From competition law to platform regulation – Regulatory choices for the Digital Markets Act

: The Digital Markets Act (DMA) brings a shift in the legal answers to the questions put forward by the amassing of power in the hands of a few digital “gatekeepers”. The DMA shall restrict this power and ensure “fairness and contestability”. Traditionally, the tools of competition law were used to set the basic rules of how to operate in markets. The shift to a different technique, regulatory law, prompts three questions: First: Why is it necessary to deviate from the path set in competition law? Secondly: What are the differences between the competition law approach and the regulatory approach of the DMA? Thirdly: What are key decision to be taken within the new regulatory framework? At the outset, DMA is presented. The DMA has not been passed as a binding law at the time of writing of this text. References in this text to the DMA relate to the European Commission’s draft of the DMA as put forward on 15 December 2020. While amendments will happen in the legislative process that is expected to be concluded in 2022, the key ideas and the general working mechanism of the DMA seem to stay in the pattern foreseen by the European Commission.

1 The Digital Markets Act (DMA) In December 2020, the European Commission proposed the DMA to answer widespread concerns regarding the power and market behavior of the operators of digital platforms (European Commission, 2020; for a more extensive analysis, see Chirico, 2021;De Streel, 2020;Ibáñez Colomo, 2021;Podszun et al., 2021a;Schweitzer, 2021).Should Alphabet be allowed to use the Google search engine to place their own services in better positions than those of other firms (selfpreferencing)?Can Meta require users of its virtual reality tools to sign up for a Facebook account?Can Amazon withhold transaction data from retailers on its marketplace, even if these data concern transactions of these retailers themselves?Should Apple be obliged to treat app developers fairly and in non-discriminatory ways in its app store?These are just some of the questions that will be answered by the DMA.
In line with the EU law-making process, the Commission coordinated its position with the Member States (represented in the Council) and the European Parliament.After the negotiations of the three parties (the so-called trilogue) came to a conclusion in the first half of 2022, the DMA entered into force on 1 November 2022.In May 2023, the phase of designating the addressees (so-called gatekeepers) started.It is expected that the obligations take practical effect in early 2024.While some of the details were controversial, politicians in the EU are overwhelmingly in favor of stronger regulation of digital platforms.
The DMA targets "gatekeepers" that operate "core platform services" (CPS), such as online intermediation services, online search engines, networking services, or operating systems (listed in Art. 2 (2) DMA).Gatekeepers are designated by the Commission based on three factors: their significant impact on the market in the EU, their operating of a CPS which is an important gateway for business users to reach end users, and the entrenched and durable position in the operations now or in the foreseeable, near future (Art.3 (1) DMA).It is presumed that these criteria are fulfilled once certain quantitative thresholds are met, including a certain number of users.The following key figures constitute the threshold: annual european economic area turnover of € 7.5 bn. or market capitalization of € 75 bn., 45 million monthly active end users and 10,000 yearly active business users, for the last 3 years, see Art. 3(2) DMA.
At the time of writing, seven companies had notified the European Commission that they reach the quantitative thresholds, namely Google/Alphabet, Amazon, Facebook/ Meta, Apple (formerly known as "GAFA"), Microsoft, Samsung, and ByteDance (TikTok).Whetherover timeother companies will fall under the gatekeeper-regime is an open question; yet the number of companies targeted by this law will not exceed 10-20 (cf. DMA Impact Assessment Report, 2020, p. 14).
If a company is designated as a gatekeeper, it has to respect specific obligations in its business affairs.These obligations are laid out in Art. 5, 6, and 7 DMA.The obligations shall be "self-executing" once the gatekeeper status is reached.It is neither required to have an activation nor an individual specification by the competent authority.For Art. 6 obligations, the European Commission may give individual directions to gatekeepers in a formal procedure (Art. 8).Gatekeepers also need to inform more extensively about their merger activities (Art.14) and consumer profiling (Art.15).
The DMA aims at ensuring "contestable and fair digital markets" (Art. 1 (1) DMA).As regards substance, the obligations mostly deal with the treatment of business users while some provisions are meant to empower end users.Business users shall have access to some data; some exclusivity clauses and other restrictions to competition are forbidden.There shall not be a mandatory use of ancillary services provided by the gatekeeper.Several provisions aim at interoperability.Self-preferencing of own products or services or those of subsidiaries in rankings is not allowed.Conditions for business users in app stores, search engines, and social networking services need to adhere to FRAND standards (fair, reasonable, and non-discriminatory).
Regarding end users, the combination of data from different services requires that users have a real choice and give specific consent.This will make it harder to collect and combine data for profiling.Also, end users need to be allowed to un-install pre-installed apps and change default settings.Messenger services of gatekeepers must ensure interoperability so thate.g.-WhatsApp users can communicate seamlessly with users of competing services.
The provisions are detailed in their wording so that disputes regarding the content of the rules are limited.They are complemented by a rule on anti-circumvention (Art.13) and requirements for internal compliance structures in the companies (Art.28).There is an enforcement regime put into place including investigations and sanctions such as fines (cf.Art.20 et seq.).The enforcement body will be the European Commission.In cases of systematic non-compliance (Art.18: three non-compliance decisions in eight years) structural measures may be imposed, including the divestiture of companies.
The DMA also contains the possibility to update and extend existing obligations in a quicker procedure than normal (Art.12) and to do a market investigation into new services and practices that may culminate in a legislative proposal (Art.19).Further provisions concern the cooperation with EU Member States and the relationship to other laws, e.g., from the field of antitrust.
The structure and content of the DMA make it clear that this is new and gatekeeper-specific regulation.This prompts the question why the more general competition law addressing all companies does not suffice for digital markets.

Digital Gatekeepers and Competition Law
The rise of companies operating digital platforms over the past two decades to superstar firms has triggered antitrust investigations throughout the world.Antitrust law (mostly called competition law in Europe) is the traditional field of the law that provides a general framework for market powerrelated concerns in the economy.It provides for a prohibition for undertakings with a dominant position in the market to abuse this power.Companies may not strengthen their position through agreements with which they use their joint power to the detriment of customers.External growth through mergers and acquisitions is subject to a review system that shall hinder high concentration.
In all three fields, abuse of dominance, collusion, merger control, activities by digital platform companies triggered antitrust proceedings.Yet, the application of competition law came to its limits as is briefly described in the following passages.In the European Union, the European Commission as the enforcement agency was not able to react in an adequate way to the challenges for free competition posed by the GAFAs.This failure of competition law is the background for the move to regulation.Part of it is ascribed to the role of economics in antitrust.

The Role of Economics in Antitrust
Competition law is the field of the law where the connection to economics is most sophisticated.Economic concepts and models are looked at during the law-making process and are often integrated (with varying degrees) in individual cases.When assessing the difficulties to enforce competition law vis-à-vis digital gatekeepers, the role of mainstream competition economics is particularly interesting.European competition law had followed a "more economic approach" ever since the early 2000s.The idea was to update rules with a view to modern economic theories and to integrate economic evidence better in enforcement (cf.Schmidtchen et al., 2007).Rules that contained outright prohibitions of corporate behaviour, so-called per se-rules, were replaced with rules that foresaw an individual assessment of the effects on a case-by-case basis ("effects based approach" instead of "form based approach").This was in line with a school of antitrust economics called "post Chicago School".The aim of enforcement was defined as enhancing efficiencies and consumer welfare (while before the focus had been on securing the competitive process by guaranteeing certain structural elements).In important cases, parties and enforcers nowadays work with economists, modeling economic effects of practices and of intervention, trying to calculate the consumer benefit in concrete monetary terms (cf.Lianos & Genakos, 2013).Open notions in competition laws such as "abuse" or "significant impediment of effective competition" are now interpreted with the help of economic models and counterfactual scenarios.
This approach was needed after a period of habitual competition law enforcement that had been detached from new economic insights.The new approach to assess effects for consumers on a case-by-case basis came at the cost of legal certainty.Compliance for companies became more difficult, and so did enforcement for competition agencies since they had to engage in more complex economic tests.The pendulum swung from easily enforceable but unsophisticated rules to very sophisticated rules that required analysis in deep detail.The first limitation of this effectbased approach was the resources needed for its application and the legal uncertainty coming with it.
A second limitation was the limited use for the new challenges to competition by platforms (cf.Koenig, 2019).The introduction of the more economic approach in European law coincided with the rise of the platform economy at the beginning of the twenty-first century.Lawyers adopt economic thinking only with a certain time lag and when theories are tested and seen as robust.Accordingly, the models and theories introduced into the competition law arena were based on knowledge on pre-digital sectors while economic thinking, e.g., on the phenomena of platforms, had already advanced, but not yet to point that made it possible to integrate it into legal contexts.
Enforcers and courts were confronted with more and more digital cases.Platforms proved to be particularly challenging for competition since many platforms tend toward monopolization due to network effects, economies of scale and scope, and the use of data (cf.Crémer et al., 2019).Silicon Valley business models were designed to exploit these factors quickly so as to establish monopolies on tipped markets.The famous words of investor Thiel (2014) that "competition is for losers" reflect these new business strategies that entail a renunciation of the competition paradigm of the market economy.
Network effects of platforms, the connections of market sides on multi-sided markets, the tipping of markets, the lock-in effects for users, the massive use of data and algorithms, the modular design of platforms, the economies of scale and scope, the growing integration of marketsall these aspects were not part of the standard toolkit that economists were able to provide to law enforcement in the 2000s.While these phenomena as such were not new, their combined power in digital platforms was unprecedented.When the integration of the "more economic approach" into competition law enforcement started, it was not up to date with the parallel rise of platform operators.It also suffered from a lack of integration of behavioral economics, institutional or evolutionary economics, and other strands of economic thinking.
A third limitation of post-Chicago economics was the one-sided focus on a narrowly understood consumer welfare standard (that deviated from the previously held more normative standard of preserving a certain market structure or competition as a process) (Podszun & Rohner, 2023).Consumer welfare was interpreted as meaning lower prices for end consumers.This led to a focus on productive efficiencies that were well calculable in models.Dynamic efficiencies, innovation, and potential future developments fell out of the picture, not least for the difficulties of finding data and putting these into models.Goals going beyond an efficiency-oriented understanding of markets, e.g., consumer choice or the ability of all market actors to decide independently and sovereign, were not visible at all.
Even more so, non-economic, political, or normative ideas that had been traditionally attached to antitrust rules, such as the restriction of power in society and the protection of democratic values, were beyond the reach of economics.New goals that nowadays are sometimes linked to antitrust enforcement from a fundamental rights perspective (particularly free speech and data privacy) were also not an issue for the post-Chicago school.The "theories of harm" developed under this umbrella were narrower than before and made interventions by competition agencies and courts more difficult.
The "more economic approach" thus had a narrow focus.It used costly economic tests and models that were based on many assumptions.The turn to this approach by the European Commission over time made it hard to deal with new phenomena in digital markets.It may well be argued that this trend from economics resulted in a severe under-enforcement of competition law (cf.Baker et al., 2018).

Abuse of Dominance
The hopes and challenges of digital antitrust law can be illustrated with the most prominent abuse of dominance case, decided on the basis of Art.102 TFEU, Google Search (Shopping).The European Commission took a decision against Google on 27 June 2017, ordering a stop of "self-preferencing" of Google-owned services in the ranking of the search engine and fining the company € 2.42 bn (Google Search (Shopping) (Case COMP/AT.39740)[2017] OJ C9/11).In the display of search results, Google had privileged its own price comparison service, Google Shopping, over competitors such as Idealo that had been ranked down.Google used its strong position in internet search to acquire an equally strong position in the market for price comparison portals.This practice is known as "leveraging of market power".According to the Commission's finding, Google's rivals, ranked low, had severe difficulties to reach consumers who tend to look at the first search results only.
These proceedings had several shortcomings.First, the European Commission failed to provide a clear-cut "theory of harm."The term "theory of harm" in competition law stands for the economic reasoning why a behavior is deemed anti-competitive.The Commission shied away from establishing a clear-cut concept of favoring (cf.Bostoen, 2022).
To understand the relevance of this, it is important to understand how lawyers work with laws and how economics fit into this: Competition law sets broadly defined legal standards ("do not abuse a dominant position").These standards are filled in practice over time by looking at past behavior (e.g., certain rebate strategies) and categorizing it into established types of abuse (e.g., abusive loyalty rebates).Such a categorization is necessary to satisfy the requirements of the rule of law: Market actors need to have legal certainty to be able to act in a compliant way, and agencies need clear interpretations of the broad general clauses so that arbitrary decisions are impossible.Economic reasoning is used to justify firm behavior or the prohibition of it.Legal certainty depends on the establishing of robust "theories of harm" underlying the types of abuse.
The Commission faced a dilemma in Google Shopping: It found that self-preferencing practices by the dominant search engine impeded market access for rivals, yet the case did not fit properly into one of the established types of abuse since these were developed for more traditional sectors.Squeezing the case into an existing "theory of harm" runs the risk of a judicial review that finds that facts and law do not match.Establishing a new type of abuse runs the risk of the judiciary pointing at missing precedent.In Google Shopping, the European General Court remedied the problem by boldly establishing a theory of harm on its own (EU General Court, 10 November 2021, Case T-612/17 -Google Shopping; cf.Bostoen, 2022).In other landmark cases, the General Court overturned Commission decisions for a lack of convincing economic evidence or theories of harm (EU General Court, 26 January 2022, Case T-286/09 RENV -Intel; EU General Court, 15 June 2022, Case T-235/ 18 -Qualcomm; also parts of the analysis in another Google case were struck down, cf.EU General Court, 14 September 2022, Case T-604/18 -Google Android).
So, in theory, the general clauses in competition law are very open for integrating new developments, yet, in practice, it is hard to tackle new phenomena after an economic revolution such as the platform revolution with the established rules.This conceptual problem became visible in the Google Shopping case.
A second flaw was the inability of the European Commission to design a suitable remedy.In a two-sided market, there are incentives for the platform operator to foreclose access so as to exploit the other market side (cf.Katz, 2017;Shapiro, 1999).Antitrust agencies finding an infringement of the law hand down a cease-and-desist order, fines, and also specifications on how to remedy the unlawful behavior.Fines are not a good deterrent for companies commanding the liquid means of Google.Designing appropriate remedial action proves to be very difficult (cf.Kathuria & Globocnik, 2020).The remedial action taken after the prohibition of self-preferencing by Google is largely criticized as having very little impact or as even more favorable to Google than before (cf.Höppner, 2020;Marsden, 2020).
The third and most obvious flaw with antitrust enforcement in this landmark case is its duration.The Commission started investigating the case in 2010, acting upon complaints by rivals.The decision was taken no earlier than 2017.While it became immediately enforceable, it is still not fully legally valid.In 2021, the European General Court reviewed the case and sided with the Commission (EU General Court, 10 November 2021, Case T-612/17 -Google Shopping).The case now rests with the European Court of Justice.So, 13 years after the first procedural steps, there is still no final result.Obviously, such an enforcement saga lags behind the dynamics and the spiraling effect of platform power.In the interim, durable harm may occur that is hard to overcome.Interim measures are possible but rarely used.

Restrictive Agreements
Restrictive agreements (forbidden under Art.101 TFEU) were the subject of control for hotel booking portals.Booking.com had established best price clauses in its contracts with hotels, so that hotels were not able to offer better conditions on other portals or in their own distribution channels (online or offline).Several national competition authorities held some of these clauses to be anti-competitive (cf.Lamadrid, 2021;Podszun & Rohner, 2022).These proceedings took a long time, again, and the different national approaches led to a European mosaic of different decisions (DMA Impact Assessment Report, 2020, Annex 5.4, p. 110 et seq.).
In the control of collusive practices, antitrust authorities in Europe have not yet taken full issue with concerns that the GAFA companies themselves restrict competition amongst them.There have been indications for such a division of markets in the digital world, and the Italian competition authority fined Apple and Amazon for colluding.Investigations are burdensome, however, particularly for national enforcers facing deep pocketed companies.

Merger Control
In merger control, the lack of meaningful enforcement is particularly striking.The European Commission for instance unconditionally cleared the acquisition of WhatsApp by Facebook (Facebook/WhatsApp (Case COMP/M.7217,C(2014) 7239 final)).It did not find a "significant impediment of effective competition," and it thereby enabled Facebook to build a global communication and messaging universe.Most transactions by the GAFA companies were not even subject to EU merger control so that Alphabet, Meta, and others were able to extend their reach to ever new markets by buying up companies (this problem is partly cured by transaction volume thresholds for merger control in some EU Member States and by a new (but controversial) reading of the EU Commission of Art.22 of the EU Merger Control Regulation, also in conjunction with Art. 14 DMA (cf.Podszun, 2023)).
In a turn of this enforcement policy, the European Commission scrutinized the Google acquisition of Fitbit.It did not oppose this deal outright (despite competitive concerns for Google's taking of the health data market), but it approved it conditionally (Google/Fitbit (Case COMP/M.9660)C (2020) 9105 final).The conditions, including a ban for using Fitbit data for advertising in the European Economic Area, run for a period of 10 years.It may well be questioned whether the monitoring of company activities for 10 years is feasible at all and a suitable remedy in merger control (cf.Spagnolo et al., 2020, p. 8 et seq.).

The Failure of Antitrust
In several platform markets, competition has basically collapsed, and markets have "tipped" in favor of one operator.
Google Search, the Android and Apple operating systems, the Meta networking and communication space, or Amazon's position as the number one retailer are proof that competition law did not live up to its promise of guaranteeing a vibrant competitive culture in markets.The list of shortcomings in EU antitrust cases highlights some of the underlying problems: Investigations and judicial review take too long, the very open and vague standards set in antitrust law make it difficult to decide in a quick and foreseeable manner, and the requirements set in line with the "more economic approach" are hard to meet when new phenomena are at stake.The design of proper remedies is difficult.
Other problems add to this: The European Commission aims at establishing a digital single market and therefore favors an antitrust policy that levels national differences.The firms instrumental for this are large global players that tend to concentrate market power in one hand.
The focus on abuses of market power may have blinded enforcers for the dangers of monopolized markets where abuses were not clearly visible at once or first had to be identified in new groups of cases.
The institutional setting of antitrust rules and their substantive content were established for a business world where coal and steel and cars were the key products in the EU.The tools were not designed and trained to deal with network effects, data, and digital ecosystems (cf. the reports by ACCC, 2019; Crémer et al., 2019;Furman et al., 2019).Other regulatory failures, in particular far-reaching non-liability rules for platforms (that enabled them to externalize costs) and under-enforcement of data protection rules, added to the power of the gatekeepers that were able to build their digital ecosystems, locking in business users and consumers alike within a short period of time.Competition law achieved too little too late during the digital revolution of markets (Schweitzer & Gutmann, 2021, offer a different reading of competition law enforcement).The path of ex post analysis of company's behavior with broadly defined standards in general clauses proved to be inefficient in "taming the tech titans" (The Economist, 18.1.2018).

Characteristics of Regulation
In reaction to this perceived failure, some national legislators, such as Germany, turned to amendments of their antitrust laws (on the different models to react to the digital revolution, cf.Botta, 2021).Germany introduced a specific competition law provision for "undertakings with paramount significance for competition across markets" that can be designated as such by the Bundeskartellamt (section 19a of the German competition act) (cf.Franck & Peitz, 2021).These undertakings have to comply with more detailed provisions of forbidden abuses but may still invoke an efficiency defense.Alphabet was the first undertaking to be named one with paramount significance for competition across markets in 2021; Meta and Amazon followed in 2022.
The European Commission refrained from amendments in competition law and instead drafted a completely new law, the DMA, that leaves the realm of competition law (on its significance, cf.Andriychuk, 2021;Budzinski & Mendelsohn, 2021, p. 15 ff.).It is a part of "regulatory law" that shall now compensate some failures of antitrust laws.Both fields, specific regulatory law and general antitrust law, are in a close relationship (cf.Hovenkamp, 2020;ICN, 2004;Shelanski, 2011).Yet, competition law is distinct from regulatory law (Akman, 2022;De Streel, 2008;Dunne, 2015).Regulation as a legal term means specific provisions for certain companies that act as providers of an infrastructure.Regulation is based on the assumption that competition can no longer control markets so that a market failure is permanent.This failure is usually due to a natural monopoly, i.e., an infrastructure with so high entry barriers (due to costs or network effects) that there is no actual or potential competition.This is the case for railway networks or energy pipelines, for instance.The operator of the infrastructure can extract profits above the competitive level.It also comes under a stricter control by government agencies since the provision of the service is often fundamental for other market actors (Shughart, 2008, p. 458).Regulatory law does no longer aim at establishing competition in the first place, but holds the company liable for societal aims, often with the yardstick of scenarios that imitate competition (as-if-competition).Turning to regulation in this sense is a disruptive step for each market concerned: The legal regime changes from general oversight with the tools of competition law to a more specific regulation of the undertakings active in the sector.
In the DMA's recital 10, competition law rules are defined as rules that are based on an individualised assessment of market positions and behaviour, including its actual or potential effects and the precise scope of the prohibited behaviour, and which provide for the possibility of undertakings to make efficiency and objective justification arguments for the behaviour in question.
The quote highlights distinctive features of competition law: the individualized assessment, the look at effects, and the possibility of justification on the basis of efficiency.
On a meta level, the decisive difference between the paths of competition law on the one hand and regulatory law on the other is the level of government involvement.
Under competition law, the state (or in the constellation of the European Union: the sovereign public actor) provides a general framework for all undertakings in which these undertakings can act freely.The role of government agencies is reduced to curb back excesses and to guarantee a structure that enables market actors to pursue their economic aims autonomously.The practices and results of such a market understanding take everyone onto a "discovery procedure" (as Hayek, 1968, the advocate of spontaneous market orders, put it).Freedom of market actors is only restricted to secure the freedom of others (in line with the "freedom paradox").
Under regulatory law, a government agency acts in a much more prescriptive way.Government does not rely on the free exchange in the market once a certain structure is guaranteed but it sees the structure as so flawed (and noncorrectable) that firm behavior has to be steered and certain results aimed at by targeted intervention.The freedom of businesses is restricted; certain practices are banned or prescribed from the outset so as to achieve the preferred results.The framework is much narrower, some paths within this framework are closed, and others are made mandatory for use.
With the DMA, the European Union takes a step into this kind of regulation for the digital sphere, subjecting certain platform services to a much more detailed governmental steering.
Obviously, with the various laws in place governing the economy, the distinction cannot be drawn as clear-cut as done here.The difference to other rules, e.g., requirements for product safety, working conditions, or environmental standards, lies in the purpose of the rules: Competition law, and to some extent also regulatory law, serves the purpose of making markets work.In recital 5 of the DMA, it is stated that "the market processes are often incapable of ensuring fair economic outcomes with regard to core platform services."Thus, just like competition law, the DMA concerns the working mechanism of supply and demand in general.Other rules serve specific non-economic purposes while competition law and DMA primarily (if not exclusively) aim at fixing market failures as understood by Pigou (1932).
In the following parts, some of the differences of a competition law path and a regulatory approach are described in more detail (it has been argued that the DMA is a hybrid of competition law and regulatory law, cf.Chirico, 2021).

General vs Sector-Specific
While competition law normally addresses all companies alike, regulatory law is typically focusing on one sector: energy, telecommunications, or railways, for instance.The DMA may be seen as sector-specific tootargeting digital gatekeepers.On the other hand, there is no "digital sector" as such, even though the DMA speaks of the "digital sector" in Art.1(1) (Akman, 2022, p. 18).The services provided by GAFA companies cut through all sectors, and they have operations running that may well be classified as being part of sectors such as tourism (Google Flight), retail (Amazon), media (YouTube), advertising (Facebook, Google), IT (AWS), or health (Apple Health).The definition of the CPS and the specific obligations do not only target what one may see as basic digital services for the economy, but also more specific ventures.
Still, it is a very limited number of undertakings that are addressed by the rules.This limited circle makes it possible to have tailor-made obligations instead of onesize-fits-all general rules as in competition law.
Traditional regulatory law is confined to sectors where the mechanisms of supply and demand do not function properly due to the dependency on natural monopolies or essential facilities that cannot be duplicated.Energy pipelines, the telephone network, and the railway system are examples for this.Regulatory law starts from the deficits in competition that are the consequence of relying on such infrastructures.That is why regulatory law is usually sector-specific.
Taking a regulatory approach in the DMA leads to two observations in this regard: First, the CPS operated by the digital gatekeepers can be regarded as "akin to an essential facility," as the European Court put it for Google Search (General Court, 10 November 2021, Case T-612/17 -Google Shopping at para 224).That is a turn since classic infrastructures are of a brick-and-mortar structure (natural monopolies, cf.Posner, 1999) whereas the dominance in search or the infrastructural position of operating systems of smartphones is more virtual in nature and partly relies on network effects.It is a major breakthrough to see some of these services as no longer contestable essential facilities.
Second, it is striking that the Commission does not only choose platform services of the gatekeepers that are very close to monopolistic constellations (such as operating systems) or form the backbone of e-commerce and digital businesses (such as search) but also includes operations such as video-sharing platforms or advertising services.All these services are treated the same in the DMA (with minor specifications in some obligations).That is an encompassing approach.
This stunning reach toward very different services provided by the gatekeepers may be justified by the nature of the digital world that is interconnected and integrated: Each platform service of the gatekeepers is connected with other services; together they form a "digital ecosystem," or a "walled garden," which aims at integrating more and more services (Bourreau & de Streel, 2019;Jacobides & Lianos, 2021;Podszun, 2019).Looking at just one of these structures (as in a classic competition law style market definition) would miss the point: The use of data leads to a convergence of markets and businesses that can no longer be analyzed in isolation.An e-mail-account with a gatekeeper may be the key to entering a digital ecosystem that locks in users.A regulatory approach therefore needs to overcome the narrow market definition and has to see operations of ecosystems as a complex network.

Standards vs Rules
The distinction between competition law and regulatory law can also be drawn along the lines of the legal distinction of standards versus rules (cf.Kaplow, 1992;Kerber, 2021;Witt, 2022).
Competition laws are closer to standards: The provisions are worded as rather broad general clauses that may need further specification through agencies and courts.For instance, Art.102 TFEU essentially says: An undertaking that is dominant in its market may not abuse this position.This leaves room for interpretation and effects analysis.
Regulatory law works with rules that are much more specific and prohibit or prescribe exact behavior.For example, in Art.6(5) of the DMA, it is stipulated that the gatekeeper shall not treat more favourably, in ranking and related indexing and crawling, services and products offered by the gatekeeper itself than similar services or products of a third party.The gatekeeper shall apply transparent, fair and non-discriminatory conditions to such ranking.
So, what the Commission established with a lot of effort in Google Search (Shopping) as a case of Art.102 TFEU is now framed in a much more precise fashion as a special duty for gatekeepers in the DMA.This notable difference between the wording of Art.102 TFEU and the prohibition of self-preferencing in Art.6(5) DMA illustrates the degree of detail in regulatory law to capture firm conduct.The "extensive investigation of often very complex facts on a case by case basis" (DMA, recital 5), that is at the heart of competition law, is turned into a self-executing, directly applicable rule that gatekeepers have to adhere to.
The distinction made here is often described as an ex post/ex ante-dichotomy (OECD, 2021).According to this view, competition law steps in retrospectively with agencies and courts assessing ex post whether there was a violation of the law.Regulatory law, so it is said, requires compliance ex ante, i.e., from the outset with laws and agencies telling companies what to do before they start their business operations.This is, however, a very stylized modeling of both areas.Competition laws require compliance from the beginning, it is only due to the nature of the wording that this is sometimes difficult for undertakings.Furthermore, in many fields of competition law, guidelines and case practice have established clear rules to follow that can be complied with ex ante.In both fields, agencies often come into play with an ex post review of what happened in markets.
Still, the obligations in the DMA are self-executing, do not need any further activation by an agency (other than in section 19a of the reformed German competition act), and shall be precise enough for gatekeepers to know what they are allowed to do or not to do (cf.Akman, 2022;Botta, 2021).The mechanisms of compliance that were established in companies during the past decade rely on such detailed rules and partly supersede the more traditional system of review by a government authority.The DMA requires audits and compliance within the gatekeeping companies in specific provisions, cf.Art. 8, 11, 15, 28.

Different Goals
Competition law scholars debate at length what the purpose of competition law is (cf.OECD, 2003;Zimmer, 2012).But as stated before, the considerations remain in the realm of making markets work, be it by protecting the competitive process or consumer welfare.In line with this, undertakings in a competition law regime are usually free to present an efficiency defense that allows them to justify an anti-competitive practice on economic grounds.
The DMA offers only very limited possibilities for exemptions in Art. 9 and 10 and importantly does not allow for an efficiency defense.This is consistent with the goals of the DMA.They are broader in definition.Art.1(1) DMA reads: The purpose of this Regulation is to contribute to the proper functioning of the internal market by laying down harmonised rules ensuring for all businesses, contestable and fair markets in the digital sector across the Union where gatekeepers are present, to the benefit of business users and end users.
A distortion of fairness cannot be compensated by economic efficiencies; these are different categories, as the Commission rightly points out (Recital 11 DMA).
The goals of the DMA are based on value-judgments efficiency does no longer take priority.Regulatory law is more open toward normative goals than competition law.In the field of energy regulation for instance the goals of the legal framework extend to the safe provision of environmentally friendly energy.Postal services are often regulated so that distant regions have daily postal services, despite of this being inefficient.
With this in mind, the regulatory goals pursued by the DMA are rather unspecific.The Commission could have gone even further with setting normative regulatory principles and ideas for the DMA (Larouche & de Streel, 2021;Podszun et al., 2021b).
Regulatory interventions become less necessary, the more competition is created.The transition of the telecommunication markets is a telling example: When the sector was opened up in the 1990s in Europe, the deregulation process was combined with close monitoring by regulators and many rules.Today, many telecommunication markets work well with only very limited sector-specific regulation and just normal competition law scrutiny.
For telecommunication, the European Commission considers a market to justify the imposition of regulatory obligations if high and non-transitory structural, legal or regulatory barriers to entry are present; there is a market structure which does not tend towards effective competition within the relevant time horizon, having regard to the state of infrastructure-based competition and other sources of competition behind the barriers to entry; competition law alone is insufficient to adequately address the identified market failure(s).(Art.67(1) of Directive (EU) 2018/ 1972 of 11 December 2018 establishing the European Electronic Communications Code (Recast)) Applying the criteria from telecom law, regulation is merited as long as there are high barriers to entry, a market structure that does not enable competition, and market failures that cannot be remedied through competition law.These requirements are certainly met for most digital services addressed by the DMA (but cf.Ibáñez Colomo, 2021).For the time being, many of the markets with digital gatekeepers will fulfill the criteria.It may be pointed out again, though, that thinking in markets (as neatly separated units) does not do justice to the connected business spaces of gatekeepers.The idea of deregulation in telecoms, if the criteria for regulation are no longer fulfilled, brings up the question whether the new platform regulation has an underlying idea to transform markets so that, one day, the DMA is no longer necessary.
Such a "finality" of platform regulation, or a transitory nature, is not visible in the DMA, even though the gatekeeper status is to be reviewed every 3 years (Art.4).The obligations in the DMA do not conclusively aim at breaking the power of the CPS.They primarily aim at hindering a further extension of that power through leveraging to other services.The monopolistic position of gatekeepers is only partly challenged by the DMA.The priority is the fairer treatment of users.This fits perfectly with the understanding of regulating a permanent monopoly.A change of such monopoly power will only come with new disruptive technology.

Hard Regulatory Choices
In Section 3, the distinguishing features of a regulatory approach were explained.Within such an approach, further regulatory choices have to be made.The following part deals with their costs and trade-offs.

Asymmetric Regulation
The basic model of regulation foresees that specific rules have to be complied with.The key question is the applicability of the Act to market actors: Who shall be bound by the special obligations?Whether an undertaking is or is not designated as a gatekeeper makes a fundamental difference.The DMA does not apply to all online intermediaries, but only several undertakings.This is an act of asymmetric regulation.
The number of addressees changes the possible specificity of obligations.The less corporations are targeted, the more specific the obligations may be.With a growing number of gatekeepers and CPS under the umbrella of the DMA, obligations need to be more general to capture the different business models.Thus, the level of possible intervention is determined by the number of undertakings targeted by the intervention.
The DMA does not explicitly confer specific rights upon users, yet the whole idea of the DMA is to open up possibilities for them (cf.Podszun, 2021).One controversy in this regard is whether other gatekeepers can profit from the DMA: If gatekeepers are forced under the DMA to make data accessible or to ensure interoperability, the ones profiting most from this could be other gatekeeping companies.They have the resources to exploit and use data and procompetitive loopholes in a much more strategic and efficient way than smaller businesses.This may mean that the DMA could strengthen the very companies it seeks to regulate.It is questionable whether this was intended in the first place, and it is doubtful whether such competition of gatekeepers would be a regulatory success.A potential institutional answer to this would be to take asymmetric regulation even further by obliging gatekeepers, but stripping them of the rights flowing to others under the DMA (cf.Andriychuk, 2021).However, this is not foreseen in the DMA.

Interference with Consumer Preferences
A second important decision concerns the role of consumer preferences.The companies and services that are targeted by the DMA are among the most-loved brands for consumers.Their services are hugely popular, and this is the result of high-quality services.With new choice screens, less data combination, consent buttons, or difficulties to offer certain integrated services out of one hand, the user experience may become less comfortable than so far.The users' growing frustration regarding the necessity to agree to or decline the use of cookies on every website in the wake of the GDPR may serve as a telling example for a regulation that frustrates consumers without granting them a considerable advantage (cf. a survey carried out by the German industry association BITKOM in 2020 showing that 43% of internet users are annoyed by cookie banners, https://www.bitkom.org/Presse/Presseinformation/Cookie-Banner-spalten-Internetnutzer).The DMA might interfere with some established consumer preferences.Yet, it needs to be born in mind that the DMA is not primarily oriented toward short-term consumer benefits but aims at consumer choice or consumer sovereignty and better conditions and fairness for business users.The regulatory approach of the DMA employs a longer-term perspective with broader policy aims than post-Chicago antitrust law where static, momentaneous snapshots of consumer welfare in a narrow sense are decisive.
The consumer preference argument does not seem to be particularly strong in the digital sphere anyway.Many of the gatekeepers act as information intermediaries.Their key service is the exploitation of data.The necessary information for building consumer preferences are channeled and presented by the gatekeepers to the users themselves.This means that the sources of information are limited, and the building of preferences is hampered.Over time, consumers may depend more and more on the paths laid out for them by their information intermediaries.Regulating the exploitation of data by the intermediary may therefore be seen as a prerequisite for building autonomous consumer preferences, so that the interference with consumer preferences is not a major issue.

Over-and Under-Enforcement
Each regulatory regime faces the difficulty to get the level of enforcement right (cf.Akman, 2022, p. 29).Over-enforcement may stifle innovation and efficient businesses.Underenforcement may jeopardize the goals pursued by the regulatory act.Whether there is systematic under-or over-enforcement can hardly be predicted.The "more economic approach" is today seen as having led to a more cautious enforcement approach in antitrust that will partly be superseded with the DMA.A systematic analysis of the impact of the DMA for businesses and innovation, particularly with so many different obligations at play, is hardly doable in advance.
Yet, the balance of over-or under-enforcement does not only depend on questions of substance, but also questions of the enforcement regime and its institutional design (the well-meant data protection rules in the EUwhile farreaching on substanceare an example for lack of an appropriate institutional design, cf. the report of the ICCL, 2021.For the DMA, the European Commission becomes the sole enforcement authority.Agencies from EU Member States may only assist the Commission.For the internal organization of the European Commission, the most important question will be the number of staff that will work on DMA enforcement.Their institutional standing, the experience of case handlers, their incentives and the role of judicial review of their decisions will be decisive factors for determining the level of enforcement. Private parties could take gatekeepers to court in private enforcement and thereby significantly raise the level of enforcement.The general understanding is that private enforcement in national courts is possible even though provisions on this matter in the DMA are very weak (Podszun, 2021).Users may sue for cease-and-desist orders, interim relief, or even damages.This could become an important pillar of DMA enforcement if national courts muster the courage to decide quickly and in a determined manner on DMA matters.

Speeding up Regulation
A key idea of the regulatory approach is to speed up regulation, thereby having impact in a faster way.In the DMA, this policy choice is guaranteed through time limits for all steps in the process, easy to check assumptions for interpreting legal terms, and self-executing rules that have to be complied with without previous intervention of an enforcing body.
This policy choice comes at the cost of more differentiated, more targeted obligations for specific services and a careful individual assessment of business practices as is typical of competition law.The ex ante regulation with less possibilities for companies to steer the process speeds up the implementation.
This may have two downsides: First, the quality of decisions may deteriorate since there is less time to check and double-check.Second, the rights of defense of gatekeepers could be restricted (cf. Ibáñez Colomo, 2021).If norm addressees have less time to put their position forward, if there are less hearings, shorter time frames, more non-rebuttable assumptions, and less mechanisms of judicial control, this means a loss in rights of defense in comparison to the current status in competition law proceedings.As rights of defense are also about improving the quality and precision of a decision, there might be negative effects.
On balance, it needs to be taken into account that gatekeepers usually have all the means to galvanize their legal support in a short period of time.Furthermore, it may be questioned more generally whether the rights of defense have grown out of proportion.It is hard to imagine a speedier resolution of controversial cases without cutting back some of the individual, right-based possibilities of companies to defend their actions.Speeding up regulation is a key element of the DMA, yet each reduction of individual and lengthy assessment needs to be weighed with these downsides.

Updating Obligations
Each new regulation faces the problem that it needs to be future-proof.That is all the more true in an environment that the Commission itself describes as "very rapidly changing" and of "complex technological nature" (DMA recital 30; cf.Blockx, 2021).How can it be ensured that the DMA is not out of tune with modern developments in the digital sector within a short period of time?How can laws keep abreast with business developments in the industries?The obligations in the DMA are modeled according to known business practices most of which found the attention of regulators (Caffarra & Scott Morton, 2021).But what if the digital gatekeepers come up with new practices that may harm contestability or be unfair toward business users or consumers?
Traditionally, there are three answers to this: The first solution would be a new legislative initiative with a reform of the rules.This, however, takes very long in the usual EU proceedings and is hardly conceivable.
The second solution would be to rely on general clauses that allow to capture different behavior due to its general wording.However, this is exactly the approach taken by competition law that proved to be unsuccessful in the digital sphere.
Third, the law could simply run out of relevance and be forgotten.
The DMA offers a new idea to address this difficult problem of updating rules: According to Art. 12, the Commission may pass delegated acts as a quicker tool of legislation to update existing obligations and to make them applicable to other services.According to Art. 19 (3), it has the power to start market investigations into new services and new practices and advise the legislators to pass new amendments to the DMA after the investigation.With these instruments, the Commission may speed up the necessary legislative process at the EU level.Delegated acts according to Art. 290 TFEU would put the Commission itself into the position of amending the DMA in a faster way.This is an important measure to avoid that the relevance of the DMA vanishes when new practices are established (cf.Larouche & de Streel, 2021).
The choice in favor of some flexibility in this amendment process comes at the cost of a power shift: Once the DMA is in force, it is no longer the Member States or the European legislature that take key decisions on new obligations in a formal procedure.The Commission, the executive body itself, determines practices that lead to unfairness or incontestability.The executive branch is further empowered while Parliament and Member States lose some influence.

Conclusion
The European Commission decided in favor of a regulatory approach when proposing the DMA.The DMA represents the lessons learned from the failures of antitrust enforcement in the digital sector.The application of competition law achieved too little too late vis-à-vis the digital gatekeepers.
The regulatory approach can be distinguished from competition law through its focus on ex ante rules (not standards that are enforced ex post).It is targeted at specific companies only and includes a variety of goals that differ from competition law goals.By shifting from general competition law to a regulation for digital gatekeepers, the European Commission recognizes that gatekeepers entertain and command "core platform services" that look like an infrastructure of the Internet.Their enormous relevance for more and more sectors in business and in society is acknowledged.
The decision for a regulatory approach is not the end of substantive, institutional, or procedural debates.In contrast, hard regulatory choices have to be taken, e.g., regarding the number of gatekeepers to be addressed by the DMA or the focus on business users instead of consumers.The level of state intervention in Big Tech will largely depend on enforcementthe institutional design of the sanctioning regime will decide the question whether the DMA becomes a powerful tool or not.The DMA is a model in speeding up proceedings and in providing an option for updating obligations, yet this may harm the rights of others.
In this text, some of the distinctions and choices were explained that make a difference when regulating the digital world.As is often the case for the law, the devil is in the details.
One element of the DMA is that it does not rely on economic evidence and even economic concepts to the same degree as twenty-first century competition law.For the relationship of law and economics, the DMA is a signal that the role of economic reasoning in individual case assessment may not fit in the current legal framework.Economics has not lost its power to inform law-making on the legislative stage, however (Fletcher et al., 2023).This means that political decisions need to be taken again and again: What is the intended level of state intervention?What are the expectations of society vis-à-vis digital gatekeepers?The DMA is the beginning of answering these questionsbut not the end of the debate.