The European and U.S. Approaches to Antitrust and Tech: Setting the Record Straight – A Reply to Gregory J. Werden and Luke M. Froeb’s Antitrust and Tech: Europe and the United States Differ, and It Matters1
By Maria Coppola & Renato Nazzini2
In the October 2019 CPI Chronicle, Greg Werden & Luke Froeb identify ten differences between the European and American approaches to single firm conduct enforcement. These differences, they claim, explain how Europe systemically differs from the United States in ways that affect enforcement against the tech giants. Werden & Froeb “lift the hoods on the antitrust enforcement machines in US and EC” in order to identify “hard-wired differences” in each jurisdiction’s approach to single firm conduct. While acknowledging that their article may be seen as overly simplistic, they identify ten meaningful differences as “essential truths.”
There is no doubt that competition agencies in Europe have brought many single firm conduct enforcement cases against the tech giants, and some of the cases would be unlikely to be brought under U.S. law.3 While Werden & Froeb make some key points with which we agree, in unpacking the “meaningful differences” to identify their asserted “essential truths,” Werden & Froeb’s selective viewfinder results in a mischaracterization of important areas of European law and court decisions. This reply addresses, in order, all ten differences identified by Werden & Froeb.4 This reply is not a line by line rebuttal, nor a refutation of every point that Werden & Froeb have identified. Instead, it offers a more fulsome discussion of the issues they raise, with the goal of providing a more accurate and complete view of European competition law and policy as it compares with U.S. law and enforcement practice.
Difference 1. The European system is driven by competitor complaints
Werden & Froeb posit that complaining to the European Commission pays off, and that the European Commission is driven by competitor complaints, whereas the U.S. antitrust enforcement agencies often leave it to the complainants to bring cases on their own. The differences between the U.S. and the European private enforcement systems, which remain significant even after the implementation of the Damages Directive in the European Union,5 and material cultural divergences in understanding the role of government in policing the economy, undeniably influence the allocation of resources between public and private enforcement.6 Some observers argue that the U.S.’s particular system of private rights that are afforded in litigation, including the availability of treble damages, fee shifting, contingency fees, liberal discovery, and class actions, has increased the bar for bringing monopolization cases precisely because the judiciary seeks to limit the availability of these powerful rewards by modifying doctrine governing liability standards to evaluate the worthiness of private claims.7
However, it is not clear what Werden & Froeb really mean by “driven” by competitors. If it means that enforcement action often starts with one or more complaints against an allegedly dominant firm, this is arguably a feature showing that the European system is working well. Particularly in an environment without an active culture of private actions, it is necessary for complainants to “police” the business environment and bring potential infringements to the attention of the competition authorities. This is why formal complainants have been given important rights in the European Union (“EU”), including access to evidence and a non-confidential version of the European Commission’s charge sheet (statement of objections), appearance at oral hearings, and, most notably, standing to challenge the European Commission’s final decision (or the European Commission’s rejection of the complaint). Incentives to become a formal complainant are influenced by these rights. This, however, works in the public interest. Complainants report alleged infringements that may be impossible or excessively difficult for the competition authorities to detect without any input from competitors or customers.
If, however, “driven” by competitors purports to suggest that the European Commission is “captured” by complainants or that the European Commission cannot run cases unless there are well resourced complainants, we believe this is incorrect. Suggesting complainants are the driving force for the cases against the tech giants ignores much of the European Commission’s experience with these cases. In the AdSense case, for example, Microsoft withdrew its complaint more than two years before the European Commission issued an infringement decision. Cases such as Qualcomm had no formal complainants, nor, to our knowledge, are there formal complainants in many of the current tech investigations. It is also worth noting that in some of the tech enforcement actions, the European Commission’s decision bore little resemblance to the original complaints.
Many of the European Commission’s tech cases are ex officio, driven by the Commission’s enforcement agenda; or they began, as many U.S. investigations do, with grievances by competitors, and then are developed by the enforcers who are informed by engagement with the target and third parties, including competitors.
Anecdotally, it appears that informal complaints involving conduct that affects the U.S. and the EU are frequently lodged with agencies on both sides of the Atlantic. However, complainants may choose to focus their efforts in Brussels because Article 102, the EU’s single firm conduct provision, makes a broader swath of conduct illegal.
Difference 2. The European system is run by politicians
Werden & Froeb claim that the European decision-making structure seems “calculated to elevate politics above technical merit.” The European Commission ultimately takes enforcement actions, acting upon the recommendation of its Competition Commissioner, who has sworn an oath of impartiality, and other Commissioners almost never interfere with the recommendation of the Competition Commissioner. The Competition Commissioner’s recommendation is based upon the work of and submissions by the Competition Directorate, DG Competition, as well as review by a cabinet partially staffed by very senior career competition officials from DG Competition. However, we do not deny that the Competition Commissioner, who could be compared to the head of an antitrust agency, may well be, effectively, a politician, which is the case of the current Commissioner, and that, formally, enforcement decisions are taken by a college of politically appointed Commissioners with different portfolios and different agendas. The position is different in the Member States, where most competition agencies are, and all will soon be required to be, independent of the government, though the heads of authority and commissioners are sometimes political appointees without any technical competition expertise. This position, which we agree may not be entirely satisfactory, is not, however, specific to single firm conduct or tech markets. It applies across the board, including to merger control. Furthermore, in practice, historically Competition Commissioners have shown repeatedly a willingness to stand up to parochial interests. Most recently, in 2019, the European Commission, at the recommendation of Commissioner Vestager, blocked the Siemens/Alstom merger against the wishes of the two national EU heavyweights, France and Germany.8
Looking at the history of the federal enforcement agencies in the U.S., most, but not all, agency decision-makers had strong antitrust backgrounds, but many also came from the political world (more frequently from the legislative branch but also the executive), and others without antitrust expertise were nonetheless excellent at their jobs.9
While there is a degree of “politicization” in EU institutional arrangements, on a closer analysis this is not unique to the EU, nor have Werden & Froeb provided evidence that this has had a systematic, or indeed any, impact on past cases.
Difference 3. The European system was conceived of as regulation, not as law enforcement
Werden & Froeb state that the European system was conceived of as regulation, not as law enforcement. It is unclear whether the debate as to whether competition enforcement is law enforcement or regulation sheds insight on understanding the differences between the U.S. and the EU approaches to tech. The debate appears to us to be academic, at least to a large extent, but we nevertheless believe that it would be inaccurate to describe the EU system of competition enforcement as “regulation.”10 Like U.S. legal prohibitions, the EU antitrust prohibitions are general in terms of their fundamental structure. Article 102 TFEU, for example, prohibits “any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it.” It is true that Article 102 goes on to set out four examples of abuse, but the EU case law unequivocally states that this list is illustrative, and what matters is whether conduct is, or is not, an abuse of a dominant position, not whether it fits one of the four categories of abuse.11 Section 2 of the Sherman Act prohibits any “monopolization, attempted monopolization, or conspiracy or combination to monopolize.” If one compares the prohibition of “any abuse of a dominant position” with the prohibition of any “monopolization, attempted monopolization, or conspiracy or combination to monopolize,” the prohibitions are similar in terms of their generality. Under EU law, the key concepts are “abuse” and “dominant position,” whereas in the U.S. the test is what “monopolize” means. 12 The way the case law has construed and applied the prohibition is also, on both sides of the Atlantic, broadly similar in method: the U.S. and the EU Courts have interpreted broad, open-textured prohibitions in light of the policy and economic thinking prevailing at the time, while being responsive to counsel’s arguments on the facts of each individual case. Of course, U.S. and EU case law differ in their approaches to given problems. We provide two examples to illustrate this point: predatory pricing and margin squeeze. In the U.S., for a predation claim to succeed, the plaintiff must show a dangerous probability of recoupment,13 whereas in the EU recoupment is a relevant factor in the assessment of a predatory strategy, but not an element of the infringement.14 Also, in the U.S., margin squeeze is not a standalone infringement,15 where in the EU it may be, under certain circumstances.16 What one could draw from these differences, however, is not any support for the claim that U.S. antitrust law was conceived of as “law enforcement” and EU competition law was conceived of as “regulation.” These differences show, rather, different perceptions and underlying beliefs about the relative importance of type 1 versus type 2 errors and, more generally, different doctrinal approaches by different courts in different jurisdictions.
It is also instructive to recall, from a historical perspective, that in one of the drafts of the Treaty of Rome, the section on competition opened with a general prohibition on discrimination on grounds of nationality. That prohibition was then removed from the draft on the request of the German delegation, with the agreement of the other delegations, because the view was taken that discrimination on grounds of nationality had nothing to do with competition law.17
It would also be inaccurate to say that EU competition law is “regulation” because it is enforced by the European Commission and national competition authorities (“NCAs”), which are, by and large, administrative authorities rather than courts. The European Commission and the NCAs are required to enforce the competition laws through individual enforcement mechanisms, not to regulate markets. Furthermore, courts have jurisdiction to apply the very same prohibitions regardless of any prior decision by an administrative authority. A firm aggrieved by an alleged abuse of dominance may complain to the European Commission or go straight to court. The European Commission or the court will apply the law, and the decision-maker should not affect the result. Also, when the European Commission or an NCA adopts an infringement decision, that decision is reviewable by a court of full jurisdiction, that is, by a court having the power to review all aspects of the decision, in fact and law.
Finally, to extrapolate – as Werden & Froeb do – that EU competition law is regulation because it prohibits exploitative abuses, including excessive prices, and excessive pricing cases turn the enforcer into a price regulator, is methodologically incorrect. Exploitative abuses are uncommon, as Werden & Froeb rightly recognize, and very much the exception rather than the rule.18 One cannot draw a general conclusion from an “exception.” Furthermore, while we recognise that the very idea of an exploitative abuse is controversial,19 as a matter of principle the test remains a legal test. It is not the case that the European Commission or NCAs can intervene in markets when they believe there is a market failure and regulate the price. What the agencies can do is to determine whether a given price falls afoul of the prohibition or not. This may be complex and certainly debatable as a matter of policy and economics – but does not necessarily turn EU competition law into price regulation as a general proposition, although it is theoretically possible to argue that the exploitative abuse branch of EU competition law has elements of regulation. It is noteworthy, however, that none of the enforcement cases against tech companies is an exploitative abuse case. Werden & Froeb’s argument on this point is, therefore, in any event moot.
Difference 4. The European system is grounded in skepticism of markets
Werden & Froeb argue that the European system is grounded in skepticism of markets. In fact, the European system is grounded in the creation of a single market, a concept firmly grounded in the belief that free trade and an open and competitive economy maximize overall welfare, against the background of a social and economic history that reflects a wariness of conduct by former state monopolies within their traditional areas of dominance. In the EU, from the outset, state monopolies, state-granted benefits and privileges, and state power coopted by private interests posed serious challenges to free and competitive markets. This history of state-run or directed economies impacts the European Commission’s enforcement priorities: more than half of the European Commission’s single firm conduct cases of the past two decades involve telecom, financial services, rail, and electricity, reflecting this economic past. This difference identified by Werden & Froeb is generalized, explaining that the U.S. protects competition, leaving unsaid the decades-old refrain that inevitably follows this statement: that Europe protects competitors at the expense of consumers. Scholars have noted that EU competition law protects the competitive process, stressing market access and the right to contest markets on the merits.20 Guided by Post-Chicago school economics,21 as Professor Eleanor Fox notes, EU competition enforcement reflects the belief that lowering barriers to entry and keeping a clear path for challengers is likely to make the market more dynamic and thus serve consumers better.22 We believe that enforcement in Europe is not guided by a skepticism of markets, but by a marginally different set of beliefs about what rules and forms of intervention make markets work better and tend to optimize overall welfare.
It is also noteworthy that while U.S. law exempts certain forms of state action from the reach of antitrust law, by contrast, the activities of public entities in the EU are fully subject to competition law. EU competition law, through Articles 106 and 107, does not exempt state-owned enterprises or state granted privileges from the application of competition law. In the Greek Lignite case, for example, the European Court of Justice (“ECJ”) repeated its well established case law, according to which a Member State may be found to have infringed Article 106(1) if its measures create a situation in which a public undertaking or an undertaking on which it has conferred special or exclusive rights, merely by exercising the preferential rights conferred upon it, is led to abuse its dominant position, or when those rights are liable to create a situation in which that undertaking is led to commit such abuses.23 For state action to be compatible with EU law, the exclusive or special rights must be proportionate and not go beyond what is necessary to achieve the public interest objective.
In general, in the area of single firm conduct, the European Commission and Member States have been more inclined to intervene against single firm conduct, whereas in the U.S., single firm conduct enforcement has been more influenced by concerns about the effects of erroneous government enforcement. This does not mean, however, that Europeans are driven by a skepticism of markets or that, in the U.S., government enforcement against single firm conduct is absent. 24
Difference 5. The European system lacks the process of U.S. court proceedings
The institutional structure and enforcement process in the EU and the U.S. differ significantly, but both systems allow for a full review by independent courts. In the EU, the European Commission investigates a case, issues a statement of objections, and gives the targets the right to access to evidence the European Commission has relied upon (access to file), the opportunity to submit a written defence, and to request a hearing. The hearing before the European Commission is, of course, nothing like a trial before an impartial court. There is no cross-examination of fact and expert witnesses and no adversarial arguments. The hearing is before the same body that started the investigation, issued the statement of objections, and will ultimately decide the case. However, under EU law, this system requires that any aggrieved party must have a right of appeal before a court of “full jurisdiction,” with the power to review the correctness, not just the lawfulness, of the decision on the facts and on the law. This is not the same as a common law trial, but it is a de novo review.25
At the Federal Trade Commission (“FTC”), when the Commission finds “reason to believe” that a party has engaged in an “unfair method of competition” and that a proceeding would be in “the interest of the public,” staff can litigate cases either through administrative adjudication before an Administrative Law Judge (“ALJ”) and the Commission (“Part III”),26 or through federal court litigation before district courts. In the case of Part III adjudication, the first appeal, available to either complaint counsel (FTC staff) or respondents or both, is an appeal from the ALJ’s initial decision to the Commission. The Commission reviews the ALJ’s decision de novo, meaning that the Commission is not bound by the ALJ’s evidentiary rulings, findings of fact or conclusions of law. The next appeal is only available to the respondent, who can appeal the Commission’s decision to the federal courts. The Commission’s findings of law are reviewed de novo, but they may be given some deference to the extent they involve interpretation and application of the FTC Act. The Court of Appeals is bound by the Commission’s findings of facts, so long as the facts are supported by evidence. The courts have interpreted this language to mean that a “substantial evidence” test applies. This test asks whether the Commission’s findings are supported by “such relevant evidence as a reasonable mind might accept as adequate,” not whether the reviewing court “making its own appraisal of the testimony, picking and choosing for itself among uncertain and conflicting inferences” would reach the same conclusion. The test is deferential to the FTC.
At the Antitrust Division, the Division acts only as a prosecutor, going to federal court to seek enforcement action. Unless the parties choose to settle the case before trial, enforcement actions will require trial in a federal court. Under the Tunney Act, federal courts review proposed consent decrees in civil antitrust cases filed by the DOJ to ensure that the remedy proposed in the consent is in the public interest, but this does not usually involve a full trial on the merits.
Thus, the U.S. and the EU procedures differ, although it is notable that these differences are not specific to antitrust. They reflect the different legal systems of continental Europe and the United States. They share in common, however, that competition decisions are subject to a full review by independent courts.
One aspect not addressed by Werden & Froeb in this section, but referred to only in passing near the conclusion, is the length of the European Court proceedings. The authors reference the European Commission’s Intel case, with the original decision in 2009, and, as of this writing, is still waiting for a final decision by the General Court after being remanded in 2017 by the ECJ. We note that it is possible for U.S. cases to be lengthy. In the FTC’s Actavis matter, the FTC filed its original complaint in January 2009, which was dismissed by a district court, a decision then reversed by the Supreme Court in 2013, and remanded. The case concluded with a consent decree in February 2019, on the eve of trial, which was scheduled to begin in March. However, while we have not done an empirical examination to compare the timing between the original enforcement action and the final appeal in all cases, we agree that it appears in general to be considerably longer in the EU, which can impact a defendant’s decision on whether or not to pursue an appeal, and thus may impact how defendants view, as a practical matter, their rights of defense.
Difference 6. The European system lacks the burden of proof of an adversarial system
As Werden & Froeb accept, the EU system does have a doctrine of the burden of proof,27 even if, as they correctly state, in the administrative proceedings it is the European Commission itself that needs to conclude whether its own burden of proof has been discharged or not. From the perspective of a common law trial, this appears puzzling. In an administrative system, however, it is not unusual for the administrative authority to decide whether, following its own investigation, there is sufficient evidence that allows a determination whether the law has been infringed. The enforcer still must be satisfied to the required legal standard that an infringement has occurred, mindful that the evidence must be persuasive to the courts since the decisions can be appealed to the General Court, and then, on points of law, the ECJ.28 If a decision is appealed, while it is for the applicant to raise specific grounds of appeal, it is for the European Commission to satisfy the court that it has discharged its burden of proof. If the court, at the end of the appeal, believes that there is a doubt as to whether the applicant has infringed EU competition law, the appeal is allowed and the European Commission’s decision annulled.29 One may, of course, prefer a common law trial as the best method to test the evidence and arrive at the truth – as we do – but it is necessary to recognize that more than half the world, including, but not limited to, the EU and most of its Member States, has administrative authorities that are empowered to find law infringements, not only of antitrust law but of many other sorts of laws, subject to review by the courts. Introducing a common law-style trial at the EU level only to address competition infringements would be out of kilter with the legal system overall. As the U.S. adversarial system is not specific to antitrust, neither is the EU administrative system. Both are well embedded in their respective legal frameworks and traditions and both have advantages and disadvantages. The EU system may have disadvantages, but one of them is not that it lacks the burden of proof of an adversarial system. It lacks a common law trial, but this is not the same thing.
Difference 7. The European system does not impeach unsound theories
In Werden & Froeb’s article, this criticism of the European system boils down to the proposition that in the EU there is no mechanism to exclude unreliable economic evidence as there is in the United States under the Daubert rule.30 It may be that all the world should adopt the Daubert rule, and we are open to a debate about its pros and cons. But the jump from being enthusiasts of the Daubert rule to saying that the European system does not impeach unsound theories is itself unsound. Many common law systems do not have a rule similar or equivalent to Daubert. For example, England, the country of origin of the common law, does not have it. Under English law, the reliability of the evidence, including expert evidence, is a matter for the jury or the judge if the trial is by a judge alone. Whether evidence is reliable or not is an issue of fact, not an exclusionary rule. It is not surprising, therefore, that an administrative system such as the European one does not have a Daubert-type of rule. But unsound theories can be and are impeached.
There are several stages in the administrative procedure that provide an opportunity for parties to challenge the European Commission’s economic theories. There are internal checks and balances, such as advice by the Legal Service of the European Commission (completely separate from DG Competition), or by the Chief Economist’s office (which provides an independent view). But, more importantly, whatever economic theory the European Commission relies upon, it must be able to withstand judicial review. While judicial review of matters of complex economic assessment has, historically, been deferential, it is now accepted that the General Court has unfettered jurisdiction to review economic evidence and assessments and set aside the decision under appeal if it disagrees with the assessment of the European Commission. This has happened, recently, in the Servier pay-for-delay case, in which the General Court disagreed with the European Commission’s market definition and set aside the decision in part.31 And it happened, famously, in three mergers cases in the early noughts, when the EU Courts set aside three merger decisions in short sequence in Airtours, Schneider, and Tetra Laval.32
There is also the important role of the Advocate General (“AG”) in criticizing what might be considered to be “unsound” ideas. In many cases, for example in the AG opinions in Coty and Intel, the AG provided trenchant analyses of the state of the literature in order to attack specific ideas, and to guide the ECJ toward a better line of reasoning. While the Court is ultimately free to disregard the opinion, the AG is nonetheless an influential voice about arguments that should be retained and those that should be discarded.
Finally, in some Member States, economic experts retained by the parties have a special duty to the court to present their views in a way that improves the court’s understanding of the matter and not to engage in simple advocacy. Put another way, the expert has a responsibility to the court and not simply to the client. European judges and economic consultants explain that this tempers advocacy, and discourages the presentation of arguments and interpretations that might in some sense be considered “unsound.”33
While cross examination in court may be a more effective mechanism for impeaching unsound theories (as Werden & Froeb argue), it is not the only way to ensure that enforcement action is economically sound. There are a number of ways in which agencies and courts in Europe can weed out unsound theories.
Difference 8. The European system maintains a low bar for anticompetitive effects
We accept that the effects threshold required to prove an infringement is a difference between the U.S. and the EU systems.34 The EU generally requires a lower bar for foreclosure, but there is a diversity among the types of conduct that Werden & Froeb ignore. For example, in Generics (UK) Ltd v CMA, the ECJ found that a branded pharmaceutical company’s overall contractual strategy could be capable of restricting competition and of producing exclusionary effects going beyond the specific anticompetitive effects of the individual agreements. The Court in essence said that the cumulative effect may be greater than the individual parts: the Court accepted that an abuse of a dominant position could occur even if the dominant firm’s agreements did not, individually, breach Article 101.35 In the Intel case, the 2017 ECJ decision clarified the role of effects in an analysis of rebate schemes, holding that the dominant firm can overcome the presumption of illegality with substantiated arguments and evidence that the conduct was not capable of restricting competition.36
The effects required depend on the type of conduct at issue. In Google Shopping, the European Commission applied a test of conduct “capable of having, or likely to have, anticompetitive effects,” finding liability based on a demonstration of anticompetitive effects focusing on traffic diversion away from third-party comparison shopping websites, and discussed in less detail the potential anticompetitive effects in the national markets for general search services.37 In Google Android, the European Commission’s decision explains that it is sufficient that the conduct tends to restrict competition or is capable of having that effect, regardless of its success.38 The European Commission’s recent approach in tech cases is, however, not the final word on the matter. There is case law from the Grand Chamber of the ECJ that establishes that, at least in certain circumstances, the anti-competitive effects test should be likely foreclosure of as efficient competitors causing consumer harm. In Post Danmark I, the Court ruled on the standard of foreclosure in predation and held that when the price applied by the dominant undertaking is lower than its average total costs but above its average incremental costs, “it is necessary to consider whether that pricing policy, without objective justification, produces an actual or likely exclusionary effect, to the detriment of competition and, thereby, of consumers’ interests.”39 The European Commission presumably interprets this case as applying only to pricing practices. This is not, however, the only possible interpretation, and it remains to be seen what the EU Courts will say in the Google appeals, for example.
The test for anticompetitive effects under Article 102 is still unclear, and while arguably problematic for some types of abuses, the case law is moving in the direction of an effects-based approach. We agree with Werden & Froeb that the standard in the EU, at least insofar as it is only “capability” to foreclose in certain cases, is low. But we disagree in that the EU case law is much more complex and other tests are applied as well, such as a test of actual or likely foreclosure of “as efficient competitors,” to the detriment of consumers.
Difference 9. The European system is receptive to leveraging theories
Werden & Froeb discuss leveraging in the European tech cases, arguing that leveraging theories in the U.S. are legally cognizable only when monopoly is threatened in the second market, whereas in Europe it is sufficient that competition is distorted in the second market. Most European Commission tech cases involving leveraging theories involve refusals to deal or tying. For the former, the standards for finding liability are demanding.40 For tying cases, the description of the EU standard as only requiring a distortion of competition without any threat of monopolization of the tied market creates the wrong impression that the European Commission intervenes on the basis of a low threshold. In tying cases, the rationale for intervention is concern about potential monopolization in the tied market. While under case law the European Commission needs only to show foreclosure in the tied market, if one examines the cases actually brought, they are consistent with a likelihood of monopolization of the tied market. While a Section 2 tying case in the U.S. may be more demanding with respect to requirements for the secondary market, in the Microsoft cases (2004, tying of media player; 2009, tying of Internet Explorer) and the 2018 Android case, the European Commission considered market power in the tying market (i.e. where the conduct takes place) and the effects in the tied market (i.e. where the effects occur and the likelihood of foreclosure needs to be established).
Difference 10. The European system does not recognize competition on the merits
The European Courts have not defined in a conclusive way what competition on the merits means, and instead use it as a benchmark to define what is not competition on the merits. While there is no hard-edged definition of competition on the merits, the ECJ’s 2017 Intel decision explains that competition by dominant firms has to be efficiency related, which means, for example, lower costs or better quality. The Court relies on the “as efficient competitor” test, concluding that a dominant company can engage in conduct that is more efficient than its competitors and EU competition law does not protect less efficient competitors.
In describing the U.S. approach, the authors claim that in the U.S., courts hold that product improvement is “lawful competition on the merits.” For this claim authors likely rely on cases such as the Ninth Circuit’s Allied Orthopedic v. Tyco. However, while, in the United States, the Microsoft Court noted that U.S. courts “are properly very skeptical about claims that competition has been harmed by a dominant firm’s product design changes,”41 the court also refuted the idea that product design is a safe harbor from antitrust rules.42
In Europe, product improvement arises in Article 102 cases as an objective justification. While this justification has failed generally before the EU Courts, it has proved successful in Article 102 cases in Member States, for example, in Streetmap.com v. Google.43
In addition to these ten differences, Werden & Froeb also note other aspects that could influence Europe’s ability to bring tech cases: the transparency of decision-making when the agencies decide not to take action and the possibility of a challenge to that decision;44 the length of time for court decisions; and the inability for the court’s decisions on remedies to reflect new facts. These elements certainly can influence the cases against the tech giants. They conclude their article, however, noting that in the U.S., the judicial acceptance of a singular focus on consumer welfare guards against the calls for action against tech companies that invoke myriad social issues. We applaud the focus on consumer welfare in the U.S., but question the implication that the European Commission’s actions against the tech companies – the focus of the article – has succumbed to anything other than its two enforcement goals: consumer welfare and market integration.45 On the contrary, so far the European Commission has been a staunch advocate for keeping other goals separate – cautioning against the use of antitrust enforcement for other purposes, such as environmental protection, labor considerations, consumer protection, or data protection violations.46
This reply does not seek to suggest that Europe and the U.S. apply the same law or have the same procedures, particularly as regards to single firm conduct. At the margin, in the United States, there is a greater faith in market forces, and U.S. antitrust doctrine recognizes narrower circumstances in which dominant firms can be held liable for competitive abuses. At the same time, we acknowledge that the past two decades have seen greater transatlantic convergence, including the movement away from a form-based approach to a greater focus on effects in European law.47
1 Gregory J. Werden & Luke M. Froeb “Antitrust and Tech: Europe and the United States Differ, and It Matters” CPI Antitrust Chronicle, October 2019.
2 Maria Coppola is counsel in the U.S. Federal Trade Commission’s Office of International Affairs and an adjunct professor of European competition law at George Mason’s Antonin Scalia Law School. The views represented here are her own, and do not reflect those of the Federal Trade Commission or any individual Commissioner. Renato Nazzini is professor of law at King’s College London and a partner at LMS Legal LLP, London.
3 We note, however, that some of the U.S. monopolization cases, such as Intel and Qualcomm, have been more expansive than parallel enforcement actions taken by the European Commission: Agreement Containing Consent Order, Intel Corp., No. 9341 (August 4, 2010); Federal Trade Commission v. Qualcomm Incorporated, Case No. 17-CV-00220-LHK (N.D. Cal. May 21, 2019).
4 The italicized text describing the differences is from the original Werden & Froeb article.
5 Directive 2014/104/EU of the European Parliament and of the Council of 26 November 2014 on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union  OJ L349/1 (“Damages Directive”), now implemented in all Member States.
6 Standalone private enforcement actions under Article 102 have been rare. One reason is that national jurisdictions, except for the UK and to a lesser extent Germany and the Netherlands, did not provide for discovery. However, in 2014, the European Commission adopted the Damages Directive, which seeks to create an effective European private enforcement regime through the pursuit of damages actions before EU Member State courts. The Directive requires all EU member states to provide for the recovery of damages in competition cases, and several provisions in the Directive, such as those concerning the binding effect or evidentiary value of NCA’s decisions and other provisions related to discovery, may facilitate the bringing of single firm conduct private actions. The Directive required Member States to transpose the Directive into national law before the end of 2017 and has now been implemented in all Member States. As noted in the American Bar Association’s Antitrust Law Section Final Report of the Task Force on International Divergence of Dominance Standards (September 1, 2019) (hereafter “ABA Report”), “Given the dearth of Article 102 private enforcement cases litigated before national courts, it is too early to tell whether the European Union’s push for more private enforcement might in the future also impact the development of substantive competition law doctrine, similar to what some have observed in connection with Section 2.” The lack of a culture of private rights in most EU countries suggests this is unlikely to happen in the near future.
7 Kovacic has explained “U.S. antitrust doctrine would have assumed a more intervention-oriented character if the power to enforce the American competition statutes were vested exclusively in public enforcement authorities, or if the private right of action had been circumscribed…”: William E. Kovacic, Private Participation in the Enforcement of Public Competition Laws, speech at British Institution of International & Comparative Law’s Third Annual Conference on International and Comparative Competition Law: The Transatlantic Antitrust Dialogue, May 15, 2003. He has specifically addressed how private rights have shaped the standards for single firm conduct cases in the U.S. in William E. Kovacic, The Intellectual DNA of Modern U.S. Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix, 1 COLUM. BUS. L. REV. 1-80 (2007). In that article, Kovacic argues that judicial fears that the U.S. style of private rights of action – with mandatory treble damages, asymmetric shifting of costs, broad rights of discovery, class actions, and jury trials – excessively deter legitimate conduct and have spurred a dramatic retrenchment of antitrust liability standards, most evident in the progression toward more lenient treatment of dominant firm conduct.
8 Case M.8677 – Siemens/Alstom, February 6, 2019, OJ C 300, 5.9.2019, 14–20.
9 From the Federal Trade Commission (FTC), for example, Chairman Jon Leibowitz (2009-2013, also Commissioner from 2004-2009) joined the FTC after a long career in the U.S. Senate, and Chairman Janet Steiger (1989-1995) did not have an antitrust background when she joined, and she ably led the FTC.
10 Regulation may be defined as “imposition of rules by government, backed by the use of penalties that are intended specifically to modify the economic behaviour of individuals and firms in the private sector”: Glossary of Industrial Organisation Economics and Competition Law, compiled by R. S. Khemani and D. M. Shapiro, commissioned by the Directorate for Financial, Fiscal and Enterprise Affairs, OECD, 1993. At a very general level, all law may be “regulation” but, generally, regulation refers to sets of rules that are either sector specific (e.g. telecoms or energy markets) or aimed at addressing a particular, well-defined problem (e.g. money-laundering). Law is a more general concept, which encompasses and authorizes regulation, but includes all norms that purport to have a given form of authority over their addressees.
11 Case 6/72 Europemballage Corp and Continental Can Co Inc v. Commission  ECR 215; Case C-333/94 P Tetra Pak International SA v. Commission  ECJ I-5951; Case C-95/04 P British Airways plc v. Commission  ECR I-2331.
12 The test for monopolization, articulated in United States v. Grinnell Corp., consists of two elements: “(1) the possession of monopoly power in the relevant market and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.” United States v. Grinnell Corp., 384 U.S. 563, 570–71 (1966).
13 Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 222, 224 (1993).
14 Case T-83/91, Tetra Pak Int’l SA v. Comm’n, 1994 E.C.R. II-755, ¶¶ 143‑44, 150. While both the U.S. and the EU require some showing of prices below a measure of costs, and both consider similar factors, only the United States requires a specific showing of the possibility of recoupment but factors important to such a showing (e.g. high barriers to entry) are considered by both. EU case law arguably infers the possibility of recoupment from the fact of dominance. ABA Report at 59-60.
15 Pacific Bell Telephone Company v. linkLine, 555 U.S. at 453
16 Deutsche Telekom AG v. Commission, 2010 E.C.R. I-9555.
17 See R Nazzini, The Foundations of European Union Competition Law: The Objective and Principles of Article 102 (Oxford OUP 2010) 126 – 131.
18 The ABA Report (supra) notes: “Thus, while the exploitative dimension of Article 102 is broad, to date, prosecutorial discretion has tended to hold it at bay.” ABA Report at 26.
19 For a critical perspective, see R Nazzini, The Foundations of European Union Competition Law: The Objective and Principles of Article 102 (Oxford OUP 2010) 275 – 282.
20 See, for example, Prepared Statement of Eleanor M. Fox, U.S. Senate Comm. on the Judiciary, Subcomm. on Antitrust, Competition Policy, and Consumer Rights, Hearing on A Comparative Look at Competition Law Approaches to Monopoly and Abuse of Dominance in the U.S. and EU 11 (Dec. 19, 2018), https://www.judiciary.senate.gov/meetings/a-comparative-look-at-competition-law-approaches-to-monopoly-and-abuse-of-dominance-in-the-us-and-eu. Professor Fox also observed that “[b]oth the U.S. and the EU, by their antitrust laws, are trying to protect the market . . . . Both laws welcome the winds of competition and sustained low pricing.” Id. at 9. Professor Fox commented that the European Union “never adopted the Chicago School premises. It does not assume markets work well. It does not admonish: Trust the market—especially not when the market is concentrated and dominated by a single firm. It does not presume that antitrust intervention is likely to mess up the market and chill competition and innovation.” At 6. On the objectives of the EU prohibition of abuse of dominance, see also The Foundations of European Union Competition Law: The Objective and Principles of Article 102 (Oxford OUP 2010) 107 – 154.
21 See discussion of Chicago versus Post-Chicago school economics in William E. Kovacic, The Intellectual DNA of Modern U.S. Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix, 1 COLUM. BUS. L. REV. 1-80 (2007) at 22-25.
22 See, for example, Prepared Statement of Eleanor M. Fox, U.S. Senate Comm. on the Judiciary, Subcomm. on Antitrust, Competition Policy, and Consumer Rights, Hearing on A Comparative Look at Competition Law Approaches to Monopoly and Abuse of Dominance in the U.S. and EU (Dec. 19, 2018) at 6. “EU competition law stresses market access and the right to contest markets on the merits. It is sympathetic to access to networks. It is hostile to dominant firms’ using leverage to take advantages for themselves at the expense of competitors, thereby “unleveling the playing field.” It does not aim to protect inefficient competitors. Rather, EU precedents safeguard a clearer path of the outsider to access markets on the merits, free from obstructions by dominant firms. Nonetheless, from the point of view of detractors who worry about excessive enforcement against dominant firms, the EU approach does protect competitors.” Id at 8.
23 Case AT.38700 – Greek lignite and electricity markets, 17 April 2018, OJ C 245, 13.7.2018, 3.
24 For example, since 2001 the FTC has brought 27 Section 2 cases. This is about half the number the European Commission brought during the same time period.
25 R Nazzini, “Judicial Review after KME: An Even Stronger Case for the Reform that Will Never Be” (2015) 40 European Law Review 490 – 508; R Nazzini, “The evolution of the ‘full jurisdiction’ of the European Courts in Article 101 and 102 matters” in R. Cisotta & M. Marquis, Litigation and Arbitration in EU Competition Law (Cheltenham, Edward Elgar Publishing Ltd, 2015) 123 – 140.
26 Once the Commission issues a Part III complaint, the Commissioners are considered to be in an adjudicative role and they are no longer involved in the investigation and prosecution of the matter. FTC staff who conduct the investigation or the prosecution of the case are prohibited from participating in the Commission’s decision or review of the matter. In addition, ex parte contacts about the matter between FTC staff and respondents and the ALJ or the Commissioners are strictly prohibited.
27 On burden of proof in single firm conduct cases under European law, see R Nazzini, ‘The wood began to move: an essay on consumer welfare, evidence and burden of proof in Article 82 [now 102] cases’ (2006) 31 EL Rev 518-539.
28 See Prepared Statement of William E. Kovacic, U.S. Senate Comm. on the Judiciary, Subcomm. on Antitrust, Competition Policy, and Consumer Rights, Hearing on A Comparative Look at Competition Law Approaches to Monopoly and Abuse of Dominance in the U.S. and EU (“Kovacic Testimony”) (Dec. 19, 2018), “Critical Judicial Oversight” at 23-24.
29 The case law is well established. See, e.g. Joined Cases C‑403/04 P and C‑405/04 P Sumitomo Metal Industries and Nippon Steel v. Commission, EU:C:2007:52, para 52; Case T‑110/07 Siemens v. Commission, EU:T:2011:68, para 44 and Case T‑445/14 ABB v. Commission, EU:T:2018:449, para 39.
30 Under the Daubert rule, a trial judge, before admitting expert evidence, must satisfy herself that the evidence rests on a reliable basis and will assist the trier of fact to understand or determine a fact or issue: Daubert v Merrell Dow Pharmaceuticals Inc 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469. The test has been refined and clarified in subsequent cases, for example, General Electric Company v K Joiner 522 U.S. 136, 118 S.Ct. 512, 139 L.Ed.2d 508.
31 Case T-691/14 Servier SAS v. Commission ECLI:EU:T:2018:922, paras 1372 – 1592, under appeal to the ECJ in Case C-176/19 P Commission v. Servier and Case C-201/19 P Servier v. Commission.
32 Case No IV/M.1524 – Airtours/First Choice, September 22, 1999, OJ L 93, 13.4.2000, 1–33; Case No COMP/M.2283 – Schneider/Legrand, January 30, 2002, OJ L 101, 6.4.2004, 134–148; Case No COMP/M.2416 – Tetra Laval/Sidel, January 30, 2002, OJ L 43, 13.2.2004, 13–87. More recently, in the UPS/TNT merger, the ECJ reinforced the need for the Commission to provide its economic analysis to the parties in a detailed and timely fashion, finding that the Commission failed to provide UPS with the opportunity to comment upon a revised econometric model, which was the basis for prohibiting the proposed merger. European Commission v. United Parcel Service, Inc., 16.01.2019, ECLI:EU:C:2019:23.
34 See Kovacic Testimony, “[T]his inclines the U.S. agencies to demand, perhaps, a greater quantum and quality of evidence before deciding to prosecute” at 17.
35 Generics (UK) Ltd v CMA ECLI:EU:C:2020:52 144 – 172.
36 Aspects of the Intel case, including about the as efficient competitor test, continue to be litigated as of this writing.
37 In the decision there are tables suggesting that generic search results reflected a large proportion of traffic to competing comparison shopping websites. Case No. AT.39740 – Google Search (Shopping), Tables 24 and 25 at 163–67. The decision is under appeal in Case T- 612/17 – Google and Alphabet v Commission OJ C 369, 30.10.2017, 37-38.
38 “The Commission is not required to demonstrate in a general manner that ‘there would have been greater competition’ absent the tying of the Google Search app with the Play Store. Rather, the Commission is required to demonstrate that the tying is capable of restricting competition on the relevant markets.” Case COMP/AT.40099, Google Android, at 201. See also Google Android at 261. For a similar statement on the antifragmentation agreements, see Google Android at 244–45.The decision also noted: “Concerning the effects of the dominant undertaking’s conduct, while they must not be of a purely hypothetical nature, they do not necessarily have to be concrete. It is sufficient that the conduct tends to restrict competition or is capable of having that effect, regardless of its success. The Commission is not therefore required to demonstrate that a particular practice has actual anti-competitive effects.” Google Android at 164–65. The decision is under appeal in Case T-604/18 – Google and Alphabet v Commission OJ C 445, 10.12.2018, 21–22
39 Case C-209/10 – Post Danmark A/S v. Konkurrencerådet ECLI:EU:C:2012:172 (Post Danmark I), para 44 and operative part.
40 Case C-7/97, Oscar Bronner GmbH & Co. KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG, 1998 E.C.R. I-7791 and Case T-201/04 and Microsoft Corp. v. Comm’n, 2007 E.C.R. II-3619.
41 United States v. Microsoft Corp., 253 F.3d 34, 65 (D.C. Cir. 2001) (en banc).
42 “Judicial deference to product innovation does not mean that a monopolist’s product design decisions are per se lawful.” MS at 65, citing Foremost Pro Color, 703 F.2d at 545; see also Cal. Computer Prods., 613 F.2d at 739, 744; In re IBM Peripheral EDP Devices Antitrust Litig., 481 F.Supp. 965, 1007–08 (N.D.Cal.1979). Also, in New York v. Actavis PLC, the court found that regardless of whether the design change to a pharmaceutical product was an improvement, the withdrawal of the older version amounted to exclusionary conduct and the court found no non-pretextual competitive justification. New York v. Actavis PLC, 787 F.3d 638 (2d Cir. 2015)
43 Streetmap.EU Ltd. v. Google Inc.  EWHC 253 (Ch) (February 12, 2016).
44 Many scholars and members of the antitrust community have suggested the U.S. agencies should consider adopting additional transparency measures to explain why the enforcer declined to take action. See, for example, former FTC Chairman William Kovacic’s recent article on the FTC’s Google decision: “By too often saying nothing, or very little, the US agencies have missed important opportunities to improve public understanding of, and confidence in, how they formulate antitrust policy.” William Kovacic, Roads Not Taken: The Federal Trade Commission and Google, in The Concurrentialiste, March 9, 2020, available here. The U.S. agencies do occasionally issue statements about the decision to close an investigation or not to challenge a merger, but these are few in number and not as detailed as the European decisions.
45 Following an e-commerce sector study, the European Commission has brought a series of cases against conduct involving restrictions on online businesses, invoking the single market imperative, an essential element of its mandate. We do note that some of the recent tech enforcement actions by the Member States involve novel application of national antitrust laws (which can be more restrictive of single firm conduct than Article 102) that is arguably a move away from traditional antitrust, including the Bundeskartellamt’s case against Facebook, and the Autorité de la Concurrence’s recent cases against Google and Apple.
46 We note that in the Von der Leyen Commission there have been increasing calls to expand the boundaries of competition law enforcement, as there has been in the U.S. Whether and how Executive Vice President and Competition Commissioner Vestager and DG Competition respond to these calls remains to be seen.
47 As the ABA Report (supra) notes, “The Task Force finds far more similarity than difference between the current U.S. and EU approaches to competition enforcement against dominant firm behavior. To a large extent, convergence is reflected not only in agency and court enforcement actions but also frequent consensus among competition agencies not to act against conduct over which many have jurisdiction. It is now reasonable to treat recent U.S. and EU enforcement against dominant firm behavior as somewhat of a singular “modern mainstream” approach, although with some important divergent currents.” ABA Report at 18.