Three/O2 – all-you-can-eat merger control?

On May 28, 2020, the General Court of the European Union delivered an important judgment that goes to the heart of how EU merger control functions, comprehensively overturning a 731-page merger prohibition the European Commission (Commission) had issued in Three/O2 in May 2016. Before Three/O2, the EU courts had not been asked to rule on whether the creation of a non-dominant firm would give rise to a “significant impediment to effective competition” (SIEC). We discuss whether the judgment lowers the bar for merger approval and paves the way for future industry consolidation.

To merge or not to merge

The market economy, central to the European project, has served consumers well. A key feature of this system is the freedom for firms to merge or divest in response to changes in market conditions. Although state intervention has taken a less prominent role, consumer welfare remains protected by an EU-wide merger control regime. Under the 1989 regime, the Commission could only block mergers if it could show (a) dominance and (b) a reduction in competition. An academic debate raged as to whether there was an “enforcement gap” that stymied the Commission’s legal ability to challenge problematic mergers. On May 1, 2004, the modernized EU Merger Regulation came into effect and reformulated the old two-limb approach in a single SIEC test. This was understood, at least by the Commission, to extend its ability to block mergers beyond the concept of dominance if it could prove anti-competitive effects would flow from the unilateral behaviour of a non-dominant firm.

Across Europe, tight oligopolies are common in the telecommunications industry, among others. In 2006 in T-Mobile/tele.ring, the Commission began to rely on the broader SIEC test when it scrutinized a four-to-three telecom deal in Austria. Even though Mobilkom was the largest mobile network operator (MNO) in Austria at the time, the Commission identified an SIEC as the acquisition by T-Mobile of tele.ring (with 10 to 20 percent market share) would remove a “maverick” firm with an aggressive marketing strategy that had doubled its market share in recent years. The deal was only cleared on the basis of a mast and spectrum divestment. Similarly, under Commissioner Almunia, four-to-three telecoms deals were subjected to lengthy investigations in Hutchison/Orange in Austria in 2012, in Three/O2 in Ireland in May 2014 and in Telefonica/E-Plus in Germany in July 2014. These were only cleared on the condition that the merged entities granted varying degrees of wholesale access to mobile virtual network operators (MVNO).

The UK deal, which was structured as a GBP10.25 billion acquisition by Three of O2, would have seen two of the four MNOs combine to form the largest player on a market already characterized as a tight oligopoly. Post-transaction, Three/O2 would have held about 30-40 percent, with BT/EE holding slightly less and Vodafone holding about 20-30 percent. While each case is unique, the UK deal was anticipated to receive close scrutiny – but, ultimately, approval – as Three committed to grant one or two MVNOs access to certain of its network capacity and divest O2’s stake in Tesco Mobile. However, after a nine-month investigation, the Commission blocked the deal on the basis it would (1) reduce competition on the UK market for retail mobile telecommunications, (2) hinder the development of UK mobile network infrastructure, and (3) reduce the MNOs willing to offer wholesale access to MVNOs. During the review in Three/O2, UK stakeholders pressed to block the deal (the CMA and Ofcom were publicly opposed and unsuccessfully sought jurisdiction to review the deal) and, notably, the Commission was under new political leadership. Following the prohibition in May 2016, it was unclear whether there was a “magic number” for telecoms deals. Indeed, a conditional clearance in Hutchison 3G Italy/Wind/JV in Italy in August 2016 saw Iliad enter the Italian market as a full-blown fourth MNO on the basis of a remedy package, and while the four-to-three T-Mobile NL/Tele2 NL deal in the Netherlands in November 2018 obtained unconditional clearance, it involved the acquisition of a competitively weak “flailing firm.”

Role of important competitive force

In its 2004 Horizontal Merger Guidelines, the Commission suggested that one of the factors that may indicate whether a merger creates an SIEC is whether it eliminates an “important competitive force.” Application of this element has been controversial. In its prohibition decision, the Commission portrayed Three as an important competitive force that would be eliminated by the merger. Three claimed that this overestimated the constraint it exercised in the retail market, noting that its market share was consistently below 10 percent and it had weak subscriber growth prior to the deal. On appeal, the General Court agreed. It held that the Commission confused three distinct concepts: (i) the concept of an SIEC as the legal test for prohibition, (ii) the concept of an “important competitive constraint” referred to in recital 25 of the EU Merger Regulation, and (iii) the concept of elimination of an “important competitive force” used in the Horizontal Merger Guidelines. This mischaracterization led the Commission to water down the meaning of an important competitive force to effectively justify it finding an SIEC in any horizontal merger, which was an unlawfully broad interpretation of the test. Rather, the Commission must prove (i) a reduction in competitive pressure, and (ii) the merger eliminates “important competitive constraints” that the merging parties exert upon each other. Notably, the General Court held the Commission may only prohibit mergers where the market effect is equivalent to a position of individual or collective dominance enabling the merged entity to “become a price maker instead of remaining a price taker.” This is a dramatic reformulation and may be welcomed by industry consolidators for its elegant simplicity. It remains to be seen whether it is an acceptable operational definition.

Role of closeness of competition

In Three/O2, the Commission relied on a limited survey of approximately 100 subscribers and switching data to estimate a diversion ratio between Three and O2 to assess closeness of competition. During the hearing, however, the General Court probed this data and concluded that while Three and O2 were relatively close, they were not particularly close competing MNOs, and that factor alone was not enough to prove an SIEC. Arguing from first principles, the General Court surmised that to hold otherwise would mean that any four-to-three merger would be capable of prohibition. On the flipside, the Commission may argue on appeal that only 11 mergers have been prohibited since 2004, with few prohibition decisions ever relying on the non-dominant theory of harm such that any “floodgate”-type argument is hypothetical rather than real. A notable four-to-three exception includes the UPS/TNT prohibition, which was later annulled for procedural reasons.

Role of economic analysis

Since 1998, it has been recognized that the Commission has a margin of discretion to carry out complex economic assessments during its merger review. The General Court in Three/O2 noted that its role was to verify that the evidence relied on by the Commission was “factually accurate, reliable and consistent” and supported its final conclusions. One takeaway is the refreshing way in which the General Court sense-checked the assumptions and logic underpinning the Commission’s economic analysis. In the previous Irish and German telecoms cases, the Commission conducted a similar upward pricing pressure (UPP) analysis to estimate the effect of the merger on average prices, predicting an average price increases of 6.6 percent in the Irish case and 9.5 percent in the German case. Both transactions were conditionally cleared. Yet, in Three/O2, the Commission outlined a 65-page UPP analysis (and calibrated merger simulation) that illustrated that the deal would only increase average overall prices by 7.3 percent. The General Court found that the UPP analysis in Three/O2 lacked probative value and criticised the fact that any UPP analysis of a four-to-three merger was likely to indicate a price-rise post-transaction, and failed to account for efficiencies stemming from the “rationalisation and integration of production and distribution processes.” This is likely to have a significant effect on how the Commission conducts detailed economic analysis (with limited information) under the timetable pressure imposed by merger review.

Other concerns

The General Court overturned the Commission’s other substantive findings. First, it held that the Commission was incorrect to find Three/O2 would hinder the network arrangements between the MNOs. Sharing of telecommunications infrastructure is a common feature in the industry. Pre-merger, Three was in a net-share with BT/EE (called MBNL), and O2 was in a net-share with Vodafone (called Beacon). While it anticipated that the merger might mean that the MBNL or Beacon net-share would become redundant or the partner’s interests misaligned, the General Court criticized the notion that disruption to MBNL or Beacon was an element which alone gave rise to an SIEC. In other words, the termination, renegotiation or alteration of a net-share arrangement could not be characterised as an SIEC. Taking the argument to its logical conclusion, the Commission could block any four-to-three merger not between existing net-share partners, which the General Court rejected as an incoherent proposition. Separately, the General Court dismissed the idea that the merger could raise competition issues for MVNOs on the wholesale market, citing the fact that Three’s market share for wholesale access was less than 5 percent in the four years prior to the merger.

Commission re-think

Time will tell whether Three/O2 is as seismic as the infamous triplet of judgments annulling merger decisions in Airtours, Tetra/Laval and Schneider in 2002. Following a number of changes to the Commission’s merger control practice, including the adoption of an in-house Chief Economist team, challenging the Commission’s position on the EU Merger Regulation has been perceived as being very difficult. Recently, however, a series of judgments called this view into question: UPS/TNT (prohibition annulled for breach of procedural rights in March 2017, upheld in January 2019); Austria Asphalt (concerning when a joint venture is notifiable in September 2017); KPN (annulling a clearance decision in October 2017); Lufthansa (annulment of a refusal to review a merger commitment in May 2018); and Ernst & Young (concerning the definition of gun-jumping in May 2018). Further, the Three/O2 ruling may encourage appellants in other recent prohibitions, such as Tata Steel/ThyssenKrupp in June 2019.

In July 2020, the Commission confirmed it will appeal the General Court’s judgment in Three/O2. A key focus of the appeal will be the standard of proof. In 2008, the Court of Justice held in Impala that “the inherent complexity of a theory of competitive harm” put forward by the Commission when assessing a merger must be taken into account when assessing its plausibility. Significantly, the Court of Justice ruled that a complex theory of harm did not, of itself, have an impact on the standard of proof that the Commission was required to meet and that it was for the Commission to approve or prohibit a deal based on the “most likely” outcome. In Three/O2, however, the General Court suggested that the standard of proof to show an SIEC in a four-to-three merger is stricter than “more likely than not” or “on the balance of probabilities.” Rather, the Commission must show a “strong probability” of an SIEC, although it need not prove it “beyond all reasonable doubt.” The General Court concluded that the standard of proof for unilateral effects on an oligopolistic market is not “substantially different” from that needed to show coordinated effects. If so, this is a significant clarification and sets a very high evidential hurdle for competition authorities, particularly given the limited timelines for merger review. The clear message to the Commission is to focus its review on real issues and not on theoretical outcomes. Not all are convinced. The former Chief Economist during the Three/O2 merger review, Massimo Motta, criticized the judgment as setting the standard too high: “you cannot have a standard of proof which goes close to beyond a reasonable doubt.” This was echoed by another former Chief Economist, Tommaso Valletti, who described such a standard as “almost impossible to meet in oligopoly mergers.”

Yet, there are positive lessons for the Commission. Despite the significant volume of mergers each year, there is precious little case law. Now is an opportunity to reset and address certain concerns regularly aired by the legal community during merger review – such as ever-widening theories of harm, the fundamental right of merging parties to fair procedures, and invidious ever-lengthening information requests.

For Three, it is clear that it achieved a resounding legal victory. Yet the UK mobile market has evolved: on May 7, 2020, O2 and Virgin Media agreed to a major GBP31.4 billion joint venture. Should Three/O2 be upheld, Three will no doubt closely watch UPS’s EUR1.7 billion action for damages against the EU for the unlawful prohibition of its merger with TNT in 2013. That will have to wait. For now, Three sells its mobile subscription plans under the moniker All You Can Eat – the judgment begs the wider question whether a more permissive approach to the SIEC test will promote a wave of all-you-can-eat industry consolidation.