recommendation the commission has developed the following test for the inclusion of markets in the recommendation:

In identifying markets in accordance with competition law principles, recourse should be had to the following three criteria. The first criterion is the presence of high and nontransitory entry barriers whether of structural, legal or regulatory nature. However, given the dynamic character and functioning of electronic communications markets, possibilities to overcome barriers within a relevant time horizon have also to be taken into consideration when carrying out a prospective analysis to identify the relevant markets for possible ex ante regulation. Therefore the second criterion admits only those markets the structure of which does not tend towards effective competition within the relevant time horizon. The application of this criterion involves examining the state of competition behind the barriers of entry. The third criterion is that application of competition law alone would not adequately address the market failure(s) concerned.”

Especially in dynamic markets, where technological developments (eg,VoIP) could rapidly undermine the competitive position of incumbent operators and where ascertaining the future durability of a dominant position will be very difficult, ex ante regulation should be a remedy of last resort. The presence of a dominant position at the time of the market analysis is an insufficient basis for the imposition of ex ante regulation. This is in line with general competition law, where the sole presence of a dominant position is not prohibited. Also, a decision to abstain from imposing ex ante remedies does not affect the NCAs’ powers to act upon any abuse of a dominant position (cf, article 82 of the EC Treaty). Therefore, we believe that the fulfilment of the three-criteria test should indeed always be a prerequisite for the imposition of ex ante regulation.

The proper application of the test requires a thorough research and analysis of the relevant market and its specific characteristics, including its geographic dimension. We believe that the analysis necessary for a proper application of the test should be nothing short of what the commission (in its guidelines) requires for the market definition and analysis. We find it hard to see why the commission only requires NRAs to (explicitly) apply the three-criteria test if they define markets that have not been defined in exactly the same manner in the recommendation. Apparently, the reasoning is that for the markets which are included in the recommendation, the test has already been performed by the commission itself when it drafted and adopted the recommendation. Nevertheless, how can the commission, when drafting, conclude that there are high and nontransitory entry barriers in a particular market in each member state without conducting an in-depth analysis of the specific market (structure), including its geographic scope? It is clear that the commission has not made an assessment of the geographic scope of the relevant markets. But the commission did define the product scope of the markets included in the recommendation. We question whether the commission’s analysis, as laid down in the explanatory memorandum to the recommendation, is sufficient to conclude that the three-criteria test has been met for every member state in which the recommendation has effect.

The decision whether or not to include a market in the recommendation has a significant impact on an NRA’s ability to regulate that market. This can be illustrated by the publications pursuant to the European consultation mechanism. A study of the so-called ‘serious doubts’ letters and the veto decisions of the commission illustrates that, thus far, the commission has been very critical on NRA draft decisions containing market definitions that deviate from the recommendation. For instance, on 20 October 2004 the commission vetoed a draft-decision of the Austrian regulator (TKK) concluding that the Austrian incumbent operator (Telekom Austria) did not possess significant market power on the wholesale market for transit services. The market for transit services in the public telephone network is included in the recommendation as one of the markets susceptible to ex ante regulation.

In its draft decision, the TKK considered that self-provision through direct interconnection was included in the market for transit services. But, according to the commission, the TKK had not brought forward sufficient evidence for the inclusion of direct interconnection in the transit market. According to the commission, to the extent that direct interconnection was relevant at all, it could only be considered in the assessment of significant market power and not in connection with the market definition.

Another example of the critical attitude of the European Commission towards the definition of markets that are not included in the recommendation is the commission’s reaction on a draft measure of the Dutch telecoms regulator (OPTA) to regulate the retail markets for the supply of free-to-air radio and television (RTV) services via cable networks. In its draft decision, OPTA concluded that the market for the supply of RTV services via cable networks was not effectively competitive and that each cable operator had a dominant position in its own provision area. For that reason, OPTA found that the relevant markets were susceptible to ex ante regulation.

On 3 November 2005, the commission expressed that it had serious doubts as to the draft measure’s compatibilty with European law. The commission indicated to OPTA that the markets defined did not satisfy the threecriteria test for ex ante regulation.

OPTA’s main argument for regulation was that cable is the predominant means of transmission and reception of RTV services in the Netherlands. For instance, in 2004 the cable infrastructure was used by more than 90 per cent of the Dutch population. OPTA also referred to a ‘regulatory imbalance’ that would arise if the cable operators were free to set their retail tariffs. This regulatory imbalance resulted from the fact that the incumbent telecoms operator (KPN) would be (relatively strictly) regulated for its (core) telephony services, while the incumbent (regional) cable operators would have more commercial freedom as regards their (core) broadcasting services.

Following the voluntary commitment of the Dutch cable operators not to raise their retail tariffs in 2006, OPTA amended its draft decision on 12 December 2005, which in turn obtained the commission’s approval on the very same day.

It follows from the Austrian and Dutch cases described above that NRAs face a much heavier burden of proof when they wish to regulate a market that is not included in the recommendation or when they define a market differently to the recommendation. When an NRA purports to regulate a market included in the recommendation, it no longer needs to demonstrate that this market meets the three-criteria test. The reasoning being in such cases the commission has already demonstrated that the market in question meets the test. This argument, however, overlooks that the recommendation does not take into account specific national circumstances, and merely contains a very general, abstract analysis why a particular market would be susceptible to ex ante regulation.

By contrast, when a market is not mentioned in the recommendation, the commission will carefully scrutinise whether the NRAs have provided sufficient (quantitative) evidence to establish that that specific market meets the three-criteria test.

The different approach of the commission towards markets included in the recommendation and those that are not included, might well be – at least to some extent – the result of a desire for harmonisation. In our view, however, (attempted) harmonisation may not lead to dissimilar treatment of similar situations in one country, as is arguably the case in the Netherlands. Nor may (attempted) harmonisation lead to regulation of markets that do not meet the three-criteria test, even if they are included in the recommendation. We therefore suggest that markets that are also included in the recommendation may only be regulated when the NRA has demonstrated that this market meets the three-criteria test. In this approach, the recommendation would really only be the point of departure for the national regulators, but the inclusion of a market in the recommendation would not relieve them of the task of carrying out – and demonstrating in their (draft) decision – the three-criteria test.

In this context it is interesting to refer to a judgment of the Dutch High Administrative Court for Trade and Industry of 29 August 2006. In this judgment, the court annulled an OPTA decision regulating the markets for call termination on individual mobile networks, markets defined wholly in accordance with the recommendation. The court held, inter alia, that OPTA had not (sufficiently) investigated whether general competition law would be sufficient to redress alleged (pricing) problems on the markets for call termination on mobile networks. As the deficiency pointed out by the court coincides with one of the criteria of the commission’s test, it follows from this judgment that the inclusion of a market in the recommendation does not relieve a national regulator from applying (at least part of) the three-criteria test.

Significant Market Power

In addition to what has been noted above in relation to market definition, the decisions of the commission pursuant to article 7 of the Framework Directive have also revealed what seems to be an inconsistency in the interpretation of the concept of significant market power (SMP). For instance, with respect to the market for publicly available fixed international telephone services, the commission vetoed a draft decision of the Finnish regulator (Ficora) concluding that there are no operators with significant market power on said markets. The commission considered that a market share in excess of 50 per cent is, in the absence of exceptional circumstances, a presumption of a dominant position, which can only be rebutted by means of clear and convincing evidence. The commission further stated that Ficora had not submitted sufficient facts or reasoning to rebut the presumption created by the market shares of 50 and 55 per cent, respectively.

A few months later the commission again vetoed another draft decision of Ficora, in which Ficora considered that TeliaSonera had significant market power on the market for access and call origination on public mobile networks in Finland. This time, the commission argued that, despite the market share of TeliaSonera being in excess of 60 per cent, other factors should have been been taken into account. No mention was made of a presumption of SMP created by a certain market share. According to the commission, Ficora had not sufficiently analysed the dynamics of the market, especially that external service providers had recently entered the relevant market. The commission also found that Ficora had not adduced sufficient evidence as regards high switching costs and the absence of countervailing buying power.

In the Dutch cable case mentioned above, OPTA had determined that each cable operator had a market share of 100 per cent within its own provision area. Nevertheless, the commission argued that OPTA had not provided sufficient proof that there were high and non-transitory entry barriers for a determination of SMP. With respect to the market for call termination on fixed networks, however, the commission vetoed a draft decision of the German regulator (RegTP) in which it had found the alternative network operators did not possess significant market power despite their market share of 100 per cent.

From the publicly available facts, the above-mentioned decisions and serious doubts letters of the commission seem hard to reconcile. Especially with respect to the significance of market shares in excess of 50 per cent, the commission’s decisions seem to lack consistency. At least they could benefit from somewhat more transparency, which would also help NRAs and courts alike to apply the current regulatory framework in a consistent and harmonised manner while doing justice to its best characteristic: the ability to tailor regulation to the specific needs of the market concerned.

It is not possible to buy entry into Who's Who Legal

terms and conditionscontact us