Oecd. SLC TEST SPAIN. UK. EU

The SLC test [for applying many of the restrictive trade practices * prohibitions] depends on whether the conduct, practice, provision of:

a contract, arrangement or understanding (CAU)*
or transaction,

has the purpose, effect or likely effect of substantially lessening competition (SLC) in a market.

OECD.ORG/COMPETITION

THE CURRENT TEST: the SLC/SIEC Test

Most competition authorities rely on one of two main tests:

(i) the dominance test and

(ii) the significant lessening of competition (SLC) test.

iii. the hybrid test : eg.EU

(i) the dominance test

Under the dominance test a merger is anticomp if it strengthens or creates a dominant position in the market. dominance is where a market leader with a degree of independence from competitive pressures is created. Dominance can be interpreted either narrowly whereby it covers only situations where the merged firm becomes dominant or more broadly as covering also collective dominance, i.e. situations where the merger affects the competitive structure of the market in a manner that is conducive to creating a coordinated equilibrium among competitors.

II. The SLC test.

Under the significant lessening of competition test, a merger has anti-competitive effects if it is likely to substantially lessen competition in the market. In comparison with the dominance test, the SLC test focuses on the effects of the merger on the market and on the loss of competition among firms rather than on threshold structural issues such as market shares. Under the SLC test, the investigation and assessment of a merger are more concerned with whether prices are likely to rise after the merger is consummated.

iii. the hybrid test,

Under the hybrid test, a merger is anticompetitive if it significantly impedes effective competition in the market in particular through the creation or strengthening of a dominant position. This is the test currently in force in the European Union. By listing the creation or strengthening of dominant position as one of the ways in which effective competition may be impeded, the hybrid test combines the standards of both SLC and dominance. Doing so may allow countries that change from dominance to the SLC test to maintain clear continuity with past decisional practice and case law. Generally, the hybrid test is viewed as being nearly identical to the SLC test and hence is treated as part of SLC family.

There may be a difference in the scope of the dominance and SLC standards, whereby the assessment of certain situations could lead to different outcomes depending on which test is used.

Whether there is a difference between the scope of the dominance and SLC tests, and how large that difference is, depend on how the concept of dominance is interpreted.

Horizontal mergers can have two types of effects, unilateral and coordinated effects. Unilateral effects are those that result from the strengthening of a market position of the merged entity, which as a consequence can act to some extent independently of its competitors. Coordinated effects on the other hand arise when, as a result of the merger, the structure of the market is changed in a way that favors tacit or express collusion among the remaining competitors.

There is no doubt that the SLC test can cover both unilateral and coordinated effects. However, whether the dominance test is capable of doing the same depends on how the notion of dominance is interpreted. If interpreted narrowly, i.e. to a certain extent literally because the wording is usually in singular (creation or strengthening of a dominant position), the dominance test does not reach coordinated effects and as such does not allow for proper assessment of many potentially anticompetitive mergers. If on the other hand, dominance is interpreted broadly as extending also to situations of collective dominance, such as the case was in the EU, the reach of the dominance test is nearly identical to that of the SLC test.

Nevertheless, even if dominance is given a broad economic interpretation, there may still be mergers leading to potentially anti-competitive unilateral effects that could escape scrutiny under the dominance test. This may occur with respect to mergers that lead to non-collusive oligopolies or vertical and conglomerate mergers. Indeed, several countries have mentioned cases from their practice, which could potentially have led to different outcomes if assessed under the dominance test as opposed to the SLC test.

There has been a clear move away from dominance and towards the SLC standard over the past seven years. There have been various reasons for switching :

In Australia, for example, a potential enforcement gap presented coordinated effects problems.

Other jurisdictions have moved to the SLC test principally to eliminate the uncertainty over the reach of the dominance standard, for example, whether it extended to situations where horizontal mergers would lead to unilateral effects without creating a clear market leader. This was the case of the EU, while other countries (e.g. Czech Republic and Poland) have switched to the SLC test to adapt their standard to that of other countries or jurisdictions.

In the Netherlands, which have always endorsed a broad notion of dominance, the change to an SLC standard has brought very little to the assessment of mergers. Because these countries had already gradually introduced an increasingly economic, effects-based approach to merger review under a dominance test, the adoption of the SLC test has mainly aligned the wording of the test with the practice.

Countries like Denmark have also mentioned that the SLC test has contributed to enhancing the role of economic analysis in their merger review and to a better understanding of the assessment by the parties and the courts. In the experience of some countries, such as Canada, the SLC test has proven to be sufficiently flexible to capture a spectrum of anti-competitive effects.

The EU and other jurisdictions also noted that the SLC test allows them to properly assess mergers that would have been problematic to evaluate under the dominance standard, such as non-horizontal mergers. In such cases, a merger between two companies that are active on different markets, neither of them dominant, may still lead to an increase in prices as a result of vertical integration.

Some countries argued that the SLC standard allows for a more flexible and appropriate assessment of some mergers because it reduces the reliance on a formal market definition. In the UK experience, for example, there are cases in which it is not necessary to formally define the market because at a “quick look” it is clear that the merger is not anti-competitive regardless of how one defines the market. In such cases, the merger review can be much faster under an SLC standard, since formal market definition often takes a significant amount of time.

the dominance test provides bright line rules and therefore offers firms a higher degree of legal certainty.

the SLC standard can provide a comparable level of legal certainty only if accompanied by the adoption of guidelines explaining in detail how the test is applied.

After the February 2009 meeting, the Chair of the Competition Committee requested delegates to express their views on which topics should be included in the agenda of Working Party n. 3 (WP3) for June 2009 (see letter COMP/2009.31). Based on the responses received by the Secretariat, it was decided that in June 2009, WP3 would hold a roundtable discussion on “The standard for merger review, with a particular emphasis on country experience with the change of merger review standard from the dominance test to the SLC/SIEC test”. The Chair asked the Secretariat to identify in a short Issues Paper the main questions related to this topic that could be addressed in the country contributions and in the WP3 roundtable discussion of June 9. In his letter of 25 March 2009, the Chair invited Committee members and observers to submit written contributions on this topic by no later than Friday, 16 May 2009.

The substantive criteria used for the assessment of mergers has already been the topic of a Committee roundtable in October 2002¹. The 2002 discussion revealed that while a number of different tests can be used to assess mergers, there were two main tests in use: (i) the dominance test, where a merger is considered anti-competitive if it creates or strengthens a dominant position and (ii) the significant lessening of competition test (SLC), where mergers are anti-competitive if competition is likely to be significantly impaired after the merger is consummated. Some countries reported to have hybrid standards and others to have substantive tests based also on other public interest considerations. The WP3 discussion will follow- up on that roundtable, focussing on country experiences in changing the standard of review for mergers.

The 2002 discussion indicated that there was no consensus on the overall superiority of one test over the other. However, there was considerable debate over whether or not the two tests cover the same set of anti-competitive effects that can arise from mergers. Generally, the SLC test was perceived as offering a broader coverage than the dominance test and to offer a more flexible enforcement approach to mergers. However, because of the greater flexibility in the SLC test, delegates concluded that a change of legal standard from dominance to SLC may introduce some legal uncertainty on the businesses as to what type of mergers are likely to raise anti-competitive effects. The change in legal standards, can therefore affect firms’ willingness to plan even inoffensive mergers. In terms of tools and instruments of analysis, the roundtable showed that market shares and concentration indices generally play a more important role when the dominance test is applied as opposed to the SLC test, which is a less structural and more economic- based tool for investigating mergers.

Over the years, and particularly following the 2002 roundtable, a number of jurisdictions have decided to change their legal standards for the review of mergers and have moved from the dominance test to the SLC or equivalent tests. Since other countries (particularly in Europe) may be considering whether to adopt the SLC or similar tests, this roundtable offers the opportunity to engage in a discussion about the experiences that countries may have had with the change of the legal test. The purpose of this issues paper is to identify the main topics for discussion at the meeting on June 9. After reviewing the meaning of the dominance test and of the SLC test and the implications each test may have in terms of merger assessment, this paper describes some of the merger reforms that have occurred in selected OECD countries.

The paper DAFFE/COMP(2003)5 briefly discusses some of the main policy issues which are involved with the choice of the merger test.

Is there a gap in the dominance test?

It is undisputed that the SLC test can be used to review both unilateral and co-ordinated effects. More controversial is to what extent the dominance test is sufficiently flexible to cover all anti-competitive mergers. The question is whether there are mergers which could harm competition and whose anti- competitive effects cannot be addressed using the existing concepts of single firm and/or collective dominance. This would be the case where post-merger, the market features are not such that co-ordination can take place and the merged firm’s market share is below the level required for establishing single dominance, but the merger nonetheless leads to unilateral effects (i.e. to a price increase).

Consider for example the situation of a so-called ‘non-collusive oligopoly’ where post-merger there would be only a few firms, none of which have enough market power to be considered individually dominant, and where collusion is also unlikely (i.e. the firms are not jointly dominant). In such a situation,

economic theory suggests that the elimination of the competitive constraints that the merging firms exerted on each other prior to the merger may lead to a general price increase in the market. The merging firms will have an incentive to unilaterally increase their prices (although the merging firms will not become dominant and there will be no co-ordinated behaviour after the merger). Also, the other market players will benefit from the reduction of competitive pressure that results from the merger, since the merging firms’ price increase may induce the switching of some demand to the rival firms, which, in turn, may find it optimal to increase prices. This might happen in particular in differentiated product markets. This is the so- called “gap” in the dominance test: mergers which allow firms to unilaterally raise prices but do not create or reinforce a single or collective dominant position cannot be prohibited.

The dominance test does not inherently extend to co-ordinated effects⁸. In most jurisdictions which have (or have had) a dominance test, collective or joint dominance is not expressly referred to in the statutory language. For example, the European Union, New Zealand and Australia all had statutory dominance tests which made no explicit reference to collective dominance. Similarly, in many European countries whose substantive merger test is dominance, collective dominance is not expressly mentioned in the text of the law. Some courts, as in the European Union, have developed a consistent case law interpreting the merger rules as applicable to collective as well as single-firm dominance but that has not been the case in all jurisdictions. In Australia and New Zealand, for example, the dominance test applied only to single firm dominance. The result was that mergers could not be blocked simply because they created a strong likelihood of co-ordinated effects. This was one of the reasons why the competition authorities in both countries argued in favour of replacing their dominance tests with an SLC test.

Addressing the ‘gap’ in the dominance test

In the debate on which test is better suited for assessing mergers, a number of suggestions short of new legislation have been put forward to address the problem of the ‘gap’ in the dominance test. For the sake of legal certainty, however, some countries have decided to amend their merger statutes and change the legal standard.

Filling the gap in the dominance test without changing the legal test

A number of solutions have been put forward to address the gap in the dominance test without the need to change the test itself. These proposals aim at finding more flexibility in the dominance test so that it can be used to catch all harmful mergers that could be stopped under the SLC test. Some of these proposals, which are discussed in the Background Note to the 2002 OECD roundtable include:

1. - weakeningthedefinitionallinkbetweenmarketpoweranddominance;

1. - varying the approach to market definition depending on the type of merger being reviewed, i.e.adopt particularly narrow markets in the case of mergers likely to produce unilateral effects inmarketsfeaturingdifferentiatedproducts;

1. - consistently adopting particularly narrow market definitions;

1. - adopting different dominance thresholds for single and collective dominance;

1. - lowering the market power threshold required to find dominance; and/or

1. - ensuringcollectivedominancecoversanti-competitiveoligopolisticinter-dependencefallingshortofco-ordinatedeffects.

These suggestions however may have some important policy implications. On the one hand, they could ensure more flexibility and a broader scope of intervention against anti-competitive mergers in those jurisdictions which apply the dominance test. On the other hand, however, there is a risk that some uncertainty will be introduced into the merger system. For example, the first option could reduce certainty on how merger rules are enforced by diminishing the utility of the jurisprudence developed in enforcing the other parts of a competition statute. Similarly, the second option may introduce some uncertainty by requiring a prediction as to what approach to market definition will be used in any particular merger. It could also create suspicions that merger review in particular cases may be driven more by a desire for a particular result than exclusive attention to proven facts. Adopting consistently narrow market definitions could also mean that many mergers involving substitute products are analysed as conglomerate mergers making it necessary to treat horizontal effects as various kinds of portfolio effects

More generally, courts may resist endorsing some of these proposals. For example, courts could prove very reluctant to extend collective dominance to block mergers where “non-cooperative” oligopolistic inter-dependence (i.e. behaviour neither explicitly nor tacitly involving collusion) will tend to result in higher prices post-merger. In addition, courts may resist de-linking the definitions for single-firm and collective dominance and, partly because of that, resist as well lowering market power thresholds associated with either collective or single-firm dominance. Courts will presumably be reluctant to open the door to finding that more than one different sized firm enjoys single-firm dominance in a properly defined antitrust market. More importantly, courts may be unwilling to de-link the definitions of dominance applied in merger review and in abuse of dominance cases, as doing so could upset the balance incorporated in abuse of dominance prohibitions.

Legislative changes in the standard of review of mergers

In light of the existence of a possible gap in the dominance test and the difficulties to fill the gap without clearly changing the test, many have argued in favour of changing the merger statutes and adopting a more flexible test, such as the SLC test, which would catch without doubt all possible anti-competitive effects of mergers. Over the years, a number of jurisdictions have changed the legal test for the review of mergers and moved from the dominance test to the SLC test or to equivalent tests. This was the case for example in Australia (1992), New Zealand (2001), the UK ( 2002) and the European Union (2004). Other European countries have more recently followed the EU and also moved to a SLC-type of test. This was the case in Belgium, France, Spain, and Poland. Other countries may be considering similar reforms for the future.

In 2002, the United Kingdom decided to change its test for reviewing mergers and moved to a significant lessening of competition standard. Unlike the previous examples, the UK did not have a dominance test before 2002, but a broad ‘public interest’ test. In particular, Section 69 of the 1973 Fair Trade Act (FTA) required the Competition Commission to consider whether the merger “operates, or may be expected to operate, against the public interest”.

Section 84 of the FTA provided that, in making this assessment: “the Commission shall take into account all matters which appear to them in the particular circumstances to be relevant and, among other things, shall have regards to the desirability (a) of maintaining and promoting effective competition between persons supplying goods and services in the UK; (b) of promoting the interests of consumers, purchasers and other users of goods and services in the UK in respect of the prices charged for them and in respect of their quality and the variety of goods and services supplied; (c) of promoting, through competition, the reduction of costs and the development and use of new techniques and new products, and of facilitating the entry of new competitors into existing markets; (d) of maintaining and promoting the balanced distribution of industry and employment in the UK; (e) of maintaining and promoting effective competitive activity in markets outside the UK on the part of producers of goods, and suppliers of goods and services, in the UK”.

According to the UK submission to the 2002 OECD Roundtable on Substantive Criteria Used for the Assessment of Mergers,¹³, despite the fact that non-competition considerations such as public security, regulatory concerns or environmental considerations, could be part of the assessment of mergers, in practice the UK competition agencies were almost entirely concerned in their investigations, with the effects of the merger on competition. The competition element therefore still carried substantial weight in practice. Under the Enterprise Act of 2002, the new merger system focused expressly on competition concerns and consumer welfare, by introducing a substantial lessening of competition test. However, if the merger is a public interest case or a special merger situation, the Secretary of State may still take other factors into account (national security and other public security concerns) in deciding whether to deviate from the decision of competition agencies.

1. 1. European Union

The European Union adopted its first merger statute (the Merger Regulation¹⁴) in 1989 and the substantive test for all merger cases was the dominance test. Article 2(3) of the Merger Regulation stipulated that a concentration “which creates or strengthens a dominant position as a result of which effective competition would be significantly impeded in the common market or in a substantial part of it shall be declared incompatible with the common market.”

In 2004, the Merger Regulation was amended¹⁵ and a new substantive test for mergers was adopted. According to the merger regulation itself,¹⁶ the reform of the merger test was necessary to fill the perceived gap in the dominance test. According to recital 25 of the new Merger Regulation, “in view of the consequences that concentrations in oligopolistic market structures may have, it is all the more necessary to maintain effective competition in such markets. Many oligopolistic markets exhibit a healthy degree of competition. However, under certain circumstances, concentrations involving the elimination of important competitive constraints that the merging parties had exerted upon each other, as well as a reduction of competitive pressure on the remaining competitors, may, even in the absence of a likelihood of coordination between the members of the oligopoly, result in a significant impediment to effective competition. The Community courts have, however, not to date expressly interpreted Regulation (EEC) No 4064/89 as requiring concentrations giving rise to such non-coordinated effects to be declared incompatible with the common market. Therefore, in the interests of legal certainty, it should be made clear that this Regulation permits effective control of all such concentrations by providing that any concentration which would significantly impede effective competition, in the common market or in a substantial part of it, should be declared incompatible with the common market […]”.

Today, the European Commission applies a SLC-type test for all mergers and acquisitions. According to Article 2.2 of the Merger Regulation, “a concentration which would not significantly impede effective competition in the common market or in a substantial part of it, in particular as a result of the creation or strengthening of a dominant position, shall be declared compatible with the common market”¹⁷. The ‘significant impediment to effective competition’ (SIEC) is considered to be equivalent to a ‘significant lessening of competition’ and is interpreted as extending, beyond the concept of dominance, to the anti- competitive effects of a concentration resulting from the non-coordinated behaviour of undertakings which would not have a dominant position on the market concerned.

¹⁵ See Council Regulation Nr. 139/2004 of 20 January 2004 on the Control of Concentrations between Undertakings.

¹⁶ See Recital 25.

¹⁷ In practice, the new merger test reverses the two limbs of the old dominance test and the creation or strengthening of a dominant position becomes now only one possible theory of harm under which a merger can be considered incompatible with the common market.

Broader policy issues connected to the dominance test and the SLC test

The debate over the advantages and disadvantages of the dominance test versus the SLC test identified a number of broader policy issues which are closely related to the type of test that is chosen for assessing mergers.

Structural versus economic analysis

Some of the supporters of the SLC test have argued that reviewing mergers under a significant lessening of competition standard is more suited to an economics-based approach used to assess the effects of mergers on competition. The SLC allows emphasis to be place on rivalry between firms, on empirical evidence and on economic analysis¹⁸. This is particularly the case for the analysis of mergers in oligopoly markets, where the SLC test appears to be more suitable to review the impact of the merger on competition in a way that the dominance test cannot. The analysis under the dominance test follows more of a structural approach that places more emphasis on market definition and on market shares and is therefore less suited to measure inter-firm competitive dynamics which may or may not lead to a loss of competition as a result of the merger in oligopoly markets and in markets with differentiated products.

Legal certainty and predictability

Some authors, however, have warned that the SLC test may require the use of increasingly more sophisticated quantitative and econometric analysis to measure unilateral effects¹⁹. The concern is not simply related to the ability of the reviewing agency or of the parties to handle complex economic evidence and lengthy quantitative analysis²⁰, when the data is available. There is also the concern that the switch to the SLC test may introduce a degree of uncertainty and unpredictability as to how mergers will be assessed, which may ultimately discourage firms from planning pro-competitive mergers in the first place.

According to Heimler²¹, the adoption of the SLC test for mergers would add too much flexibility to the merger standard with the risk that the test for assessing mergers will become too broad, leaving the competition agency with too much discretion. He argues that this may be contrary to the basic objective of legal provisions, which is to provide an effective guide to enterprises and individuals. Heimler is less concerned with the application of the SLC test by “capable and rigorous” enforcers than by the risks that a more flexible test can pose when applied by less sophisticated jurisdictions.

Over-enforcement and under-enforcement

The choice between a more rigid or a more flexible test also has policy implications on the type of merger regime that a jurisdiction wishes to put in place. The trade off is between more rigid rules, which may provide more certainty to firms but runs the risk of letting some anti-competitive mergers go through,

Voelcker, however, points out that complex econometric analysis requires agencies to hire the necessary expertise and resources to handle such complex matters. This may be a challenge for some agencies. Voelcker is particularly sceptical that courts reviewing merger decisions have the ability to use the complex econometric and qualitative evidence that the SLC test may require in its application. In addition, referring to the EU proposed change to the SLC test, he adds: “As long as there is a perception that the benefits (as well as the limitations) of using econometric evidence are not well understood at both staff level and within the DG COMP hierarchy, many companies may be hesitant to introduce such evidence”.

and more flexible rules, which can potentially catch all anti-competitive mergers but bears the risk of prohibiting some pro-competitive mergers. Indeed both arguments can be used. For example, those who have argued against the dominance test have emphasised its rigidity and the fact that it can lead to too many prohibitions. Others have argued that the dominance test does not catch all types of anti-competitive mergers.

Links between the merger test and unilateral conduct rules

Another policy issue that has been flagged in the debate on the dominance and the SLC tests is whether it is necessary to link the merger test with the legal standard for unilateral conduct. To the extent that mergers are found in other sections of competition statutes, it has been argued that there can be overlaps in jurisprudence with potentially important effects on merger review and on how other parts of the competition statute are applied²².

The concepts of single and collective dominance developed in the review of mergers can affect how abuse of dominance provisions are applied and vice versa. This could be quite unhelpful, particularly if dominance test jurisdictions wish to lower the threshold of what constitutes dominance in order to block certain anti-competitive mergers, especially those having unilateral effects. Before doing so, however, one has to consider how this will tend to widen the scope of their abuse of dominance prohibitions. Moreover, the finding that a merger either leverages or strengthens a previously unidentified dominant position could have important ramifications for subsequent application of the abuse of dominance prohibition to the merging parties. Such a finding of dominance could presumably trigger increased scrutiny of the firms’ conduct under abuse of dominance prohibitions. This point applies as well to mergers where the dominance is of a collective nature. In that case the ramifications affect not just the merging parties, but the whole group of firms sharing the collectively dominant position.

Those in favour of moving from a dominance test to an SLC test for mergers have also argued that: “Not only is there no logical or necessary link between the substantive test for mergers and the control of abusive behaviour, there is much to be said for detaching the two tests from one another. Such a detachment could serve the useful purpose of making clearer that merger control is about maintaining effective competition in markets, and not about predicting future abusive conduct”²³.

International co-operation

One other policy concern that has been flagged as relevant in the debate on the change of the substantive test for mergers is the fact that a greater homogeneity in the standard of review of mergers among jurisdictions would facilitate international cooperation. This point could support the adoption of either the dominance or the SLC test depending on the group of countries for which co-operation and convergence is deemed particularly important. As noted above, one reason that New Zealand adopted the SLC test was to align its merger regime with the one in Australia, its most important economic partner.

While the similarity of tests may be important to facilitate cooperation between agencies, it has also been emphasised that to facilitate international convergence there are other factors that could be just as or even more important than the substantive test that is applied to mergers²⁴. Differences in objectives, the threshold applied for judging how much market power is too much, the analytical approach used to analyse

market power, and the way in which efficiencies are factored into merger review are equally important areas to explore in order to secure greater convergence in results.

‘gap’ cases

The literature has identified a number of cases that could potentially illustrate the so called ‘gap’.

These are merger cases where unilateral effects could arise below the threshold of dominance.

Heinz/Beech-Nut

The so-called “Baby Food” merger case²⁵ in the US is often cited as an example of this situation. The case involved the merger between Heinz and Beech-Nut, which after the merger would face competition only from Gerber, which remained the market leader. The case was investigated by the US Federal Trade Commission which applied the SLC test to assess the effects of the merger. The merger was prohibited as it substantially impaired competition. This case is often referred to as a case where the dominance test would have failed to properly capture the unilateral anti-competitive effects of the merger. Although the number of players in the market would be reduced from three to two players, the merger only involved the second and third players in the market with the remaining player retaining the leading position even after the merger. The facts of the case would have made it difficult for an agency applying the dominance test to argue that the merger would have created or strengthen a dominant position on the market.

Oracle/PeopleSoft

Another case that has been identified as a possible illustration of the gap is the merger between Oracle and PeopleSoft²⁶. The transaction was investigated both in Europe (under the dominance test) and in the US (under the SLC test). Oracle and PeopleSoft are vendors of enterprise application software and compete with SAP (the market leader) and a number of smaller players. From the Commission decision it appears that the European Commission investigated whether unilateral effects would arise from a three-to-

²⁵ FTC v H.J. Heinz, 116 F.Supp. 2d 190 (2000 U.S. Dist.), revised 246 F.3d 708 (2001 U.S. App.).

²⁶ See Commission Decision of 26 October 2004, case COMP/M.3216.

two transaction, but it concluded that such effects were not likely to arise because evidence that was obtained after the Oral Hearing indicated that markets were broader and included a larger number of software vendors²⁷. Commentators of this decision argued that the Commission relied upon unilateral effects and the analysis was devoid of any single-firm dominance considerations²⁸. The Commission actually tried to close the enforcement gap in the application of the dominance test in this case by stretching the concept of collective dominance to non-coordinated effects on oligopolistic markets. The market structure, however, was not conducive to collective dominance and the fact that SAP and Oracle/PeopleSoft had similar market shares in an innovative market with differentiated products would render a credible allegation of single-firm dominance difficult. The merger was ultimately approved by the European Commission, but many argued that because of the dominance test the Commission was not able to properly address the unilateral effects that the merger raised²⁹.

T-Mobil/Tele.ring

After the EU changed its merger test from the dominance test to the SIEC test, the first case which was investigated by the Commission showed the greater flexibility of the new test. The merger involved the merger between T-Mobil and Tele.ring, two Austrian mobile network providers³⁰. The transaction offered the opportunity to the Commission to apply the new SIEC test to a merger which would not create a market leader. After the merger, the merged entity would have had a market share slightly smaller than that of the leading mobile provider in Austria (Mobilkom). The merger lead neither to a single dominant position nor to collective dominance, but significantly reduced competition resulting in higher prices by eliminating a competitive constraint on the incumbents. Under the new merger test, the Commission concluded that “especially with the elimination of the maverick in the market and the simultaneous creation of a market structure with two leading, symmetrical network operators, it is likely that the planned transaction will produce non-co-ordinated effects and significantly impede effective competition in a substantial part of the common market”³¹. The Commission focussed its attention on the likely effects that the merger would have on prices in the Austrian end-customer market and concluded that even if prices would not rise in the short term, the elimination of Tele.ring as a pricing constraint would make it unlikely for prices to continue falling significantly as previously. The merger was however ultimately approved by the Commission, subject to the commitments offered by T-Mobil.

An hypothetical ‘gap’ case for discussion – A bank merger

This last section of the paper presents a short hypothetical exercise. The purpose of this exercise is to discuss whether the application of different standards of review for mergers may lead to different assessments of the same transaction. This could offer the opportunity to those delegations which have not had experience with both tests to contribute their perspective.

This hypothetical case concerns the proposed acquisition of the Bank of Investment (BoI) by the Bank of Commerce (BoC), both active in retail banking services to personal consumers and small and medium enterprises (SMEs) within the country of Oceanica.

BoC is currently the second largest bank in Oceanica and it offers a wide range of retail banking services with a particular emphasis on services to SMEs. The BoI is the fifth largest retail bank in that country and is widely seen as a relative newcomer in the market, having started its operation 15 years ago. It has grown rapidly through acquisitions as well as organically and has developed a reputation of something of a “maverick” that charges lower prices to consumers than its competitors. According to consumer surveys, BoC and BoI are the closest competitors for the supply of several bank services.

Both banks offer the whole range of retail banking services to private consumers, SMEs, large firms and institutions. They also possess a country-wide network of branches that gives them a presence all over the national territory of Oceanica. Due to their size, the range of services offered and the breadth of their branch network, they are considered to be “national banks”. As such, they belong to the group of the Big Five, all national banks that dominate the bank system in Oceanica.

Within the Big Five we have also Bank One, the historical market leader in Oceanica, just ahead of the second largest bank (the BoC). Together, the Big Five banks account for over 85% of the banking business in Oceanica. The remaining 15% are controlled by a multitude of smaller banks that specialize in different areas and have no national presence. The largest of these is much smaller than the BoI.

The sector has evolved substantially in the last 20 years. After a period of general broad expansion, where several banks entered the Oceanica market – and some of them achieved a significant presence in the market like the BoI – the sector has been consolidating over the recent years. Several acquisitions have occurred leading to a growing concentration of market shares in the hands of the Big Five banks.

Notably, BoC has expanded its presence in the market through a series of high-profile acquisitions during the past years – it was not the only one, other banks in the Big Five group have also achieved this position through acquisitions. Throughout the whole period, Bank One has remained the leader of the market, with only BoC starting to challenge its position recently.

For the purpose of this exercise, we will assume that the HHI analysis identifies a selected number of markets where the merger could raise concerns for competition. As shown in the Table below, Bank One is the leader of all the relevant markets. With the proposed merger, however, the new entity would become the market leader for credit for SMEs. For the remaining markets of concern, Bank One would still remain the leader – although by a narrow margin.

Distortion of competition would be outweighed by economic benefits of the merger. Thus, the substantive test was not based solely on competitive criteria, but also enabled the Office to approve a concentration leading to the distortion of competition in case that implementation of this concentration was in the public interest. For the rest of the cases a concentration was not cleared by the Office. The Office could also attach to its decision conditions and obligations intended to ensure that the parties to a concentration comply with the commitments they have entered into vis-à-vis the Office necessary for the protection of competition. The above mentioned substantive test provided the Office with a great degree of administrative discretion and thereby did not ensure a sufficient degree of legal certainty for the parties to a concentration.

By adoption of the new Competition Act (Act No. 143/2001 Coll., on the Protection of Competition) the Czech legal framework came close to the existing EC framework, both in respect of the implementation of the objective notification thresholds based on undertakings´ turnover and in respect of the substantive test, which was close to the dominance test. By applying this new approach the Office would approve a concentration if the merger did not result in the creation or the strengthening a dominant position, as a result of which effective competition would be significantly impeded in the relevant market. If the proposed concentration led to the creation or the strengthening of a dominant position and the Office’s competition concerns were not eliminated by the commitments or other measures proposed by the merging parties, the Office would not approve such a concentration.

Following the Czech Republic’s accession to the EU in 2004, an amendment of the Competition Act modified the application of the substantive test. Based on the modified test, a concentration which would significantly impede effective competition in the relevant market in particular as a result of the creation or the strengthening of a dominant position of the merged entity would not be approved by the Office.

The current substantive test

Based on the above, the Competition Act (currently in force) has introduced the so-called SIEC (substantial impediment to effective competition) test. This substantive test reflects both elements of the

U.S. concept of the SLC (substantial lessening of competition) test as well as the elements of the dominance test. The key element of this test is whether there is a significant impediment of competition as a result of a concentration. The new test therefore considers dominance as only one possible cause of a significant impediment to effective competition, and thus widens the net to catch other situations.

Market shares and the overall level of concentration in a market normally give useful first information about the competitive importance of both the merging parties and their competitors. The Office considers that post-merger market shares are calculated on the assumption that the post-merger combined market share of the merging parties is the sum of their pre-merger market shares. The Competition Act provides that concentrations which, by reason of the limited market share of the merging undertakings, are not liable to impede effective competition may be presumed to be compatible if a combined market share of undertakings concerned does not exceed 25%. In order to measure concentration levels, the Office applies the Herfindahl-Hirschmann Index (HHI) and the change in the HHI from pre-merger to post-merger (“delta”) as first indications of the change in competitive constraints in the market following the merger, similarly to the practice of the European Commission.

The assessment of impact of concentration differs depending on the type of a concentration. The most frequent type of a merger and, from the standpoint of possible effects on competition, is a horizontal merger, i.e. concentration of actual or potential competitors in the same relevant market. The Office targets two main ways in which horizontal mergers may significantly impede effective competition: by eliminating important competitive constraints on one or more firms, which consequently would have increased market power (“non-coordinated/unilateral effects”), or because it makes anticompetitive coordination between the remaining firms more likely or more effective (“coordinated effects”).

First, as mentioned above, the non-coordinated effects occur in situations where, as a result of a concentration, significant competitive constraints on one or more undertakings are eliminated and consequently the market power of these undertakings is strengthened without coordinating their activities in the market. Generally, a merger giving rise to such non-coordinated effects would impede effective competition by creating or strengthening the dominant position of a single firm and cases of non-collusive oligopoly.

Regarding the creation or the strengthening of the dominant position of a single firm, the newly- created undertaking will face either no or at least no significant competition after the merger and will be in a position to act to a considerable extent independently of its competitors, their customers and, ultimately, of consumers (as stated by the ECJ). The other alternative is represented by the situation where a concentration restricts the competition in oligopolistic markets where undertakings do not coordinate their activities so far, and where the investigation did not prove that the concentration itself could facilitate such collusion, i.e. the creation or the strengthening of a collective dominant position (or coordinated effects respectively).

Even in such situations, elimination of one single competitor can cause distortion of the market equilibrium, increased price and reduced output. Usually such situations are often called as “unilateral effects” or non-collusive monopoly. In such situations a concentration carried out in the oligopolistic markets leads to neither the creation of the dominant position of a single firm nor the creation of a collective dominance (so it does not facilitate collusion among biggest players in the market), nevertheless the level of competition is reduced as a result of elimination of significant competitive constraints.

Second, anticompetitive effects of the merger may be represented by “coordinated effects”, i.e. in situations where a merger in a concentrated market may significantly impede effective competition, through the creation or the strengthening of a collective dominant position, because it increases the likelihood that firms are able to coordinate their behavior and raise prices.

Aforementioned coordinated effects include so-called tacit collusion, a situation that has been in the framework of the dominance test covered by the collective dominance doctrine. The probability of coordination of competitors is predominantly increased by factors, such as a comparable market position, stable and non-elastic demand, high degree of transparency in the market, product homogeneity, similar cost structure and also the existence of structural and contractual links between the undertakings in the market.

Using the substantive test, the Office also considers some factors outweighing the increase in market power of merging undertakings as a result of a concentration, such as the countervailing buyer power.

When assessing possible impacts of a concentration it is also necessary to take into account the efficiencies, that may counteract the effects on competition and in particular the potential harm to consumers that the merger might otherwise have. Examples of these benefits are allocation efficiency, which optimizes the allocation of product to customers e.g. by means of rationalization of distribution network allowing the customer to obtain demanded product at a lower price; production efficiency, whereby a producer reaches optimal output of the product e.g. through a rationalization of administrative or management processes, which due to the decrease of variable costs consequently results in lower prices for customers; and dynamic efficiency, i.e. optimal level of innovation, development and creation of new products contributing to the consumer welfare (e.g. through rationalization of research and development expenses, which as a consequence prevents doubled expenses to find innovative solution in one area and which leads to redirecting saved funds to other fields).

The concept of Failing Firm Defence is another element affecting economic assessment of impact on competition. The aim of this concept is to protect the company facing serious economic difficulties. The application of the Failing Firm Defence concept leads to clearance of a concentration, which should otherwise be prohibited. The Office has to prove that the economic situation of the acquired undertaking is so serious that the undertaking is likely to bring its activities to an end and to exit the affected market.

Besides horizontal mergers, the Office reviews vertical and conglomerate mergers where the risk of negative impact on competition is not so high. Vertical mergers could raise serious competition concerns where one of the merging undertakings has significant market power and the vertical integration could lead to the increase of barriers to entry in the relevant markets. Therefore, the Office, when reviewing such concentrations, focuses on the assessment of barriers to entry and relating market foreclosure risk. As regards conglomerate mergers, the Office assesses especially issues related to portfolio power. So far, the Office has not reviewed any conglomerate merger which would raise serious competition concerns.

SPAIN

Since merger review was first introduced in the Spain under the 1989 Competition Act¹, mergers have been assessed taking into account whether the transaction may hinder the maintenance of effective competition in the market. Therefore, from the beginning, the test applied in Spain has been closer to the SLC test than to de dominance test, since the analysis is not solely based on the creation or reinforcement of a dominant position.

It is true, however, that most of the prohibited or conditioned mergers in the past were cases where dominance was the issue. Nevertheless, through the years, the standard for merger review in Spain has evolved to a more economic-based approach, more coherent with the flexible test provided in the 1989 Competition Act and especially with the trend which is being followed by most developed countries.

Recently, the Spanish competition regime has undergone an ambitious reform which culminated in the adoption of a new Competition Act² that came into force in September 2007. As far as the standard for merger review is concerned, the same substantive test is applied under the new Act although two relevant changes have been introduced. First, the new legal framework limits the government’s role in merger review, which is foreseen only under exceptional circumstances. Second, it specifies the criteria of substantive assessment that will guide the decisions of both bodies (the new National Competition Commission –CNC- and the government). In effect, the 2007 Competition Act clearly separates those factors that will guide decision-making by the CNC, focused on the maintenance of effective competition in the markets in line with the previous Act, from those on which government’s intervention may be based, related to the protection of public interest. In this respect, the new legal framework provides a non- exhaustive list of specific criteria which may guide the decision of the Council of Ministers, different from those based on competition concerns.

Consequently, the recent legislative changes have contributed to clarify the elements that the Spanish Competition Authority takes into account in merger review, with the systematisation of, among others, those that have been considered to date in the reports of the extinct Competition Service and of the Competition Court³. At the same time, some aspects of the substantive test have been particularly highlighted, such as the treatment of business efficiencies and the assessment of cooperative aspects.

The purpose of this paper is to put forward the key elements of the substantive test for merger review in Spain and to describe its application with a very recent example, the Gas Natural/Unión Fenosa Merger Case, where the test was carefully applied not only to identify the competition concerns resulting from the merger but also the suitability of the proposed commitments.

¹ Ley 16/1989, de 17 de septiembre, de Defensa de la Competencia.

² Ley 15/2007, de 3 de julio, de Defensa de la Competencia (B.O.E. de 4 de julio de 2007).

³ The application of the 1989 Competition Act was entrusted to two administrative bodies: the Competition Service (Servicio de Defensa de la Competencia), in charge of handling the proceedings, and the Competition Court (Tribunal de Defensa de la Competencia), with the functions of legal ruling in antitrust cases and of forwarding its report and proposal to the Minister of Economy in merger control.

The substantive test for merger review in Spain

Merger review under the 1989 Competition Act

The 1989 Competition Act introduced a preventive merger control system purported to avoid that mergers may hinder the maintenance of effective competition in the market and foster practices resulting in a diminished consumer welfare or competitiveness of the Spanish economy and, thus, damaging public interest

The substantive test therefore focused on the way a merger would possibly affect the market structure and hinder effective competition, and not just whether a merger would create or strengthen a dominant position. Yet, unlike the SIEC or SLC tests which explicitly refer to “substantial” effects, the Spanish test did not make reference to the extent of such harm nor to the concept of “effective competition”. However, it is clear that what it was pursued was that the merger did not give rise to the exertion of market power by the merging parties in detriment of competition in the affected market. Therefore, the assessment required more than just looking at market shares and dominant positions. In fact, from the beginning, the test has been applied in a restrictive manner, limiting prohibitions and conditional approvals to situations in which the merger “could mean a significant obstacle for effective competition according to the principle of proportionality”

Indeed, the 1989 legal framework stated that the decision as to whether a merger could hinder the maintenance of effective competition on the market should be made by analysing its potential or actual restrictive effects, paying special attention to several circumstances: definition and structure of the relevant market, the possibilities of choice offered to suppliers, distributors and consumers or users, the economic and financial power of the companies involved, the evolution of the supply and the demand and external competition.

Moreover, it foresaw that the Court could also consider the contribution that a merger could make to improve the production and marketing systems, promote technical or economic progress, boost international competitiveness of the national industry or ensure the interests of consumers and users, and if such contribution compensated for its restrictive effects on competition. In the case of joint ventures, the Court had to carry out a special analysis of the possible restrictive effects on competition derived from the presence of the participated undertaking and of the parent undertakings in the same market or in vertically integrated or close markets.

Legislative changes in merger review introduced by the 2007 Competition Act

As it has already been mentioned, the recent 2007 Competition Act has allowed a clarification of thetest without changing the criteria of substantive assessment, which is still in line with the SLC and SIECtests. The new legal framework provides in section 10 that “the National Competition Commission shall assess the economic concentrations in light of the possible impediment to the maintenance of effective competition in all or part of the national market”.

Specifically, the CNC shall adopt its decision taking into account, among others, several elements which include, in addition to those mentioned in the 1989 Act, the real or potential competition of undertakings located either within or without the national territory, any barriers to entry in the relevant markets, the countervailing power of the demand or of the supply and their capacity to offset the position in the market of the affected undertakings, the economic efficiencies derived from the merger and, in particular, the extent to which these efficiencies are transferred to the intermediate and ultimate consumers, specifically, in the form of a bigger or better supply and of lower prices.

Besides this clarification of criteria, the new Act provides that the Council of Ministers, on an exceptional basis, may assess economic concentrations in light of criteria of general interest other than protecting competition. In particular, it puts forward some examples: defence and national security, protection of public security or public health, free movement of goods and services within the national territory, environment protection, promotion of technological research and development and guarantee of adequate maintenance of the objectives of sector regulation. It is, however, a non exhaustive list.

Brief summary of some key features of merger review in Spain

In order to give some practical guidance on the key elements of the substantive test which has just been explained, this section focuses on the most relevant factors which are taken into account in merger review, giving a brief summary of the factors assessed in each case, on the basis of illustrative cases where such factors were considered.

Market shares

Even though the test applied in Spain is not a dominance test, the creation or strengthening of a dominant position is one of the competition concerns which must be assessed by the Spanish Competition Authorities, thus benefiting from the broad jurisprudence on this issue.

On a general basis, and according to the practice of the Spanish Competition Authorities since 1989, market shares around 10% are not considered to be harmful for competition⁶. Likewise, combined shares below the notification threshold (25% under the 1989 Competition Act and 30% under the new Competition Act) are considered to be acceptable. Even high market shares may be acceptable if there are no horizontal or vertical overlaps⁷ or if those overlaps are small in comparison with the market share of the acquiring company in the affected markets⁸.

In the analysis of market shares, it is also important to assess whether the parties’ market shares have been increasing or decreasing in the last years or if there is a high degree of volatility. This is especially relevant in markets where innovation plays a key role⁹ or where liberalisation is taking place¹⁰.

Moreover, in line with EU practice, HHI indexes are sometimes considered to assess the degree of concentration in the affected markets¹¹.

Real and potential competition and barriers to entry

Even if the horizontal overlap resulting form a merger is small, competition concerns may arise if the transaction eliminates a competition restraint on the merged entity (i.e. with the acquisition of the maverick firm on the market¹² or of a company which exerts a high degree of competition on the market¹³).

At the same time, coordinated effects are assessed according, in general terms, to the same elements of assessment identified by the Commission in its Guidelines for horizontal mergers¹⁴.

Besides real competition, potential competition must be taken into account in order to get the real picture of the foreseeable post-merger scenario. The possibility of new competitors accessing the relevant market and their ability to discipline the behaviour of the incumbents are two of the main aspects taken into account when assessing the risk of hindering effective competition. The threat of entry may effectively limit the behaviour of the incumbents, if it is probable, it may take place in the short term and has significant impact on competition.

In this sense, it is useful to focus on factors such as imports or recent successful entries to the market¹⁵.

In any event, entry of new competitors will depend substantially on the assessment of entry barriers to the market which may influence their decision to enter the market to the extent of eliminating the discipline they may exercise on the incumbents.

Many barriers to entry have been identified in past cases: economic barriers¹⁶, legal barriers¹⁷, technological barriers¹⁸, barriers from intellectual property rights¹⁹ or barriers arising from brands or investments on advertising²⁰.

Vertical considerations: possibilities to switch suppliers and customers

Another crucial factor in merger review is the possibility to choose between alternative suppliers so that the considered merger does not result in the expulsion of a competitor due to the lack of access to suppliers.

Likewise, vertical effects in downstream markets must be assessed taking into account if the merger will harm a competitor chose sole customer is the acquired company which will be vertically integrated after the merger.

Again, these vertical considerations have a lot to do with the degree of concentration in upstream (suppliers) and downstream (customers) markets²¹.

Countervailing power of demand or supply

The possible impact of a given merger on the supply structure of the affected market may be counterbalanced by demand negotiating power capable of preventing the appearance of anticompetitive practices. Countervailing power depends on the relative strength of demand when setting the price and other contractual conditions and is basically determined by the nature of clients, their concentration degree, demand-price elasticity, the importance of brand loyalty, the distribution characteristics, the negotiating procedure to determine the contractual conditions and the foreseeable evolution of demand.

Countervailing power is especially important in markets with a monopsony structure or where demand is concentrated in big customers such as the Administration²².

This analysis will refer to the countervailing power of supply when the concentration basically affects the demand of goods and services.

Efficiencies

In line with the trend followed by most competition authorities, the assessment of efficiencies has been highlighted under the new Competition Act. This does not mean that under the 1989 Act efficiencies were not taken into account to clear a merger. Quite on the contrary, the Spanish Competition Authorities have reflected in the reports of several relevant mergers the assessment of the alleged efficiencies by the notifying party, following the criteria set out in the Commission’s Guidelines on horizontal mergers²³.

Yet, the advantage introduced by the new framework is the clarification of the legal and economic test which must be applied in order to assess such efficiencies: (i) efficiencies must derive from the merger which is being assessed, (ii) they must compensate the identified competition concerns, (iii) they must contribute to the improvement of the production or commercialisation systems and to business competitiveness, (iv) and they must be transferred to the final consumer in the form of more or better supply and of lower prices.

Others

There are many other relevant factors which the Spanish Competition Authorities have taken into consideration in merger control in the past: structural links with competitors resulting form the merger, coordination between parent companies through the creation of joint ventures, portfolio effects and network externalities, etc.

The Gas Natural/Unión Fenosa case

This section describes the recent merger between two of the main Spanish energy groups, Gas Natural and Unión Fenosa, which was cleared last February, subject to commitments offered by Gas Natural. It is both a most recent and interesting case to illustrate the substantive test described above from the enforcement angle, not only because of the number of competition concerns identified in the markets involved, but also for the implications in the assessment of the commitments which were finally adopted.

The Gas Natural/Unión Fenosa merger started with the formal notification by Gas Natural on the 3^rd September 2008 of its proposed acquisition of Unión Fenosa.

Gas Natural and Unión Fenosa are the respective parent groups of two vertically integrated energy groups, operating in nearly all gas and electricity markets in Spain and worldwide.

Gas Natural, jointly controlled by Repsol and La Caixa, is the main Spanish operator in the gas sector. It is the biggest gas supplier to Spain (60% of total gas supply in 2007), retail supplier to end customers (more than 50% of gas consumption in 2007) and the main gas distributor. Furthermore, despite being a newcomer in the electricity sector, Gas Natural ranks fourth in generation due to its combined-cycle gas turbine (CCGT) power plants and has a big growth potential in retail supply due to its experience and commercial network in the gas market. Repsol, the leading company in the Spanish oil markets, is also active in gas production and supply as well as in electricity generation.

As for Unión Fenosa, it is primarily active in the electricity sector, where it is ranked third behind Iberdrola and Endesa in generation, distribution and retail supply. In recent years, Unión Fenosa entered successfully the gas markets, mainly through Unión Fenosa Gas, S.A. (UFG), a vertically integrated company active in all gas markets which Unión Fenosa jointly controls with the Italian ENI, S.P.A.

On the 7^th November 2008, the CNC Council decided to initiate phase two of proceedings in order to carry out a more detailed analysis of the transaction. The main competition concerns raised by the takeover which were identified by the Spanish Competition Authority following the substantive test described in section II of this paper were the following:

The increased market power of the merged entity in the Spanish gas markets, mainly in supply and retail distribution, due to its greater size and vertical integration, lessening the supply-side choices for its rivals and resulting in the disappearance of Unión Fenosa as an independent competitor which had been acting as a maverick in the last few years.

Adecreaseincompetitioninthewholesaleelectricitymarket(orpool) as a result of the strengthened position of the merged entity, its increased symmetry with main rivals Iberdrola and Endesa and the higher level of concentration in the markets, fostering the incentives and capacity both for unilateral and coordinated effects.

The integration of two close competitors in the gas and electricity retail supply markets. GasNatural’sstrongsimultaneouspresenceingasandelectricity,togetherwiththeresultingoverlap

in the gas and electricity distribution networks of the two undertakings, could place the merged entity in a highly advantageous position to strengthen its market power in the supply to small customers due to vertical integration between distribution and retail supply as well as portfolio effects.

Thecreationorstrengtheningofstructural links with competitors in the gas end electricity markets through joint-ventures and financial participations.

In addition, the CNC report highlighted the fact that one of the most relevant entry barriers in the gas sector is the need to achieve a minimum efficient scale in terms of customers (CCGTs, large industrial consumers and domestic consumers) in order to ensure stable gas contracts in the upstream markets.

Under the current Competition Act, and in case a merger gives rise to obstacles to the maintenance of effective competition, the notifying parties, at their own initiative or at the request of the CNC, may propose commitments to remove those obstacles, without prejudice to the CNC’s authority to establish further conditions if it believes the proposed commitments are insufficient or inadequate to solve the problems detected. It is also foreseen the possibility of launching a market test among competitors and other undertakings in order to test their opinion as regards the sufficiency and adequacy of the proposed commitments, in line with the procedure followed by the European Commission.

Therefore, on 21^st January 2009, Gas Natural submitted a first proposal of commitments to solve the competition problems identified by the CNC. That proposal was followed by intense negotiations with the CNC which led to the filing of its final proposal on the 10^th February. Commitments were subject to public consultation twice during this negotiation phase, and all submissions were assessed and taken into account by the CNC, leading to subsequent reviews of the first proposal.

At its meeting of 11 February 2009, the Council of the CNC cleared the way for the merger between Gas Natural and Unión Fenosa, subject to the commitments presented by Gas Natural, having found that those commitments could solve the competition problems that were previously identified.

The commitments finally accepted by the CNC’s Council were assessed under the principles of adequacy (the commitments must eliminate the anti-competitive risks generated by the merger), proportionality (their aim is not to offset all competition problems that may exist in the market, but only those arising from the specific transaction) and least intervention (in case there are several options, the one that is easiest to implement should be chosen).

A key feature of this operation was that it affected a great number of vertically and horizontally interrelated markets. Therefore, the adequacy of the commitments presented by Gas Natural was analysed considering them as a whole, with all simultaneous effects that each commitment could have in other affected markets in order to mitigate one or more of the detected potential risks for effective competition. What is more, the assessment of commitments had to bear in mind all the elements considered under the substantive test and not just horizontal or vertical overlaps, but also including the abovementioned entry barrier to gas markets and the structural links with competitors.

Accordingly, the final solution designed included divestitures that would favour the entry or strengthening of operators so as to counteract the disappearance of Unión Fenosa as an effective competitor and contribute to ease the barriers which would otherwise hinder the entry of the acquirer of the divested assets.

As a result, the CNC Council Resolution accepted the following commitments proposed by GAS NATURAL, considering they were adequate and proportionate to solve the competition concerns detected:

To sell complete gas distribution networks accounting 600,000 distribution points (equivalent to 9% market share at a national level).

Toselltheattached portfolio to those distribution points (accounting approximately 600,000 small and domestic consumers).

To sell 2,000 MW of combined-cycle gas turbine (CCGT) power plants.

ToimplementmeasuresinordertoensurethatUnión Fenosa Gas (jointly controlled with ENI) may continue operating independently as a gas supplier in Spain.

TosellitsstakeinEnagás (the gas transmission network operator) and to reduce its ties with

Cepsa, the main competitor of Repsol in oil markets.

The sale of 600,000 distribution points more than offset the size of the distribution network acquired by Gas Natural and the shedding of the attached 600,000 customers far exceeded the acquired market share of gas consumers (94,000 customers aprox.). Furthermore, the divestitures entail complete networks, which will facilitate their autonomous management and make them more attractive to prospective buyers. The CNC concluded that this divestiture would boost the buyer’s competitiveness vis-à-vis the Gas Natural group in carrying out the gasification of new areas and the development of new distribution networks. In addition, the commitment included supply of gas by the merged company to the buyer during a transitory period which was estimated to be enough to permit the latter negotiate supply arrangements with other suppliers on the basis of the assets acquired.

With regard to the power plants divestiture, the commitments implied the release of 2,000 MW of CCGT plants, which would offset not only the horizontal overlap in generation but also the vertical strengthening between gas supply and electricity production. CCGT plants are the ones where the parties to the merger overlapped, and have a strategic relevance given that they set the wholesale (marginal) market price and cover eventual shortfalls of renewable plants. The divestiture was therefore considered adequate because it reduced the strategic position of the new entity in this technology.

The commitments also reduced significantly any post-merger risks of coordinated effects with Endesa and Iberdrola (main players in electricity in Spain). In particular, divestiture of combined-cycle assets would help to maintain an asymmetry in generation portfolios of the new company compared to those of its rivals.

Finally, Gas Natural presented three commitments aimed at mitigating the possible risks arising from the creation or strengthening of certain structural links that the CNC found that could hinder effective competition in the markets. First, the sale of Gas Natural’s stake in Enagás (the gas system operator in Spain) prevented any interference by the merged company in infrastructure planning and management and strengthened the independence of the system operator. Second, the withdrawal from Cepsa’s Board of Directors and the commitment to build Chinese walls for sensitive commercial information of Cepsa were also considered adequate for avoiding any intrusion by Repsol, Cepsa’s main competitor and the biggest operator in the liquid hydrocarbons market.

The merger was finally cleared subject to these commitments on 17^th February 2009, once the Government rejected to intervene on the basis of public interest criteria other than protecting competition.

To sum up, the assessment of competition concerns in the different affected markets of the Gas Natural/Unión Fenosa Case carefully followed the substantive test which has been described in section II of this paper, taking into consideration a great deal of key features far beyond horizontal overlaps and successfully providing a rigorous assessment of the commitments adopted.

Final remarks

The experience in Spain with the substantive test for merger review introduced in 1989 and recentlyclarified under the new framework, has been very positive, allowing the Spanish Competition Authority to apply a flexible test more coherent with the increasing importance of economic analysis.

It seems important to recall that during the discussion meetings prior to the adoption of Regulation 139/2004, Spain, together with France, were the two leading supporters of the change to a closer SLC test at EU level. The French-Spanish proposal, which was formally presented in September 2003, was finally taken into consideration by the Commission and therefore served as basis for the final wording of the current substantive test under Regulation 139/2004. Thus being a much more similar test to the one already applied in Spain since 1989.

THE UK

The submission goes on to consider whether the differences between the UK substantial lessening of competition (SLC) test and the European Community Merger Regulation (ECMR) significant impediment to effective competition (SIEC) test give the UK Authorities greater flexibility on whether or not to conclude on market definition in certain cases. For example, in analyzing the hypothetical ‘gap’ case between the Bank of Investment and the Bank of Commerce under the SLC test, the Authorities may not have found it necessary to conclude on market definition, instead considering several possible market definitions.

Finally, we briefly consider whether other ‘gaps’ may exist, with reference to sequential mergers and minority shareholdings, and note that the flexibility of an SLC/SIEC test makes the Authorities better placed to deal with these issues as they arise than a dominance test.

Objective of UK merger control

The objective of the UK merger control regime is to ensure that a merger does not substantially lessen

competition. The Authorities’ joint draft Merger Guidelines¹

explain that competition is viewed as a

process of rivalry between suppliers (i.e. firms) seeking to win customers’ business over time. Rivalry may take various forms. For example, firms may seek to undercut each other on price, increase output more than rivals, or outperform each other on quality, productivity or innovation to create new or improved products or markets. For customers, rivalry can therefore have many beneficial effects, for instance by driving down prices and improving quality and variety. Any merger will be considered by the Authorities in terms of how rivalry is likely to be affected over time. A merger giving rise to an SLC will reduce the beneficial effects of rivalry, creating one or more ‘adverse effects’ for consumers.

the OFT and CC issued draft joint guidelines on how they assess the competitive impact of mergers – http://www.oft.gov.uk/shared_oft/consultations/OFT1078con.pdf . The publication revises and expands guidance material currently contained in several publications issued separately by the two authorities after the introduction of the Enterprise Act 2002; in particular it will replace the following OFT publications: Mergers-substantive assessment guidance (OFT516), Guidance note revising Mergers- substantive assessment guidance (OFT516a) and Revision to Mergers: substantive assessment guidance- exceptions to the duty to refer markets of insufficient importance (OFT 516b); and the CC publication: Merger References: Competition Commission Guidelines, CC2.

The application of the UK substantive test

In 2002, with the passage of the Enterprise Act (the Act), important changes were made to the assessment of mergers in the UK. The status of the OFT and CC as independent decision-making authorities was confirmed and the broader ‘public interest’ test was abandoned in favour of a new competition test, namely a ‘substantial lessening of competition’ test. Decisions are taken by the Authorities, with a provision for a public interest test retained for certain circumstances.

1. The UK substantive test

The assessment of mergers in the UK is conducted as a two-phase process, giving distinct but interrelated roles to the OFT and the CC. At Phase 1 the OFT assesses whether the merger has resulted or may be expected to result in an SLC. If it considers this is the case it must refer the merger to the CC²; otherwise the merger is cleared. At Phase 2, the CC decides whether a merger has resulted, or may be expected to result in an SLC. If it finds this is the case, the CC may prohibit the merger, order divestiture in a completed merger or apply other remedies. Otherwise, once again, the merger may be cleared. In exceptional cases (see below) it is the Secretary of State (SoS)³ who decides.

1. Theories of harm

‘SLC’ is not defined in the Act, but a merger which is considered to substantially lessen competition over time may be expected to reduce the beneficial effects of rivalry, creating one or more ‘adverse effects’ for consumers. The Authorities would not find an SLC without an expectation of adverse effects for consumers.

The Authorities base the central analytical framework for an inquiry upon identified ‘theories of harm’⁴. Theories of harm are the hypotheses that are tested during an inquiry: the ways in which a merger (or the features of a market) could give rise to consumer detriment typically fall into the following categories:

1. - Unilateral effects: a worsening of the competitive offer by the merged firm or other firms in the market without the need for coordination, due to the loss of an independent competitive constraint. In order to assess the extent to which a merger may give rise to unilateral effects, the Authorities will consider a wide range of factors including the number of firms in the market and their market shares, the closeness of competition between the merging parties, the choice of alternative suppliers and evidence of potential competition⁵.

1. - Coordinated effects: a worsening of the offer by a number of firms within the market, because the merger creates or strengthens the conditions under which they might collude tacitly (or

The Authorities will consider evidence of pre-existing coordination as well as how the merger affects the ability to reach and monitor terms of coordination and the internal and external stability of coordination.

1. - Vertical or conglomerate effects: a lessening of competition that harms consumers, resulting from a merger of suppliers of products or services which are not substitutes for one another. These may be inputs to one another (a vertical merger) or complements for, or unrelated to, one another (a conglomerate merger). In vertical merger cases, the Authorities will consider how the merger affects the merging parties’ incentives and ability to foreclose customers and/or inputs. Similarly, for conglomerate mergers the Authorities will consider how the merger affects the parties’ ability and incentive to foreclose the market through tying and bundling.

In some cases, the Authorities may consider several concurrent theories of harm⁷. The Authorities do not carry out their analysis against these generic theories of harm but develop and apply them on a case by case basis based on the evidence available to identify specifically how consumer harm may arise.

1. The ‘public interest’ test

Before 2003, merger control under section 69 of the Fair Trading Act 1973 (FTA) was based on a ‘public interest’ test allowing the Authorities to take into account, in addition to competition matters, certain non-competition matters such as public security, regulatory concerns and environmental considerations, although in practice most mergers were decided on competition grounds, with the final decision taken by Ministers.

A mechanism was retained under the Act for dealing with special public interest cases, involving a principal role for Ministers and an advisory role for the Authorities. In each case the role is specific and closely delineated. For certain specified and limited areas of public interest, Ministers may intervene in mergers to require that investigations cover issues other than competition matters. In such cases, whilst the Authorities will investigate and report on the effects of the merger on competition, Ministers retain ultimate decision-making power after taking into account both competition and public interest considerations. Until 2008 this power to intervene was limited to cases concerning national security (which includes public security) and certain media issues—freedom of speech, accuracy of presentation of news in newspapers, and media plurality. In 2008 the interest of maintaining the stability of the UK financial system was added to the list of public interest considerations.⁸ Ministers may add further public interest criteria, but only by passing secondary legislation subject to approval by Parliament.

These powers to intervene to protect the public interest have most often been used in mergers involving the defence industry, although the powers have recently been used in relation to issues of media plurality in the case of BSkyB/ITV⁹ and in relation to the stability of the UK financial system in the case of Lloyds/HBOS¹⁰.

1. Effect of the change of test

The change from a ‘public interest’ test to an ‘SLC’ test has not resulted in significant changes in the way mergers are assessed in the UK. This is due to the way the public interest test was implemented. In 1984, Norman Tebbit, then Secretary of State for Trade and Industry¹¹, stated, in what subsequently became known as ‘the Tebbit doctrine’, that his policy would be to make merger references primarily on competition grounds. This was reconfirmed by subsequent Ministers and final decisions were similarly based.

Enforcement issues connected to the dominance and the SLC test

1. Structural versus economic analysis

Under both a dominance test and a SIEC or SLC test a wide range of similar factors will be taken into consideration before an adverse finding is made, such as choice of alternative suppliers, switching costs and the importance of the competitive constraint being removed. However, the choice of test makes a difference in the roles played by market definition and concentration data in merger assessment.

The identification of a dominant position within a defined market is an essential part of the dominance test. The SIEC test used in the ECMR moves away from focussing solely on market definition to focussing on the actual loss of competitive constraint and the resulting effect on consumers. This enables the ECMR to capture those mergers which may result in consumer harm as a result of a reduction in competition, but where the market share of the merged firm does not reach the threshold for dominance – the so called ‘gap’ cases. However, the SIEC test under the ECMR continues to refer specifically to the creation or strengthening of a dominant position as a particular example of how competition may be impeded as a result of the merger, indicating that a structural analysis may still carry significant weight in the final SIEC decision.

The UK SLC test and the ECMR SIEC test are framed differently in that the UK SLC test makes no reference to the creation or strengthening of a dominant position as a specific way in which the reduction in competition may arise. The UK’s new draft joint Merger Guidelines indicate that defining the relevant market is intended to provide a helpful framework for assessing the relevance of different constraints, but that it is not an end in itself and unilateral effects do not necessarily turn on a particular market definition. As such the Authorities may decide not to conclude on market definition and may instead consider several alternative markets as part of the investigation.

Specifically, in differentiated goods markets, drawing precise market boundaries runs the risk of either overstating or understating the degree of competitive constraint between specific suppliers. Instead the Guidelines state that the emphasis should be on loss of rivalry, focussing on closeness of competition. The analytical process of considering market definition may help to identify the closeness of competition between the merging parties, but the Authorities will also consider all the available evidence to assess the loss of competitive constraint resulting from the merger.

Whilst most UK merger cases still do define a relevant market, there are a number of recent UK examples where the Authorities have not concluded on market definition, focusing instead on the direct rivalry between the merging parties. In Lovefilm/Amazon¹² (proposed acquisition of Amazon’s online DVD rental subscription business), the OFT cleared the acquisition noting that the relevant frame of reference may be wider than only on-line DVD rental, but given the various alternative means of accessing video content it was not necessary to conclude on a precise definition. More recently, in a merger that has been referred to the CC (NBYT Europe/Julian Graves¹³), the OFT considered that whilst the evidence available to them, including critical loss analysis, pointed to a narrow market definition of nuts, seeds and fruit from specialist retailers, it was not necessary to conclude on market definition given the closeness of competition between the merging parties. Instead the OFT placed particular importance on evidence of consumer behaviour, including estimates of diversion ratios, and assessed whether the alternative choices of customers would be sufficient to constrain the merged entity such that it would be unable to raise prices or reduce non-price factors (quality, range and/or service). This case is currently being investigated by the CC.

In Hamsard/Academy Music¹⁴ the CC did not conclude on the precise product or geographical market for live music venues in London but looked at the competitive constraints from other venues on each of the venues owned by the parties. Similarly, in Vue / A3 Cinema¹⁵ the CC did not conclude on whether

multiplex and non-multiplex cinemas were in the same market or on the precise boundary of the geographic market and considered the competitive constraint faced by individual cinemas on a case by case basis.

Although it may not be necessary to decide on the precise boundaries of the relevant market when the Authorities would reach the same conclusions as to the effects of the merger under different market definitions, this is less likely to be the case when the CC concludes that a merger has resulted or may be expected to result in an SLC and remedial action is required.

‘Gap’ cases

The Authorities consider that the focus of the dominance test on single firm conduct or groups of firms acting collectively to imitate the policy of a monopolist makes it unsuitable for dealing with non- collusive oligopoly mergers, in particular where the full anti-competitive effects are not confined to the merging parties. For example, where two producers of close substitutes in a differentiated product market merge and the loss of this competitive constraint results in the merged entity being able to increase prices. Other firms in the market respond by increasing their prices and this process or reaction and counter- reaction by the merged entity results in a new equilibrium price which exceeds the direct effect on the merging parties. Such cases may be considered to result in consumer harm, despite the merging firms only having relatively small shares of supply on the wider market. These are considered to be ‘gap’ cases¹⁶.

1. Examples of ‘gap’ cases

In its 2002 submission to the OECD the UK identified the Lloyds/Abbey banking merger as a case which would not have met a single firm dominance test but would have been caught under the SLC test. The merger¹⁷ was prohibited by the SoS on the recommendation of the CC on grounds of a reduction in competition in the personal current account market which was expected to result in adverse consequences for consumers. It would not have met a single firm dominance test¹⁸; however, the CC noted that the four leading banks would have had a combined market share of 77 per cent following the merger, that competition in the market was lacking and that the merger would have removed one of the main sources of competition to them. The CC did not consider the issue of collective dominance.

The Authorities have not observed any other clear ‘gap’ cases since, but note that these cases may be considered to be less observable, as ‘gap’ cases involve Type II errors (an anti-competitive deal being let through) where there may be fewer protests, compared to Type I errors (a deal wrongly prohibited) where the merging parties would have clear incentives to protest.

1. A hypothetical ‘gap’ case for discussion – A bank merger

The hypothetical bank merger outlined in the Issues Paper¹⁹ between Bank of Investment (BoI) and Bank of Commerce (BoC) displays strong similarities to the prohibited UK merger between Lloyds TSB and Abbey National in 2001 and the permitted merger between Lloyds TSB and HBOS in 2008²⁰.

In the hypothetical BoC/BoI merger, BoC and BoI would cease to be distinct and the estimated market shares for the provision of banking services is in excess of 25 per cent. As such, the share of supply test in section 23(4) of the Act would be satisfied.

The UK Authorities would be likely to consider unilateral effects arising from the elimination of actual and potential competition as well as potential coordinated effects. Market definition would be analyzed, but it may be the case that the Authorities would not conclude on the precise market boundaries. In this case the findings would be unlikely to turn on market definition and the Authorities would most likely consider a range of possible market definitions. The framework for analysis of competitive effects would be unlikely to differ substantially whether individual banking services are considered or whether the Authorities consider a wider ‘banking services’ market.

In assessing horizontal unilateral effects, factors likely to be considered include the number of firms in the market, market share (potentially across a range of markets), the closeness of competition between players in the market including rivals’ reactions to past events, the competitive strength of the target firm, customers’ choices of alternative suppliers, and the scope for potential future competition. In particular, estimations of diversion ratios, either through econometric analysis or consumer surveys, combined with gross margins would give a strong indication of closeness of competition. The Authorities would also consider evidence of customer switching and competitor monitoring by the merging parties.

In assessing coordinated effects, the Authorities would consider whether, following the merger, conditions in the market would be such as to create or increase the ability and/or incentive for firms in the market to coordinate their behaviour.

We foresee three potential issues in relation to the so called ‘maverick’ reputation of BoI which would require further consideration. Firstly, BoI’s current market share may underestimate the competitive constraint it exerts on BoC and other players in the market which would also feed into the potential competition assessment. Secondly, the merger may make coordination more likely by removing a maverick that would otherwise disrupt it.²¹ Finally, however, the relatively recent entry and strong growth of BoI may also indicate that barriers to entry and expansion are relatively low. The Authorities may therefore also want to look at whether the entry and growth path of BoI could be easily replicated in the relevant timeframe.

Broader policy considerations

There remain a number of areas where it is difficult to ensure that all mergers which in the long-run may be harmful to consumers are captured; however, we consider an SIEC or SLC test is better placed to capture these cases as it offers a greater degree of flexibility than a dominance test. Two broad issues are discussed below.

1. Dynamic merger review

Merger proposals are endogenous and occur over time. Approving a currently proposed merger will affect the profitability and welfare effects of potential future mergers, the characteristics of which may not yet be known.

Under the SLC test mergers are typically considered on an incremental basis. There are questions about how Authorities should consider waves of mergers, especially in the current economic climate, as a stream of small mergers may at least in principle result in an overall detriment to consumers. It will usually be possible to identify a particular transaction within a series which results in an SLC. However, a situation might arise where a series of small consolidations, each only causing a limited lessening of competition, gives rise to a substantial detriment to consumers without any particular transaction in the series resulting in an SLC.

To date there has not been any need for the Authorities to address the issue head on as the Authorities’ processes have been sufficiently flexible to adapt appropriately to all circumstances that have arisen. However, where markets are evolving and changing rapidly and transactional activity is high, it may be appropriate for competition authorities to take a more dynamic approach to merger review – although we note the results in Nocke and Whinston (2008)²² which suggest pairwise comparisons may be sufficient.

1. Minority shareholding²³

The UK merger jurisdiction recognises three levels of ownership interest: (i) a controlling interest, understood as de jure control by acquisition of greater than 50 per cent of the voting rights; (ii) the ability to control policy, understood as de facto control with shareholdings below 50 per cent; and (iii) the ability to materially influence policy of the target firm (material influence) including the strategic direction of a company and its ability to define and achieve its commercial objectives.

The threshold for material influence is flexible, with no specific threshold for determining whether a shareholding would confer material influence or not. For minority shareholdings below 25 per cent, the Authorities consider other factors such as interlocking directorships, cross-shareholdings and asymmetries in shareholdings.

For example in BSkyB/ITV²⁴, the CC considered whether the absolute and relative size of BSkyB’s shareholding would give BSkyB the ability to block special resolutions and whether the industry knowledge and standing of BSkyB, combined with the absolute and relative size of its shareholding, might give it the ability materially to influence the strategy of the ITV board.

Under the SLC test, the UK Authorities have indentified unilateral, coordinated and vertical theories of harm arising as a result of the acquisition of less than full control. Unilateral effects have arisen where the acquisition results in a direct loss of an independent competitive constraint, reducing the acquirer’s incentives to compete; where the shareholding prevents or influences the strategic decisions of the target firm (‘strategic’ unilateral effects); and where the shareholding prevents a full acquisition by a rival firm. It would generally be expected that any efficiency gains that may be obtained through a full acquisition are unlikely to be realised by a minority shareholding.

The potential for unilateral effects to arise through the frustration of rival bids or by reducing the ability of the target firm to act independently in deciding its short- or long-term strategy may indicate that small minority shareholdings may still have the potential to result in consumer harm and may not require the acquiring firm to hold a dominant position in an overlap market. This may support a tougher assessment of the criteria considered in reaching a finding of material influence under an SIEC/SLC test.

UNITED STATES

In the United States, mergers have been challenged under two laws. Section 7 of the Clayton Act of 1914, described in detail below, specifically addresses anticompetitive acquisitions and has long been the primary basis for merger challenges. Section 1 of the Sherman Act of 1890 prohibits certain agreements and also can be used to challenge mergers. We address Section 1 as part of a discussion of changes to the

U.S. substantive merger standard over time, later in this submission.

The substantive merger standard in the United States

Section 7 of the Clayton Act currently provides that:

No person engaged in commerce or in any activity affecting commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital and no person subject to the jurisdiction of the Federal Trade Commission shall acquire the whole or any part of the assets of another person engaged also in commerce or in any activity affecting commerce, where in any line of commerce or in any activity affecting commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.¹

Section 7 was intended to serve as “an effective tool for preventing” anticompetitive mergers.² The federal agencies that share merger enforcement responsibilities—the U.S. Department of Justice (DOJ) and the Federal Trade Commission (FTC) (collectively referred to as the Agencies)—believe that Section 7 can and does serve as intended. Section 7 covers “the entire range of corporate amalgamations”³ as well as all anticompetitive effects flowing from them.

The substantial lessening of competition (SLC) standard in Section 7 prohibits mergers and acquisition reasonably likely to produce significant anticompetitive effects.⁴ All mergers and acquisitions are “tested by the same standard, whether they are classified as horizontal, vertical, conglomerate or other.”⁵ “Merger enforcement, like other areas of antitrust, is directed at market power.”⁶ “The lawfulness

EC

This submission, firstly, describes the standard merger review test applied under the EC Merger Regulation by the European Commission and the background to the reform (section 1) and, secondly, provides an overview on the Commission’s merger review practice before and after the introduction of this standard test (section 2). It then provides some general conclusions (Section 3).

The change of the test: from dominance to SIEC test

In December 2002, the Commission decided to reform the EU merger control system. It followed ayear of consultation and debate on the basis of a Green Paper¹ which, among other issues, launched a reflection on the merits of the substantive test enshrined in Article 2 of Council Regulation (EEC) No 4064/89 (“the old Merger Regulation”).² In particular, Article 2(3) of the old Merger Regulation provided that the European Commission (“the Commission”) must appraise mergers with a view to establishing whether or not they would “create or strengthen a dominant position as a result of which competition would be significantly impeded in the common market or a substantial part of it“. The test used at the time was therefore the so-called dominance test.

The Green Paper launched a debate on the effectiveness of the dominance test compared with the substantial lessening of competition (“SLC”) test used in several other jurisdictions (in particular, in the USA). Respondents to the consultation argued both for and against change. Those who pleaded for a change to a SLC-type test mainly argued that the test would be more appropriate for dealing with the full range of competition problems that mergers can give rise to, and in particular that there may be a gap in the scope of the test in Article 2. In their view, it was clear that the dominance test applied to mergers creating or strengthening the leading market player (single firm dominance). It was also clear that the Commission could intervene under the notion of collective dominance against mergers in oligopolistic markets. However, they considered that the scope of application of the dominance test could not extend to concentrations giving raise to non-coordinated effects in the absence of single firm and/or collective dominance.

Following this debate, the solution retained by the EU legislator was to adopt the so-called substantial impediment to effective competition (“SIEC”) test whilst retaining the creation or strengthening of a dominant position as a principal example of the application of the test. Article 2(3) of Council Regulation (EC) No 139/2004 (“the Merger Regulation”)³ therefore provides that the Commission must assess whether

EC Conclusion

The application of the SIEC test in the Merger Regulation and the analytical framework provided by the Horizontal and Non-Horizontal Merger Guidelines has proven to properly cover the great variety of merger scenarios encountered with and to assess the resulting competitive effects.

The introduction of the SIEC test in the Merger Regulation confirmed the Commission’s already undertaken shift from a somewhat structural approach based on market shares and concentration levels towards an effects-based approach. This approach is the result of a gradual development which commenced well before the introduction of the SIEC test into the Merger Regulation. This can be seen from an examination of cases before and after the introduction of the SIEC test and not the least from available statistics which indicate a similar degree of intervention before and after its introduction.

FINAL Conclusions

The Competition Act does not encompass general procedures for defining markets or rules for identifying and measuring factors of economics-based analysis. However, it is apparent from the commitment and consistency of both the FNE, in the rigorous application of its first guideline, and the TDLC, in its decisions, that despite the absence of a binding regulation detailing the substantive analysis that the Tribunal may undertake when assessing merger transactions, that the underlying guiding principles and factors that will be taken into consideration can be ascertained.

Moreover, it may be said that the substantive analysis performed by the Chilean Competition Authorities, although probably not strictly fitting within the traditional tests (i.e., SLC and Dominance), encompasses elements of both, which make it sufficiently sound and efficient as a tool to perform merger reviews.

The legislative changes to the Competition Act, the last of which took place in 2004, and although none of these changes specifically affected the merger review procedures, are a strong indication of the shift towards a more sophisticated economic approach, which has clearly influenced the way merger reviews are performed.

Merger control is based mainly on voluntary consultations by the merging parties. Therefore, there are no big risks of over enforcement.

SUMMARY

Frederic Jenny opened the roundtable discussion on the standard for merger review, with a particular emphasis on country experience with the change of merger review standard from the dominance test to the SLC/SIEC test, which attracted a lot of interest among the delegations.

Before the discussion with the delegates, the Chair introduced the three experts who were invited to give initial presentations. Nicholas Levy, partner at Cleary Gottlieb Steen and Hamilton; John Boyce, partner at Slaughter and May; and Alberto Heimler, professor of economics at the Italian School of Government. Mr Levy and Mr Boyce discussed their experience with EC merger control and the changes in the substantive test. Mr Heimler presented a paper with his views on how the SLC test should work in practice.

According to Mr. Levy the standard of review of mergers is important as it distinguishes mergers that should be prohibited from those that ought to be approved. In addition, the legal standard used to assess mergers gives an indication as to how the analysis should be carried out by the competition authority.

The first EU merger regulation was adopted in 1989 after a vigorous debate, which did not focus much on whether the merger test should be based on the creation or strengthening of dominance or on the merger significantly lessening competition, but rather on whether social or industrial considerations should be taken into account. At the time, the European Commission was adamant that mergers should be assessed only on competition criteria and in that respect succeeded in both formulating the EC Merger Regulation and its subsequent application appropriately. Indeed, in the early years the Commission was watched as to whether it would be subject to political interference in its application of the Merger Regulation but successfully resisted any political pressure.

As regards the substantive test, in the early days there was also uncertainty as to whether the dominance test would be capable of capturing all anti-competitive transactions. The debate at the time was not on whether there were so-called “gap cases” (i.e. cases of anti-competitive mergers which would not be caught by the dominance test) but whether the dominance test could cover coordinated effects (or joint dominance cases) because the test was expressed in the singular – “the creation or strengthening of a dominant position”. By taking a logically expansive interpretation, the European Court of Justice confirmed, 10 years later, that coordinated effects were indeed captured by the dominance test.

However, over the years, controversy continued as a result of various contested cases. But the debate was not on the substantive test, rather it focussed on the framework within which the Commission assesses evidence, the system of internal checks and balances to which the Commission’s decisions are subject, and the concerns with the Commission being at the same time investigator, prosecutor and decision maker.

As part of the review of the Merger Regulation, there was a debate between 2001 and 2004 on the substantive test and whether to replace the dominance test with an SLC type test. The European Commission took the position that dominance was sufficiently broad to capture all transactions that could raise competition issues; however, to eliminate all remaining doubts, the European Commission acknowledged that the dominance test could be changed. After considerable discussion, the European

Commission adopted a hybrid test – “significant impediment to effective competition in particular through the creation or strengthening of a dominant position” (the SIEC test). In doing so it took the view that a hybrid test encompasses the best of both worlds. On the one hand, it ensured continuity of years of case law on the dominance test and, on the other hand, it makes it clear that the focus of the merger review will be on the loss of competition from the transaction.

After five years of application of the SIEC test, Mr. Levy noted that it is appropriate to ask whether this experiment has worked. There are four questions that arise: (i) Are there material differences between the dominance and SIEC standards? (ii) Is there an “enforcement gap” between dominance and SIEC? (iii) Did the change in substantive test materially alter the EU merger review system? (iv) Has the new EU merger standard facilitated international convergence in merger control?

In Mr. Levy’s opinion, the differences between the dominance and the SLC standards are theoretical as much practical. The risk with the concept of dominance is that the notion of independence on which it is based is insufficient to distinguish a dominant firm from a non-dominant firm. Even a monopolist is not entirely independent, because unless demand is constant it runs the risk that any rise in prices will lead to loss in sales. If, on the other hand, dominance means just strength per se, than any increase in strength could be taken as the strengthening of a dominant position. Mr. Levy praised the Commission for having applied a flexible and dynamic approach to dominance, one which was not unduly focused on structural factors alone. In addition, according to Mr Levy, the Commission understood the importance of unilateral effects and the assessment of the closeness of competition between the parties already under the dominance test. However, the virtue of an SLC standard is that it focuses on the degree of change in the dynamics of competition, which is what merger control in essence is.

Is there an “enforcement gap” between dominance and the SLC standards? In early 2000 there was a lively debate about the so-called “gap cases”. The hypothetical discussed in the Issues Paper of the Secretariat is an example of a potential gap case. In practice, not many transactions fall into this category but there have been some, such as the Heinz/Beech-Nut case. But there were always ways to capture the transactions falling in the “gap”, either by adopting a sufficiently narrow market definition, or by developing a theory of harm based on the strengthening of dominant position or possibly developing a coordinated effects theory. However, given the uncertainty on whether courts would endorse these theories the Commission acknowledged there could, at least in theory, be an enforcement gap, and therefore considered the possibility of changing the merger review standard.

Did the new substantive test materially change the EU merger review system? The EU merger control system experienced many changes over the past five years: an increasing reliance on sound economics and hard evidence, a more forensic and rigorous review of that evidence and a more consistent focus on unilateral effects cases. In Mr. Levy’s opinion, these changes were not caused by the change in substantive standard but simply accompanied it. In fact some of the other developments were actually more important, such as the adoption of horizontal and non-horizontal merger guidelines, which have clarified the standard applicable to conglomerate and vertical effects of mergers, the high standard of proof to which the Commission was subject by the courts, the appointment of a chief economist, the increased focus on unilateral effects, the greater emphasis on quantitative analysis and the increased internal scrutiny (checks and balances). Mr. Levy suggested that these developments had a more significant impact on EU merger control than the change in the substantive test, which was an important development that helped address possible (but rare) gap cases. However, at the end of the day it did not make as much of a difference as some of the other changes.

Mr. Levy noted that the question of whether the change in the substantive test facilitated international convergence is a particularly interesting one to be addressed at the OECD. Certainly, in recent years, there has been more convergence in merger decisions. However, there is no evidence that that this trend is due to

the adoption by the EU of a substantive test more similar to that used by other agencies. Rather, Mr Levy noted that there has been an increased alignment of analytical tools and economic theories as to what merger control should be about. Diverging outcomes – the extreme example of which is the GE/Honeywell decision where there were analytical differences in the approach to conglomerate effects – were not the result of different substantive tests. According to Mr Levy, the co-existence of different substantive tests is not as important as the increasing convergence on analytical tools and economic analysis. For a practitioner, differences in jurisdictional thresholds across different countries may have far more significant implications.

To sum up, according to Mr. Levy the differences between dominance and SLC standards are in practice primarily differences of emphasis and intellectual interest. There is a small enforcement gap between dominance and SLC standards in non-collusive oligopoly cases, but those are rare and there are ways to captured them under the dominance standard. The change in and of itself had only modest practical effects on EU merger control. However, together with other changes there has been an important evolution. International convergence in merger control has been occurring notwithstanding the differences in substantive tests.

John Boyce said that the changes in the UK and the EU, which have occurred since the mid-80s, were more evolutionary than revolutionary. For most of this period, the UK actually had a public interest test but in practice decided cases on competition grounds. Therefore, the change of the SLC test with the 2002 Enterprise Act was to adapt the word of the law to the practice. A more significant change was perhaps the greater independence that the Enterprise Act granted to the OFT and the Competition Commission, which now can take their own decisions rather than merely advising the Secretary of State. However, in the UK it is also recognized that there may be wider public interest issues, which could justify intervention in mergers.

At the EU level, there have been also significant changes since the mid-80s. Initially, the European Commission had no merger review powers, yet occasionally it intervened to sanction mergers under Article 81 and 82 EC. The first merger regulation gave the European Commission express powers to review and approve mergers under the dominance test. At the outset, there was some uncertainty as to the scope of the concept of dominance, but as was mentioned by Mr. Levy, the Commission, with the backing of European courts gradually stretched the notion of dominance to cover situations involving joint dominance or coordinated effects. Later, a series of cases lead to the debate on the so-called “gap” which lead to the change of legal standard in favour of the hybrid SIEC test. Mr. Boyce noted that as a matter of fact, the European Commission successfully extended the concept of dominance to reduce the gap to a minimum. In practice, gap cases have been very few or, according to some, non-existent. Nevertheless, the change brought clarity particularly because in some jurisdictions, courts were not supporting an extensive interpretation of the concept of dominance. In Mr. Boyce’s opinion, the new test certainly gave the European Commission and its Chief Economist team more freedom to use various econometric techniques in applying the new SIEC test.

Mr. Boyce mentioned that another significant change in merger control in the past two decades has been its internationalization. In the mid-80s even the largest global transactions had to be notified only in a few countries. Presently, significantly more transactions are subject to multiple notifications in an increasing number of jurisdictions, which delays the clearance process and increases legal and regulatory costs. Mr. Boyce noted that overlapping reviews are also likely to require significant resources within the competition agencies involved. This is an area where closer international cooperation in the future could help reduce the time and costs of the review.

Mr. Boyce emphasized that just as traffic rules are not an end in themselves but exist to achieve public welfare objectives so merger rules are there to reduce the risk of consumers being harmed by

anticompetitive mergers. Merger rules need to be sufficiently clear to provide merging businesses with legal certainty and be enforced sensibly. It is reasonable that competition authorities should have a certain degree of discretion when enforcing merger control rules. However, agencies’ discretion needs to be subject to proper checks and balances. In terms of harmonization, it is not necessary that all countries have identical rules, but in an increasingly global environment, it is sensible for authorities to understand each other. Having a similar standard can certainly contribute to that.

Looking back to 2002, the “annus horribilis” for the European Commission when three merger cases were lost in court, it was said that for a speedy and effective merger review, there was a need for a proper legal framework, clear merger guidelines, a properly staffed administration and a rapid and independent review process. It is notable that a few years later, in a presentation summarizing the changes in the new EU merger regulation, Mr. Boyce devoted significantly more space to the changes in jurisdictional thresholds and procedural changes than to the change in the substantive test. At the time, Mr. Boyce was concerned that the European Commission may attempt to stretch the boundaries of the SIEC test as it did in the past with the dominance test, making the outcome of the merger review process uncertain and unpredictable. The European Commission has not done that and today the merger control regime in the EU is reasonably predictable. With the next review of the EC Merger Regulation coming up, the European Commission is about to issue a report on how the Merger Regulation is operating in practice. The fact that no radical procedural or substantive developments are expected implies that the system has worked relatively smoothly.

Mr. Boyce concluded by saying that the evolutionary changes in the UK and EU merger review towards an SLC type test have been a success. The SIEC test in the EU is more “intellectually honest” than the Commission’s previous practice when it stretched the concept of dominance beyond what some would see as its limits. By adopting the SIEC test the merger control standard can be applied independently of the Article 82 EC standard, which uses the word dominance as well but which concerns the control of monopoly power. Finally, keeping the reference to dominance in the hybrid test has facilitated the transition from the old system to the new and ensured continuity with a developed body of case law. Increased harmonization within the European Competition Network (ECN) is desirable as it facilitates cooperation, including case referrals under the EC Merger Regulation. At a more global level there is scope for further cooperation, and the alignment of substantive standard can help with that. Moving to the SLC test can also increase the credibility of a competition authority in the eyes of other authorities that already have an SLC standard. That is what happened when the UK moved away from the public interest test to the SLC standard.

Alberto Heimlerbegan his presentation by mentioning that his remarks refer to his paper entitled“Was the Change of the Test for Merger Control in Europe Justified 4 Years after the Introduction of the SIEC Test?”. The paper was a response to an article on the same topic published by Greg Worden in the same issue of the European Competition Journal.

In the debate on whether to change the merger test from dominance to SLC, Mr. Heimler took the minority position of defending dominance quite strongly. However, Mr. Heimler noted that when looking forward, one is often proven wrong by subsequent events, which happened to both himself and to those who promoted the SLC standard. His presentation focused on why he was wrong and the reasons for which it would have been better to stick with the dominance standard, which are however different than those he had defended originally.

Mr. Heimler mentioned the exchange he had with those who argued that the dominance standard does not adequately encompass all possible competition concerns, particularly with respect to non-collusive oligopoly mergers. Mr. Heimler argued that an SIEC test adds very little in terms of reducing false

negatives but greatly increases the risk of false positives. However, Mr. Heimler stated, both views were proven wrong.

As regards dominance, Mr. Heimler thought that in Europe there was a problem with its definition. The European Court of Justice in United Brands defined dominance in terms of independence of behaviour. But not even a monopoly is independent as it is bound by the demand curve and it cannot raise prices without a limit. The notice on the definition of the relevant market, adopted by the European Commission in 1997 endorses a more economic approach of dominance. In that context dominance is the ability to raise prices by a small but significant non-transitory increase in price (SSNIP). The notion of dominance was hence shifted from independence to market power. A merger to monopoly can lead to a smaller increase in price than a SSNIP and this was one of the reasons why Mr. Heimler believed at the time that a change to SIEC standard was dangerous.

With respect to unilateral effects in the market for homogenous products, Mr. Heimler sees no problem in applying the dominance test because in a Cournot-type competition a merger always increases prices unless it is associated with price efficiencies. The dominance standard therefore works well when a merger is large enough.

In a differentiated product market, mergers among close competitors (that internalise sales) may very well increase prices. If a SSNIP test shows that a merger would not lead to the requisite change in price, then the merger should not be blocked. Mr. Heimler referred to a recent paper by Baker and Shapiro, which suggests that when identifying an anticompetitive merger one should look not at market shares but rather at the diversion ratio between the merging firms. This is analogous to a SSNIP test and therefore, if dominance is defined in terms of market power – as the ability to raise prices – then it is capable of addressing all potentially anticompetitive transactions in differentiated product markets.

Mr. Heimler noted that in the EU no merger short of dominance was blocked since 2004. Hence from an empirical point of view there has been no need to change the test. An additional concern was that the broad concept of dominance in the merger review could have created repercussions for abuse of dominance cases (Article 82 EC). However, in Mr. Heimler’s opinion, the main issue in the EU is that there has been a truncated analysis in the assessment of an abuse, i.e. the finding of abuse has often been based on the existence of dominance alone. Hence, there is not a problem with the breadth of the concept of dominance itself but rather with the way abuses are analysed.

Mr. Heimler then noted that in the EU there is a problem with collective dominance, which has become a real issue after the 2002 Airtours/First Choice judgment of the Court of First Instance. In that case, according to Mr. Heimler, the court set a very stringent test for collective dominance, which basically requires a showing that firms have a common policy and present themselves in the market “as a single entity”. In Mr. Heimler’s view, this is a very hard standard to meet, and therefore no merger has been prohibited under a collective dominance theory since then. Mr. Heimler went on to note that in the US coordinated effects are considered in a somewhat less structured manner and would cover, for example, the acquisition of a maverick. The acquisition of a maverick in the US is considered under coordinated effects, while in Europe the analysis is made under unilateral effects.

Frederic Jenny thanked the panellists for their very interesting presentations and opened the floor to discussion with the delegates. He suggested that the discussion should focus on five areas: (i) the scope of dominance and the SLC standards, (ii) examples of “gap” cases, (iii) enforcement issues under both tests,

(iv) the role of economic analysis under both tests and (v) other policy considerations and international cooperation.

The scope of the dominance test and of the SLC test

On the scope of the two substantive tests, Mr. Jenny noted that the SLC test can cover both unilateral and coordinated effects. It is however more controversial whether the dominance test allows this. On this issue, he noted that the submission from Switzerland, which applies the dominance test, reported on the conclusions of the 2007/2008 evaluation of the Swiss Cartel Act, which indicated that a risk exists that anticompetitive mergers may be approved. Mr. Jenny asked the Swiss delegation to discuss these conclusions and briefly comment on the Swissgrid and BZ/20 Minuten cases.

Switzerland began with a preliminary remark on the relatively high jurisdictional thresholds applicable in Switzerland. This means that mergers caught in Switzerland are mainly international and are reviewed also by the European Commission. In such situations, the Swiss authority cooperates with the European Commission if the parties to the merger agree, which they generally do. However, this also means that many smaller but potentially anticompetitive mergers escape review. Unfortunately, this is something than only the legislature can change by amending the competition act and it shows that the substantive test itself is not the only problematic issue in this area.

Under the applicable merger test, the Swiss Competition Commission can prohibit a merger which creates or strengthens a dominant position and is capable of eliminating effective competition in the relevant market. This is a wording that can be interpreted either along the lines of an SLC test or more in accordance with a strict dominance standard. The Competition Commission uses the former interpretation. It defines the relevant market but also looks at neighbouring markets, dynamic effects and so on.

In two cases this analysis led to prohibition decisions, which were challenged on appeal. The reviewing court applied a strict interpretation of the test, and ruled that effective competition was not eliminated where there was some competition left. Therefore, the question now is what “effective competition” means. In the BZ/20 Minuten case the court found that since there was no competition prior to the merger there would be no change post merger. Therefore, its interpretation of the test may be quite fact specific and hence not indicative of the general reach of the current wording.

However, following these two rulings there are in principle two possibilities for the Competition Commission to pursue. It can either continue to apply the test as before and distinguish the two cases on the facts, or it can pursue a legislative change of the test, which is a possibility endorsed by the evaluation group that drafted the 2007/2008 report. For the sake of legal certainty, the latter option would be preferable. However, it is not clear that the legislator is ready to amend the Cartel Act. In any event, the Swiss experience so far shows that the wording of the test in itself is not as important as the way it is applied in practice.

The Chair thanked the Swiss delegation and mentioned that some countries have a test that is between the dominance and SLC test. One of these countries is Spain where the test is close to the SLC standard but not the same. He asked Spain to explain what test is applied to mergers in Spain and why it was recently changed.

Spain stated that it applies a test that is slightly different from the SLC test. The first Merger Act adopted in 1989 applied a standard to mergers, which looked at whether the transaction hindered the maintenance of effective competition. The new 2007 Merger Act created a new independent Competition Commission and slightly changed the wording of the test. The new test focuses on possible impediments to the maintenance of effective competition. In contrast to the test in the EU it does not contain references to the creation of dominance or references to substantial effects or harm. Nor does it define what is meant by effective competition. However, in practice the test is applied restrictively, i.e. only mergers which present serious competition concerns are challenged.

The other changes introduced in the 2007 Merger Act include clarifications of several important factors in the assessment of mergers, such as real or potential competition, entry barriers, countervailing power of demand or supply, efficiencies and so on. In addition, it provides that only for reasons of public interest, such as national security, public health or the protection of the environment, the Council of Ministers can overrule the decision of the Competition Commission. Therefore, unlike the previous Competition Act, Government intervention can not be due to competition concerns and it is only likely to happen in very exceptional circumstances.

The Chair passed the floor to the UK delegation which in its contribution said that it applies an SLC test. However, the test applied in the UK is slightly different than the SIEC test applied by the European Commission because it contains no reference to dominance. The Chair invited the UK delegation to comment on whether this was only a matter of wording or whether there could be differences in the enforcement and specifically to discuss the Lovefilm/Amazon and Hamsard/Academy Music cases to illustrate how the test is applied in the UK.

The UK noted that prior to the adoption of the 2002 Enterprise Act, there was neither dominance nor an SIEC test in the UK, but rather a public policy test, which was interpreted on competition grounds. The important development in the last six to seven years in the UK, as in other jurisdictions, was a shift from a structural analysis of mergers toward a more economics-based approach. Greater attention is now devoted to defining the actual theory of harm, the counterfactual to the merger, as well as looking at internal documentation to establish how the parties view the market. When assessing mergers the focus is now on rivalry, specifically the closeness between the merging parties and their rivals and the competitive constraints they face. As a consequence, in the UK there have been a number of cases where it was not necessary to define the relevant market.

The Hamsard/Academy of Music case, for example, involved live concert venues and a precise product or geographic market in this instance was not thought to be not necessary. In that case, there were two services involved, live music production and live music venue management, and the reviewing agency thought it more helpful to look at the characteristics of individual concert venues and the constraints faced in each instance. In some cases, just by looking at these individual characteristics it was clear that the merged entity would have had a significant ability to raise prices. The UK delegation noted that this case illustrated well the proposition that looking at competitive constraints can be a much more direct way of assessing a merger. The Lovefilm/Amazon case was interesting as a number of factors were looked at and eventually through survey evidence and internal documents it became clear that the main competitive constraint the parties faced came from the development of new products.

The Chair turned to the Korean delegation, whose submission stated that Korea revised its Merger Guidelines in 2007 and adopted the SLC standard, while in a 2002 submission to a previous OECD roundtable, Korea had defended the dominance standard. The Chair asked the reasons behind this change of views.

Korea first described the three main changes made to the Merger Guidelines in 2007, apart from the adoption of the SLC standard. First, the HHI was introduced as a way to measure market concentration. Second, all market share presumptions were removed, allowing the KFTC to analyse the relevant SLC factors on the basis of economic evidence. Finally, pressure from foreign producers and the possibility of diversion of export volume to domestic market were added as elements to consider when assessing competition restrictions. As regards the change in the substantive standard, Korea noted that its position in 2002 was that the SLC and the dominance standards can be used to supplement each other. However, Korea recognized the benefits of the SLC test and therefore decided to revise its merger guidelines accordingly.

“gap” cases

The Chair moved to the next point for discussion, i.e. whether there are in fact any gap cases. He noted that Mexico is in a unique position to address this point since it has both the dominance and the SLC standard. However, in recent years, a number of mergers have exposed the weakness of the dominance test. In light of this, the Chair asked the Mexican delegation to comment on these developments and on the Coca Cola/Jugos de Valle merger.

Mexico explained that the Coca Cola/Jugos de Valle merger was a good illustration of the limits of the dominance test. While Coca Cola had been found to be dominant in the soft drinks market in an earlier investigation, its merger with Jugos de Valle, an important juice producer, would be difficult to block under the dominance test because the impact of the transaction was on a different market (the juice market). Therefore the application of the SLC test in this case was considered more appropriate to properly assess the effects of the transaction.

The Chair noted that several country contributions mentioned the existence of gap cases. The Czech Republic, where the test was changed to the SIEC after the accession to the EU in 2004, is one of them. He invited the Czech Republic to comment on the Telefonica/Deltax Systems merger and to describe how the SIEC test was applied there.

Czech Republic started by describing the evolution in its attitude to the change in the substantive test. At first, the Czech Competition Authority was sceptical about the necessity of changing the dominance test and to adopt the SLC test. In its opinion, the dominance test was capable of capturing the vast majority of problematic mergers. However, soon after the standard was changed, it realized that the SIEC test allowed more flexibility and the possibility to reach cases that they could have not reviewed under the old dominance test.

The Telefonica/Deltax Systems merger presented particular challenges which, although possible to deal with under the dominance test, were more easily addressed under the SIEC standard. Telefonica is a major telephone operator in the Czech Republic, while Deltax was a medium sized company, which provided information and communication technologies to other companies. Both parties operated on different markets, however, the contentious issue was that Deltax supplied information systems to the Czech Telecommunication Regulatory Agency and was responsible for their maintenance. The concern was that through the acquisition of Deltax, Telefonica would gain access to the information that its competitors in the telecommunications market submit to the regulatory authority. In the end, the merger was approved subject to the transfer to a third party of the service and maintenance contract with the Czech Telecommunication Regulatory Agency.

The Chair moved to the submission from Hungary, which discussed how Hungary changed to the SLC standard in June 2009 after having argued for years that both the dominance and the SLC test were substantially the same. He asked what were the reasons for this change and in particular he asked if this development was prompted by the Competition Authority coming across a gap case, such as the HTTC Matel case described in the written submission.

Hungary answered that the HTTC Matel case was instrumental in the process of moving towards an SLC-type test. This was a 4 to 3 merger in the telecommunications market. After the merger, the parties would have a combined market share of a little over 25%. There was therefore no ground to intervene on the basis of dominance or unilateral effects, and coordinated effects would have been hard to prove. There were, however, concerns about the effects of this merger due to its 4 to 3 character. Therefore, a bidding study was carried out, and while the study dispelled any concerns about this particular merger, it showed the weakness of the dominance test in the face of a potential gap case, which the HHTC Matel certainly

was. According to the Hungarian delegation, even if there are not many gap cases, from time to time they occur, in particular with respect to non-collusive oligopolies as is mentioned in the Secretariat Issues paper. Hungary also stressed the benefits of the SLC test from an analytical point of view noting that its Chief Economist can apply it with much less difficulty than the dominance standard. Lastly, another reason for the change to the SLC standard was Hungary’s desire to align its merger policy with that of the EU, which had adopted the SLC test several years ago.

The Chair noted the discrepancy between the position of the country contributors and that of the panellists. Most delegations seem to point out that there are a number of gap cases, hence underscoring the significance of moving towards and SLC test. On the contrary, the experts in the presentations have taken the position that the change in the standard makes very little difference. He asked the experts to comment on this perhaps only apparent discrepancy.

Mr Levy responded that in his opinion there are differences between the two tests but he also noted that if one looks at broadly similar cases that were reviewed under the two different tests, such as the France Telecom/Orange and T-Mobile Austria/tele.ring cases, it is fair to say that in both cases the same result was reached, although the France Telecom/Orange case was reviewed under the dominance standard and the T-Mobile Austria/tele.ring case under the SIEC test. What was perhaps most troubling about the dominance test was that it was not certain whether the European Courts would endorse its application in a non-collusive oligopoly scenario, which is why it was appropriate to change the legal test. Mr. Levy did not agree with the suggestion that the success of a test should be determined by the number of transactions blocked. Indeed, after the adoption of the SLC test there were concerns that the European Commission may be over-expansive in its application of the new standard. However, these concerns have been dispelled over time in practice.

As a last remark Mr. Levy referred to the Coca Cola/Jugos de Valle case mentioned by the representative of Mexico and noted that, over the past year, broadly similar transactions involving acquisitions by Coca Cola of local juice producers had been reviewed in three jurisdictions (i.e., Mexico, China, and the U.K.) under broadly similar substantive tests. What is interesting is that, notwithstanding the application of equivalent substantive tests to broadly similar fact patterns, the conclusions of the reviewing agencies were quite different (in the U.K., the transaction in question was approved unconditionally; in Mexico, the relevant transaction was approved conditionally; and in China, Coca- Cola’s acquisition of a local juice producer was blocked). These cases suggest that the question of which substantive test is applied matters less than the emphasis placed on the available evidence and the theories of harm that an agency chooses to pursue.

Mr Boyce agreed with Mr. Levy that there is a difference between the SLC and the dominance tests and that there may be gap cases. However, how great this difference is depends on how dominance is interpreted. The European Commission interpreted it very broadly, rendering the gap very small. Australia, on the other hand, was held by its courts to a very narrow definition of dominance, which naturally lead to the gap being significantly larger. In response to the Hungarian experience, Mr. Boyce voiced a concern that the newly adopted test may be taken too far. In this respect, his experience shows that under the SLC standard agencies develop sophisticated theories of harm, which require the analysis of vast amount of data. This may complicate the review and may slow the process down. He urged the competition authorities to consider the impact that such analysis may have on the speed of the review because in the end, for the merging parties time is money.

Mr Heimler stressed that dominance is a better concept because the concept of dominance is more easily understood by the reviewing courts who in most instances are not experts in economic analysis. In 2002 when the European Commission lost three merger cases in court, there were concerns that by adopting the SLC test, the European Commission may attempt to become over-expansive in its analysis.

While acknowledging that none of these concerns materialised, Mr. Heimler remained worried about the potentially extensive reach of the SLC test. He noted that there have been significant developments over the years in the application of Articles 81 and 82 EC, which has moved towards a more economic approach. Similarly, the concept of dominance in the merger context could have been de-formalized in favour of a more economic approach. Mr. Heimler concluded that the sophistication of the economic and legal analysis is more important than the change of the test itself.

The Chair thanked the experts for their comments and turned to the next question for discussion, which was the hypothetical case presented in the Secretariat’s Issues Paper. The Chair invited Finland to comment on the hypothetical case.

At the outset, Finland reported that it applies the dominance test but is considering the adoption of an SLC test for a number of reasons. First, in a review of past cases there have been some potential gap cases identified. Second, there is a clear intention to align the Finnish merger review to the European system.

As regards the hypothetical case, the Finnish delegation identified it as a clear gap case, which would be difficult to assess under the current dominance standard as applied in Finland. Some proponents of the dominance test have noted that the test’s applicability to such challenging cases also depends on the way in which the dominance test is applied. Referring to such remarks the Finnish delegation noted that in order to apply the dominance test to this hypothetical case, the notion of dominance would require significant stretching. In this respect, the Finnish delegate agreed with Mr. Boyce’s reference to the SIEC test being more intellectually honest. In a final remark Finland clarified that its 2002 submission to a previous roundtable, about the lack of difference between the dominance and SLC standards, related to the issue of legal certainty. Indeed there is very little difference in terms of legal certainty between the two legal tests because while the SLC test may at first sight appear more vague, the dominance test also carries a degree of uncertainty with respect to defining the relevant market and assessing the potential harm of a given merger.

The Chair then asked the Japanese delegation to describe how the Japanese merger test, which is very close to the SLC test, would be applied to the hypothetical case.

Japan began by outlining the merger test applicable in Japan. The test is based on a substantial restraint of competition from both a unilateral and a coordinated effects perspective. It went on to describe in detail the analysis, laid out in its written submission, of how this test would be applied to the hypothetical case. Japan noted, if Bank One, as the competitor of the merged bank, does not have any particular limitation of its ability in terms of excess capacity and the substitutability for the products of the merged bank, it would be a factor to prevent the merged bank from controlling the market from the perspective of unilateral effects. At the same time, as the market structure will become more oligopolistic, they would want to gather more information about the characteristics of the market and analyze from the perspective of coordinated effects more in detail. In short, in reviewing the hypothetical merger the Japanese authority would look not only at market share but also at additional factors, such as the competitive condition of the market, the presence of a maverick, barriers to entry and so on.

The Chair then asked the delegates if there are any technical arguments that can be made in favour of the SLC test. For example, Canada’s contribution explains Canada’s experience with respect to an SLC standard in merger review, including the analysis of efficiency considerations. On this specific point, he asked the Canadian delegation to describe how efficiencies are addressed in Canada.

Canada described in detail its approach to efficiency considerations in the context of mergers and how they are assessed. In Canada, there is a bifurcated process whereby the Competition Bureau first determines whether there are SLC concerns. If there are, it is for the merging parties to argue and prove

efficiency justifications. Interestingly, apart from the usual efficiency arguments linked to synergies and price reduction, the Canadian courts also require that wider considerations, such as overall savings to the economy, are given proper weight in merger review. As regards the assessment of efficiency arguments under SLC and dominance standards, Canada expressed its doubt as to whether efficiencies can be properly evaluated under a dominance standard, such as when they result in making the merged firm more competitive.

The Chair noted that the point made by the Canadian delegation is on point and one which leads to a question for the Turkish delegation. The Chair asked how Turkey applies the dominance test to mergers and in particular what structural and non-structural factor it takes into account in its assessment of potentially anti-competitive mergers.

Turkey stressed that as a first step it looks at structural elements such as market shares, concentration index and entry barriers, but that afterwards it also takes into account various additional non-structural factors, such as countervailing buyer power, history of the behavioural elements of the undertakings, participation in joint ventures, interlocking directors and contacts in more than one market.

The Chair then turned to another technical issue, i.e. the importance of carrying out a market definition under an SLC test. The UK in its contribution mentioned that in a number of recent transactions it did not define the market but instead focussed on the direct rivalry between the merging firms. On the other hand, in its submission, Ireland stresses the importance of market definition under an SLC standard. However, later in its submission it also seems to suggest that it does not always defines the market. The Chair gave the floor to the Irish delegation.

As a preliminary point Ireland noted that it always strives to review the cases sooner than before the 30-day initial review period. Market definition can be time consuming but it is often not necessary under the SLC test because the circumstances of the case show that whichever way the market is defined, there is not an SLC concern. In such circumstances, it is indeed not necessary to go through the complete market definition exercise. Often, it is more important to understand how the market works, who the competitors are, what the competitive constraints are and so on. Indeed, when a conclusion is reached that there are SLC concerns, it is necessary to present the results of the analysis on the basis of a formal market definition. However, Ireland cautioned that is very important to check whether the conclusion rests on overreliance on market definition. If a slight change to how the relevant market is defined undermines the conclusion about the existence of SLC concerns, then it is advisable to review whether the overall analysis is sound.

The Chair thanked Ireland for its contribution. He then moved to the next issue and mentioned that the Danish contribution focuses on the difference between the dominance and the SLC tests in terms of standard of proof and invited Denmark to comment on this issue.

Denmark described its experience with the adoption of the SIEC test and the change in analysis that it has brought about when compared with the dominance test. Under the SIEC test, the importance of economic analysis and of the assessment of the actual effects of a merger gained in importance. This change had been under way for some time but the switch to the SIEC test in 2005 has accelerated its pace.

With respect to the effect on the outcomes of merger reviews, Denmark was somewhat sceptical about the real impact of the change to the SIEC test. In Denmark there was no gap case. Moreover, dominance is retained as part of the SIEC test and it is not clear whether the standard of proof with respect to SIEC should be the same as under the dominance test or not. The question is whether some extra proof, such as hard econometric evidence, is required in the case of an eventual gap case.

The Chair then referred to the German contribution, which indicated that Germany applies the dominance test. As regards the standard of proof, it suggests that the standard is identical under the SLC and the dominance tests, although there is a difference in what is considered to be the harm to competition. The Chair invited Germany to comment on the general trend towards the adoption of the SLC test and whether in Germany there is still a debate about moving to an SLC standard.

Germany agreed with previous comments suggesting that the SLC and the dominance tests are not that far apart, even though they may look at mergers from different standpoints. In Germany, the dominance test is interpreted as covering both single and collective dominance. Gap cases may exist but they have not been encountered yet. In practice, there have been no cases that could not have been addressed under the dominance test.

As to the issue of economic analysis, Germany emphasized that the dominance standard does not prevent the use of sophisticated economic techniques in merger review, as the practice of the Bundeskartellamt shows. As is in other jurisdictions, economic theory plays an increasing importance in German merger review.

With respect to whether there is a debate about changing to the SLC test, Germany answered that extensive discussion took place before the last amendment of the relevant law in 2005. The conclusion of this discussion was to retain the dominance test. The main reason for doing so was the need to preserve legal certainty and continuity of 36 years of case law in merger review. However, as the EU and other jurisdictions gradually move to the SLC test, there will certainly be another debate because, as Finland noted, the convergence in standards is certainly a very important consideration.

The role of economic analysis under both tests

The Chair turned to the Netherlands, which in its contribution stated that the change in test makes no real difference since it only shifts the emphasis from legal characterization to economic analysis. The Chair asked the Dutch delegation to discuss the role of economic analysis under the SIEC test, particularly in view of what other jurisdictions have noted about the SIEC test allowing greater room for economic analysis.

Netherlands explained that an economics-based approach to assessing mergers plays a central role in the NMA analysis. This has not changed with the adoption of the SIEC test. However, the wording of the new test may better reflect what already happened in practice as well as provide easier access to an economics-based analysis. In that sense the Dutch delegation noted that the shift to SIEC was helpful when discussing cases with the parties and also in defending cases before the courts.

The Chair then asked the European Commission to comment on the recent developments in its approach to merger review under the SIEC test.

The European Commission (EC) mentioned that in 2002 the debate was on whether the dominance test would cover cases where horizontal mergers would lead to unilateral effects without creating a clear market leader. At that point in time, on the basis of the Courts’ jurisprudence, the EC believed that dominance could cover such cases. However, there was a degree of uncertainty, which the change in test helped to dispel. In that sense, the change to the SIEC test and the adoption of Guidelines that interpret it in detail have positively contributed to legal certainty.

With respect to gap cases, the EC noted that although rare, they nevertheless occur as has been shown in two instances: the TeamMobile/Telering merger and the BASF/Ciba merger. In general, the impact of the change in the applicable test has been modest, particularly because the move towards a more effects- based approach had already begun. There may have been more emphasis on unilateral effects and

quantitative analysis in recent years and the change in test was perhaps a contributing factor to that. The EC emphasized that what certainly has not changed is the intervention rate, contrary to some of the concerns voiced prior to the adoption of the SIEC test.

As regards coordinated effects, the EC stressed that even though there have been cases such as Airtours/First Choice or Sony/BMG in which the court disagreed with the EC’s analysis, these concerned the application of the concept of coordinated effects to the specific facts of individual cases rather than the concept itself. While there have been no recent prohibitions under coordinated effects theory, there have been a number of cases in which concerns over coordinated effects were resolved through remedies. The adoption of the SIEC test has not changed much in this area.

Last, the EC pointed to an area, which had not yet been discussed, non-horizontal mergers. In these types of cases there are companies that may have a degree of market power but are not necessarily dominant. By vertically integrating, they may have the ability and the incentive to raise prices in another market. Such a change may have negative effects on consumers even if dominance is not reached in the second market. There have been two cases, Tom Tom/Tele Atlas and Nokia/Navteq, that illustrate this point. In both cases, the EC developed a theory of harm based on unilateral effects; however, in the end, the evidence found did not support the theory. The EC noted that experience with these types of cases has shown that the SIEC test may have contributed to finding the appropriate analysis as well as adding legal certainty, in particular after the non-horizontal merger guidelines were issued.

The Chair noted that the EC contribution was the first to raise the issue of vertical mergers. Another issue that had not been discussed thus far was the question of whether the two tests may lead to under- or over-enforcement. This question was addressed in the Australian submission.

Australia described its experience with the dominance test as perhaps out of date because it moved to the SLC test in 1992. However, prior to that date, dominance was interpreted very narrowly, unlike in some other jurisdictions, and therefore virtually only covered single firm dominance. This situation led to serious under-enforcement, the effects of which remained until today, in particular in the media and grocery sectors. The change to the SLC test eliminated this deficit without leading to over-enforcement. There was a slight increase in prohibitions right after the adoption of the SLC test, but that has stabilized over time and now the prohibition rate in Australia is in the range of 2-3%. As a final remark, Australia responded to some of the questions regarding the relative lack of legal certainty under the SLC standard. It remarked that like other jurisdictions it adopted detailed merger guidelines and that the AAAC regularly releases assessments and statements of reasons in relation to particularly important mergers. This ensures sufficient legal certainty and there have been no complaints from the business community so far.

International cooperation

The Chair turned to the last topic of this roundtable and asked the delegates to comment on the question whether convergence on substantive tests facilitates cooperation among competition authorities. He asked the US to present its views on this issue.

The US stated that in its experience diverging standards have not hampered international cooperation. Given the developments which were already discussed, many authorities around the world use an approach based more and more on economic theory irrespective of whether they operate under the dominance or SLC standard. There have been many cases in which authorities that use different tests arrived at the same conclusions. Even in cases such as Heinz/Beech-nut and Peoplesoft/Oracle, which could technically be considered gap cases, the authorities on both sides of the Atlantic reached a similar result. For example, the US cooperates with the Bundeskartellamt quite often on merger cases and the fact that Germany still applies the dominance test has not prevented the two agencies from reaching a sound common position.

However, the US also stressed that having the same standard may ease cooperation with other authorities as it enables them to use the same vocabulary and to focus on commonalities.

The US noted that international cooperation has evolved significantly in the past decade due to the growing understanding of the importance of focusing on economic analysis, which is relatively independent from the applicable substantive test. One area where there is more work to be done is possibly the area of vertical mergers as was mentioned by the EC. There it seems possible that different results could be reached by authorities applying different tests. Therefore, this is an issue that should be further discussed in the future. In this respect, the US emphasized the importance of various international fora in which experiences and best practices are exchanged for international cooperation.

The Chair noted that the issue of benefits arising from use of the same vocabulary is also discussed in the submission from Romania. He asked the Romanian delegation to present its view on what the arguments are in favor of increased cooperation through convergence.

Romania noted that it still applies the dominance test but that a change to the SIEC test is contemplated. It went on to state the reasons why it considers international convergence to be important. First, having a common standard across different jurisdictions is beneficial for the international business community as it brings more predictability with respect to the outcome of cases. Second, within the EU, convergence can lead to the establishment of common principles and also facilitates eventual reallocation of cases within the European Competition Network.

The Chair then invited Poland to presents its view that convergence of tests is not necessary for international cooperation.

Poland submitted that it agreed with the US position with regard to international cooperation: it is far more important to have a common analytical approach, which does not require a complete convergence of the legal tests. However, it noted that there may be areas in which convergence could further facilitate international cooperation, such as procedural rules and exchange of information.

The Chair concluded the roundtable by briefly summarizing the discussion. He noted that first, there seems to be a clear move towards the SLC test; second, there appears to be quite a few gap cases at the national level; and third that dominance is an encompassing test only when given an economic interpretation. He also noted that one of the arguments in favor of dominance focused on the fact that judges more easily accept the concept of dominance than that of SLC. However, as dominance was gradually interpreted in an expansive manner it also became increasingly economic, which is a development that reduces the appeal of this argument. This issue also plays out in the area of legal certainty, where there were arguments from both sides. Indeed the SLC test may be a more complicated concept; however, dominance has lost its precision through its expansive interpretation. The Chair further noted that one key take-away from this roundtable is that most of the countries that have switched to an SLC test have been satisfied with the change.

He concluded the roundtable by thanking all the contributors, interveners and the Secretariat for a very stimulating Issues Paper and a very interesting hypothetical case.

Is there a gap in the dominance test?

Addressing the ‘gap’ in the dominance test

Filling the gap in the dominance test without changing the legal test

Legislative changes in the standard of review of mergers

Broader policy issues connected to the dominance test and the SLC test

Structural versus economic analysis

Legal certainty and predictability

Over-enforcement and under-enforcement

Links between the merger test and unilateral conduct rules

International co-operation

‘gap’ cases

Heinz/Beech-Nut

Oracle/PeopleSoft

T-Mobil/Tele.ring

An hypothetical ‘gap’ case for discussion – A bank merger

The current substantive test

SPAIN

The substantive test for merger review in Spain

Merger review under the 1989 Competition Act

Legislative changes in merger review introduced by the 2007 Competition Act

Brief summary of some key features of merger review in Spain

The Gas Natural/Unión Fenosa case

Final remarks

THE UK

Objective of UK merger control

The application of the UK substantive test

The UK substantive test

Theories of harm

The ‘public interest’ test

Effect of the change of test

Enforcement issues connected to the dominance and the SLC test

Structural versus economic analysis

‘Gap’ cases

Examples of ‘gap’ cases

A hypothetical ‘gap’ case for discussion – A bank merger

Broader policy considerations

Dynamic merger review

Minority shareholding23

UNITED STATES

The substantive merger standard in the United States

EC

The change of the test: from dominance to SIEC test

EC Conclusion

FINAL Conclusions

SUMMARY

The scope of the dominance test and of the SLC test

“gap” cases

The role of economic analysis under both tests

International cooperation

Leave a Reply Cancel reply

Minority shareholding²³