European Union: Article 102 TFEU and the European Commission’s 2025 Guidelines: what to expect and what to ask for

In summary

In March 2023, the European Commission published a communication foreshadowing amendments to its Article 102 TFEU Guidance and acknowledged that its current guidance no longer reflected the commission’s approach in determining whether to pursue a case as a matter of priority. This article discusses the proposed 2025 Guidelines against the commission’s recent decisional practice and the European jurisprudence.

Discussion points

• Article 102 TFEU and the European Commission’s 2009 Guidance
• European Commission’s 2025 Guidelines
• Better safeguarding the rights of defence of parties under investigation through the 2025 Guidelines

Referenced in this article

• European Commission’s 2025 Guidelines
• Intel v Commission
• Google Shopping
• Qualcomm Inc v Commission
• Unilever Italia v AGCM
• Google Android
• Servizio Elettrico Nazionale, ENEL SpA, and Enel Energia SpA v Autorità Garante della Concorrenza e del Mercato and Others
• Bronner
• Lithuanian Railways
• Towercast
• Regulation 1/2003

Article 102 TFEU and the European Commission’s 2009 Guidance

Article 102 of the Treaty on the Functioning of the European Union (TFEU) governs practices by companies able to act independently of their competitors, customers and, ultimately, consumers. It prohibits not only those practices that directly cause harm to consumers (eg, exploitative abuses), but also practices that cause consumer harm through their impact on competition (eg, exclusionary practices).

In the European Union, article 102 TFEU is enforced by the European Commission (EC) and the national competition authorities (NCAs). To guide companies on its enforcement priorities in applying article 102 TFEU to abusive exclusionary conduct by dominant undertakings, the EC adopted guidance in February 2009 (EC’s Guidance).

In last year’s edition of this publication, we suggested that, based on recent decisional practice, jurisprudence and the evolving concept of exclusion in Europe, the EC’s Guidance may require updating. Recent theories of harm and anticompetitive practices, such as self-preferencing, data leveraging, naked restrictions, excessive pricing and anticompetitive disparagement, have not yet been adequately addressed by the EC’s Guidance. In addition, we argued, traditional forms of anticompetitive practices that are referenced in the EC’s Guidance, such as fidelity-inducing loyalty rebates, may require revision following the case law of the Court of Justice of the European Union (CJEU), including in relation to the as-efficient competitor (AEC) test and the EC’s standard of evidence.

In March 2023, the EC published a communication foreshadowing its amendments to the EC Guidance. In its Amending Communication and Policy Brief, the EC acknowledges how its Guidance no longer reflects the approach in determining whether to pursue as a matter of priority certain cases of exclusionary conduct. The EC expects to adopt new guidelines in 2025 (the 2025 Guidelines).

2025 Guidelines: what to expect and what to ask for

In its Policy Brief, the EC accepts the premise that its 2025 Guidelines will need to focus on an effects-based approach and codify the CJEU’s recent judgments. The EC is likely to avoid an overly rigid implementation of the effects-based approach in the 2025 Guidelines, but will retain a sufficiently descriptive approach to facilitate enforcement.

Based on recent CJEU jurisprudence, this article explores additional guidance and clarifications that the EC may seek to address in its 2025 Guidelines.

Reflecting the effects-based approach

The CJEU has confirmed and endorsed the main elements of an effects-based approach to exclusionary conduct by dominant undertakings in numerous recent judgments, including in Intel and Google Shopping, and even more recently in Qualcomm and Unilever.

In these judgments, the CJEU established that it will be more difficult for the EC to characterise certain types of abuses, including fidelity-inducing rebates, exclusivity restrictions and exclusivity payments as automatic infringements under article 102 TFEU. Where a company under investigation submits arguments and supporting evidence that this conduct did not result in an anticompetitive effect, the EC must assess the effects of the conduct, and to find an abuse, the EC must establish that the practice of a dominant undertaking has ‘at least potential – anti-competitive effects in the relevant market’.

In this section, we assess the CJEU’s recent jurisprudence and provide recommendations on how the EC may seek to address and reflect recent judgments in the 2025 Guidelines.

Fidelity-inducing loyalty rebates are not automatically unlawful

As set out in last year’s edition of this publication, in January 2022, the General Court (GC) handed down its (revised) judgment in Intel, finding that where a dominant company implements a system of fidelity-inducing loyalty rebates and demonstrates that the rebates do not restrict competition or have foreclosing effects, the EC must take the dominant undertaking’s defence into account. This means that fidelity-inducing loyalty rebates are not automatically illegal under article 102 TFEU and it does not ‘relieve the Commission in all cases of the obligation to examine whether there were anticompetitive effects’.

Exclusivity clauses are not automatically unlawful

Unilever also narrows the scope, finding that exclusivity clauses do not automatically infringe article 102 TFEU. In January 2023, the CJEU held in this case that exclusivity clauses imposed by a dominant company do not automatically amount to an abuse of dominance under EU law. Instead, a competition authority must:

• apply the criteria established in Intel to assess exclusionary effects; and
• examine the economic evidence advanced in defence of the dominant undertaking.

The facts of the case were as follows: through its network of distributors, Unilever financially incentivised ice cream sellers in Italy to exclusively buy Unilever’s individually packed impulse ice cream. Following a complaint, the Italian Autorità Garante della Concorrenza e del Mercato (AGCM) investigated this behaviour and fined Unilever about €60 million for abusing its dominant position, in a decision published on 31 October 2017.

When a regional Tribunal in Italy confirmed the AGCM’s decision, Unilever appealed to Italy’s Council of State, arguing (inter alia) that the AGCM had wrongfully ignored Unilever’s economic analysis, which demonstrated that the exclusivity clauses had no restrictive effects on the market. Prior to ruling on the case, the Italian Consiglio di Stato referred certain questions to the CJEU, including whether the competition agency’s standard of proof in an abuse of dominance case, implemented through exclusivity clauses, requires it to carry out an AEC test or assess economic studies provided by the dominant company relating to whether the clauses had restrictive effects.

In response to these questions, the CJEU held that exclusivity clauses must be subject to an effects-based analysis and may be objectively justified. The CJEU confirmed that Intel applies not only to cases involving rebates, but also to exclusivity provisions. As such, where a dominant company submits evidence during an investigation supporting that its conduct was not capable of restricting competition and of producing the alleged exclusionary effects, the competition agency must analyse:

• the extent of the company’s dominant position;
• the share of the market covered by the exclusivity; and
• the potential existence of a strategy designed to keep any competitors that are as efficient from entering the market.

Exclusivity payments are not automatically unlawful

In Qualcomm, the GC held that exclusivity payments do not automatically infringe rticle 102 TFEU. It held that the EC should have analysed the anticompetitive effects of the exclusivity payments and demonstrated that these payments are capable of affecting competition and foreclosing an AEC.

On 15 June 2022, the GC handed down its judgment in Qualcomm’s appeal, dismissing the EC decision of 24 January 2018, fining Qualcomm just below €1 billion for abusing its dominant position. By way of background on the case, Qualcomm manufactures chipsets that enable smartphones and tablets to connect to cellular networks and are used for both voice services and data transmission. Chipsets can be integrated or stand-alone and can be compatible with one or more cellular communication standards, such as the global system for mobile communications (GSM), the universal mobile telecommunications system (UMTS) and the long-term evolution (LTE) standards. Qualcomm sells its chipsets to original equipment manufacturers (OEMs), such as Apple, which incorporate them into their devices. This case concerned LTE, UMTS and GSM standard chipsets sold by Qualcomm to Apple, and the agreement between Qualcomm and Apple relating to the delivery of chipsets, which provided for incentive payments to be made by Qualcomm to Apple on the condition that Apple sourced all its LTE chips from Qualcomm.

In its appeal against the EC’s decision, Qualcomm submitted that the EC had not provided evidence that the agreements concerned were capable of leading to the anticompetitive foreclosure of competitors and that the EC had failed to prove that Intel or any other rival undertaking was capable of satisfying Apple’s exacting technical and scheduling requirements. Moreover, Qualcomm submitted that the contested decision had acknowledged that no competitor was capable of supplying LTE chipsets for use in iPhones through the relevant period and had not included any allegation that the payments concerned had any impact on Apple’s supply decisions for iPads launched in 2011, 2013 and 2016. Instead, the contested decision sought to demonstrate the foreclosure of a single competitor (Intel) regarding supply for iPads to be launched in 2014 and 2015, which represented less than 1 per cent of the relevant market and suggests that the payments concerned could not foreclose any competitor.

The GC found that merely reducing Apple’s incentives to switch to Qualcomm’s competitors (and characterising the payments as exclusivity payments) was insufficient. Instead, the EC should have analysed ‘whether the exclusivity payments were capable of having an anticompetitive effect and foreclosed at least as-efficient competitors’.

How the 2025 Guidelines may seek to reflect this effects-based approach

The Amending Communication acknowledges that the CJEU has confirmed an ‘effects-based approach to Article 102 TFEU as well as clarified the meaning and scope of certain concepts included in the Guidance on enforcement priorities’.

In its Amending Communication, the EC sets out that there is no longer a need to demonstrate that actual or potential competition has been excluded or marginalised, but instead, conduct can be unlawful if it ‘adversely impacts an effective competitive structure’ that allows ‘the dominant undertaking to negatively influence, to its own advantage and to the detriment of consumers, the various parameters of competition’. In this respect, the EC acknowledges that conduct is only anticompetitive if it benefits the dominant undertaking and adversely impacts consumers (although the EC contends that this conduct no longer needs to be demonstrated as profitable for the dominant undertaking).

This is already a helpful clarification, but the EC may consider reflecting and implementing the CJEU’s recent jurisprudence, which establishes that where a dominant undertaking submits that its fidelity rebates, exclusivity restrictions or exclusivity payments have no anticompetitive effect, the EC shall carry out an effects-based analysis to reach a finding of abuse. In these circumstances, this conduct will only constitute an abuse of article 102 TFEU if it is capable of causing consumer detriment and relays an advantage on the dominant undertaking. Dominant companies and their trading partners will be looking to the 2025 Guidelines for practical examples of the types of conduct resulting in both consumer detriment and advantage. In addition, the EC may want to consider reflecting the CJEU’s test as per Intel and Unilever in its 2025 Guidelines and provide guidance on how dominant undertakings may rebut allegations of anticompetitive effects.

AEC test in abuse of dominance cases

In essence, an AEC test requires an analysis of the price level charged by the dominant undertaking and a comparison with its costs, with the aim of assessing whether a hypothetical as-efficient competitor could profitably compete against a pricing practice by the dominant undertaking.

Based on the EC’s recent Amending Communication and Policy Brief, the EC is likely to choose to use the AEC test rather more selectively in the future. However, as we explore in this section, the EC may not escape the application of the AEC test so easily.

When the EC applies the AEC test, it must do so rigorously

In Google Android, the GC held that the EC had erred in its application of the AEC test (for the reasons set out below) and, consequently, about the alleged exclusionary effect of the portfolio-based revenue share agreements (RSA) on a hypothetically as-efficient competitor. The GC, therefore, annulled the EC’s decision insofar as the EC considered the portfolio-based RSAs in themselves to constitute an abuse.

By way of background of the case, the EC found that by 2018, approximately 80 per cent of smart mobile devices used in Europe and worldwide were running on the Android operating system. Google’s ownership of Android apps and services meant that original equipment manufacturers (OEMs) wishing to obtain Google apps and services, and mobile network operators (MNOs) who wanted to install Google’s proprietary apps and services on devices sold to end users, had to enter into agreements with Google. The EC found that Google had abused its dominant position by means of the following separate, but closely interrelated, anticompetitive practices:

• mobile application distribution agreements, under which Google required OEMs to pre-install its general search app (Google Search) and browser app (Chrome) for them to be able to obtain a licence to use its app store (Play Store);
• anti-fragmentation agreements, under which OEMs that wished to pre-install Google apps could not sell devices running versions of Android that were not approved by Google; and
• portfolio-based RSAs, under which Google granted OEMs and MNOs a percentage of its advertising revenue, provided that those manufacturers or operations had agreed not to pre-install a competing general search service on any device within an agreed portfolio.

Google appealed the EC decision in its entirety. As discussed above, in relation to the portfolio-based RSAs, the GC held that the EC incorrectly found that the sole pre-installation condition included in the portfolio-based RSAs was abusive, as the EC erred in its application of the AEC test.

The GC held that where the AEC test is applied, it must be conducted rigorously. To determine whether a competitor, which hypothetically is at least as efficient, would likely be foreclosed by the contested practice, the EC must examine economic data relating to cost and sales prices, and determine whether the dominant undertaking is engaging in below-cost pricing:

• if reliable data is available, the EC is required to use information on the costs of the dominant undertaking itself; and
• if reliable information on those costs is not available, the EC may decide to use the cost data of competitors or other comparable reliable data.

The GC held that in relation to the general search query share that might be contested by the AEC, the AEC test contained several errors of reasoning. The EC had failed to isolate costs that could be attributed to an AEC and just extrapolated from data contained in a third-party document. Further, the EC’s assessment of the propensity of mobile devices already in circulation to generate lower revenues than those of newer mobile devices was incomplete.

When the dominant undertaking submits an AEC test, the EC must assess it

Already in Intel, the CJEU held that when a dominant undertaking puts forward economic evidence, including an AEC test, the EC must assess it. The CJEU again confirmed this defence strategy in SEN, where it held (in response to a request for a preliminary ruling under article 267 TFEU from Italy’s Council of State) that competition authorities must examine the evidence adduced by the dominant undertaking to prove that its conduct did not produce restrictive effects. And most recently in Unilever, the CJEU held that a competition authority must assess the probative value of economic analysis, such as the results of an AEC test, submitted by a dominant undertaking suspected of exclusivity clauses abuse during the investigation.

The EC’s approach in the Amending Communication

In relation to price-based exclusionary conduct, the EC amended its 2009 EC Guidance to reflect that the AEC test is optional and only one of several methods of assessing price-based exclusionary conduct. The EC states that the outcome of the AEC test would not conclusively elevate the allegation against the dominant undertaking, should the EC find that an AEC could compete effectively with the pricing conduct of the dominant undertaking:

When analysing data to assess whether an equally efficient competitor can compete effectively with the pricing conduct of the dominant undertaking, the Commission will integrate this analysis in the general assessment of anti-competitive foreclosure (see Section B above), taking into account other relevant quantitative and/or qualitative evidence.

To date, the EC’s direction seems to suggest that it is sidelining the AEC test in all abuse of dominance cases, except those relating to exclusionary pricing conduct where the CJEU has held that the EC was required to carry out an AEC test, for example in predatory pricing and margin squeeze cases. However, in Google Shopping, the CJEU only referred to predatory pricing and margin squeeze cases as an example, not limiting the test to these practices. In SEN, the CJEU has already found that the AEC test can also be relevant in assessing non-pricing exclusionary practices. While the complexity and difficulty in carrying out the AEC test is apparent from recent cases, stakeholders would benefit if the 2025 Guidelines clearly set out the circumstances in which the EC would apply the AEC test (which should not be limited to margin squeeze and predatory pricing cases), what data it would consider relevant and what the dominant undertaking needed to demonstrate. Separately, if neither the dominant undertaking nor the EC have the data required to perform the AEC test readily available, the EC should use its investigatory power to obtain the requisite data (Google Android). The EC has been warned by courts not to rely on mere conjecture or hypotheses.

Further, the EC should reflect in its 2025 Guidelines that it will rigorously review the results of an AEC test in any event, if relied upon by the parties (Intel, Google Android and Unilever).

Trending developments to be reflected in the 2025 Guidelines

Types of abuse where the EC does not need to demonstrate indispensability

As set out in last year’s edition of this publication, the GC held in Google Shopping that Google’s conduct should not be analysed based on the strict indispensability condition set out in Bronner, even if the conduct ultimately resulted in an implicit refusal to supply.

The EC Guidance generally treated constructive refusal to supply in the same manner as an outright refusal to supply and applied the same criteria to both types of abuses. The EC’s Amending Communication now sets out that the EC may investigate cases where a dominant undertaking imposes unfair access conditions to a particular input, even if the input is not indispensable. While this should not mean that dominant undertakings are generally required to deal in non-essential facility cases with pre-existing supply, dominant undertakings and their trading partners would nevertheless benefit from the EC clarifying and confirming the circumstances when a refusal in relation to a pre-existing supply relationship might be abusive, including confirmation that this refusal must result in competitive harm (eg, anticompetitive discrimination or anticompetitive favouring) to be abusive. As expressed, the Amending Communication suggests that there is a lower bar for what used to be referred to as constructive refusal to supply cases, as compared to outright refusal to supply cases. This is not intuitive and not supported by IMS, Magill and Microsoft.

In addition to constructive refusal to supply cases, there are other abuse cases where the indispensability of a supply input need not be established. For example, in Lithuanian Railways, the CJEU held that where a dominant undertaking itself loses access to the benefit of its asset, this conduct did not equal a refusal to access (where the dominant undertaking reserves the asset to itself, deriving an immediate benefit), but an independent form of abuse. However, in that case, the CJEU also found that the case law and indispensability requirement arising from Bronner could not apply where the applicable regulatory framework already imposes a duty to supply on the dominant company, or where the dominant position derived from a statutory monopoly.

It would be useful for the EC to reflect these different legal tests, including the different treatment of former statutory monopoly, in its 2025 Guidelines.

New wave of disparagement cases

As foreshadowed in last year’s edition of this publication, anticompetitive disparagement continues to emerge as a new form of abuse. In addition to national investigation, the EC has at least two ongoing disparaging conduct cases.

• In October 2022, the EC preliminary found that Teva had abused its dominant position by implementing a systematic disparagement campaign targeting healthcare professionals and casting doubts about the safety and efficacy of a competing glatiramer acetate medicine and its therapeutic equivalence with Copaxone.
• In June 2022, the EC opened an investigation into Vifor Pharma, suspecting that Vifor disparaged its competitor Pharmacosmos by pursuing a misleading communication campaign about its safety, thereby delaying its market entry and market uptake.

While the facts of these two cases are yet to be made public, establishing disparagement as a novel abuse of dominance is likely be challenging for the EC. A particular warning came in February this year from the Paris Court of Appeal, which overruled a decision of the French Competition Authority, fining Novartis and Roche €444 million for implementing a disparaging communications campaign which amounted to an abuse of dominance. While the French authority found that healthcare professionals were misled in relation to the use of Avastin, a cancer drug, in the off-label treatment of age-related macular degeneration, the French Court of Appeal decided that a communications campaign may not amount to illegal disparagement if it concerns a subject of general interest, has a sufficient factual basis and is brought in a measured tone.

Indeed, these novel abuse cases may provide scope to demonstrate how the 2025 Guidelines will reflect the EC’s evolving approach to enforcement of exclusionary conduct. This can be seen in the description of anticompetitive foreclosure, which in the EC Guidance required that ‘effective access of actual or potential competitors to supplies or markets is hampered or eliminated’. By moving towards a description, which in the words of the Policy Brief: ‘takes into account the different types of exclusionary conduct that a dominant undertaking can implement’, there appears a greater likelihood for the EC to establish that a disparagement campaign has ‘adversely impacted an effective competitive structure’, which has allowed the dominant undertaking to ‘negatively influence’, to its own advantage, ‘the various parameters of competition’.

When a transaction can amount to an abuse of dominance

In Towercast, the French Competition Authority sought guidance from the CJEU as to whether it could investigate Telediffusion de France (TDF), who acquired competitor Itas, under the French equivalent of article 102 TFEU, although TDF’s acquisition of Itas was not notifiable in France and not reviewed by the EC.

In response, the CJEU held that the prior control of operations with a Community dimension introduced by the Merger Regulation does not preclude a subsequent control of concentrations that do not meet that threshold. When carrying out this subsequent control, the authority must verify whether a purchaser – who is in a dominant position on a given market and who has acquired control of another undertaking on that market – has, by that conduct, substantially impeded competition on that market.

The implications of this reference are that not only deals below the EC and member state thresholds (and not referred to the EC under article 22) can be subject to enforcement under article 102 or national equivalents. Against this background, the EC may consider reflecting in its 2025 Guidelines the circumstances and priorities under which a transaction (that has or has not been reviewed by an NCA or the EC) could be reviewed under article 102 TFEU, and how the EC or NCA would demonstrate that competition was harmed.

The CJEU’s judgment has already found practical application by the Belgian Competition Authority, who opened an ex officio investigation into a possible abuse of dominance by Proximus in the context of the takeover of edpnet.

Could the 2025 Guidelines address how the EC may better safeguard the rights of defence of parties under investigation?

The GC has recently had cause to scrutinise the EC’s procedures in article 102 TFEU investigations, including examining whether certain of the EC’s procedural errors have resulted in an infringement of the dominant undertaking’s right of defence. The observance of the rights of the defence is a fundamental principle of EU law and requires that an undertaking under investigation is given full opportunity to comment on the facts and evidence relied on by the EC.

In the Qualcomm judgment published in June 2022, the GC found that the EC’s procedural errors had resulted in an infringement of Qualcomm’s rights of defence and consequently annulled the EC’s decision on the following grounds.

• In breach of article 19 of Regulation 1/2003, pursuant to which the EC may conduct interviews for the purpose of collecting information relating to the subject matter of an investigation, the EC failed to fulfil its obligations to record meetings and conference calls held with Qualcomm’s competitors and consequently, to include these records within the EC’s case file. The GC found in favour of Qualcomm, which argued that it would have been better able to ensure its defence absent the EC’s procedural error. The GC also found that additional errors were made by the EC in failing to disclose whether meetings with certain third parties did or did not occur, as well as providing records that were too general in nature to provide an adequate basis for Qualcomm to exercise its rights of defence.
• The GC also found that differences between the statement of objections and the EC’s decision were grounds for finding of a breach of Qualcomm’s rights of defence in this case. Qualcomm argued that, in dropping its analysis in respect of alleged abuses in the UTMS chipsets market following the issuance of the statement of objections, the EC infringed Qualcomm’s right to be heard. The GC agreed on the basis that Qualcomm could have adduced different economic evidence in respect of its conduct in the LTE chipset market if it had been made aware of the EC’s change of course.

Procedural failures were also raised in Google Android. In that case, the GC examined alleged breaches in relation to record keeping by the EC as well as the EC’s refusal to provide an oral hearing following the issuance of letters of fact by the EC.

As regards the allegations in relation to record-keeping, the GC did not agree with Google’s claim that its right of defence was infringed by the EC not maintaining full and timely interview notes with third parties. Although the GC recommended that the EC record each interview it conducts with a third party (for the purpose of collecting information relating to the subject matter of an investigation) and to approve and add to its file as soon as possible, it found that Google had failed to establish that it would have been better able to ensure its defence absent the EC’s procedural failure.

However, in the same appeal, Google successfully argued that the EC infringed its rights of defence by denying it the opportunity to respond orally to the EC’s AEC test. The GC found that the EC’s letters of fact, which were sent instead of issuing a supplementary statement of objections, had materially supplemented the substance and scope of the EC’s objection and significantly altered the EC’s evidence, particularly in relation to the AEC test. The GC found that Google could have developed its submissions more easily orally and a hearing would have enabled Google to provide the EC with information that could have resolved some of the ambiguities. The EC had, thus, infringed Google’s rights of defence by refusing a hearing on the AEC test, resulting in a partial annulment of the EC’s decision against Google

Looking to its ongoing and future investigations, the EC will no doubt continue to improve its investigation procedures to avoid any part of its decisions being annulled based on procedural errors. Whether the 2025 Guidelines are the appropriate publication to take stock of procedural failures is questionable. However, in the same way that the updated Guidelines will seek to codify recent article 102 TFEU case law, the EC may also wish to update its procedural guidance to ensure that case specific learnings are taken on board. In doing so, the EC will ensure best practice is achieved not only within its own case teams, but also within NCAs that rely on the EC to shape their procedural frameworks.

Notes