The purpose of this article is to analyse whether the instruments available to combat the abuse of substantial market power are appropriate for big techs. To do so, I dissect Article 102 of the TFEU and Section 2 of the Sherman Act in the light of three paradigmatic cases: Google Search Shopping, Google Android, and Epic Games v. Apple. After introducing the legal rules and cases, I examine the problems created by the concepts of “position of dominance” and “monopolisation”. Next, I explain whether Google's and Apple's practices have been considered “abusive” or “improper” and why. Then, I briefly describe the enforcement procedures and the available remedies and highlight the main issues. Later, I present the existing legislative initiatives in the EU and USA aimed at limiting the power of digital giants. I end with the mandatory conclusions.
CEUR Workshop Proceedings
But there is a dark side: market congestion [
In this paper, I focus on the abuse of economic power as it has been the preferred weapon -but also highly controversial- to fight big techs. I look at the two sides of the Atlantic Ocean to assess its efectiveness because, despite the diferences in the literalness, they have similar rules. To analyze them I use three cases as a guide.
The two key rules are Article 102 of the Treaty on the Functioning of the European Union
(TFEU) and Section 2 of the Sherman Act (15 U.S.C. § 2). The former forbids the abuse by one or
more undertakings of a dominant position within the internal market if it afects or may afect
trade between Member States. Its second paragraph adds four practices that are considered
abusive. Article 102 can be broken down into three elements. The first one is the position of
dominance; the second, the abuse, and the third one the impact on intracommunity cross-border
trade. Section 2 of the Sherman Act forbids monopolization, the attempt to monopolize and
agreements to monopolize the trade between several States or foreign nations. Three behaviours
are prohibited: monopolization, attempt to monopolize, and conspiracy to monopolize. They
can be sanctioned with a fine of up to 100 million US dollars for corporations or 1 million US
dollars for other persons and, as it can be considered a crime of felony, also with up to ten
years in prison. The typified behaviours are composed of several elements. The monopolization
requires i) the possession of monopoly power in a relevant market, and ii) having wilfully
acquired or maintained it through improper means. Regarding the attempt to monopolize,
courts distinguish three elements: i) predatory or anticompetitive conduct, ii) the intent to
monopolize, and iii) a high possibility to achieve it [
Both legal rules deserve three considerations. Firstly, they are extremely cryptic as none
of them defines the key elements. Article 102 does not establish the meanings of dominant
position or abuse, and the practices listed in its second paragraph are considered just examples
of the last factor [
Lastly, both legal rules are designed to be applied to cross-border cases. Section 2 refers
to trade or commerce among several States or with foreign nations. Article 102 includes two
international intra-community conditions: on the one hand, the undertaking under scrutiny
shall have a dominant position in the internal market; on the other, the abuse may afect trade
between Member States. Doubts arise about international extra-community cases; for instance,
when the enterprise that has abused its dominant position does not have the nationality of
one of the twenty-seven Member States, as it normally happens with big techs. The case law
of the EU Courts favours the extraterritorial application of Article 102, but it requires that
the exploitation of the dominant position has efects in the Single Market [
This research is mainly based on three cases: two European and one American. The first deals
with price comparison services (Google Search Shopping) [
The American case is a Judgment of the Northern District Court of California and confronts Epic Games, Inc, an American videogame and software developer, with Apple Inc (Case No. 4:20-v-05640-YGR).2 The plaintif alleged that the company founded by Steve Jobs had violated section 2 of the Sherman Act, among other laws, by demanding a commission of 30% of the revenue generated by the software posted on the app store, for not allowing to create apps markets inside the App Store, for imposing the exclusive use of its payment system (“In-App Purchase” or IAP) and for not allowing consumers to be informed of alternatives to the App Store and to the IAP. The defendant raised a counterclaim alleging that the claimant had unfulfilled their contract and asked for the payment of the owed sums. Judge Yvonne Gonzalez Rogers partially dismissed the plaintif’s complaint, stating that Apple neither had monopoly power nor had abused it. Nonetheless, it held that Apple’s anti-steering clauses breached California’s Unfair Competition Law, so she ordered this company to stop enforcing them. Nonetheless, the injunction became inefective because the Court of Appeal of California upheld the request of the defendant to stay the Order. On the other hand, Judge Gonzalez Roger agreed with Cupertino’s company and ordered Epic Games to pay the commission that it would have owed if it had fulfilled the contract. The plaintif appealed against this decision.
The reasons to compare these cases are three. The most important is that the courts deal with the core elements of the abuse of dominant position and of monopolization, so they show which are the dificulties of applying Article 102 and Section 2 in digital markets. Besides, the main documents are publicly available. Finally, they afect such iconic big techs as Google and Apple. Nonetheless, it is worth stressing that there is a very important diference between the European and American cases. The formers are the result of public enforcement: the European Commission investigated the case for several years and finally issued a decision forbidding the behaviour and imposing a penalty. The General Court is deciding on the appeal of the sanctioned undertaking. The American case is a private enforcement procedure: a customer of Apple sued it because the contractual terms were abusive and the big tech counterclaimed due to the breach of the contract. There was not an administrative phase before: the American antitrust agencies did not investigate the case before or ruled on the behaviour of Apple.
In Judgment Hofmann-La Roche v. Commission, the Court of Justice of the European Union defined the dominant position as “…a position of economic strength enjoyed by an undertaking which enables it to prevent efective competition being maintained on the relevant market by 2The documentation is epic-games-inc-v-apple-inc/.
available
at
https://cand.uscourts.gov/cases-e-filing/cases-of-interest/
afording it the power to behave to an appreciable extent independently of its competitors,
its customers and ultimately of the consumers…” [
Section 2 of the Sherman Act requires monopoly power, which has been defined by the
US Supreme Court as the power to control prices or unreasonably restrict competition in a
given market on a lasting basis [
It is essential to define the relevant market to assess whether an undertaking enjoys a
dominant position or a monopoly power. None of both legal systems have legal rules on this
subject, although the European Commission published a Notice in 1997 explaining its practice
[
Two dimensions compose the relevant market: the product and the geographical ones. The
former comprises the goods or services produced by the investigated undertaking and all the
others that consumers consider interchangeable according to their features, price, use, and
qualities. The geographical market is the area where the dominant enterprise operates [
Defining the relevant market is a very complex task, not only because it requires knowledge of
economics, but also because there is a myriad of diferent products, various levels of substitution
and companies ofering very diferent goods and services [
The similarities that exist regarding the definitions of dominant position, monopolization, and relevant market fade away in the cases under scrutiny. In my humble opinion, it is not a problem of diferent outcomes -while the European Commission and the General Court held that Google had a dominant position, Judge Yvonne Gonzalez Rogers refused that Apple had a monopoly power-, but of the definition of the relevant market; particularly the dificulty of applying long-established criteria to digital markets and digital giants. For instance, while the European institutions asserted that operating systems can be the object of a relevant market, the Northern District Court of California refuses it.
The dominant position of Google seemed to raise no doubts in Google Search Shopping, as the General Court asserted that it did not dispute the considerations of the Commission in this regard. By contrast, the definition of the relevant market sparked controversy in the case 3https://competition-policy.ec.europa.eu/public-consultations/2022-market-definition-notice_en. Google Android, since Google argued that Apple exerted competitive pressure regarding the smart mobile operating system. It had not behaved autonomously but reacted to the business strategy of Tim Cook’s company. The General Court did not accept this argument and upheld the European Commission’s analysis because of three reasons. Firstly, Apple did not ofer its operating system (iOS) to original equipment manufacturers: they could obtain a license and install Google Android on the devices they manufactured and marketed, but not iOS. Secondly, Google’s investment in the development of its operating system could not be attributed to Apple’s competitive pressure, but to the belief that the mobile internet was the future. As the company directed by Sundar Pichai thought that users would prefer it to the computer one, it developed an operating system for the mobile device that allowed it to keep controlling the data source that was essential to its business model. Thirdly, there was no parallelism between the updates of both operating systems. Since Google achieved a dominant position, the Android versions lasted longer without changes.
The decision of Judge Yvonne Gonzalez Rogers was diferent in case Epic Games v. Apple: She rejected that operating systems could be the object of a market and denied that Apple had monopoly power. The key laid in the definition of the relevant market. The claimant identified a foremarket, which targeted the operating system, and two aftermarkets: one for iOS app distribution and another for iOS app payments. The defendant proposed defining a broader market, encompassing transactions of all digital games. The judge rejected both proposals and was very critical with the former one. She stated that there was no market for operating systems, but for cell phones, which are something more than just an operating system. She put apps in the spotlight and distinguished two main markets: the apps for games and the rest of the apps. The case referred to the former ones, which comprised four submarkets: apps for mobile devices videogames, for personal computer game stores, for console games stores, and for cloud gaming-based streaming platforms. Again, the first one was the relevant: the dispute between Epic Games and Apple took place in the mobile gaming apps market.
This decision leads the Judge to reject that Apple had monopoly power. After defining it as the capacity to control prices or exclude competition, she asserted that it requires at least 75% of the market share. Apple did not reach this figure: according to Yvonne Gonzalez Rogers, it just had a 57% of the mobile gaming apps market. And she added three other reasons to support her appreciation: i) there had been no decrease in transactions in this market -that is what uses to happen when an undertaking controls it-;4 ii) competitors were lurking -thus, Apple could not behave independently; and iii) the actor did not focus its attention on the relevant market. Nonetheless, she acknowledged that the commission charged by Apple was very high, that the anti-steering clause guaranteed Apple high-profit margins, and that there were high barriers to entry because potential competitors should attract developers and users at the same time. Despite the lack of monopoly power, the Judge continued assessing whether the defendant’s behavior could be considered abusive. The reason was that below the monopoly power quota, section 1 of the Sherman Act could still be applied. In my view, maybe it could have been necessary to assess whether Apple could have been attempting to monopolize too. 4I humbly do not share the argument because of the specificities of digital markets. As returns of scale are extreme, digital giants do not need to reduce their outcome to save costs or increase profits. This is why there has not been the decrease in transactions that is typical of markets controlled by dominant undertakings.
It is not enough with huge power (call it dominant position or monopoly): Article 102 and
Section 2 require that the behaviour of the investigated company is unlawful. The problem
is the lack of definition: none of these provisions describe what is abuse or which means are
improper. Thus, the case-law becomes essential. It is worth adding that an expert stresses
that the more important practices in digital markets are self-preference, predatory prices, and
exclusionary practices [
Some scholars assert that the European institutions do not use a unique single definition of
“abuse”, but several [
For the same reason, both the Commission and the Court of Justice admit that there may be
objective justifications for allegedly abusive conducts [
European institutions impose a special responsibility on dominant undertakings: they should
not prevent the existence of efective competition in the market [
Despite the diferences in the wording of Article 102 and Section 2, there are some parallels that
make their comparison easier and allow drawing conclusions. Firstly, the Sherman Act does
not forbid monopoly power per se. Although it is associated with detrimental consequences
for consumers (lower output, higher prices, and less innovation), the DoJ explains that it is
not necessarily so, because monopolies are normally the outcome of eficiency. “An eficient
ifrm may capture unsatisfied customers from an ineficient rival and this is precisely the sort of
competition that promotes the consumer interest that the Sherman Act aims to foster” [
The problem lies in qualifying a conduct as improper. And here it is another similarity with
the EU, as American courts hold a doctrine like “competition on merits”. The US Supreme Court
comments that improper means are those diferent from ofering a superior product or business
acumen [
It is dificult to decide whether the conduct of a monopolist breaches Section 2 because most market behaviours generate eficiencies and, at the same time, exclude competitors. Indeed, practices that only generate eficiencies or that are pure predatory or exclusionary are exceptions. Therefore, per se solutions are not common and the courts must carry out assessments that are highly complex and consume a lot of time and resources. To avoid them and increase legal certainty the DoJ asked for tests and proof allocation rules that eased the analysis. Unfortunately, the dificulty is so high that the Supreme Court has established specific tests only for predatory pricing and bidding. In the absence of a special rule regarding the valuation of a behaviour, courts follow a three steps procedure. The assessment starts with the duty of the plaintif to demonstrate that the monopolist harmed the competitive process and consumers. If this requirement is met, it is the turn of the monopolist to demonstrate the objective justifications for its behaviour. Whether it succeeds, it has neutralized the accusation of violating Section 2. Nonetheless, the plaintif still has the chance to show that the negative efects of the behaviour outweigh the positive ones or that there were alternative behaviours that were less detrimental. Lastly, the DoJ explains that this system imposes a special responsibility to monopolist enterprises as a certain practice can be held illegal just if carried on by a company that has monopoly power, but be considered legal if it does not have such power.
It is worth adding that Judge Yvonne Gonzalez Rogers made a particular enforcement of the antitrust rules on Epic Games v. Apple since she applied Sections 1 and 2 of the Sherman Act together. She grounded this decision in the fact that it is quite common in case law and that Apple had no monopoly power. However, she found that one of the essential elements of Section 1 was missing: it could not be considered that there had been an agreement because Apple imposed the contractual terms; it did not seek the consensus of its partners. However, as this solution could clash with the objectives of Antitrust Law, she continued with the analysis. Secondly, although she tried to refer to each behaviour separately, she often carried on a joint assessment or she referred the analysis of a particular practice to the appraisal done on the others. The cause is the application of the rule of reason. As the practices under scrutiny were not per se forbidden, Yvonne Gonzalez Rogers applied the three-step analysis characteristic of the rule of reason. Lastly, she did not limit herself to assessing Apple’s behaviour according to Sections 1 and 2 of the Sherman Act. On the one hand, she also applied State Laws, especially the California’s Cartwright Act and the California’s Unfair Competition Law. On the other, she also ruled on Apple’s counterclaim stating that Epic Games breached the license contract (Developer Product Licensing Agreement).
Self-preferencing is a common practice among big techs and consists in promoting its own products before the ones of competitors. The case Google Search Shopping represents a good example as the European Commission held that it constituted an abuse of a dominant position, and the General Court upheld its decision. It started in 2002 as Alphabet’s subsidiary began providing its own price comparison service in the framework of its general search services. In 2005 it separated them, and in 2007 renamed the former one; but it did not have success until it changed the algorithm, which applied only to rival services, and improved its service with eye-catching techniques. Since that moment, Google price comparison results appeared at the top of its general search services page, provided with rich graphical features and images. The results of competitors appeared lower on Google’s general search site, without a rich format, images, and additional information on the products. According to the European competition authority, the outcome was an increase in Google’s trafic to its service and a reduction in the rivals.
Among the diferent issues that the General Court had to deal with, we focus on three. Firstly,
it shared the view of the European Commission that Sundar Pichai’s company had abused its
dominant position. It qualified the practice under scrutiny as discrimination, which is forbidden
by letter c) of Article 102 TFEU [
In summary, regarding the qualification of self-preferencing, the General Court stressed that Google’s conduct was forbidden not only because it promoted its service, but also because it had demoted the rivals. Thus, the abuse comes from the combination of both practices: the promotion of its specialized comparing system and, at the same time, the demotion on its general results page of competing comparison shopping services. It also added that the diference in treatment was not the consequence of the diferent relevance claimed in the light of comparison criteria. Although it seems that dominant undertakings have a duty to equal treatment, it is worth highlighting that the Court took into consideration the extraordinary power of Google. It asserted that it had a superdominant position that made it a gatekeeper, and that it was protected by high barriers of entrance. Then, the doubt remains: does a dominant enterprise that is not a gatekeeper have the duty to provide equal treatment? 3.3.2. Tying The second behaviour that deserves analysis is tying, as large undertakings appeal to this practice to leverage their power to new markets. It consists of linking the product for which they wish to acquire greater power to the product over which they already have a position of dominance, so customers can only acquire the last one if they also buy the former. Google used this practice in the framework of the Mobile Application Distribution Agreements (MADAs): original equipment manufacturers that desired a licence of the Google’s app store (the famous Play Store) and/or of the general search app had to pre-install the Chrome browser in the devices they produce. Thanks to this arrangement, the company directed by Sundar Pichai ensured that smartphones and other mobile devices running the Android operating system contained its two main sources of data: the search app and the Chrome browser.
The Commission held that these practices breached Article 102 because they gave the
dominant a significant advantage over its competitors, who could not ofset. Besides, they helped
Google to maintain its market power, they deterred innovation and damaged, directly or
indirectly, consumers. The General Court upheld the Decision in the Judgment of 14 September
2022. It explained that, as it stressed in case Microsoft vs Commission, tying practices are illegal
if five requirements are met: i) the tying and tied products are two separate products; ii) the
undertaking concerned is dominant in the market for the tying product; iii) it does not give
customers a choice to obtain the tying product without the tied product; iv) the practice in
question ‘forecloses competition’; and v) it is not objectively justified [
In Google Android doubts afected the fourth requirement. The Commission and Google discussed whether it was met; particularly, whether it was necessary to prove that a restriction of competition had efectively taken place or whether it was enough with the evidence of its capacity to restrict competition. The General Court supported the last interpretation and held that the Commission had provided suficient evidence. Although, in theory, original equipment manufacturers could pre-install competing apps and services, in practice they did not do so because of the tying agreements. Therefore, it concluded that the combined efects of the dominant’s agreements and practices hindered the pre-installation of competing general search apps and browsers.
Tying also raised a concern in Epic Games v. Apple, although without major treatment.
The claimant argued that the defendant tied its payment system to the distribution of apps:
Apple forced developers to accept the In-App Purchase (IAP) if they wanted to use its store to
commercialize products. The Judge of the Northern District Court of California referred to case
Jeferson Parish [
The General Court dealt with two kinds of exclusivity practices in case Google Android and held diferent solutions. The first practice referred to Android forks, which are variations of the operating system developed by third parties. The problem laid with the variations that Google had not qualified as compatible, because it did not allow its services and apps to be used with them. Indeed, it conditioned the licence of the Play Store and of the search app to accepting anti-fragmentation obligations -id est, not using versions of the operating system that did not pass the compatibility test. The European Commission held that it was an abuse of its dominant position because these variations competed with Android and, hence, they were a menace for Google. By prohibiting their installation, the subsidiary of Alphabet hindered the development of the variations of its operating system, reinforced its dominant position in national markets for general search service, deterred innovation, and damaged consumers.
On appeal, the dominant undertaking argued that the anti-fragmentation agreements were necessary to preserve the Android ecosystem as an open-source model, from which consumers benefited. The General Court rejected this argument based on two premises. The first one was that non-compatible variations were licensing operating systems and, therefore, rivals of oficial Android. The second argument was that anti-fragmentation agreements forced original equipment manufacturers to refuse them, so developers were discouraged to make products that competed with the oficial version. The consequence was upholding the appealed Decision.
The second exclusivity practice referred to the revenue share agreements, by which Google granted original equipment manufacturers and mobile network operators a percentage of its advertising profits if they did not pre-install competing general search services in the devices they produced or commercialized. The European Commission considered these agreements were “exclusivity payments” and held them abusive because they reduced the incentives to pre-install competing services, they created or increased barriers to access to markets for general search services, they deterred innovation and they lacked objective justifications.
The General Court agreed that revenue share agreements were exclusionary practices similar to loyal rebates and they should be assessed according to the “As Eficient Competitor” test. 5 Consequently, the European competition authority should have demonstrated that the behaviour under scrutiny excluded a competitor which hypothetically was equally eficient, charged customers the same prices as Google, and faced the same costs. In this case, the European judiciary accepted the appellant’s argument and annulled the Decision on this subject. On the one hand, it criticized the competition authority because the market share afected by this practice was not relevant. On the other, the European Commission erred in applying the AEC test.
In Epic Games vs Apple the Judge examined three exclusivity practices: i) the prohibition to create app markets inside the App Store, ii) the obligation to use Apple’s payment system (In-App Purchase), and iii) the anti-steering clause. They prevented competition, so they were detrimental to consumers and developers since there was no competitive pressure on prices -this was why Apple had maintained over the years a commission of 30%, which was very high-. Nevertheless, the Judge held that the procompetitive efects outweighed the anticompetitive ones. She attached huge importance to the fact that these restrictions allowed Apple to build a secure and protected digital environment. The iOS and the Apple Store were excellent at defending against malware, ofensive content, and fraud, as well as preserving privacy. Consumers and developers benefited from this “walled garden”: as the former ones felt safe, they relied on the apps commercialized in the oficial store. Secondly, the Judge accepted the defendant’s argument that Apple’s policy fostered interbrand competition: while Apple ofered a closed ecosystem, Google’s one was open. Hence, she stated: “That distinction ultimately increases consumer choice by allowing users who value open distribution to purchase Android devices, while those who value security and the protection of a ‘walled garden’ to purchase iOS devices” (page 146). She also underlined that there had not been a decrease in the number of apps in the Apple Store; indeed they had increased a 1200 5Paragraphs 640 f. The AEC test has raised several problems in cases related with technologic markets; for instance in Intel. The EU Court of Justice annulled the Judgment of the General Court because it did not assess all of Intel’s allegations regarding that test. It explained that loyalty rebates could be considered unlawful if the undertaking had a dominant position. But, if the investigated enterprise proved that it was not capable of restricting competition or that its behaviour did not foreclose the market, the Commission had to assess the possible existence of a strategy aiming to exclude competitors that were at least as eficient as the dominant firm. Although the Commission held that the rebates meant per se an abuse, it applied the AEC test. Hence, the General Court had to examine “whether the Commission had carried out the AEC test in accordance with the applicable rules and without making any errors”. As it did not do so and it also did not carry out the alternative calculations proposed by Intel, the judgment was not in accordance with law. Vide the Judgment of 6 September 2017, cit. Four and a half years letter, the General Court decided again on the Decision C(2009) 3726 final and this time annulled it on the grounds that the application of the AEC test was flawed.
The outcome was diferent under the California’s Unfair Competition Law. Epic Games not only claimed that Apple had breached the Sherman Act but also some State Laws, and it succeeded in proving that the anti-steering clause represented an unfair practice. Section 17200 of the Business and Professional Code of California forbids “…any unlawful, unfair or fraudulent business or practice”. Judge Gonzalez Rogers held that Apple had not committed an unlawful act -she had previously rejected that it had breached Sections 1 and 2 of the Sherman Act-, but its behaviour was unfair as it collided with the policy or spirit of Antitrust Laws. Her decision was grounded on two tests. According to the tethering one, the Judge of the Northern District Court of California held that the anti-steering clause collided with the legislative policy, because it reduced the information available to consumers and developers, it decreased innovation, and it made it possible for Apple to maintain a too high remuneration. Regarding the balancing test, she rejected the defendant’s argument of the similarity with brick-and-mortar stores as digital platforms are diferent. Apple’s one was a black box: it imposed silence to control the information and to prevent consumers from getting informed as well as the substitutability between platforms. Therefore, the Judge ordered an injunction forbidding Apple to prohibit developers to include in their app functions that direct consumers to rival purchasing mechanisms or to communicate with consumers.
In Epic Games v. Apple, the Judge assessed whether a commission of 30% of the revenues generated in the app store was predatory. The percentage was quite high and lacked justification, as Apple had established it accidentally when started commercializing the app store. It harmed consumers and developers because the prices increased, and fewer resources could be applied to innovation. It had lasted for a long time without variation, despite the complaints of developers and the pressure of regulators. Notwithstanding, the Judge of the Northern District of California ruled in favour of the defendant, on three grounds. First and most important, it was a reward for Apple’s intellectual property -although Yvonne Gonzalez Rogers recognised that the defendant had not proved the suitability of the 30% commission, which seemed too high. Secondly, it was a remuneration for giving access to Apple’s audience, which had a high value. Lastly, the Judge accepted the defendant’s argument that other platforms applied similar commissions. However, she acknowledged rivals used to negotiate downward while Apple did not reduce its remuneration.
The cases under scrutiny illustrate the two enforcement possibilities that exist, with their lights and shadows. The private enforcement consists of litigation in the courts against the undertaking (or undertakings) that has (have) restricted competition. The plaintif can be a particular individual (either a consumer or a corporation) or a public agency (normally, a competition authority) looking for the cessation of the illegal behaviour or the compensation of damages. It usually has the advantage of speed, as Epic Games v. Apple demonstrates. The plaintif went to court in the late summer of 2020 and about a year later the judge issued her decision. The problem is that procedure, and especially evidence, is enormously demanding; so much so, that it seems unlikely for a lawsuit brought by a consumer or a small or medium enterprise can succeed. For instance, the definition of the relevant market is extremely dificult. In Epic Games v. Apple, Judge Yvonne Gonzalez Reyes held a definition diferent from the ones proposed by the parties, and this disparity was key to the result since Epic Games could not prove that Apple had a monopoly power or used improper means.
Public enforcement is an administrative procedure carried out by a competition authority
against one or several undertakings to put an end to an infringement, restore competition,
and deter the commission of future violations. It may also produce a judicial process if the
sanctioned enterprise is not satisfied with the administrative decision and appeals to the courts.
Moreover, it can also be followed by a private enforcement procedure if victims decide to claim
damages. It has the advantage that an administrative body, that uses to have great investigating
powers, prosecutes the restrictive practice. However, the costs of the enforcement proceedings,
especially when dealing with big undertakings, can be extraordinarily high and the resources
of competition authorities are always finite [
The remedies provided by Section 2 of the Sherman Act and Article 102 of the TFEU are diferent.
The former relies on criminal sanctions as the monopolization is considered a felony. The court
can order imprisonment for less than 10 years and a fine of 100.000.000,- US dollars in the case of
corporations or 1.000.000,- US dollars for other types of persons. Nonetheless, civil enforcement
is preferred, and the typical remedies are available to plaintifs [
Article 102 of the TFEU does not establish any remedy, so it is necessary to appeal to the
rule that develops it: Regulation 1/2003. Article 7(1) gives the Commission the power to adopt
6Advanced Micro Devices Inc (AMD) complained to the European Commission about Intel’s practices in 2000. The
European competition authority oficially initiated the investigation in 2004 and in 2009 adopted a Decision holding
that the investigated enterprise had abused its dominant position. Five years later, the General Court upheld its
decision, but in 2017 the Court of Justice annulled that Judgment based on errors in the assessment of evidence.
The case returned to the General Court which in 2022 ruled in favour of Intel and annulled the decision of the
Commission, which has appealed against the Judgment.
any measure, proportional to the infringement, to eliminate the restriction of competition.
Remedies can be structural or behavioural, but the last ones are preferred -the former can
only be adopted when behavioural measures are not so efective or when they would be more
onerous for the sanctioned undertaking. The practice of the European competition authority
reflects this preference, as it does not appeal to structural measures. The reasons could be
that splitting up undertakings is very dificult as it means allocating staf and assets that serve
the firm as a whole and it can imply eliminating eficiencies that have been very dificult to
achieve [
The analysis of the cases under consideration reflects the above explanation. Epic Games asked for the compensation of damages and an injunction to finish the unlawful practice. As the Judge mostly dismissed the action, she only ordered Apple not to use the anti-steering clause. Nonetheless, this remedy became inefective because the defendant’s appeal succeeded and the California Court of Appeals postponed the injunction indefinitely. On the other hand, Apple’s counterclaim prospered: Judge Gonzalez Rogers upheld it, declared that Epic Games breached the contract, and ordered it to pay the amounts that would have accrued under the contract.
In the three Google Decisions, the Commission ordered to cease the illegal conduct, not to engage in the same or similar behaviors in the future and imposed a total fine of around eight billion euros. The efectiveness of these measures raises doubts. Although this quantity seems very high to the average human, it is debatable whether they have a disincentive efect on an undertaking as powerful as Google. On the other hand, the behavioral obligations are aimed to put an end to the competition restrictions, but they are not restoring the original conditions of the market - the competition that existed before the forbidden conduct had efects. In my humble opinion, the Commission should have adopted more resolute remedies, like the ones it established in the Microsoft case when it ordered the dominant to disclose the necessary information for developers to make the software compatible with Windows and to ofer a version of the operating system without the Windows Media Player.
The efectiveness of current legislation raises concern. The International Monetary Fund and
some economists warn about the menace that big techs represent, so they call for keeping
markets open and competitive, avoiding high access barriers, and impeding the exercise of
market power [
The European Union has echoed these requests and has approved a regulation that imposes ex ante remedies to gatekeepers: the so-called Digital Markets Act.7 It aims to achieve contestable and fair competition in the digital sectors where gatekeepers are present (Article 1(1). The scope of application is structured around the core platform services ofered by gatekeepers to business or end users established or located in the European Union. The two main elements are core platform services and gatekeepers. Article 2(2) defines the former by listing the services that it comprises: a) online intermediation services, b) online search engines, c) online social networking services, d) video-sharing platform services, e) number-independent interpersonal communications services, f) operating systems, g) web browsers, h) virtual assistants, i) cloud computing services, and j) online advertising services. The list can be increased if the Commission considers that there are new practices that are unfair, that limit the contestability in markets, and which are not properly ruled by the Act (Article 19).
Gatekeepers are undertakings that provide core platform services, have a huge economic power and control access to digital markets and how competition develops in them. Article 3(1) defines them by three features: (i) they have a strong market influence, (ii) they provide core platform services that constitute a gateway for professional users to reach end users, and (iii) they have an entrenched and durable position. As these qualities are quite abstract and their application could raise doubts, paragraph 2 establishes the presumption that they are present when certain economic thresholds are exceeded. Nonetheless, it is a prima facie presumption that can be challenged. It is important to stress two issues. On the one hand, only the enterprises that have been qualified as “gatekeepers” by the European Commission are subject to the obligations established by the Digitals Markets Act concerning the core platform services assigned. In other words, it is necessary a specific qualification act for an enterprise to have the status quo and duties of gatekeepers. On the other hand, this qualification does not depend on the possession of a dominant position. The European Commission will establish it based on the economic thresholds provided by Article 3(2). Since it is not applying Article 102, it does not need to define the relevant market and determine whether the investigated undertaking has enough market power to operate independently within it.
Undertakings qualified as gatekeepers have a special responsibility, as they need to comply with obligations and prohibitions to guarantee open and fair competition in digital markets. They are provided by Chapter III and referred to each of the core platform services established by the Commission regarding a singular undertaking. Articles 5, 6, and 7 set up the main ones, most of which come from previous European competition cases, like the ones we have studied. For instance, the gatekeeper shall not treat more favourably, in raking, indexing and crawling, its services and products rather the similar ones of third parties -besides it shall apply transparent, fair, and non-discriminatory conditions to such rankings-; it shall allow business users to communicate and promote ofers to end users through the channels they wish; it shall not impose the use of identification services, web browser engines or payment services; it 7Regulation (UE) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828. Although it entered into force on November 1st, 2022, its legal eficiency is delayed until May 2nd, 2023 or June 25th, 2023. Cfr. Article 54. shall allow and technically enable the use of third-party software applications and software application stores using its operating system; and it shall allow providers of services and of hardware efective interoperability with their products.
Articles 8 and 12 complement them. The former provides for the possibility for the Commission to establish measures to ensure that a given undertaking complies with its obligations and that these are efective in preserving competition in the market. Article 12 empowers it to update the list of obligations. Due to the risks that mergers create, Article 14 requires gatekeepers to provide information on the concentrations they wish to carry out and the techniques used to draw up the consumer profiles they apply to their basic platform services. All these obligations are self-executing: gatekeepers should fulfil them without the necessity of a request from a public body.
As it could not be otherwise, the European Commission is vested with powers of investigation and punishment. Chapter IV establishes three possible market investigations that are addressed to the designation of an undertaking as a gatekeeper, to assess whether an enterprise has systematically unfulfilled its duties, and to decide the necessity for adding new elements to the basic platform services or for identifying new pernicious practices to combat. Secondly, the powers of investigation that the Digital Markets Act grants the Commission are similar to those provided for in Regulations 1/2003 and 139/2004. It is worth commenting that it expressly adds the possibility of analysing the algorithms used by the gatekeeper and of appointing “verification agents” to make sure that an undertaking complies with its obligations. On the other hand, some defence rights are explicitly recognized to gatekeepers, such as the right to be heard and the right of access to the file. Regarding the sanctions, they are like the Competition Law ones. For instance, Article 30(1) establishes the possibility to impose a fine of up to 10
The assessment of the Digital Markets Act is positive because it will improve competition in digital markets. One example is the rumour that Apple will allow third-party app stores on iPhones and iPads in order to accommodate to the new legislation. Nonetheless, some issues raise concern. Firstly, its complexity: it is a dificult rule to interpret and apply. Indeed, it is mainly addressed to big enterprises that have enough economic and legal resources to deal with such a challenging text. But it is worth remembering that it is a Regulation, so it can be directly enforced by national courts (for instance, in the event of actions for damages against gatekeepers that had unfulfilled their obligations). Secondly, the only competent authority to enforce the Act is the European Commission, which workload will be considerably increased. Although the overseeing of its compliance does not lay on the DG Comp but rather on the DG Connect, it remains to be seen whether the Commission has suficient resources to carry out the supervision. Thirdly, it will be necessary to draw a line separating regulation and defence of competition. The reason is that Article 1(5) forbids Member States to impose other obligations on gatekeepers aimed at ensuring contestable and fair markets. But Article 1(6) gives preference to Competition Law, both EU and domestic, over the Digital Markets Act. Thus, a problem appears: Member States cannot establish new obligations on gatekeepers, but nothing forbids them to approve new antitrust rules that can be applied to them. Lastly, Digital Markets Act provides public measures, but it does not rule on private enforcement. The question arises as to which rules will become applicable if gatekeeper breaches cause harm to users and they claim compensation.
A similar legislative proposal exists in the United States: the American Innovation and Choice Online Act (AICOA - S. 2992). The bill was introduced in Congress by David Ciccilline on June 11th, 2021 and Senators Amy Klobuchar and Charles Grassley released an updated version in May 2022. Although it has not yet been approved, I will try to briefly draw a parallel with the EU Digital Markets Act.
Regarding the purpose, the AICOA seems less ambitious, because it declares being focused on fighting “certain discriminatory conduct(s) by covered platforms”. By contrast, the Digital Markets Act aims to achieve a market in which fair and contestable competition reigns. The central element of the AICOA’s scope of applications is the “covered platform”. The bill defines it in Section 2(5) in terms of several parameters, among which are its monthly active users, the annual sales volume, and the market capitalization. These parameters are closely reminiscent of the quantitative criteria used by the Digital Markets Act to presume that a company meets the qualitative requirements to be considered a gatekeeper. Therefore, both figures can be considered equivalent, especially if we also take into consideration that only the undertakings appointed by the Federal Trade Commission and the Department of Justice are qualified as “covered platforms”. However, the EU Regulation stresses the role of gatekeepers as a gateway for business users to reach end users. On the other hand, there is no reference in AICOA to core platform services, but all the products ofered by covered platforms are subject to it.
Section 3 lists the behaviours forbidden to covered platforms. Among them, the following stand out: self-preferencing own products over those of other business users of the platform, limiting the ability of business users to compete with the covered platform, discriminating among the business users, restricting the interoperability, conditioning the access to the covered platform on the acquisition of products or services which are not intrinsic to the last one, using non-public data obtained from the platform by the interaction with the goods or services of its business users, and impeding or hindering uninstalling software applications that were preinstalled on the covered platforms. Section 3 lists some defences that exonerate the platform from liability, like protecting the safety, user privacy, or the security of non-public data, as well as maintaining or substantially enhancing the core functionality of the platform.
The bill also establishes some remedies; for instance, a civil penalty of up to 15% of the total United States revenue or the injunctions necessary to prevent, restraint, or prohibit violations of the Law. The enforcement will be the responsibility of the Federal Trade Commission, the Attorney General, and the attorney general of the States.
The two main legal provisions to fight against the abuses of big techs are Article 102 of the TFEU and Section 2 of the Sherman Act. Their efectiveness raises doubts. Firstly, they were created and developed on non-digital markets. Secondly, the dominant position and the monopoly power depend on the definition of the relevant market. The criteria traditionally used to define it are not suitable for digital markets because they do not take into considerations elements that are essential in digital markets. Thirdly, the high degree of abstraction of the “competition on merits” paradigm deprives the qualification of a behaviour as abusive or improper of legal certainty. Fourthly, the enforcement of Article 102 and Sections 2 face the problems of obtaining the necessary evidence, of errors in the procedure and, in the case of public enforcement, slowness. Lastly, it is doubtful that the remedies applied by the European Commission succeed in restoring eficient competition and deterring further abuses. Because of these reasons, the Digital Markets Act should be welcomed as it provides an ex ante control and it may alleviate the Commission’s burden of proof, since behaviours that are contrary to it may be considered abusive.
Project Reorientación de los instrumentos jurídicos para la transición empresarial hacia la economía del dato. Proyectos de Generación de Conocimiento 2021 del Ministerio de Ciencia e Innovación. PID2020-113506RB-100 (2021-2024). IP PhD José Antonio Fernández Amor.