On September 26th, the European Commission prohibited the acquisition of the online flight booking service eTraveli by Booking Holdings, the leading provider of online hotel bookings in the European Economic Area (EEA). The Commission argues that the acquisition would have strengthened Booking’s dominant position in the market for online hotel bookings by enabling it to expand and reinforce its travel services ecosystem.
The most significant aspects of the Commission’s decision were its rejection of the behavioural remedies offered by Booking, and its reliance on an ecosystem theory of harm. The prohibition also suggests that the Commission will not be swayed by political pressure to facilitate the emergence of dominant “European tech champions” through weaker enforcement of competition policy. But it remains to be seen whether the decision heralds a permanent, and necessary, strengthening of the EU’s approach to mergers. The decision also has implications for the Commission’s approach to enforcement of the Digital Markets Act.
The deal and the Commission’s concerns
In November 2021, Booking Holdings, which owns key brands including Booking.com, Agoda and Kayak, announced its intention to acquire eTraveli, a Swedish company which owns several major flight booking websites, for €1.6 billion. While Booking Holdings is headquartered in the US, its largest brand – Booking.com – was founded and remains based in Amsterdam and is considered by many to be a leading European technology company.
The deal was notified to the Commission in October 2022, followed by the launch of an in-depth investigation in November that year and a Statement of Objections in June 2023, foreshadowing many of the arguments contained in the final decision. That decision, issued on September 26th, prohibited Booking from acquiring eTraveli based on the Commission’s concerns that it would further strengthen the former’s dominant position in the market for hotel online travel agencies (OTAs). The Commission defines OTAs rather broadly as an “important intermediation service, matching demand and supply for travel services” including “accommodation, flights, car rentals, and attractions”.
Based on feedback provided by stakeholders including hotels and competing OTAs – “almost 15,000 hotels” according to Competition Commissioner Didier Reynders – the Commission provided several justifications for blocking the merger:
- Booking already enjoys a position of dominance in the EEA hotel OTA market, with a market share of over 60%. It faces little meaningful competition, resulting in higher commissions paid by hotels dependent on Booking platforms, which in turn are likely passed onto consumers. Moreover, the company benefits from growing network effects as ever more consumers and hotels use its services;
- The acquisition would have allowed Booking to “expand its travel services ecosystem” by increasing Booking’s access to potential customers, as “flight OTA services are an important customer acquisition channel for hotel OTAs”. According to the Commission, flights are the most common route to cross-selling accommodation, while “customer inertia” means that a “significant share of these additional consumers would have stayed on Booking’s platforms”;
- This strengthened ecosystem would, in turn, have reinforced the network effects Booking enjoys, while increasing barriers to entry and expansion facing rival OTAs. Notably, the Commission focused on potential rather than actual competitors, warning that “OTAs currently on a path to becoming full-fledged competitors may not be able to do so if the transaction goes ahead”;
- The strengthening of Booking’s dominant position following the acquisition would have increased its bargaining power vis-à-vis the hotels that depend on its platforms, raising the prospect of higher costs for those hotels and ultimately, for consumers.
Rejection of behavioural remedies
One of the most noteworthy aspects of the Commission’s decision was its rejection of the behavioural remedies put forward by Booking to resolve the competition authority’s concerns. While in general, the Commission has shown a strong inclination towards behavioural over structural remedies and prohibitions, in this case, it found that the proposed remedies would not preserve competition “on a lasting basis” and be too difficult to monitor effectively.
Booking’s proposed solution was to present customers with a choice screen after purchasing a flight, in which hotel offers from Booking and other hotel OTAs would be displayed. The choice screen would have been powered by Booking-owned KAYAK and displayed on both the Booking.com and eTraveli flight platforms to customers either located in or travelling in the EEA. The choice screen would rank the OTAs based on price, with the KAYAK algorithm responsible for selecting the offers and setting “technical and quality standards”.
Based on its analysis of the remedies, including testing them with market participants, the Commission found them to be unsatisfactory for the following reasons:
- The selection and ranking of competing hotel OTA offers would have been opaque and discriminatory, due to KAYAK’s control of the algorithmic process;
- The competing offers would have only been displayed following flight purchase and not via other important channels such as emails, notifications and other website pages;
- The commitments would have been challenging to monitor, due to KAYAK’s algorithm “working as a black box”.
While we do not know the extent to which the Commission negotiated with Booking over the proposed remedies, and how willing Booking was to address the Commission’s concerns, ultimately the competition authority decided that the acquisition’s prohibition would be more effective than the commitments already on the table.
As mentioned, the Commission’s decision to prohibit the deal is highly significant in and of itself. It is only the 11th time that the EC has blocked a takeover in the past decade, compared to around 3,500 approvals over the same period. This has contributed to rising market concentration in Europe, not least in the technology sector, in which acquisitions have been a key means of acquiring market power. One major reason for this has been the Commission’s historical aversion to structural remedies and prohibitions and preference for complicated behavioural remedies. This is despite growing scepticism among experts and enforcers regarding the effectiveness of (especially behavioural) remedies, given the challenges in designing, monitoring and enforcing them. This is especially true in digital markets, where asymmetries of information and expertise between competition authorities and companies are particularly stark.
Other recent developments suggest that the Commission’s prohibition of the Booking/eTraveli merger is not a one-off event, but part of a broader shift. In September 2022, the Commission blocked Illumina’s acquisition of cancer screening startup GRAIL, after rejecting behavioural remedies offered by Illumina. According to the Commission those remedies, which included providing patent licences to some of Illumina’s rivals and commitments to conclude agreements with GRAIL’s rivals, would have been insufficient to preserve competition in the market. Looking beyond merger control to antitrust, the Commission has indicated that it is inclined to pursue a divestment to address Google’s anti-competitive conduct in the ad tech supply chain. Even senior European officials themselves have publicly expressed scepticism of behavioural remedies. The Director-General of DG Competition, Olivier Guersent, was recently quoted claiming not to “know a single enforcer that likes behavioural remedies” and describing them as “ineffective (…) sensitive to asymmetries of information” and “terribly intensive in resources to monitor”.
Of course, the Commission has not abandoned behavioural remedies altogether. Over the past 12 months in the tech sector alone, the Commission has approved massive deals including Microsoft/Activision Blizzard and Broadcom/VMware subject to commitments offered by the merging parties, and ended an antitrust investigation into Amazon’s Marketplace following behavioural remedies proposed by Amazon. Yet much of the reasoning proffered by the Commission in rejecting Booking’s proposed commitments, particularly the opacity of Kayak’s “black box” algorithm, would seem to apply to many of the remedies typically offered by tech companies in competition investigations. The outcomes of ongoing and future digital merger investigations – such as Amazon/iRobot (where a decision is due later this year) – will reveal whether the dial has truly shifted.
Just as significant is the fact that the prohibition decision was – for the first time ever – based solely on a so-called “ecosystem” theory of harm. This is the notion, particularly when it comes to digital markets, that market power and competition cannot be fully understood through narrow vertical and horizontal theories of harm, but only by taking into account the sprawling “ecosystems” across which firms concentrate and exert their dominance. While there have been calls for greater emphasis on ecosystems in competition policy for some time, and the Commission itself recognised their importance in its 2019 report on Competition policy for the digital era,
Ecosystem concerns have featured to varying degrees in other recent merger investigations, including Microsoft/Activision Blizzard (which looked at how acquiring Activision’s games catalogue could strengthen Microsoft’s position in cloud gaming and operating systems) and Amazon/iRobot (where the role of iRobot’s user data in reinforcing Amazon’s marketplace dominance is currently being investigated). However, the emergence of new technologies with seemingly unlimited use cases, from generative AI to virtual and augmented reality, is only increasing the need to take ecosystem effects into account in competition investigations. For example, enforcers will need to closely monitor attempts by large digital platforms to leverage their dominance in highly concentrated markets such as search, social media and e-commerce, and over assets including computing power and data, to gain control over generative AI.
In blocking Booking’s acquisition of eTraveli, the Commission has also signalled that it will not be swayed by political pressure to facilitate the emergence of so-called “European champions” through weaker enforcement of merger control. This pressure has grown in recent years in response to anxiety about Europe’s perceived inability to produce globally competitive technology and industrial corporate giants. Yet as with its prohibition of the Siemens/Alstom merger in 2019, the Commission appears to have concluded that regional market concentration is not the recipe for European growth and prosperity. Yet DG Comp’s continued reluctance to block acquisitions by Big Tech platforms suggests that the regulator lacks the same resolve when it comes to large American acquirers, despite the recent Illumina/Grail decision.
Finally, the decision has some interesting implications for the Commission’s approach to enforcement of the Digital Markets Act (DMA). For starters, the EC’s willingness to stand in the way of Booking’s acquisition suggests it will not shy away from designating Booking as a gatekeeper under the DMA. This would represent a direct rebuttal to claims from certain quarters that the DMA is a “protectionist” piece of legislation designed to disadvantage American technology companies while giving a leg up to their European counterparts. Encouragingly, the prohibition decision also implies that the Commission does not plan to treat the DMA as another route to addressing its merger concerns, as the authority could have chosen to approve the acquisition while hoping that Booking’s subsequent designation would root out any anti-competitive practices. By prohibiting the merger, the EC has ensured it is not adding to its future enforcement workload.
It is too early to say whether the Commission’s prohibition of the Booking/eTraveli merger represents a genuine sea change in EU merger control. What is already clear is that the Commission’s rationale on behavioural remedies and ecosystems has applicability far beyond this specific transaction, in both merger control and antitrust investigations. A greater emphasis on structural intervention and ecosystem harms would go some way towards rectifying the historical underenforcement of merger control law in the EU, and ensure the Commission has the right tools to prevent further consolidation in digital markets, particularly in AI.