People
This article was first published on Global Competition Review in September 2025; for further in-depth analysis, visit the GCR Digital Markets Guide.
Introduction
Over the past 15 years, most-favoured nation clauses (MFNCs) have increasingly attracted the attention of European competition authorities. MFNCs, also known as “parity clauses”, typically ensure that a seller does not offer better terms (e.g., lower prices) to any buyer than those offered to the party covered by the clause. While historically used in offline contexts, the rapid expansion of digital markets and online platforms has seen a significant rise in the application of MFNCs in new settings.
In particular, platform MFNCs (PMFNCs), used by online intermediaries like booking websites and marketplaces, have become widespread. These clauses can prevent business users (e.g., hotels or retailers) from offering lower prices or better conditions on other platforms and even on their own direct distribution channels.[1] This article critically analyses MFNCs – especially PMFNCs – using economic research and recent European competition enforcement cases. It also examines how the Digital Markets Act (DMA) applies to MFNCs, banning them for designated gatekeepers and marking a major shift in EU digital regulation.
The article is structured as follows. In “MFNCs and forms”, we introduce and explain the concept of MFNCs and outline their different forms. In “Potential pro- and anticompetitive effects of MFNCs”, we delve into the economics of MFNCs. We discuss under what conditions MFNCs can generate pro-competitive effects, such as reducing transaction costs and encouraging investments by platforms by deterring “free-riding”. We also examine their potential anticompetitive effects, including the softening of price competition, raising platform fees, and creating barriers to entry and expansion. The analysis draws on economic literature and highlights the role of market structure in determining the net effects. In "Landmark cases on PMFNCs", we review recent decision practices by European competition authorities and courts in more detail. We consider landmark cases involving major platforms like Booking.com, Amazon and Apple, illustrating how enforcement approaches have evolved and diverged across jurisdictions. In “The DMA – a per se ban of PMFNCs for gatekeepers”, we discuss the implications of the DMA for MFNCs. In particular, we consider how the DMA codifies past enforcement concerns and its likely impact on platform business models and markets dynamics going forward.
MFNCs and forms
MFNCs have their origins in international trade. In this context, MFNCs aim to prevent discrimination by requiring that any benefit, such as a tariff reduction, granted to one trading partner be extended to all others.[2] This basic idea of non-discrimination later also found application in other areas such as investment treaties and commercial contracts.
More generally, MFNCs are contractual provisions under which one party commits to treating the other party no less favourably than it treats any other of its counterparties.[3]
MFNCs can apply to various contractual dimensions and at different levels of the supply chain, from upstream manufacturing to downstream distribution and retail. Earlier antitrust assessments of MFNCs generally focused on “wholesale” MFNCs – an MFNC relating to the wholesale price of a good – or other contractual terms in a business-to-business (B2B) relationship.
The European Commission (EC) has been assessing wholesale MFNCs since 1966, when it was notified of an agreement between Ateliers de Constructions Électriques de Charleroi (ACEC), a Belgian manufacturer of electric transmission systems, and Société des Automobiles Berliet (Berliet), a French bus manufacturer.[4] The case concerned a technical and research cooperation agreement between ACEC and Berliet for the development and commercialisation of electric transmission systems for buses. ACEC agreed to provide Berliet with the most favourable commercial terms it would offer any customer. The EC reviewed whether the agreement violated Article 85(1) of the Treaty establishing the European Economic Community (EEC Treaty), now Article 101 of the Treaty on the Functioning of the European Union (TFEU), and ultimately granted an exemption under Article 85(3) (now 101(3) TFEU).[5] Since then, several cases in various sectors have been investigated by the EC.[6]
In the past 15 years, the increasing relevance of digital markets has brought a new kind of MFNC, namely PMFNCs.[7] Those MFNCs have been prevalent in the context of online platforms that intermediate transactions in sectors such as hotel bookings, e-books, mobile apps, insurance, and other consumer goods. Unlike wholesale MFNCs, PMFNCs directly influence the prices and terms available to end-consumers. As online platforms have gained prominence, PMFNCs have faced growing attention from competition authorities and play a central role in policy debates, influencing pricing, platform rivalry, and digital market entry.
To better understand the economics of MFNCs in the context of online platforms, it is useful to examine how online platforms operate. A platform can be described as an intermediary operating in multi-sided markets, in which it seeks to facilitate interactions between different user groups (the “sides” of the market).[8] Indirect network effects are central to the functioning of a platform. These effects imply that the benefit enjoyed by a user of one group depends on how well the platform does in attracting users for the other group.[9] That is, for at least one user group, the platform becomes more attractive as the size of the other group increases.
PMFNCs typically take one of two forms: narrow or wide.[10]
Narrow PMFNCs confine the parity obligation to the supplier’s direct channel. That means that the terms of the supplier’s goods or services on the platform are required to be at least as attractive as on the supplier’s own direct distribution channel.[11] The supplier, however, may still offer better terms on other distribution channels. For instance, under a narrow PMFNC, a hotel cannot charge less on its own direct distribution channel than the lowest comparable rate it offers on the platform, but it may still grant lower rates on a competing platform or distribution channel.
Wide PMFNCs stipulate that, once a supplier agrees to specific terms on a platform, they are prohibited from offering more favourable terms through any other distribution channel, including alternative online platforms and their own direct sales channels. For example, under a wide PMFNC, if a hotel sets a room rate on a platform with a PMFNC, it is contractually prohibited from offering a better deal on other platforms or its own website.
PMFNCs typically apply to price, in which case they are known as price parity clauses (PPCs). Correspondingly, under a narrow PPC, the supplier agrees not to offer lower prices on its own website than on the platform. A wide PPC extends this restriction to all other platforms as well, including the supplier’s own site.
Potential pro- and anticompetitive effects of MFNCs
MFNCs have traditionally been considered as potentially having both pro- and anticompetitive effects, although views in this regard have evolved over time. In this section, we discuss the effects that have been raised in the academic literature, including competition authorities’ studies and guidelines.
Potential pro-competitive effects
PMFNCs may lead to pro-competitive effects (e.g., in the form of efficiency gains), under certain conditions. For example, it is often suggested that PMFNCs may protect platform investments and reduce problems in relation to free-riding, such as “showrooming”.[12] Showrooming occurs when potential customers use a platform to research products – checking availability, comparing brands and reviewing prices – but make the final purchase off-platform, usually through the supplier’s own direct channel. In this context, PMFNCs aim to internalise the value of discovery and information provision by preventing business users from free-riding on the platform's investment in attracting consumers.[13]
Further, PMFNCs may help mitigate the “hold-up” problem.[14] This problem typically arises when one party hesitates to make relationship-specific investments because they fear the other party may later exploit that dependence by renegotiating terms, switching partners or offering better terms to other partners.[15] This fear may lead to underinvestment and inefficiencies.[16] In platform markets, the hold-up problem may, for instance, occur when a platform invests in promoting a supplier’s products.[17] After gaining visibility and customers, the supplier might start offering better terms (e.g., lower prices) elsewhere.[18] This would put the platform at a disadvantage in the competition for customers, thus undermining the platform’s investment. In such cases, by guaranteeing investing platforms that they will not be worse off if better terms are later offered to others, wide and narrow PMFNCs may mitigate the hold-up problem, thereby encouraging early participation.[19] Such efficiency effects may be particularly relevant in settings where suppliers incur substantial fixed costs that are recouped through initial sales on investing intermediaries, creating incentives to offer lower prices to subsequent intermediaries in order to expand sales volume. If the problem significantly affects the intermediaries, they may hold up investments in the supplier, thereby preventing "new, better or cheaper products" from entering the market.[20]
Further, MFNCs may lower transaction costs by simplifying negotiations and reducing the need for constant price monitoring or renegotiation, particularly in markets characterised by high search costs.[21] If gathering information about prices is a costly process, PMFNCs may help to speed up the process and save costs on searching for the best deal, given that PMFNCs would effectively require the same deals across platforms and direct channels.[22]
Potential anticompetitive effects
While PMFNCs may lead to pro-competitive effects under certain conditions, it is also widely recognised that PMFNCs may raise important competition concerns: they can reduce platform competition by preventing suppliers from offering lower prices or rewarding platforms with lower commissions. This may limit the incentives of platforms to compete on commissions. Such effects can be particularly appreciable where one or more platforms have substantial market power, and a material proportion of consumers do not multi-home in the sense of comparing prices across platforms. They can also limit competition between the supplier and the platform, weakening incentives for suppliers to steer customers to their own, often cheaper, direct channels. Moreover, PMFNCs can produce foreclosure effects: by locking suppliers into uniform pricing obligations, they raise the cost of switching or using multiple platforms, increasing entry barriers for new or smaller intermediaries.[23] These risks have led multiple EU competition authorities to investigate and, in some cases, prohibit the use of PMFNCs, which we discuss in "Landmark cases on PMFNCs".
PMFNCs may soften the competition between platforms and direct channels (as regards wide MFNCs) and from direct channels (as regards narrow MFNCs), as they introduce rigidity into the competitive process by imposing uniformity across distribution channels. In terms of prices, when a supplier (e.g., a hotel or e-book publisher) agrees to a wide parity clause, it commits to offering the same price and commercial terms on all platforms. When the distribution takes place via an agency model, the immediate consequence is that the supplier cannot respond to variation in platform commission rates by offering a lower retail price on a lower-commission platform.[24] The same logic applies to room availability or promotional placement parity clauses: because any preferential term must be replicated across all channels, PMFNCs may suppress the supplier’s incentive to discriminate in favour of more efficient or competitive platforms. As a result, PMFNCs may soften price and non-price competition among incumbent intermediaries, by effectively removing suppliers’ ability to respond to commission differences through pricing. If every platform must receive the same listing terms, then platforms may have little incentive to reduce their fees, improve user experience or innovate with promotional tools, as suppliers cannot credibly reward those improvements with better terms (e.g., lower retail prices).
Further, PMFNCs may also impede market entry as they may create a barrier to entry and expansion for challenger platforms. A new entrant may wish to undercut incumbents on commission fees or introduce more consumer-friendly features, but if suppliers are bound to extend identical terms across all platforms, the entrant is unable to offer differentiated value to either side of the market. Without the ability to attract supply, a challenger platform may struggle to build a viable user base, which is essential to platforms, and its entry may be effectively foreclosed.
This foreclosure mechanism can be conceptualised in three steps. First, the incumbent platform with appreciable market power imposes a PMFNC as a non-negotiable condition of access for suppliers, such that the market coverage of MFNCs is high. Second, the supplier internalises that any superior terms granted to a rival platform must be matched on the incumbent channel, thereby eliminating any net gain from reallocation of traffic or volume to the lower-cost platform. Third, facing no prospect of commercial advantage, the supplier refrains from undercutting, and the entrant platform is left unable to differentiate its offer, regardless of its cost structure or efficiency. The result is the preservation of the incumbent’s market power and the marginalisation of potential competition.
However, competition concerns are not limited to wide PMFNCs. While narrow PMFNCs are less restrictive than wide PPCs in the sense that they do not restrict competition between platforms, they may still sustain similar economic outcome under certain conditions.[25]
Moreover, wide PMFNCs can reinforce vertical integration and deter innovation. Wide PMFNCs may discourage suppliers from granting more favourable terms to smaller or new platforms and can therefore limit experimentation with alternative business models. In such a case, PMFNCs do not just foreclose rival intermediaries but also reduce dynamic competition, as new business models or novel pricing structures are suppressed.
Furthermore, the inclusion of PMFNCs in agency agreements may facilitate upstream coordination among suppliers, who then, in turn, may raise prices as they internalise the commission in their pricing on all channels. This may result in consumer harm, as price dispersion disappears and average prices may increase.
In digital platform markets, these dynamics could be magnified by strong network effects. The market power of platforms may depend on both the existence of multiple sizeable platforms, but also the extent to which consumers shop around across platforms (i.e., multi-home), rather than using just one platform (i.e., single-home). In this regard, homogeneous pricing across platforms can reduce the incentive for consumers to multi-home to research and test alternatives, further entrenching incumbent dominance.[26] This not only consolidates user traffic but also reinforces the feedback loop of increasing returns to scale for the incumbent platform. Under such conditions, price parity clauses may act as a strategic bottleneck to some extent, which prevents competition from functioning properly.
Narrow PMFNCs can also contribute to this effect. By denying suppliers the freedom to undercut platform prices on their own websites, narrow PMFNCs prevent direct channels from becoming a competitive discipline on intermediary platform margins. The loss of this "outside option" may weaken the supplier’s ability to bypass high-commission platforms, leaving them locked into relationships where they have limited leverage and little capacity to negotiate better terms.
Balancing the effects of PMFNCs
Balancing the effects of PMFNCs typically requires a case-by-case assessment of both their pro-competitive and anticompetitive impacts. This involves evaluating whether any efficiency gains or consumer benefits are sufficient to outweigh potential harm to competition, considering the specific market context and characteristics of the agreement in question.
The EC and the UK Competition and Markets Authority (CMA) draw a distinction between wide and narrow MFNCs, with wide MFNCs having more wide-ranging potential to be anticompetitive even where a platform’s market share is no more than 30 per cent.[27]
In this regard, we note that the EC’s "Competition policy for the digital era" suggests that:
[. . .] because of very strong network externalities (especially in multi-sided platforms), incumbency advantage is important and strict scrutiny is appropriate. We believe that any practice aimed at protecting the investment of a dominant platform should be minimal and well targeted. If competition between platforms is sufficiently vigorous, it could be sufficient to forbid clauses that prevent sellers on a platform from price differentiating between platforms (i.e. a ban of “wide” MFNs) while still allowing clauses preventing the seller from offering lower prices on its own website ("narrow" MFNs). If competition between platforms is weak, then pressure on the dominant platforms can only come from other sales channels (e.g. in the case of hotel booking platforms, direct sales by hotels on their own websites) and it would be appropriate to also prevent "narrow" MFNs.[28]
Landmark cases on PMFNCs
PMFNCs have been the subject of multiple infringement decisions and investigations under Articles 101 and 102 TFEU. While wide PMFNCs are generally considered anticompetitive, the treatment of narrow PMFNCs varies across national authorities, reflecting a fragmented enforcement landscape.[29] This section reviews key cases to illustrate how competition authorities and courts have assessed PMFNCs across different markets and legal frameworks.
E-books
One of the earliest decisions providing guidance on the EC's stance towards PMFNCs was the Apple iBooks case.[30] In 2010 to 2011, Apple sought to launch e-book sales on its iBooks platform and entered negotiations with five major publishers, all of whom had existing wholesale agreements with Amazon. The parties ultimately agreed to shift to an agency model, where publishers set the retail prices and Apple acted as an agent, earning a commission on each sale.[31] The contracts with Apple included wide PMFNCs that required publishers to guarantee the lowest retail price for any given title on Apple’s platform.[32]
The EC considered that these contracts raised competition concerns "as to their compatibility with Article 101 of the Treaty and Article 53 of the Agreement on the European Economic Area". The EC took the position that the PMFNCs acted as a commitment device, aligning the publishers’ incentive to force Amazon to change from a wholesale to an agency model.[33] In turn, this generalisation of the agency model then allowed the publishers to effectively align retail pricing across platforms and facilitated horizontal coordination among e-publishers.[34] Ultimately, Apple and the publishers involved committed to terminate the PMFNC and refrain from similar practices for a period of five years, finishing in 2017.[35]
Similarly, in June 2015, the EC, on the basis of Article 102 TFEU, initiated an investigation into Amazon’s e-book distribution agreements with major publishers.[36] In 2017, the authority found a range of potentially anticompetitive provisions embedded in Amazon’s agreements including the use of wide PMFNCs.[37] In the contracts, publishers were obliged to offer Amazon terms no less favourable than those granted to any other distributor, and to notify Amazon of any superior deals, thereby triggering Amazon’s right to match or renegotiate.[38] In its decision, the EC concluded that these practices risked foreclosure of rival e-book platforms and reduced "[. . .] E-book Suppliers' incentives to support, and invest in, new and innovative business models [. . .]".[39] Amazon agreed to cease the use of PMFNCs for five years throughout the EEA.
Hotel booking platforms
Between 2013 and 2025, online travel agencies (OTAs) such as HRS, Booking.com and Expedia came under scrutiny for their use of PMFNCs.[40] OTAs are platforms that facilitate the buying and selling of travel-related services.[41] Typically, an OTA aggregates offers from multiple suppliers – such as hotels – and displays them to consumers through its online platform.
HRS
On 20 December 2013, the German competition authority, the German Federal Cartel Office (Bundeskartellamt), prohibited wide PMFNCs in contracts between HRS[42] and hotels as a form of vertical restraint.[43] According to the authority, the PMFNCs of HRS "violate section 1 GWB (Act against Restraints of Competition) and Art. 101 (1) TFEU".[44]
With respect to the commercial relationship between HRS and the hotels, the Bundeskartellamt recognised that HRS could not be considered an agent in its legal sense under EU competition law.[45] The authority signalled that "[t]he activities of HRS do not depend on the hotel partners of HRS. The MFN clauses do not restrict the conduct of the alleged agent, but rather that of the alleged principal". In addition, it was pointed out that HRS was not a dependent agent, as HRS was responsible for its own financial and economic risks.[46]
HRS appealed the Bundeskartellamt’s decision, but in January 2015, the Higher Regional Court of Düsseldorf (Oberlandesgericht Düsseldorf) upheld the finding, ruling that HRS’s PMFNCs represented a significant restriction of competition.[47] This decision had significant competition law implications on the use of PMFNCs by OTAs.[48]
Booking.com and Expedia
Booking.com and Expedia, two of the largest OTAs operating in the EU, also required hotels to adhere to similar parity obligations. These wide PMFNCs were the focus of multiple investigations by national competition authorities (NCAs) across the EU. In 2015, the French Autorité de la concurrence, the Italian Autorità Garante della Concorrenza e del Mercato and the Swedish Konkurrensverket accepted commitments from both Booking.com and Expedia to discontinue the use of wide PMFNCs and replace them with narrow PMFNCs.[49] These commitments were considered sufficient by other NCAs – such as those in Austria, Poland and Denmark, which in consequence closed their investigations in the online hotel booking sector.[50], [51], [52] The CMA closed its investigations due to administrative priorities after these commitments were given by Booking.com and Expedia.[53] The revised contractual model was implemented across the EU in July 2015 for Booking.com, and in August 2015 for Expedia.[54]
Nonetheless, in 2016, different NCAs carried out a monitoring exercise commissioned by the heads of the European Competition Network (ECN) "to measure the effects of recent changes to the parity clauses used by online travel agents (‘OTAs’) in their contracts with hotels".[55]
Despite the adoption of narrow PMFNCs across the EU, the Bundeskartellamt and German courts took a more stringent position. On 23 December 2015, the Bundeskartellamt fined Booking.com for using narrow PMFNCs in contracts with hotels in Germany.[56] The Bundeskartellamt considered that such clauses restricted competition both on the market for the provision of accommodation services, and on the market for the provision of online intermediation services by platforms to accommodation providers.[57] The Bundeskartellamt made no distinction between wide and narrow PMFNCs, finding that any such clauses significantly restricted competition by discouraging lower commissions, limiting new sales strategies and foreclosing market entry – especially when used alongside similar clauses by other platforms.[58]
In the Bundeskartellamt’s view, even the proposed narrow PMFNCs continued to infringe both German and EU competition law. The Bundeskartellamt dismissed potential pro-competitive justifications, concluding that free-riding concerns were minimal and outweighed by the negative effects on competition.[59] It also noted that less restrictive remuneration models, such as fixed or percentage-based fees, could adequately protect the OTAs’ investments.[60]
Booking.com appealed against the decision at the Oberlandesgericht Düsseldorf. The court’s judgment on 4 June 2019 upheld, in part, the appeal brought by Booking.com as it found that, although the narrow PMFNC did restrict competition, it could nevertheless, as an ancillary restraint, be deemed necessary to enable Booking.com to receive fair remuneration for its provision of services. The court therefore concluded that the possibility for accommodation establishments to transfer reservations to their own reservation systems was sufficient justification for Booking.com to prevent them contractually from engaging in free-riding activities.[61]
However, the Oberlandesgericht Düsseldorf’s judgment was annulled by the German Federal Court of Justice (Bundesgerichtshof) on 18 May 2021, hearing an appeal brought by the Bundeskartellamt. The narrow PMFNC was again found to have significantly restricted competition on the market for online hotel reservation platforms and on the market for hotel accommodation. Further, the court concluded that such a clause could not be classified as an ancillary restraint, since it had not been established that, in its absence, Booking.com’s profitability would be compromised.[62]
Following the judgments by the German courts, PMFNCs were considered further in Spain in 2024. On 29 July 2024, Spain’s Comisión Nacional de los Mercados y la Competencia (CNMC) imposed fines totalling €413.24 million on Booking.com for abusing its dominant position. The CNMC fined Booking.com for two separate infringements, an exploitative abuse that relates to PMFNCs and an exclusionary abuse.[63] According to the CNMC, the exploitative abuse consisted of imposing unfair trade terms through an asymmetric PPC.[64] Interestingly, the CNMC did not take a more stringent position on narrow PMFNCs per se, but fined the combination of a narrow PPC together with the asymmetry in pricing. However, Booking.com has agreed to abolish any form of PMFNCs under its DMA commitments.[65]
Additionally, on 19 September 2024, the European Court of Justice (ECJ) issued a preliminary ruling in response to a request from a Dutch court regarding Booking.com’s narrow PMFNCs. Booking.com had argued that its PPCs were necessary for the proper functioning of its platform and thus should be considered ancillary restraints under competition law. In this context, the ECJ clarified that PPCs cannot be considered ancillary restraints. The Court reasoned that these clauses were neither objectively necessary for the operation of the platform nor proportionate to its legitimate business objectives, such as preventing free-riding, meaning they could not fall outside the prohibition of restrictive agreements and thus fell under the scope of Article 101(1) TFEU.[66]
Further investigations – marketplaces, insurance market and food delivery
In 2012 and 2013, the UK Office of Fair Trading (OFT) and the Bundeskartellamt investigated wide PMFNCs regarding Amazon’s marketplace.[67] Both investigations were closed when Amazon decided to end the PMFNCs.[68]
Later in 2014, the UK CMA concluded its private motor insurance market investigation.[69] The CMA found that certain contracts between private motor insurance providers and price comparison websites (PCW) included wide PMFNCs, which restricted insurance providers from offering lower prices through other sales channels.[70] These clauses were found to limit price competition, hinder innovation, and create barriers to entry, ultimately leading to an adverse effect on competition. The CMA ultimately banned the use of wide PMFNCs but allowed the use of narrow PMFNCs.[71]
In a more recent decision, on 19 November 2020, the CMA considered the wide PMFNCs imposed by a PCW comparethemarket.com (CTM) in its agreements with a number of home insurers. CTM’s wide PMFNCs contractually prevented these home insurers from quoting lower prices on CTM’s rival PCWs. The CMA found that CTM’s network of wide PMFNCs had the appreciable effect of preventing, restricting, or distorting competition between PCWs and between home insurers competing on PCWs.[72]
However, the UK Competition Appeal Tribunal (CAT) decided to set aside the CMA’s decision on 8 August 2022, after hearing CTM’s appeal against the CMA decision. The CAT found that the CMA’s analysis was flawed. In particular, as regards the alleged anticompetitive effects of the wide PMFNCs, the CAT considered that a great deal of the CMA’s analysis operated at the level of theory or bare assertion, with no significant reference to quantitative evidence: (1) there was no reliable evidence upon which to conclude the existence of any adverse effect of wide PMFNCs on either insurance premiums or the commissions,[73] (2) the evidence the CMA adduced was anecdotal at best and lacked depth and consistency with the CMA’s theory of harm, and (3) it was not possible for CTM and the CAT to test the evidence relied upon in any way.[74]
In 2015, the German authority Bundeskartellamt announced that it had closed its investigation into Verivox, a price comparison platform for electricity and gas. The Bundeskartellamt closed the investigation after Verivox voluntarily removed all PMFNCs from its contracts with energy providers.[75]
More recently, in July 2023, the Bundeskartellamt, announced it had closed its investigation into Yd. Yourdelivery GmbH,[76] for its use of narrow PMFNCs in its contracts with restaurants.[77] As in previous cases, the PMFNCs raised concerns, but the authority found that current market conditions were marked by a dynamic and evolving food delivery sector with new entrants like Uber Eats and Wolt. Therefore, the Bundeskartellamt concluded that it did not "have sufficient indications [. . .] to suggest that the clause represents a serious barrier to market entry by new platforms offering differentiated services".[78]
The DMA – a per se ban of PMFNCs for gatekeepers
The DMA entered into force in November 2022. Since then, it identifies certain large platforms as gatekeepers, imposing a strict menu of ex ante obligations to prevent anticompetitive conduct in digital markets.[79] The DMA imposes an explicit ban on PMFNCs for designated gatekeeper platforms. Article 5(3) DMA prohibits a gatekeeper from preventing its business users from offering the same products or services on other platforms or through their own direct channels on more favourable terms, including lower prices. In effect, any contract term or policy that compels business users to give the gatekeeper the cheapest price or best conditions is prohibited.[80]
Recital 39 of the DMA underscores the EU’s rationale: such parity restrictions undermine inter-platform contestability and unfairly limit business users’ freedom to use alternative online channels. The DMA thus ensures that business users can "differentiate commercial conditions, including price" across channels, with the aim of enhancing competition. Importantly, the ban extends not only to explicit MFNCs but also to any measure with equivalent effect, such as punitive higher commissions or de-listing applied by a gatekeeper in retaliation for a business user offering lower prices elsewhere.[81] This broad language prevents gatekeepers from shifting to indirect or algorithmic tactics to enforce price parity.
It has been argued that the CNMC fined Booking.com for such algorithmic tactics.[82] Booking.com implemented the number of bookings on Booking.com as a ranking criterion. This criterion may have served two purposes: (1) disincentivising hotels from multihoming and offering higher prices on Booking.com compared to other platforms, and (2) cementing the price parity, because even if hotels were to offer lower prices on other distribution channels, the hotels would be penalised in terms of a lower ranking.
Conclusions
This article has analysed the competitive implications of MFNCs focusing on their application by online platforms. MFNCs can, under certain conditions, result in efficiency gains such as protecting platform investments and mitigating free-riding. At the same time, they may restrict pricing freedom, reduce inter-platform competition, and raise barriers to entry, particularly when imposed by platforms with significant market power.
The enforcement record across Europe reflects a degree of convergence in the treatment of wide PMFNCs, which have frequently been found problematic due to their broader market effects. In contrast, narrow PMFNCs have received more divergent assessments. While some authorities and courts have accepted efficiency justifications, others have emphasised their potential to restrict competition, particularly in concentrated or structurally constrained markets. Recent case law, including the ECJ’s decision on Booking.com, has clarified that such clauses do not generally qualify as ancillary restraints, reinforcing the need for a substantive effects-based analysis. The overarching lesson is that the effects of PMFNCs depend on a range of factors, including market structure, platform dynamics, and the availability of alternative, less restrictive contractual arrangements. For legal and economic analysis, this implies the need for careful case-by-case evaluation, taking into account both potential benefits and competitive risks.
The DMA’s per se prohibition of MFNCs and any measure with equivalent effect represents a significant regulatory intervention in digital markets that carries significant ramifications for digital platform markets. Legally, the DMA’s prohibition of PMFNCs creates a per se rule that supplants the previous case-by-case approach under competition law when gatekeepers are under investigation. Gatekeepers can no longer generally defend such clauses as efficiency-enhancing – the DMA reflects a policy judgement that the harms outweigh any justifications.
From an economic perspective, it remains to be seen if the end of MFNCs on major platforms will improve price competition and spur platform rivalry. Business users are now free to price their products differently across channels – offering discounts on alternative platforms or direct channels without fear of contract breach or demotion. At least in theory, this ability to multi-home and differentiate terms may help erode the lock-in power of gatekeepers. In the longer run, gatekeepers may face competitive pressure to lower their commissions or improve services to attract and retain business users if they can no longer rely on PMFNCs to keep prices uniform. If and how these effects materialise will, however, depend on the circumstances of the market.
Endnotes
[1] PMFNCs can apply broadly to suppliers’ own direct online and offline distribution channels or be limited more narrowly to a supplier’s own online channel.
[2] See, for instance, the WTO principles:https://WTO | Understanding the WTO – principles of the trading system.
[3] When a seller promises a buyer terms at least as favourable as those offered to others, this is known as a most favoured customer clause (MFCC). Conversely, if a buyer guarantees at least as favourable terms to its supplier than it provides to any other supplier, this is a most-favoured supplier clause (MFSC). See, for example, Baker, J. and Chevalier, J. (2012). “The competitive consequences of most favoured-nation provisions”, Antitrust, 27: 20.
[4] German version of EC Decision Case IV/26045-ACEC-Berliet. More precisely, the EC was investigating a wholesale MFCC. Investigations into MFSC at the wholesale MFSCs are less prevalent. The EC investigated them, for instance, in its Hollywood Studio investigations (see Press release: ec.europa.eu/commission/presscorner/detail/en/ip_04_1314). In this context, the EC began investigating MFNCs in May 2002, after finding they appeared in most "output deals" between major Hollywood studios and European pay-TV broadcasters. It closed the investigation after the studios decided to withdraw the clauses.
[5] German version of EC Decision Case IV/26045-ACEC-Berliet, Section I.
[6] For a list of cases, see, for instance, Bostoen (2017). "Most Favoured Nation Clauses: Towards an Assessment Framework under EU Competition Law", CORE (3), 223–226.
[7] See, for instance, Boik A. and Corts K. (2016). "The effects of Platform Most-Favored-Nation Clauses on Competition and Entry", Journal of Law and Economics (59), 105. PMFNCs are also known as retail MFNCs (RMFNCs).
[8] Rochet J.-C. and Tirole, J. (2003). "Platform competition in two-sided markets", Journal of the European Economic Association (1), 990. In that sense online platforms typically connect two, sometimes more, distinct user groups. Typically, business users, such as hotels, publishers or app developers are connected to end users, such as travellers, readers and mobile app consumers.
[9] For instance, the more travellers a booking platform attracts, the more appealing it becomes to hotels, and vice versa; the more readers on an e-book site, the more valuable it is to publishers.
[10] Ezrachi, A. (2015). "The competitive effects of parity clauses on online commerce", European Competition Journal, 11:2-3, 488-519, DOI: 10.1080/17441056.2016.1148870.
[11] Strictly speaking, this definition varies depending on the distribution model that the platform has in place. Distribution models of platforms may have different forms, ranging from a wholesale model to an agency model and combinations between those. In a wholesale model, the supplier sets the price charged to the online platform but not the final consumer price on the platform. Here, MFNCs help ensure the platform maintains competitive costs in the sense that other platforms are not offered lower wholesale prices by suppliers. Under an agency model, the supplier determines the final retail price on the platform, while the platform earns a commission based on agreed revenue sharing. Here, the platform acts purely as an intermediary without buying inventory. MFNCs do not affect the platform’s costs directly but guarantee to the platform in question that retail prices remain competitive for consumers. Other distribution arrangements may blend wholesale and agency features. For example, a hybrid wholesale model (or merchant model) might involve the supplier setting a wholesale price with an agreed fixed markup for the retailer. In this article we mainly focus on the agency model, and hence the definition of a PMFNC is as provided.
[12] See, for example, Wang, C. and Wright, J. (2020). “Search platforms: Showrooming and price parity clauses”, the RAND Journal of Economics, 51(1), 32–58.
[13] These free-riding arguments typically assume that: (1) other platforms or the direct channels have not made similar investments in consumer acquisition; and (2) an appreciable proportion of consumers would otherwise engage in showrooming, such that this would otherwise adversely affect investments (including in a platform’s reputation for offering low prices, such as where the platform highlights a low priced or value option, which might cause consumer complaints if consumers frequently find such products cheaper elsewhere).
[14] Strictly speaking, PMFCs may prevent rent-seeking behaviour by the supplier and thereby mitigate the hold-up problem.
[15] Such relationship-specific investments may include tailoring products, committing resources or building infrastructure.
[16] Particularly so, in long-term or high-stakes commercial relationships.
[17] This promotion may for instance be conducted through advertising, algorithmic placement or consumer acquisition.
[18] This includes other platforms as well as the supplier’s own direct channel.
[19] Wide PMFNCs prevent rent seeking by suppliers on their own channel and other platforms, while narrow PMFNCs prevent rent seeking with respect to the suppliers’ own direct channel.
[20] See discussion in EC, "Support studies for the evaluation of the VBER, Final Report", 2020, p 104.
[21] Baker, J. and Chevalier, J. (2012). "The competitive consequences of most favoured-nation provisions", Antitrust, 27: 20.
[22] For example, see discussion in Motta M., "Competition Policy: Theory and Practice", 2004, p 158; EC, "Support studies for the evaluation of the VBER, Final Report", 2020, p 104.
[23] There is an abundance of academic literature discussing potential anticompetitive effects of MFN clauses. For example, Fletcher, A. and Hviid, M. (2016). "Broad Retail Price MFN Clauses: Are They RPM ‘At Its Worst’?", Antitrust Law Journal 81(1); Boik, A. and Corts, K. S. (2016) "The Effects of Platform Most-Favored-Nation Clauses on Competition and Entry", Journal of Law and Economics 59(1); Johnson, J. P. (2017). "The Agency Model and MFN Clauses", Review of Economic Studies 84. Moreover, Section 3.4.2.3.1 of EC’s "Support studies for the evaluation of the VBER, Final Report", 2020, provides a more comprehensive literature review of MFN’s pro- and anticompetitive effects.
In addition, there are also several competition authority decisions in relation to MFNCs such as the German NCA’s 2013 prohibition decision relating to the MFNCs of the German OTA HRS. We discuss these decisions in the subsequent section.
[24] In other words, for instance a hotel cannot reward an online travel agent (OTA) that charges a 10 per cent commission by listing a cheaper room rate, nor can it penalise a platform charging 20 per cent by raising the retail price on that channel.
[25] The EC considers that "[u]nder certain conditions, in particular where the number of providers of online intermediation services is limited, narrow retail parity obligations may affect the incentives of buyers of the online intermediation services to pass on changes in the price of the intermediation services in their retail prices. This may lead to a softening of competition between the providers of online intermediation services which is similar to the effect of across-platform retail parity obligations". (EC, "Guidelines on vertical restraints", 2022/C 248/01, 30 June 2022, para 371).
[26] For example, if every hotel offers the same price on all OTAs, consumers may have no reason to search beyond the most prominent platform.
[27] For example, the new EC VBER guidelines consider that across-platform retail parity obligations are more likely than other types of parity obligation to produce anticompetitive effects and have excluded wide MFNCs from the scope of safe harbour while keeping narrow MFNCs in it. (Section 8.2.5 "Parity obligation" of EC, "Guidelines on vertical restraints", 2022/C 248/01, 30 June 2022). Meanwhile, wide MFNCs and any measure with the same effect qualify as hardcore restrictions under Article 8 f of the UK’s "Vertical Agreement Block Exemption Order" (VABEO).
[28] EC, "Competition policy for the digital era", 2019, pp 5–6.
[29] An exception may be the CMA’s investigation into Comparethemarket, an insurance comparator platform that has been overturned by the CAT due to insufficient evidence for anticompetitive effects. See also Section "Further Investigations – marketplaces, insurance market and food delivery".
[30] EC Decisions Case COMP/AT.39847-E-BOOKS.
[31] id, para 30.
[32] id, para 31. This agreement also covered Amazon’s e-book platform, although there was a wholesale agreement in place with Amazon and not an agency model. This implies that if Amazon lowered the price for a title on its platform, the publishers had to follow suit in lowering this title’s price on Apple’s platform.
[33] id, para 38.
[34] id, para 57.
[35] id, 5.1.2 and 5.2.2.
[36] Press release: https://ec.europa.eu/competition/antitrust/cases/dec_docs/40153/40153_1359_6.pdf.
[37] EC Decision Case AT.40153 E-book MFNs and related matters (Amazon), para 2.
[38] id, section 4.5.
[39] id, para 78.
[40] At first glance, these platforms may appear to operate under an agency model, just like the Apple’s iBooks arrangement, where hotels set the retail room prices and the OTA receives a commission for each booking. However, the legal classification of this model under EU competition law is more complex. Under EU competition law, an agency relationship is only considered genuine if the agent assumes no or only insignificant commercial or financial risks (e.g., inventory costs, marketing expenses, payment processing). In this regard, OTAs often bear significant commercial and financial risks, which challenges the notion that they qualify as genuine agents within the meaning of EU competition rules. See for instance, the English version of the Bundeskartellamt Decision Case B 9 - 66/10, paras 147–149.
[41] These services include, among others, hotel rooms, car rentals and vacation packages. Such platforms fall into the narrow functional category of "travel booking" presented by the OECD in "An Introduction to Online Platforms and their Role in the Digital Transformation", https://www.oecd.org/en/publications/an-introduction-to-online-platforms-and-their-role-in-the-digital-transformation_53e5f593-en.html.
[42] HRS-Hotel Reservation Service Robert Ragge GmbH is a German OTA.
[43] English version of Bundeskartellamt Decision Case B 9 - 66/10, p 3–4.
[44] id, para 137. Because HRS held a market share exceeding 30 per cent in value terms during the relevant time period, the agreement did not qualify for exemption under the Vertical Block Exemption Regulation (VBER) (see English version of Case B 9 - 66/10, paras 188–189).
[45] id, paras 147–148.
[46] id, para 148.
[47] See OLG Düsseldorf judgment Case VI-Kart 1/14(V), 9 January 2015.
[48] Press release: https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2015/09_01_2015_hrs.html?nn=55030.
[49] Press release: https://en.agcm.it/en/media/detail?id=42f88c3c-d668-409f-b604-40581a05c97b.
[50] BWB Report 2015, https://www.Tätigkeitsbericht BWB 2015, p 41.
[51] https://uk.practicallaw.thomsonreuters.com/2-621-3867?transitionType=Default&contextData=(sc.Default).
[52] The Danish Competition and Consumer Authority terminates investigation of hotel booking portals’ price clauses (MFN clauses) | Antitrust Alliance.
[53] Summary Decision: Hotel online booking investigation: Case closure summary.
[54] Press release: https://en.agcm.it/en/media/detail?id=42f88c3c-d668-409f-b604-40581a05c97b, 25 June 2025, according to the Report on the monitoring exercise carried out in the online hotel booking sector by EU competition authorities in 2016, https://competition-policy.ec.europa.eu/document/download/c50b5e61-1ea3-47bc-ac4c-57d35f8f77fa_en?filename=hotel_monitoring_report_en.pdf Expedia decided to apply a narrow parity clause EU-wide, from August 2015. Shortly after the French competition authority accepted the commitments, any form of price parity clause has been deemed illegal under "Loi Macron" (Loi n° 2015-990 du 6 août 2015). In November 2016, Austria amended its competition rules via legislation to prohibit both wide and narrow PMFNCs in contracts between OTAs and hotels. Similarly, any type of price parity clauses by OTAs has been banned in Italy in August 2017. In July 2018, Belgium followed suit with the publication of the Act on Pricing Freedom for tourist accommodation that bans all forms of MFNCs.
[55] Report on the monitoring exercise carried out in the online hotel booking sector by EU competition authorities in 2016, para 1.
[56] Bundeskartellamt Decision Case B 9-121/13.
[57] See, for instance, Bundeskartellamt Decision Case B9-121/13, para 6.
[58] id, pp 73–94.
[59] id, pp 98–110.
[60] Bundeskartellamt Decision Case B 9-121/13, pp 114–115. The Bundeskartellamt found that Booking.com failed to convincingly demonstrate why realistic, less restrictive alternatives to its business model based on narrow MFNCs would be significantly less efficient. Instead, Booking merely reiterated earlier claims dismissing models such as cost-per-click payments or user fees, without addressing viable alternatives already used in hospitality (e.g. Airbnb), B2B supplier platforms (e.g. Wer liefert was), and real estate portals (e.g. ImmoScout with membership models, and Immowelt/Immonet with ad-based pricing). The authority emphasised that these pricing models are established in comparable digital markets, and Booking.com’s rejection of them lacked credible explanation.
[61] OLG DUS judgment Case VI-Kart 2/16 (V), 4 June 2019, paras 58 following.
[62] ECJ Decision Case C-264/23, 19 September 2024, paras 59–63.
[63] The exclusionary abuse restricted competition from using the number of bookings on Booking.com as a ranking criterion thus encouraging hotels to concentrate their sales on Booking.com and using a hotel’s profitability on Booking.com as a criterion to have access to the Preferred Partner and Preferred Partner Plus programmes. See CNMC Summary Decision, p 2, available at https://www.cnmc.es/sites/default/files/5555287.pdf5555287.pdf.
[64] The asymmetry allowed Booking.com to unilaterally lower prices on its platform via its BSB programme.
[65] See Booking.com, "Understanding the Digital Market Act", available at: https://Understanding the Digital Market Act | Booking.com for Partners.
[66] See eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:62023CJ0264.
[67] According to the OFT, in October 2012, it had launched an investigation following numerous complaints, citing concerns that the policy could be anticompetitive by raising platform fees, discouraging new market entrants, and leading to higher prices for consumers.
[68] For the Bundeskartellamt, see press release: https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2013/27_08_2013_Amazon-Preisparit%C3%A4t.html. The UK Office of Fair Trading (OFT) closed its investigation on grounds of administrative priorities (see https://webarchive.nationalarchives.gov.uk/ukgwa/20140402160400/http://oft.gov.uk/news-and-updates/press/2013/60-13.)
[69] https://assets.publishing.service.gov.uk/media/5421c2ade5274a1314000001/Final_report.pdf.
[70] Private motor insurance market investigation, para 9.
[71] id, para 63.
[72] UK CMA, "Price comparison website: use of most favoured nation clauses" Case 50505, 19 November 2020.
[73] In other words, there was no reliable evidence upon which to conclude the existence of any adverse effect of wide PMFNCs on either prices charged by insurance providers to consumers or prices charged by price comparison websites to insurance providers.
[74] UK CAT, Case No: 1380/1/12/21, 8 August 2022.
[75] Press release: https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2015/03_06_2015_Verivox.html. Although the authority found no competition concerns with Verivox’s current services, it emphasised that such clauses could hinder competition by restricting suppliers’ pricing freedom and limiting inter-platform rivalry.
[76] Yd. Yourdelivery GmbH is also known as Lieferando.
[77] Thus, the clauses required that prices on its platform match those offered through the restaurants’ own channels.
[78] Press release: https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2023/12_07_2023_Lieferando.html.
[79] Criteria for being designated a gatekeeper are laid out in Article 3 of the DMA.
[80] Thus, this prohibition applies to narrow and wide PMFNCs alike.
[81] DMA, recital 39.
[82] Hausfeld,Algorithmic enforcement of price parity .
