Internationally, there is growing discomfort about the market position of some large digital players that serve as matchmakers and gatekeepers, controlling entire ecosystems. In Europe and the U.S., “Big Tech” is associated with the names of Google, Amazon, Facebook, Apple, and Microsoft – now widely famous under the acronym “GAFAM” – and possibly a few others. Those Big Tech players are accused of foreclosing or absorbing potential competitors, erecting barriers to entry and leveraging their entrenched market positions. While the immediate effect on consumers is often difficult to assess, the claim is that there is long-term harm to innovation and consumers.
The European Commission has proposed the Digital Markets Act (DMA) as a regulatory tool that is meant to complement EU competition law, in order to guarantee contestable digital markets. However, from a policy point of view, the current self-restriction to behavioural remedies in competition law and merger control, as well as the focus on behavioural ex ante regulation via the DMA, is at best a half-hearted and at worst a misguided way to effectively address the Big Tech challenge.
We argue in favour of a competition law toolkit with extended options to use structural measures to tackle entrenched market dysfunctionalities: expanded and strengthened merger control; extended possibilities to respond to infringements of competition law and equivalent provisions with structural remedies; and the availability of forced divestiture, possibly after a market investigation.
The Digital Gatekeeper Challenge
An increasing part of economic and social activity is facilitated by digital players and channelled through the internet, and a small number of firms have taken key “gatekeeper” positions. GAFAM have become private regulators dictating terms and conditions to participants in their ecosystems. Even if users are uncomfortable with those terms, there are often few viable alternatives to some of the services offered. Increasingly, GAFAM’s position looks to have become entrenched for a number of services. Strong network effects increase the value of a digital service for consumers and business users, and, because of coordination problems and inertia, switching to newcomers is unattractive. It is a bit like a meritocracy trap: Big Tech firms make more-attractive value propositions and become more sophisticated in extracting rents, and challengers have to overcome more and more hurdles.
The EU’s Regulatory Response
Regulating firms with power relative to other businesses in a vertical relationship or entrenched market power is a natural response. This is indeed the approach the EU has taken. The EU lawmaker has established across-the-board transparency rules in its Platform-to-Business Regulation 2019/1140. These are meant to improve business users’ position vis-à-vis the digital platforms. The EU lawmaker also targets a small number of particularly powerful players with an entrenched market position. With the proposed DMA, it seeks to prohibit certain business practices when adopted by them.
To illustrate, if a platform obliges its business users not to offer lower prices elsewhere, business users cannot divert consumers through price. This limits platforms’ incentives to compete on fees or other conditions they offer to their business users. End users may suffer from higher prices or lower quality. This concern goes beyond Big Tech and applies to other platforms in strong positions with respect to specific user groups, such as hotel booking platforms vis-à-vis independent hotels or event ticketing platforms vis-à-vis concert organizers. Thus, the rationale for singling out a few firms is not obvious in the case of some of the proposed obligations and prohibitions in the proposed DMA. The raison d’être of an intervention is not contingent on a gatekeeper position, as presupposed by the DMA proposal.
Things look different if the risks of exploitation and foreclosure increase with the scope of operations and market entrenchment, thanks, in particular, to overwhelming network effects. Here, the DMA approach is broadly suitable – that is, to specifically address gatekeeper platforms and to implement remedies that aim at keeping markets open or opening them up.
Behavioural Remedies and Their Shortcomings
At the EU level, we see, first of all, fines and behavioural remedies by the Commission based on findings of an infringement of Article 102 TFEU (abuse of market dominance) against Google (Alphabet) (Google Shopping, Android, and AdSense), as well as pending investigations against Apple (App Store Practices (music streaming), App Store Practices, and Mobile payments), Amazon (Amazon Marketplace and Amazon – Buy Box) and – again – Google (ad tech and data-related practices). Moreover, merger control may give the option to regulate market conduct via behavioural remedies so that combined resources (data) may not readily be used to erect new market barriers. In the Google/Fitbit merger proceedings, the Commission made extensive use of this. For instance, Google had to commit to not using the health and wellness data collected from Fitbit devices for Google Ads (including search advertising and display advertising).
Lawyers consider structural remedies – as opposed to behavioural remedies – to be the more intrusive measure. They may be seen as surgery instead of permanent drug treatment. Yet, structural remedies appear to be more in line with the idea that the state trusts market forces within an economic order it has formed, to guarantee the functioning of markets. Thus, in aggregate, “surgery” may be regarded as preferable, as it ultimately amounts to more economic freedom for all market players – including those that are subject to regulation due to their economic power.
In line with our view, the British Competition and Markets Authority (CMA), the Australian Competition and Consumer Commission (ACCC) and the Bundeskartellamt have issued on 20 April 2021 a “Joint statement on merger control enforcement”, stating that the “increasing complexity of dynamic markets and the need to undertake forward-looking assessments require competition agencies to favour structural over behavioural remedies.”
One structural remedy that is hotly debated in the U.S. is the “breakup” or, to use a term that sounds less dramatic, forced divestitures. Sometimes, such forced divestitures simply correct a merger that has turned out to be problematic. While a reinforced merger policy clearly cannot deal with the problems from the past, it may help in the future. Thus, a discussion of structural remedies in response to competition problems should start with merger policy
Merger Policy for Big Tech
Big Tech has acquired a large number of start-ups over the last decade. In digital markets it is difficult to foresee how these start-ups would have developed if they had stayed independent or had been acquired by some other firm. Currently, to block a merger, the Commission has to argue a case showing that the notified concentration will have anti-competitive effects. Given the uncertainties and the competition authorities’ lack of information, this is often an almost impossible mission.
Strengthening competition authorities’ power to prohibit mergers could mean to lower the required standard of proof, or giving them the power to reverse the burden of proof with regard to the expected effects on competition in clearly specified scenarios – for instance, if one of the merging firms has a powerful entrenched position. Such a position may be identified by a market investigation or by applying criteria as established in the proposed DMA or section 19a of the German Competition Act.
Various reform approaches point in this direction. In France, the Senate has approved a legislative initiative, currently pending in the National Assembly, that would shift the burden of proof in merger cases involving (designated) large digital gatekeepers (“entreprises structurantes”): When the competition authority initiates an in-depth examination of a notified transaction, it is the undertaking that must provide evidence that the transaction is not likely to harm competition. Likewise, in the U.S., the House Judiciary Committee recently approved a bill that would ban acquisitions by (designated) large digital platforms, unless they can prove that the merger will not harm actual or potential competition.
The French, German and Dutch governments (“Friends of an Effective Digital Markets Act”) lament that Article 12 of the DMA proposal lacked ambition. They demand that the existing EU merger framework should be modified for DMA gatekeepers: Acquisitions of low-turnover but high-value targets should be captured and the substantive test should be adapted to more effectively address cases of “potentially predatory acquisitions”. This is water to our mill.
Market Structure in Digital
Ultimately, the choice over whether we live in an environment with a few firms controlling large ecosystems or a more fragmented digital world is political. Since innovations yet unknown will benefit unforeseen digital activities, our working hypothesis is that a more fragmented world is likely to deliver more innovation than a world with few ecosystems, controlled by heavily regulated firms.
Absent the ability to impose divestiture obligations and to run a stricter merger policy, the DMA and its envisaged regulatory approach toward Big Tech may still help in opening up space for new and independent digital players. By analogy, this happened in a number of regulated industries, for example telecommunications, in which privatized incumbent firms were subject to more stringent regulation than newcomers were.
Regulation via the DMA or, possibly, via behavioural remedies imposed on the occasion of merger proceedings, seems to be a very indirect way of achieving this outcome. Enabling the European Commission to impose divestiture obligations as the result of a market investigation and, looking forward, to block digital conglomerate mergers more easily are more direct paths to a healthier digital world. In light of institutional constraints, this may be wishful thinking, but at least an open discussion in the Member States would be helpful. The option of stricter merger control and forced divestiture should not be taken off the table.
Forced divestitures are not unheard of in Europe. In West Germany, after World War II, mandated unbundling, such as that of IG Farben, was a powerful beacon to herald in the new paradigm of guaranteeing competition by means of regulatory intervention – if necessary, also by intervention in the market structure. This post-war period of competition law in West Germany was based on Allied decartelization laws inspired by U.S. antitrust law.
In the U.S., divestiture obligations are accepted as a legitimate remedy in the competition toolbox and its availability under section 2 of the Sherman Act has been recognised by the U.S. Supreme Court. Certainly, this remedy is not exactly routinely used, but it is an option. Very recently, in its complaint filed against Google on 20 October 2020, the U.S. Department of Justice requests the court to “enter structural relief as needed to cure any anticompetitive harm”. Moreover, politically, “unbundling Big Tech” is not seen as a far-fetched objective of competition policy. Rather, after recent debates in the House on a package of six tech-focused bills, Dan Bishop, a Republican U.S. Representative for North Carolina, remarked, “I will tell you, I’m not 100% there to break up Big Tech, but I’m close.”
Under German law, similar to EU law, structural remedies are hardly ever used, as they can only be imposed in case of competition infringements “if there is no behavioural remedy which would be equally effective, or if the behavioural remedy would entail a greater burden for the undertakings concerned than the structural remedies”. Neither in the EU nor in Germany can divestiture be ordered as an (objective) instrument of market regulation. The last attempt to add such an instrument to the competition toolbox – initiated by the Federal Ministry of Economic Affairs and supported by the Monopoly Commission – got bogged down in 2010. The debate at the time focused on conceivable targets (remarkably, Big Tech was not really on the radar yet), on fundamental questions of competition policy, and on uncertainties about what leeway EU law and fundamental rights left to the German legislature.
Since 2010, however, the rise of the digital platform economy and of concentration within it has continued. With this ongoing fundamental transformation of our economy, it would seem careless not to have an open mind to reconsidering structural instruments that were previously rejected as being too harsh for the addressees or too burdensome for the authorities to be implemented.
What Can We Ultimately Expect from the DMA?
In response to structural competition problems, it is only natural and indeed consistent to consider structural remedies, including breakups of digital conglomerates. Time will tell whether GAFAM will have to divest some of their activities. The more immediate question in the EU is who is willing to invest political capital to initiate a serious debate on strengthened merger control and forced divestitures as a regulatory instrument. A cautious move in this direction is the attempt by the rapporteur of the European Parliament to amend Article 16 of the DMA proposal in such a way that, first, the choice between behavioural and structural remedies is to be made based not on “proportionality” but on “effectiveness” and, second, that for structural remedies to be imposed it need not be shown that “equally effective behavioural remedies” are either not available or “more burdensome for the gatekeeper concerned”. In addition, there may still be hope that the initiative of the “friends of an effective DMA” to strengthen merger control will find fertile soil. However, we fear that the EU will, in the end, continue to restrict itself to playing games of behavioural remedies and regulation. We would be happy to be proved wrong.
Jens-Uwe Franck advises the Federal Ministry for Economic Affairs on issues related to the DMA.