The recent judgment of the CJEU in Booking.com represents yet another development in the long series of cases concerning price parity clauses in the platform economy. In Booking.com’s case, the judgment represents the end of the line for its parity clauses. In its greater context of applying EU competition law in the digital economy, the judgment offers new insights into applying two important concepts: ancillary restraints and market definition. In the case of ancillary restraints, the Court’s findings reveal that putting forward such an argument in the case of digital platforms entails demonstrating that the restrictive measure under scrutiny is indispensable to the platform’s existence as a legal and economic construct rather than to its business model. In the case of market definition, the judgment helps enhance the evidential value of market definitions performed across different jurisdictions and procedures.
Background to the case
On 14 September 2024, the Court of Justice delivered its judgment in the preliminary ruling procedure in the Booking.com case. The case, launched before the District Court of Amsterdam, was an attempt by Booking.com to obtain final clarification on the legality of its parity clauses, particularly the narrow ones. These narrow price parity clauses prohibited hotels from having lower rates for their rooms on their website than on Booking.com. The position of Booking.com has always been that it needed such clauses to prevent free-riding by hotels, which could occur when consumers search for rooms on Booking.com but then make their reservations on the hotel website because of the lower rates. If the reservations do not go via Booking.com, it cannot charge a remuneration for its intermediary service. In the past these clauses were also accompanied by wide price parity clauses that prohibited hotels from having better offers on competing hotel room booking platforms.
The implementation of these price parity clauses by Booking.com has led to numerous cases across the EU and outside of it. Overall, the wide price parity clauses were considered to be harmful to competition across platforms and were eventually also abandoned by Booking.com. The legality of narrow parity clauses, however, did not enjoy this consensus across jurisdictions. However, all the MFN-related cases concerning Booking.com took place at the national level over the years, which meant that EU courts have not yet provided their view on the legality of such clauses. Therefore, the launch of the preliminary reference procedure by the Amsterdam Court offered an opportunity to obtain clarity on the matter throughout the EU. A favorable decision from the CJEU would have meant a significant win for Booking.com in the context of both public and private enforcement.
The importance of a favorable finding for Booking.com stems to a great extent from the case practice on its parity clause in Germany, where the German Federal Court of Justice ruled against Booking.com. According to the federal court, Bookings’ parity clauses significantly restricted competition and could not constitute ancillary restraints. Therefore, the clauses were found to infringe Art. 101 TFEU, thereby opening the door to private enforcement claims from hotel owners that were subject to these terms.
Due to the primacy of EU law, if EU courts were to find that the narrow price parity clauses of Booking.com constitute ancillary restraints and thus do not infringe art. 101 TFEU, this would absolve Booking.com from all the decisions taken against it at the national level. Accordingly, all the grounds for fining Booking.com for this component of the infringement and follow-on damages claims would have been entirely void. The cross-border nature of Booking.com’s practice in this respect would have also prevented diverging findings based on the national equivalents of art. 101 TFEU as dictated by art. 3(2) of Regulation 1/2003.
The two preliminary questions referred by the District Court of Amsterdam covered two of the main angles through which the legality of Booking.com’s parity clauses could be (re)established. The first question addressed to the CJEU tackled the permissiveness of narrow parity clauses directly by requesting the CJEU to indicate whether such clauses can constitute an ancillary restraint under art. 101 TFEU. The second question sought to obtain guidance on how the market definition for Booking.com should be performed to apply the Vertical Block Exemption Regulation (VBER). The answer to this question would then indirectly impact the permissiveness of Booking.com’s parity clauses by determining whether Booking.com would fall under the VBER. For this to happen, a relevant market in which Booking.com does not exceed the 30 % market share threshold would be required. In such a case, the permissiveness of the narrow parity clauses would be presumed as long as the VBER remains applicable, as they do not constitute hardcore or excluded restrictions under this regulation.
In its judgment, however, the CJEU did not deliver Booking.com with the answers it desired. Some aspects of the CJEU’s answer may have even shut down more avenues than expected for Booking.com’s ongoing and future cases in the context of private enforcement.
However, as with any CJEU case, the Court’s findings are not limited in relevance to the undertaking concerned. The CJEU’s answers to the preliminary questions will inevitably have consequences for two matters that commonly pose a challenge in the context of digital platforms (and outside of it): the concept of ancillary restraints and market definition.
Ancillary restraints and digital platforms: legal constructs v. economic activity
The legal and economic rationale behind the concept of ancillary restraints is relatively straightforward. Actions that may restrict competition to a certain extent but are necessary to facilitate pro-competitive operations of (much) greater magnitude or otherwise neutralize such detrimental effects at least may be excluded from the prohibition of art. 101 (1) TFEU. The inclusion of a non-compete clause in various types of sales agreements facilitating acquisitions of other undertakings (or various aspects thereof, such as IP rights) constitutes the most evident of such instances. This approach is sensible and can be considered entirely neutral to the context in which it would apply. Accordingly, on the face of it, it is hard to see why the application of this concept in the context of digital platforms should pose any challenges. A closer look at the Court’s answers reveals a mismatch between the concept of ancillary restraints and digital platforms whose governance rules often include measures that aim to limit multi-homing and switching in order to ignite and continuously fuel the network effects they rely on.
When the Court was asked to indicate whether Booking.com’s narrow parity clauses could constitute an ancillary restraint, doubt arose as to whether this question was suitable for the Court to answer in the context of a preliminary procedure. This is because preliminary procedures do not go into factual discussions and leave the final application of EU law to the case’s specific circumstances in the hands of the referring court. Accordingly, for the CJEU to answer the preliminary questions it received, it must be able to decide on the matter in a rather abstract manner without having to go into the factual aspects of the case in a similar length as a national court would. The question was, therefore, whether this would be the case with the narrow parity clauses of Booking.com.
In its judgment, the CJEU engaged with this question and concluded that assessing whether Booking.com’s narrow parity clauses could constitute ancillary restraints did not require an in-depth factual assessment and could, therefore, be assessed in the abstract. Initially, this answer could be seen as yet another formalistic argument made by the CJEU to ensure its jurisdiction concerning the interpretation of EU (competition) law. When delving into the actual meaning of this answer, however, it becomes clear that its implications go far beyond the matter of jurisdiction over the interpretation of EU law.
The indication that the CJEU is able to assess whether a certain practice can be considered as an ancillary restraint in the abstract means, in essence, that in order for the practice to be ancillary, it should be indispensable almost entirely irrespective of the specific circumstances of the respective case. Consequently, this possibility would also be predominantly limited to (pro-competitive) non-context dependent goals or, better said, legal constructs, such as a SPA, selective or exclusive distribution, or a franchise agreement. This is because assessment of whether a specific type of restriction is needed to allow for such constructs to be utilized in commercial practice does not require per se, looking (far) beyond the functional boundaries of the respective construct. Such an assessment can be done to a large extent by mapping out the various conflicting interests at play in the context of such constructs from a theoretical economic perspective (e.g., the conflicting interests and incentives of parties to an SPA when it comes to the market re-entry possibility of the seller post transaction).
By contrast, applying this concept to restrictive actions that help facilitate the profitable continuation of a specific ongoing economic activity is far less suitable since the respective market conditions, which cannot be incorporated in the abstract assessment, determine the need for such actions. In other words, in this latter situation, the question of whether a given practice can be considered necessary (and thus potentially ancillary) may be a matter of timing. Practices that may be indispensable for starting a business may lose this status once the business is mature.
This contrast is of particular importance in the case of digital platforms, where initially igniting the network effects that fuel their growth may require certain restrictive practices that are no longer needed or as important once that goal has been achieved. However, this time-dependent aspect of indispensability cannot be taken into account if the assessment of the practice is to be done in the abstract without going into the factual context of the case in depth. Accordingly, the natural growth pattern of digital platforms cannot truly be accounted for in the ambit of such assessments, which may require many platforms to redesign their governance that often incorporate rules that seek to limit, for example, switching and multi-homing in an attempt to ignite and/or fuel the network effects at play.
Therefore, for certain practices to be considered ancillary restraints, they should be indispensable for the platform’s existence, meaning such action should be indispensable to the construct of a platform as such from an economic perspective. An example of such a practice would be implementing a skewed pricing structure and corresponding measures to prevent arbitrage, which are vital for creating a platform, regardless of its business model. If a platform cannot make different value propositions to its various customer groups (i.e., implement a skewed pricing structure), it will not be able to get different kinds of customer groups, which have different degrees of demand for the platform service, to join the platform and thus fail to launch. While not every iteration of a skewed pricing structure will align with (EU) competition law, the ability to implement such a structure is nevertheless indispensable. Accordingly, if a platform prohibits its (commercial) customers, for example, from passing on their platform-associated costs to consumers, this may be considered ancillary to sustaining such a structure, which is principal to the existence of a platform.
By contrast, both wide and narrow parity clauses do not fulfill this type of indispensability requirement because the need for their implementation varies across (platform) business models. The same applies to actions that attempt to mimic or replicate the effect of MFN clauses without actually imposing them directly, such as demoting the ranking of platform customers (e.g., hotels or retailers) that offer better prices on their websites. The inclusion or omission of such actions is not, as such, connected to the economic construct of a (multisided) platform and thus should not be considered as ancillary under the scope of art. 101 TFEU.
When viewed in this context, the finding of the CJEU that Booking.com’s narrow parity clauses cannot qualify as ancillary restraints was inevitable. As mentioned, such clauses are not necessary for the creation of a platform, and at this moment in time, they are also not indispensable for Booking.com’s viability. Whether that was ever the case in the past is irrelevant to this conclusion from a formal legal perspective. Therefore, there would be no room for an argument by Booking.com to limit the scope of its liability by attempting to claim that its parity clauses constituted an ancillary restraint for a certain period. Any other conclusion than the one provided by the CJEU would create conflicts with the functioning of art. 101 TFEU. Consequently, for follow-on damages claims in Germany and elsewhere in the EU, the decision of the German Federal Court of Justice stands as (at least prima facie) evidence of infringement.
Individual exemption under Art. 101(3) TFEU: the impact of (increased) market power
The strictness of the ancillary restraints test will push most discussions on the pro-competitive effects of investigated practices to art. 101(3) TFEU, which is more suited for this purpose when dealing with context-specific matters. This also seems to be the stance taken by the CJEU in this case. Generally speaking, this test would be considered less stringent as the element of indispensability is associated with specific efficiencies claimed rather than (economic) viability in the abstract. In this respect, the free-riding argument persistently put forward by Booking.com may have had (theoretically) a better chance of success. Nevertheless, such an argument has been overtaken by events in the case of Booking.com. By the time the CJEU addressed the preliminary questions, Booking.com had been designated as a gatekeeper under the DMA, which prohibits the use of parity clauses altogether in art. 5 (3), and had arguably reached a position of dominance, according to the Commission. In Spain, the practice of narrow price parity clauses was found to constitute an abuse of dominance by Booking.com.
While all of this does not exclude the ability to rely on art. 101 (3) TFEU, formally speaking, it certainly reduces the prospect of success drastically. This is because the increase in market power directly relates to the assessment criteria of art. 101 (3) TFEU.
The increase in market power impacts, in essence, two elements of a potential justification argument. First, it increases the magnitude of the anticompetitive effects that need to be compensated by the claimed efficiencies. Secondly, the indispensability of the restriction likely diminishes as more market power would equip the undertaking concerned with more (less restrictive) means of obtaining such efficiencies. This is true even when the magnitude of claimed efficiencies may grow in proportion to the increase of market power. Accordingly, even if Booking.com were to be in a position to put forward an art. 101(3) TFEU justification, it would face an uphill battle (at the very least).
The preliminary procedure and market definition
The only way out of the above conundrum depended, in essence, on the second preliminary question concerning the correct market definition for the purpose of applying the VBER. However, the nature of the question and its timing make this way out unattainable.
When it comes to the nature of the question, the topic of market definition is one where the interaction with EU Courts is rather limited, particularly in the context of preliminary procedures. In such procedures, providing a specific market definition in a given case is not an action the CJEU can take. The purpose of the preliminary procedure under art. 267 TFEU is to provide clarity on the application and legality of EU law. In the context of market definition, this means the CJEU can provide in its answer a general description of the required methodology according to established case law and Commission practice but not a specific delineation of a market or a view of the potential substitution between one or more specific products or services. Accordingly, the most that Booking.com could have achieved in this respect was to obtain guidance that would allow it to put forward a relevant market that consists of non-platform actors.
In practice, however, that did not happen. When asked how the market definition in the case of Booking.com needs to be defined, the CJEU recapitulated the approach of the Commission in this matter as indicated in the new Notice on the Market definition almost in its entirety. Such an approach is, however, entirely neutral when it comes to the outcome, thus leaving the final decision on this matter to the District Court of Amsterdam. While the District Court of Amsterdam is free to reach its conclusion, the commercial and legal context of this case dictate a relatively straightforward outcome for consistency.
The consistency of the market definition and relevance of increased market power
For Booking.com to fall under the VBER, a broad relevant market would have to be defined consisting of more types of actors than hotel room booking platforms so that Booking.com’s market share would not exceed the 30% threshold of art. 3 VBER. Such an outcome is, however, practically unthinkable when considering the established increase of market power by Booking.com in the EU Commission’s practice. First, in the blocked acquisition of Booking/eTraveli, the Commission indicated that Booking.com has a dominant position in the OTA market. This was, in essence, the foundation for the competitive concern identified for the concentration. Secondly, this provisional finding of dominance was later complemented by a designation as gatekeeper under the DMA. While there is no perfect overlap between the concept of dominance under art. 102 TFEU and gatekeeper under the DMA, the idea that an undertaking that meets the quantitative benchmarks of the DMA and is found to hold an entrenched durable position would possess less than 30% market share is, at the very least, questionable. Furthermore, the fact that the market definition in the case of Booking.com has already been performed multiple times and resulted in a market consisting solely of other platforms, it is not clear why, in the absence of entirely different market conditions, any other outcome should be expected. Reaching such a diverging finding would go, to a certain extent, against the rationale of the CJEU remark concerning the cross-jurisdictional value of market definition.
The cross-jurisdictional value of the market definition and digital platforms
According to the CJEU, the market definition delineated by a competition authority in the process of a competition law procedure (in this case, under 101 TFEU) can be used in the context of a different procedure in another Member State. While the cross-border use of evidence in competition law cases (particularly follow-on procedures) is not new, the evidentiary value is extended to market definitions performed in the context of diverging types of procedures. In the case of Booking.com, this means both market definitions done in the context of the MFN cases in the EU and the context of (EU) merger control have (at least) a prima facie evidentiary value. Accordingly, in the absence of entirely different market conditions, the District Court of Amsterdam should follow such findings. While the DMA designation decision does not fall under this category of evidence, it nevertheless indicates that Booking.com possesses a significant degree of market power that goes beyond what would be expected from parties falling under the VBER to hold.
In the greater context of the digital economy, where multiple procedures are conducted across different jurisdictions, such a cross-jurisdiction use of market definition findings can be very valuable. First, to make the initiation of ‘twin’ cases via public or private enforcement easier. Take, for example, the case against the Apple Appstore in the Netherlands concerning dating apps, which could, in principle, easily be replicated in other jurisdictions and applied to additional categories of apps as the conditions of iOS app distribution are identical across the board, and thus, so are the possibilities of substitution that consumers and developers have. The same would now also, in essence, apply to the case of Booking.com, where the practice in Germany could serve as evidence for follow-on claims in other member states. Secondly, this also allows for an easier initiation of new cases by taking the market definition from one procedure to another. This is particularly the case when it comes to merger control, which would typically offer more findings on market definition than antitrust procedures because they occur far more often.
At the same time, this possibility creates some friction in private enforcement cases since it may, at times, require the national judge to perform a task that would normally not do, namely to analyze the value and relevance of various market definitions. Arguably, this may be more than the judicial review by EU Courts for the market definition produced by the Commission in light of its complex economic nature. Luckily, however, such friction would likely be minimal as NCAs are increasingly better at coordinating their practices within the ECN framework.
Conclusion
With this judgment, the debate on the legality of Booking.com MFN clauses can be said to have come to an end, at least in the EU. With this judgment, it also becomes clear that relying on the ancillary restraint doctrine when it comes to the ongoing business practices of platforms will be highly challenging. For future arguments to have a chance at success, they need to relate to the fundamental elements that are essential to the existence of platforms as such and not solely important to one or more platform-based business models. These latter arguments are better suited for an individual justification in the context of art. 101(3) TFEU. The individual justification allows for a more case-specific analysis of the pro- and anti-competitive effects of the investigated restriction.
When it comes to market definition, the evidentiary value awarded to market definitions performed in the context of other procedures may help increase the number of cases that would otherwise be considered very burdensome to start in the context of both public and private enforcement. This possibility can, in a way, soften the unfortunate reality that the EU Commission cannot pick up all the relevant cases in the digital market context. Furthermore, it can strengthen the role of NCAs in the enforcement of (EU) competition in the case of digital platforms if the Commission’s focus in this respect is diverted to the DMA.