By Kayvan Hazemi-Jebelli (Computer & Communications Industry Association)1
You may not have noticed, but there’s a new antitrust revolution brewing in Europe. In February, the European Commission formally announced its intention to update competition rules to ensure that markets where large technology companies are present remain “fair and contestable.” I must confess that I assumed this meant accelerating the Commission’s ongoing revision to decades old guidance on assessing market power, or the safe-harbor rules for vertical and horizontal agreements, or perhaps that the Commission would facilitate strict enforcement of the recent, but still yet to apply, Platform-to-Business Regulation.2 I was wrong. In June the Commission revealed what it meant, with three major antitrust “pillars” towards fairness and contestability:
- Vigorous enforcement of existing competition laws;
- Ex ante regulation of digital platforms; and
- A new market investigation tool with the power to impose structural and behavioral remedies absent any finding of competition infringement.
One might call such far-reaching proposals ambitious reforms. But a critical look under the hood suggests a return to form-based antitrust presumptions; a counter-revolution to 50 years of jurisprudence focused on consumer welfare, economic effects, and causality.
Vigorous (Burden-Shifted) Enforcement
Vigorous, a-political, objective, and evidence-based antitrust enforcement is good, and should remain the standard that each jurisdiction expects from its global trade partners. It’s what we’ve worked decades to achieve and what we push for in international trade agreements. But the European Union has been criticized for sometimes putting the interests of its businesses ahead of consumers. At the ABA Spring Meeting this year, which had to be delivered virtually due to the COVID-19 crisis, many astute ears justifiably pricked up during the enforcer panel with European Commission Executive Vice President for “A Europe Fit for the Digital Age,” Margrethe Vestager. She stressed that investigators shouldn’t face excessive hurdles to proving their cases. Rather, they should be able to rely on presumptions that certain kinds of behavior were likely to harm competition. It’s not the first time the Commission has advocated for shifting the burden of proof. Presumptions and burden shifting can promote sound judicial and administrative efficiency, if supported by economic theory and a history of enforcement experience confirming the pernicious effects of the underlying conduct. But as some have pointed out, where the existing burden of proof is on the Commission to show only a capability of anticompetitive effect, is such burden shifting efficient?
More likely, it would be counter-productive, chilling competition by presuming consumer harm for a range of ostensibly procompetitive conduct. The Commission’s self-preferencing case against Google’s shopping comparison service is the most prominent example. In that case, the Commission condemned Google’s decision to improve product search results by drawing from its own quality-controlled product comparison database instead of from competitors’ unknown, inaccessible and unverifiable databases. For any product developer, Google’s decision would seem eminently more reliable, consistent and straightforward. Or, as the Commission itself once said “a company’s production of its own requirements is not in itself an abnormal act of competition.”3 When closing its related investigation, the FTC warned that “[p]roduct design is an important dimension of competition and condemning legitimate product improvements risks harming consumers.”4 The Commission nevertheless found Google’s conduct discriminatory, and imposed an eye-watering multi-billion euro fine. This has led to a series of follow-on damage claims, even though an appeal of the Commission decision is pending at the European Court of Justice,5 and despite courts in the UK and Germany having already rejected similar claims on similar facts. One might think it too early for a broad presumption of consumer harm under the “self-preferencing” banner, particularly where it encroaches on legitimate product design choices. One might consider that opening up such conduct to antitrust liability would chill procompetitive conduct and reduce consumer welfare. But this chilling effect is just one aspect of this drift towards “fairness and contestability.”
Ex Ante Regulation (of Bigness)?
The second pillar of the new antitrust revolution is a new ex ante market regulation for “very large online platforms acting as gatekeepers.” Unsurprisingly, the Inception Impact Assessment (“IIA”) launching the consultation states as one policy option an outright prohibition on self-preferencing, citing it as an example of conduct that is “potentially market-distorting or entrenching economic power of the large online platforms.” But confirming the Commission’s existing posture is not as revolutionary as what appears to be a new policy objective to limit bigness.
The IIA identifies particular obstacles facing its overarching policy objective to “increase the innovation potential and capacity across the online platform ecosystems in the EU’s single market.” In particular, it aims to tackle market dynamics that “reduce the social gain from innovation,” pointing to the variety of ecosystem services, platforms operators’ resources and scale, and their ability to enter adjacent markets as factors that are “keeping innovative market operators from expanding or entering the market in the first place.” The IIA notes that “Europe’s estimated 10 000 online platforms are potentially hampered in scaling broadly and thereby contributing to the EU’s technological sovereignty, as they are increasingly faced with incontestable online platform ecosystems.”
The IIA states that increasing the “contestability” of online platform ecosystems “would have a positive impact on innovation and research, technological development and growth of the digital economy.” It is based on an assumption that increasing “contestability” would mean better outcomes for consumers. But an overbroad interpretation of “contestability,” i.e. limitations on ecosystem growth, would more likely have a negative impact on the innovation, research, technological development, and digital economy growth that the Commission seeks to support. As discussed further below, this is because (a) large digital ecosystems drive aggregate demand upon which adjacent innovations can flourish; (b) their size contributes positively to the impact of innovation; and (c) their size is inherent to their innovation producing business models.
First, large digital ecosystems have driven the growth of the digital economy. Network effects happen because consumers and/or businesses prefer to be on platforms that other consumers or businesses are using. In the digital environment, where choices are abundant, consumers often prefer convenience, consistency, and a one-stop-shop experience. These efficiencies lower transaction costs, increase consumer welfare, and drive overall technological adoption. In the words of one well-known technologist, “it just works,” and it’s this intense focus on user experience and customer satisfaction that has dramatically increased adoption of new technology around the world. As the CMA’s recent market study acknowledged, “[i]Integration of a wide range of products and services can deliver efficiency savings and can also improve the consumer experience overall, by increasing the ease with which a range of different services are accessed.”6 A digital ecosystem that offers reliable solutions to a variety of user needs, that does a number of jobs conveniently and well, will often be more important to the marginal user considering a digital alternative than the “social gain” of having to navigate through a swamp of “innovative alternatives” as foreseen by the IIA. That’s why attempts to fragment these ecosystems are likely counterproductive, a result of overlooking the value that broad digital ecosystems provide to their users, and their contribution to the overall growth of the technology industry.
Second, larger platform ecosystems make it easier for innovators to reach massive scale, to the benefit of consumers. According to Apple, the App Store ecosystem supported $519 billion in billings and sales globally in 2019 alone. That’s a lot of opportunity for innovators. Network effects mean not only that users benefit from the size of a network, but also that innovations are more quickly disseminated throughout that ecosystem. New tools, features, and functionalities, new solutions to existing problems, spread more quickly through large ecosystems than fragmented ones.
Lastly, digital intermediaries often need size. Even if in specific cases increasing “contestability” were to increase investment incentives on the edges, it could also dramatically reduce the investment incentives of the platforms and their competitors. That’s because the business model of large digital intermediaries, particularly those based on software, is premised on significant up-front capital costs, very low marginal costs, and massive scale. Start-ups and SMEs in this sector often operate at a loss while attempting to achieve sufficient scale to become profitable. Sometimes that scale comes from expanding into adjacent markets. And allowing this has led to massive innovation in a very short period of time. According to the Commission’s own estimates, the five largest “gatekeeper platforms” (Amazon, Alphabet, Microsoft, Apple, Facebook), spent a combined €72.8 billion in R&D in 2018, nearly twice the amount spent by the entire EU ICT sector (approx. €42 billion).7 These companies are competing to create revolutionary new technologies and push the boundaries of innovation in computing devices, entertainment, mobility, productivity tools, the retail experience, and more. Efforts aimed at increasing ecosystem contestability that would result in reducing potential scale or scope (or increasing marginal costs) could make such ventures terminally unprofitable. This would ultimately result in less up-front investment, less technological development, and less growth for the digital economy.
The Commission’s apparent suspicion of bigness and market structure signals a departure from the existing antitrust framework. As Furman Inquiry Panel member, and antitrust rapper, Philip Marsden has said “the way forward is not by making non-evidence based decisions condemning market structures.” But it is this very suspicion of the efficiencies of scale and scope that lies at the heart of the Commission’s ex ante proposals. In this context, the Commission’s desire for “contestability” could risk much of what has gone right with the existing competition framework. One can legitimately wonder whether the proliferation of technological innovation that surrounds us would be as prevalent under a system that prioritized “social innovation” or “contestability.”
Ex Ante-Ante Intervention (or, the “New Competition Tool (‘NCT’)”)
The third pillar of the new antitrust revolution is a new market investigation tool to address “certain structural risks for competition” with a particular emphasis on markets with network, scale, and data effects (e.g. technology markets). However, the NCT consultation identifies several “structural competition problems” which do not, under the traditional framework, appear to be competition problems at all. In particular:
- “Tipping markets” are not problematic. Many investments are made in start-ups on the basis that they can be recovered once sufficient scale is achieved. The characteristic of “tipping” is itself related to the increased value that users derive from the network effects. Where such characteristics are present, preventing tipping would mean worse outcomes for consumers.
- “Leveraging” into adjacent markets is not inherently problematic. Where a firm has transferable skills and resources, leveraging those for market entry leads to increased efficient competition in those adjacent markets and hence better outcomes for users. The Commission offers no distinction between procompetitive and anticompetitive leveraging.
- “High entry barriers” are a characteristic common to several industries and are not inherently problematic unless a competitor with a superior business model or product cannot contest an incumbent, or when the threat of such entry is insufficient pressure to ensure efficient market outcomes.
By raising these three as “structural” concerns, the Commission again appears to presume anticompetitive effects from conduct or market circumstances that would normally be part of a case-by-case assessment and require some showing of causality. Under the traditional competition assessment, unilateral conduct by non-dominant undertakings is unlikely to be problematic by definition. This is because a company that has the ability to harm competition is likely dominant, and a company that is not dominant is unlikely to have the market power sufficient to unilaterally cause an anticompetitive effect on the market. Also, the concept of dominance is flexible, and increasingly applied to narrow market segments such that most firms can be found dominant where their anticompetitive conduct is liable to harm competition. The on-going revision to the pre-Google Market Definition Notice will no doubt help in this respect. So why go after non-dominant companies specifically?
The basis for this new intervention power not only presumes that certain unilateral conduct can cause harm (as in the ex ante tool), but it makes a further presumption that such conduct would cause harm even absent market power. This form of treatment has usually been reserved only to the most “hardcore” categories of restraints. So, not only does the Commission make an ex ante assessment of the conduct and presume harm, but it would make a further ex ante assessment of future impact, presuming a future effect. Such broad discretionary power to impose structural and behavioral remedies on speculative future oriented presumptions signals a revolutionary return to form-based antitrust enforcement over economic assessment and causality. That is of course, if legal certainty is to be expected. The alternative would be the Commission picking winners and losers via speculative predictions of future harm.
The Commission’s shift to a new “fair and contestable” paradigm is as yet undefined. The substance of the Commission’s proposals, and the nature of the presumptions incorporated therein, suggest a shift away from the existing international and multilateral framework focused on economic efficiency and an evidence-based assessment of causality, towards one more focused on market structure, on “fairness,” “choice,” and “contestability,” and the Commission’s predictions as to the future. This revolution certainly appears to be aligned with the push from certain Member States who advocate for a competition policy facilitating the creation of “European Champions.” After all, who will benefit when bigness is punished on the grounds of supporting “innovative alternatives”?
Of course, there’s no need to be overly alarmed by the possibility of this new antitrust revolution. We are still in the early stages of an open consultation and the Commission is meeting with industry representatives and considering their views alongside the enforcement experience of the Commission and national competition authorities, “findings and proposals from the worldwide reflection process on the need to amend the competition law framework” including “dozens of recent reports on platform power by regulators as well as by academic experts,” advice from DG COMP’s Economic Advisory Group on Competition Policy, consumer associations, the EU’s Platform Observatory and related work on platform regulation, and additional research papers commissioned by the EC “concerning the economic power of large online platforms.” The Commission will weigh all the available evidence and at the end of summer, once the consultation period closes, quickly have to finish its proposals to complete the intra-Commission review and approval, before formally proposing it in Q4 2020. Those who are concerned by the more revolutionary aspects of the Commission’s proposals will have their opportunity to speak, though there are clear political winds pushing the Commission to “do something” and only limited time for experts to engage in the debate before the main window of opportunity closes.. Ultimately, we will have to wait and see how much the Commission’s proposals will be able to preserve the more economic approach, and thereby competitive incentives and efficiencies of scale and scope.
To the extent that policy makers intend to preserve innovation and consumer welfare as policy objectives, they will need to hear the nuance, and reflect on their presumptions. Technology advocates will do our best to convey these points in the time allowed, because we all benefit from a system that rewards efficiency, investment, and vigorous competition.
1 Competition & Regulatory Counsel for the Computer & Communications Industry Association, written in his personal capacity. A complete list of CCIA’s membership, its source of funding, is available on its website: https://www.ccianet.org/.
2 Set to apply from July 12, 2020.
4 “[c]hallenging Google’s product design decisions in this case would require the Commission – or a court – to second-guess a firm’s product design decisions where plausible procompetitive justifications have been offered, and where those justifications are supported by ample evidence.” See Statement of the Federal Trade Commission Regarding Google’s Search Practices In the Matter of Google Inc. (FTC File Number 111-0163, January 3, 2013), available here. See also Competition Bureau statement regarding its investigation into alleged anticompetitive conduct by Google, (Canadian Competition Bureau, April 19, 2016), available here.
5 CCIA intervened before the European Courts.