logo-image
Effect-based Approach in Merger Control
Author: Iryna Fomina, Anastasia Usova
Source: The Ukrainian Journal of Business Law. – 2014. – July-August. – p.30-32
Download pdf

In today's dynamic world competition has become a basis for economic development, which has so far been the best booster of economic growth. The last two decades have witnessed an impressive growth of fast developing sectors such as IT, pharmaceuticals, energy, with markets facing a significant intensification of competition. The Antimonopoly Committee of Ukraine (the AMCU) plays an important role in protecting effective competition with a particular focus on merger control, which is evidenced by a notably large number of merger reviews that grows year by year.

The expected harmonization with the EU legal standards provided under the Association Agreement1 entails harmonization of Ukrainian competition laws with the EU framework, including in the field of merger control. Initially the Ukrainian merger control system derived from the European one; however, the last decade has demonstrated appreciable differences in assessment of mergers by the European Commission and the AMCU, especially in application of theories of harm. In view of actively debated harmonization with EU standards it is important to understand the key differences of the two models and how Ukrainian merger control could be changed following the alignment with the EU.

To find the answer to this question let us have a look at the approaches currently used in Ukraine and in the EU for assessing mergers; the factors which served as a push for Europe to introduce new standards of merger review and their practical implications on business.

How merger assessment works in Ukraine

Ukrainian merger control is based on two concepts — dominance standard (i.e. form-based approach) and significant impediment to economic competition (i.e. effect-based approach).

In particular, Article 25 of the On Protection of Economic Competition Act of Ukraine (the Competition Act) stipulates that the AMCU "shall grant its approval for a concentration where it does not lead to monopolization2 or significant restriction of competition in the entire market or in a substantial part of it". In practice, however, the dominance standard is prevailing, which means, that if as a result of the concentration the applicants' market share exceeds 35% (presumption of dominance), it is practically impossible to have such concentration cleared within Phase I review — with almost 100% certainty such transaction will raise serious competition concerns on the part of the AMCU and trigger a Phase II in-depth investigation.

Even if the transaction is further cleared within Phase II, the in-depth investigation may, as such, be quite burdensome for the applicants due to considerable amounts of data to be provided with the authority and substantial human resources to be involved during the investigation, furthermore, it may put at risk timely implementation of the transaction. In particular, the Phase II review period is limited to three months, however, in practice the investigation may take much longer — if additional documents, information and/or expert analysis are required, the term may be suspended or even restarted, in which case a new three-month period commences from the date on which these documents, information and/ or expert analysis have been filed with the AMCU. This essentially means that the AMCU may extend the phase II review period for as long as it deems necessary.

Therefore, due to the prevailing dominance standard, even the parties to those concentrations which do not have any appreciable effects on competition automatically make a Phase II path as soon as the dominance threshold is triggered.

European reality

In the EU the merger control system based on the "dominance test" similar to the current Ukrainian one existed under the Council Regulation 19893 until 2004. The European Commission first assessed possible dominance of an undertaking and only in the second place the probability of impediment of competition in the common market; thus there was a high likelihood for a merger to be blocked already at the stage of finding dominance without further analysis of the potential effects on competition. In the late 1990s it was becoming more and more evident that such a system of merger control needs to be revised and put in line with business reality in order to avoid unjustified faultfinding in mergers that, although leading to dominance did not, in fact, create significant risks for impediment of competition.

The revision gained its momentum in connection with the events of 2002 when for the first time the Court of First Instance (the CFI) reversed the Commission's prohibition decisions in three major cases (Airtours4, Tetra Laval5, Schneider Electric6).

In Airtours v. Commission the CFI reversed the decision of the Commission, which had blocked acquisition of First Choice by Airtours assuming creation of a collective dominance which would significantly impede effective competition. This was the first case where the CFI overturned the Commission's decision to block a merger claiming that the Commission has failed to establish that post-merger the three remaining large tour operators would have an incentive to cease competing with each other or the level of competition would have been any different. A similar decision on the reversal of the Commission's blocking was reached by the CFI in Tetra Laval v. Commission where the Commission was heavily criticised on factual findings and analysis. The CFI concluded that the effect of most conglomerate mergers was neutral rather than anticompetitive; otherwise the Commission had to provide respective empirical evidence to the contrary. In the same manner the CFI upheld Schneider's appeal in Schneider v. Commission and annulled both the Commission's merger prohibition and its divestment order, once again underlining the Commission's failure to provide sufficiently clear theory of harm in its statement of objections which later led to a blocking decision.

The above cases marked important shortcomings in EU merger control and served as an accelerator for a comprehensive reform, which resulted in adoption of a new regulation in 2004 — Council Regulation (EC) No.139/2004 (the Regulation 2004). The Regulation 2004 introduced, inter alia, a more economic approach to assessing mergers based on profound effect analysis (the Theory of harm), meaning a move from "dominance test" to the "significant impediment to effective competition" test.

Theory of harm in merger control

The theory of harm was made a cornerstone of merger assessment which is based on harm principle. Harm principle itself derives from a theory of crime where an action can be banned only if it causes harm to someone. In terms of merger control the theory of harm is applied to verify whether a merger is likely to prevent, restrict or distort competition. A consistent theory of harm is built on a coun-terfactual analysis — what would have happened if the merger did not take place. Under Regulation 2004 the Commission should compare "the foreseeable impact of the merger"7 with the situation that would have occurred without the merger (i.e. counterfactual). It is impossible to prove a merger to be anti-competitive without knowing what the alternative and its effects were, and how competition could be prevented, restricted, or distorted in each particular case. An appropriate counterfactual is a fundamental basis for a well-grounded theory of harm.

The EU case law underlines that a coherent theory of harm should ensure the following:

— articulation as to how competition and, what is more important, consumers will be harmed with respect to an appropriate counterfactual;

— internal logical consistency;

— consistency with the incentives of the notifying parties to impede competition; and

— consistency with available empirical evidence.

As a result of this switch, after 2004 the Commission cleared a number of mergers which would be hard to imagine in 1990s before the reform8. For example, in Lufthansa/Austrian Airlines case, where post-merger combined market share would have amounted to 50-60%, the Commission concluded that in the absence of the acquisition by Lufthansa, Austrian Airlines could have been potentially acquired by Air France-KLM. Therefore, the effects of the given transaction were compared with those of the hypothetical merger with Air France-KLM.

The theory of harm may be particularly pertinent to review of mergers in the IT sector, which is currently one of the markets most prone to competition concerns. In Intel/McAfee case9, for instance, McAfee's competitors — security solutions producers were concerned that post-merger Intel, with its own market share of around 70-80%, might give more advantages to McAfee, thereby impeding effective competition. Despite the Commission's findings that Intel could potentially foreclose other software providers, due to the expected merger-related efficiencies (e.g. synergies of combining soft/hardware for security technologies) the merger (even though conditionally) was cleared.

Importantly, the theory of harm also enabled the Commission to capture potentially anti-competitive mergers (Friesland Foods/ Campina10, Ryanair/Aer Lingus11) that did not fall within the market dominance thresholds.

Regulation 2004 proved to be a big step forward in the development of the EU merger control, first aimed at equipping the Commission with better tools for assessment of potential impact on consumer welfare avoiding over-reliance on structural parameters, and secondly, providing merging companies with improved guidance to better anticipate and assess potential competition concerns in relation to the notified transactions. The most important is that it allowed the Commission to substantively weigh potential competitive risks without creating unreasonable pressure on merging companies.

Prospects of switching to EU standards

In this context harmonization with the EU standards in Ukraine would be beneficial both for Ukrainian business and for foreign companies (in 2013 approx. 71% of the concentrations were notified on behalf of non-residents). Switching to a more economic approach would allow saving time and resources for the parties to non-problematic transactions and would provide for a higher predictability of the decision-making process.

However, it should be understood that if the EU path is to be taken in order to fully enjoy the said advantages a set of measures is required to be implemented, such as introducing amendments to the Competition Act, issuance of respective guidelines by the AMCU outlining requirements and recommendations for self-assessment by the notifying parties, public record of the AMCU's decisions similar to that of the European Commission, possibility of informal consultations with the AMCU.

It is worth noting that even without such regulatory changes one may already observe a positive move in the AMCU's practice in the direction towards alignment with the EU approach — within a number of merger cases the AMCU has accepted for its consideration references to EU case law on issues such as market definition, the economic rationale of a merger and its potential efficiencies, as well as the Commission's preliminary findings or clearance decision in relation to the merger under review by the AMCU. Furthermore, in its own conclusions the AMCU is making more and more references to the Commission's decisions and rulings adopted by EU Courts.

In light of the above, we hope that the aforesaid steps already being taken by the AMCU towards the European effect-based approach will ensure a smooth transition to more sustainable EU standards in merger control, thus creating a fertile ground for business to comply with competition law and for the AMCU to safeguard fair competition in Ukrainian markets.

1 Association Agreement between the European Union and the European Atomic Energy Community and their Member States, of the one part, and Ukraine, of the other part

2 Under "monopolization" the Competition Act provides for creation, maintenance or strengthening of monopolistic or dominant position, which means that in Ukraine the concept of monopolization is equivalent to that of dominance.

3 Council Regulation (EEC) No.4064/89 on the Control of Concentrations between Undertakings, OJ L 395, 30/12/1989.

4 Case T-342/99, Airtours Plc v Commission, [2002] E.C.R. II-01785

5 Case T-5/02, Tetra Laval B.V v Commission,

[2002] E.C.R. II-04381

6 Case T-310/01 , Schneider Electric SA v Commission, [2002] E.C.R II-04071

7 Horizontal Merger Guidelines, op.cit., para.13.

8 Case COMP/M.4864 TomTom/Tele Atlas, 2008. Case COMP/M.5335, Lufthansa/SN Airholding, 2009. Case COMP/M.5440, Lufthansa/Austrian Airlines, 2010.

9 Case COMP/M.5984 Intel/McAfee, 2011.

10 Case No COMP/M.5046 Friesland Foods/ Campina, 2008.

11 Case No COMP/M.4439 Ryanair/Aer Lingus,

2007.