Beyond exclusion: revisiting exploitative abuses in digital platform markets

Abstract

With the rapid expansion of digital platforms, concerns regarding exploitative practices have gained prominence, necessitating a closer examination of the possibilities available to competition authorities to address them. In light of the higher error costs of enforcement compared to exclusionary abuses, scholars have come up with structured approaches to identify in which situations exploitative abuse actions are justified. However, these approaches are mostly geared towards excessive prices and have not been adapted to digital platform markets, where customers and consumers are more likely to encounter unfair trading conditions. The aim of this article is to examine the extent to which the structured approaches developed in the literature are applicable to cases of unfair trading conditions by digital platforms. In order to do so, we first review the literature to establish the most widely supported conditions under which intervening against excessive price is justified. We identify the following conditions: (i) the presence of high and non-transitory barriers to entry; (ii) an enhanced dominant position; (iii) the absence of sector-specific regulation; and (iv) enhanced consumer harm. We then apply those conditions to exploitative abuse cases in digital platform markets and propose adjustments to ensure they are appropriate to guide competition authorities’ intervention.

1. INTRODUCTION

Article 102 TFEU prohibits the abuse of a dominant position, which includes practices that are directly exploitative of consumers, as well as conduct that, through the erection of artificial barriers, can exclude other players from the market, thereby reducing competition and its benefits for consumers. The European Commission’s enforcement priorities have traditionally lied on exclusionary abuses, and as such the case law of the Commission and the European Union (EU) Courts has not provided much guidance on how to deal with exploitative ones.¹ This is rooted, among other things, in the understanding that in the absence of barriers to entry, a dominant undertaking making super-competitive profits will attract new entrants to the market.² If entry into the market is possible, the competitive process is considered to be a better way to remedy excessive prices.³ Moreover, a dominant undertaking’s ability to engage in exploitative conduct, which generally occurs when the firm secures a significant advantage at the expense of its customer or consumers,⁴ may indicate that the Commission or national competition authorities (NCA) have under-enforced Article 102 by not identifying exclusionary practices in the first place.⁵

However, with the rapid expansion of digital platforms, concerns regarding exploitative practices have gained prominence, necessitating a closer examination of the possibilities available to competition authorities to address them.⁶ The characteristics of the digital market make exploitative abuses potentially more harmful to consumers. These characteristics include strong network effects, high fixed and low variable costs, and the key role played by data, which together result in markets being dominated by a few giant platforms.⁷ In a ‘winner-takes-most’ market, users are often locked in and new entry is difficult.⁸ Thus, customers and consumers may have no choice but to accept unfair terms or excessive prices for a prolonged period of time. Furthermore, paying with data creates information asymmetries, meaning that consumers are less aware when they are overpaying and, consequently, less likely to seek out alternatives. This benefits dominant platforms, which can use their position of power to impose unfair terms without fear of retaliation.

In recent years, a number of NCAs have shown a willingness to bring exploitative abuse cases.⁹ The German authority brought a case against Facebook,¹⁰ the French authority investigated Google’s conduct in the Google Ads Rules¹¹ and Google News cases,¹² and the Dutch authority looked into Apple’s treatment of online dating app providers.¹³ The European Commission, nonetheless, does not envision changing its prioritization of exclusionary over exploitative abuses, despite the increasing threat of exploitation by digital platforms, as shown in its forthcoming guidelines on the enforcement of Article 102.¹⁴ Before the recent App Store Practices case,¹⁵ the Commission has very rarely brought an exploitative conduct case, let alone one about unfair trading practices.¹⁶

There have been calls for the Commission to pay more attention to potential exploitative abuses, as these are clearly meant to be covered by Article 102.¹⁷ In particular, it is argued that in cases of serious market failure and in the absence of regulatory oversight, bringing an exploitative abuse case may not only be appropriate but also necessary to prevent harm.¹⁸ In light of the higher error costs of enforcement compared to exclusionary abuses, various scholars have come up with structured approaches to identify in which situation it may be justified for competition authorities to open exploitative abuse proceedings. However, these approaches are mostly geared towards excessive prices and have not been adapted to digital platform markets, where customers and consumers are more likely to encounter unfair trading conditions.

The aim of this article is to examine the extent to which the structured approaches developed in the literature in relation to excessive prices are applicable to cases of unfair trading conditions by digital platforms. The intention is to tailor the existing approaches to new economic insights about digital ecosystems. The article is not concerned with identifying whether a price is excessive or a trading condition is unfair, but rather informing the authorities’ enforcement priorities. While setting enforcement priorities is different from establishing criteria for identifying abusive exploitation, both are relevant for minimizing enforcement errors. Furthermore, competition law in the digital market is now complemented by the Digital Markets Act (DMA),¹⁹ which imposes obligations on gatekeepers, that is large digital platforms providing core platform services. However, this article will not consider the role of the DMA, as we believe that abuses of dominance, including exploitative ones, will continue to play a role in digital platform markets in the future and should continue to be studied.²⁰

Our suggested approach is valid irrespective of whether it is applied by NCAs or the Commission. Still, we argue that it is particularly the Commission that could rely on this approach in bringing actions against digital platforms, when they exploit customers and consumers across several Member States. The intention of the authors is that by creating more certainty around cases in which intervention in the form of exploitative abuse actions is justified and necessary, we might see more of these cases in the future.

The article has a backward- and a forward-looking component. To lay the foundation for our analysis, we start by giving a short overview of exploitative abuses under Article 102 (Section 2). Secondly, we review the literature to establish the most appropriate conditions under which intervening against exploitative behaviour is justified (Section 3). Finally, we apply those conditions to cases dealing with exploitative abuses in digital platform markets and propose adjustments to ensure they are appropriate to guide competition authorities’ intervention (Section 4).

2. EXPLOITATIVE ABUSES UNDER ARTICLE 102 TFEU

Article 102 TFEU prohibits any abuse of a dominant position, including ‘directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions’.²¹ The wording of Article 102 is clearly intended to cover exploitative abuses of dominance.²² The rationale for intervening against exploitative behaviour is that the undertaking is using its market position to ‘reap trading benefits which it would not have reaped if there had been normal and sufficiently effective competition’.²³ It is worth noting that in some jurisdictions, notably under US antitrust law, exploitative practices are not deemed anticompetitive. Moreover, some scholars have criticized the enforcement of exploitative abuses under EU competition law.²⁴

While it is unambiguous that Article 102 applies to exploitative abuses, assessing them comes with a number of challenges, both conceptual and practical.²⁵ Conceptually, it is far from clear what exploitative conduct entails, be it in the form of excessive prices or unfair terms.²⁶ Practically, the question is how to distinguish pro-competitive from anticompetitive conduct, for example competitive and excessive prices.²⁷ For example, intervention in relation to excessive prices may be problematic in markets in which a high-profit margin is required for further investment and innovation.²⁸ Another practical problem from an enforcement perspective is to decide which exploitative abuse cases should be investigated. Potentially going after any case where consumers overpaid would cause excessive deterrence and burden competition authorities’ resources. It is this practical problem that we seek to address in this article.

Exploitative abuse cases are mainly brought in the form of excessive pricing and, more rarely, unfair trading conditions.²⁹ Excessive prices, according to the European Court of Justice (ECJ), are prices that have ‘no reasonable relation to the economic value of the product supplied’.³⁰ The provision against unfair trading conditions relies on the same reasoning as in the case of excessive pricing, and basically prevents a dominant firm’s ‘imposition of conditions on its customers that directly harm them, usually by an onerous restriction of their freedom of action’.³¹ According to the case law, a dominant undertaking’s trading condition is ‘unfair’ under Article 102(a), if (i) it is unfavourable or detrimental to the interest of the dominant undertaking’s trading partners or third parties, including consumers, that are affected by the trading conditions,³² and (ii) it is not necessary for the achievement of a legitimate objective or in any event not proportionate for that purpose.³³ Arguably, the assessment of unfair terms boils down to the question of whether a dominant firm’s terms are necessary and proportionate since these are the terms that you would be expected to find in a competitive market.³⁴ Given the difficulty in determining whether terms are unfair, exploitative abuses have predominantly been brought in the form of excessive pricing cases. However, as explained in the next section, this may not always be applicable in digital platform markets.

3. PROPOSALS FOR A STRUCTURED APPROACH TO IDENTIFY APPROPRIATE EXPLOITATIVE ABUSE CASES

A key challenge for the law on Article 102 TFEU, particularly since the arrival of the more-economic approach, is to formulate administrable rules that provide sufficient clarity but at the same time avoid excessive legal formalism.³⁵ This challenge also relates to the problem of keeping error costs of enforcement to a minimum. As mentioned above, the risk of false positives with respect to excessive prices is high. Intervention by competition authorities should therefore only occur in exceptional circumstances.³⁶ In the Latvian Copyright Society, Advocate General Wahl suggested three limiting factors in bringing excessive price actions, namely, (i) the existence of high barriers to entry or expansion, (ii) the absence of a sector regulator, and (iii) the absence of countervailing buyer power.³⁷ Many years earlier the literature had presented somewhat similar limiting principles.³⁸ Various scholars have used those limiting principles to devise structured approaches to determine when an excessive pricing action under Article 102(a) is warranted.³⁹ However, at the moment, there is no widely accepted approach.⁴⁰ The more pressing concern is that those structured approaches are mostly geared towards excessive price actions. Yet, on digital platforms, due to zero-price and multi-sided markets, unfair trading conditions are more common. In this section, we seek to develop a comprehensive and structured approach to guide enforcement action in terms of exploitative abuses on digital platforms.

The literature on EU competition law has produced at least eight different structured approaches to excessive pricing.⁴¹ At the outset, all authors agree that a competition authority should only intervene against excessive prices in very exceptional circumstances. In fact, some authors consider the individual steps of their approaches to be ‘limiting principles’ that are meant to prevent competition authorities from intervening in unjustified cases. The idea behind a structured approach is that competition authorities can use it as a screen to select appropriate cases for their investigations.

All commentators argue that intervention is only justified in markets characterized by very high and long-lasting barriers to entry. The other characteristic, one that typically goes hand-in-hand with the former one, which most authors agree on, is that mere dominance does not suffice but there must be ‘enhanced dominance’. This principle is by far the most complex one and has been interpreted differently by various commentators. Some consensus exists among the authors on three further limiting principles, namely that there is no (sector-specific) regulator; that there is a lack of effective means for a competition authority to eliminate the entry barriers; and that there is consumer harm ‘distinct from, and additional to, the harm that derives from the mere payment of an excessively high price’.⁴² We will elaborate on the above limiting principles one by one.

Presence of high and non-transitory barriers to entry

A main objection to the use of excessive price action follows from the premise that a sector, governed by competition law rather than regulation, allows market forces to operate freely, and one anticipates the competitive process to function reasonably well.⁴³ However, in some sectors, for various reasons, this premise may not hold true. The first condition for using excessive pricing action is the presence of high and non-transitory barriers to entry. Barriers to entry (and expansion) describe any material difficulties that an incumbent’s actual or potential competitors face when seeking to enter the market or expand within it.⁴⁴ While high barriers to entry generally shield a dominant firm from competition, it is argued that the bar for intervention should be set higher than requiring an ‘ordinary’ dominant position; this is where ‘enhanced dominance’ comes into play.⁴⁵

Enhanced dominance

An argument for limiting the use of excessive price action stems from the idea that high prices and profits should be seen as the reward and incentive for firms’ investments and innovations.⁴⁶ Thus, it is suggested that we need to limit market interventions where the competitive harm of exploitative abuses is highest, while the distorting effect of price regulation is lowest.⁴⁷ Motta and de Streel distinguish between (i) earning dominance due to risky investments and business acumen, ie competition on the merits and (ii) gaining dominance due to current or past special rights or un-condemned past exclusionary anticompetitive practices, that is not resulting from competition on the merits.⁴⁸ In other words, the dominant position needs to be qualified with respect to its source.

In the first instance, which may be applicable to industries with network effects, the dominant firm is likely the winner of the competition for the market, where the initial stage of the competition is characterized by low prices and intense competition to capture the market, and a subsequent stage where the market has tipped in favour of a firm, which then dominates the market.⁴⁹ In such instances, high prices in the later stage can be seen as a normal outcome of the competitive process and motivation for the firms to compete for the market in the initial stage.⁵⁰ For a competition authority to bring an excessive price action may not be justified, as it may distort the functioning of the market and the normal dynamics of competition with a competitive edge.⁵¹

In the second case, where the dominant firm has been shielded from competition, it is difficult to argue that high prices are a result of past investments or efforts.⁵² In such cases taking action against excessive pricing may be warranted.⁵³ Motta and de Streel divide the second scenario into two alternatives. First, dominance may stem from current or past barriers to entry (eg, scarce resources or natural monopoly characteristics) or the firm had gained access to the market in an unfair and discriminatory manner (eg lobbying efforts to get legal protection and create an economically unjustified rent).⁵⁴ Secondly, dominance may be the result of un-condemned exclusionary practices in the past.⁵⁵ Such situations could arise either because the firm did not hold a dominant position at the time of engaging in the exclusionary conduct (which is therefore not actionable under EU law), or due to a type II error or false negative (ie where a competition authority failed to intervene when it should have).⁵⁶ Röller calls the former situation a ‘gap case’ and the latter a ‘mistake case’,⁵⁷ and argues that this enforcement gap can be closed by interventions against exploitative abuses that are based on the attainment of dominance through anticompetitive exclusionary conduct.⁵⁸

Other commentators support the idea of limiting antitrust action to instances where exploitation is made possible by exclusionary conduct.⁵⁹ A very prominent case on post-exclusionary exploitation is United Brands where the Court found that Article 102 is breached if a firm uses its dominance to obtain commercial benefits that it would not have obtained in a competitive market.⁶⁰ This post-exclusionary exploitation test requires a causal link between dominance and abuse.⁶¹ Nazzini argues that ‘[t]he cause of the exploitative abuse could only be the actual exercise of the dominant undertaking’s ability to exclude rivals.⁶² Therefore, exploitation in United Brands was made possible by exclusion.’⁶³

While there have been various attempts to qualify the enhanced dominant position of the firm charging excessive prices, they all seem to have in common that the ‘road to dominance’ is not one based on competition on the merits but rather due to either special rights or prior exclusionary conduct.

Absence of sector-specific regulation

The first two conditions arguably often apply to regulated network industries or liberalized industries, such as the telecommunications and energy sectors. In such instances, the more suitable entity for intervention is typically the sector-specific regulator and not a competition authority.⁶⁴ There is some agreement that excessive price action by competition authorities is only warranted if there is no regulator or if the regulator is not operating effectively.⁶⁵ Such regulators are likely to have superior knowledge of the sector and typically have the right to intervene with a lower burden of proof.⁶⁶ While the existence of a regulator does not exclude the application of competition law, it is argued that it should decrease the enforcement priority of bringing an abuse case.⁶⁷

Some commentators contend that exploitative abuse action should be limited to liberalized industries as a supportive measure during the liberalization process. The rationale behind introducing this industry-specific standard is rooted in the two abovementioned conditions, namely that market dominance in liberalized industries often arises from government intervention and special rights rather than superior efficiency or risky investments, and secondly, that these industries typically feature significant barriers to entry, preventing new players from competing away supra-competitive profits.⁶⁸ O’Donoghue and Padilla criticize this condition for various reasons. For one, Article 102(a) makes no distinction between recently liberalized sectors and others, raising concerns about potential discrimination if enforcement policies were explicitly targeted exclusively at the newly liberalized ones.⁶⁹ Secondly, this condition appears largely superfluous, given that the majority of liberalized industries are already subject to measures controlling prices or profits.⁷⁰

According to Röller ‘antitrust intervention remains possible, not least because industry-specific regulators are more likely to be subject to regulatory capture (and hence may fail to protect consumer welfare) than a competition authority, whose competence extends horizontally across most or all economic sectors’.⁷¹ Yet, conflicts may arise in certain situations between the regulator and the competition authority, where the former considers a firm’s price to be appropriate while the latter contends that the price is excessively high.⁷² Some commentators are therefore in favour of a clear division of competences between competition authorities and sector regulators to avoid overlapping enforcement.⁷³

Enhanced consumer harm

Charging prices above competitive levels inherently causes harm to consumers. As mentioned above, potentially investigating too many cases where consumers overpaid or felt unfairly treated would not only cause undue deterrence but also strain competition authorities’ resources.⁷⁴ Therefore, some authors propose that for excessive price action to be justified, there needs to be consumer harm ‘distinct from, and additional to, the harm that derives from the mere payment of an excessively high price’.⁷⁵

This condition serves as a safeguard that the cost of a mistaken acquittal (‘false negative’) is considerably greater than the cost of a mistaken conviction (‘false positive’).⁷⁶ Such enhanced consumer harm could, for example be the risk that the excessive prices may prevent the emergence of new goods or services⁷⁷ or competition in adjacent markets.⁷⁸ This example is derived from the case law on refusal to license IP rights, where the abusive practice prevented the development of a new or differentiated product for which there was a potential consumer demand.⁷⁹ Hence, consumers would be harmed where excessive prices discourage investment in production technologies or impede product development by downstream firms.⁸⁰

This condition also aligns with the post-exclusionary exploitation condition above and appears to be consistent with the case law.⁸¹ Pursuant to Gal ‘[m]ost of the cases brought by the Commission on grounds of excessive pricing were essentially of exclusionary rather than of exploitative nature’.⁸² For instance, in General Motors⁸³ and British Leyland,⁸⁴ the Commission considered the high fees imposed by the manufacturers to be a strategy aimed at preventing parallel importation of vehicles. This does not mean that the case could equally be framed as an exclusionary abuse case, but that the potential harm of the exploitative conduct is exacerbated by its exclusionary effects.

In this section, we have identified and explained the four conditions of a structured approach to identify appropriate exploitative abuse cases. In the next section, we will apply this approach to exploitative abuse cases in digital platform markets and assess whether it is suitable to direct competition authorities’ enforcement efforts.

4. APPLICATION TO DIGITAL PLATFORMS

Based on the above discussion, we can establish a structured approach against excessive pricing that consists of four cumulative conditions. These are (i) the presence of high and non-transitory barriers to entry; (ii) an enhanced dominant position; (iii) the absence of sector-specific regulation; and (iv) enhanced consumer harm. The conditions not only find support in various scholarly works but also in the case law of the ECJ.⁸⁵ However, it is important to note that most of those sources dealt with excessive pricing. The question is whether the conditions above are applicable in instances of unfair trading practices in digital platform markets. We seek to answer this question in this section.

The Meta case

In February 2019, the German competition authority (Bundeskartellamt (BKA)) ordered Facebook to limit its processing of user data, upon finding that it was imposing exploitative terms under the German equivalent of Article 102 TFEU.⁸⁶ Facebook was found to have abused its dominance, because it effectively forced users to agree to its terms and conditions, under which it could combine data collected outside of the Facebook website with users’ Facebook profiles.⁸⁷ The BKA argued that ‘there is no effective consent to the users’ information being collected if their consent is a prerequisite for using the Facebook.com service in the first place’.⁸⁸ The absence of valid consent was tied to Facebook’s dominance and the lack of alternative social networks in the market. Furthermore, the BKA argued that Facebook’s conduct deprived consumers of control over their personal data, constituting a violation of the right to informational self-determination.

Presence of high and non-transitory barriers to entry

The existence of high entry barriers is evident when it comes to digital platform markets. As explained at the beginning of this contribution, network effects and the role of data significantly contribute to barriers to entry.⁸⁹ As a social media platform, heavily relying on personalized advertisement to monetize its service, Meta clearly benefits from strong direct and indirect network effects and from its control over considerable user data. These barriers to entry significantly contribute to entrenching its dominant position.

Enhanced dominance

This case can be seen as partially resulting from the Facebook/WhatsApp merger, since it is only by acquiring WhatsApp that the problem with combining the user data across the services, and gaining more control over data, arose. This can be considered a mistake case, since arguably the merger should have been blocked in the first place.⁹⁰

In Facebook/WhatsApp, when analysing the closeness of competition between WhatsApp and Facebook Messenger in the market for communications apps, the European Commission noted several differences between the platforms.⁹¹ In relation to privacy policies, it stated that ‘contrary to WhatsApp, Facebook Messenger enables Facebook to collect data regarding its users that it uses for the purposes of its advertising activities’.⁹² It then indicated that ‘WhatsApp and Facebook Messenger have been reported as being the two main consumer communications apps simultaneously used by the majority of the users in the EEA’.⁹³ The fact that users kept on using both platforms regardless of their differences was taken to indicate complementarity rather than competitiveness between the firms. As a result, the merger could not undermine competition between them.

The conclusion that WhatsApp and Facebook Messenger were not close competitors is questionable for a number of reasons; first, the Commission assumed that ‘markets involving personal data function efficiently and reflect consumer preferences’⁹⁴ instead of asking whether the market characteristics allow for competition on data protection to take place. Furthermore, the fact that WhatsApp did not get a massive shift of Facebook Messenger users (at least before the merger took place) could also indicate that some consumers did not value data protection as highly as others or that they were not in a position to assess the level of data protection they were getting.⁹⁵ In addition, multi-homing might have been necessary for users in order to be in touch with their contacts from both platforms, because of the large network effects, and does not necessarily have to constitute an expression of users’ data protection preference.⁹⁶

Besides these questionable stances taken in the merger review process, after having cleared the merger, the Commission fined Facebook €110 million for providing misleading information about the merger.⁹⁷ Specifically, Facebook had communicated to the Commission that it would be unable to establish reliable automated matching between Facebook users’ accounts and WhatsApp users’ accounts. It went on doing exactly that after the merger and the Commission found that Facebook staff were aware of this technical possibility. Nonetheless, according to the Commission, this incorrect or misleading information would not have had an impact on the outcome of the decision.

Absence of sector-specific regulation

The role of sector-specific regulation is relevant in this case, since this conduct also falls under data protection authorities’ competency area. As a matter of fact, actions prohibiting WhatsApp from sharing consumer data with Facebook were initiated by national data protection authorities (DPAs) in Germany,⁹⁸ France,⁹⁹ and the UK.¹⁰⁰ When it comes to unfair trading practices as opposed to excessive pricing, this limb of the test necessitates closer scrutiny. While it is fairly straightforward to determine whether a regulatory regime that regulates prices is in place, it is not always immediately apparent whether a regulator is competent to control the terms that are considered exploitative under competition law.

The ECJ got to rule over this aspect of the case in Meta¹⁰¹; the German court asked the ECJ whether, under the GDPR, an NCA can, when investigating an abuse of a dominant position, find that a dominant undertaking has breached the GDPR. It also asked whether this is possible even if the same terms are being investigated by the competent DPA.¹⁰² The ECJ submitted that, in the GDPR or otherwise, there are no specific rules governing the cooperation between NCAs and DPAs and that the GDPR does not prevent NCAs from finding that a form of data processing carried out by an undertaking that abused its dominant position does not comply with the GDPR.¹⁰³ It then stated that compliance with the GDPR or a lack thereof may be a ‘vital clue among the relevant circumstances of the case’.¹⁰⁴ Accordingly, in some markets it may be necessary for NCAs to assess whether rules other than competition law are being complied with, including the GDPR.¹⁰⁵

However, the ECJ also emphasized the distinct objectives of the two frameworks and pointed out that an NCA that finds a breach of the GDPR does not replace the DPA. What an NCA does is ‘merely noting the non-compliance of a data processing operation with the GDPR’.¹⁰⁶ At the same time, the ECJ acknowledged that this could create problems in terms of the coherent application and enforcement of the GDPR. It addressed this issue by invoking the duty of sincere cooperation.¹⁰⁷ The ECJ stated that an NCA cannot depart from the decision of a DPA and must consult and cooperate with the latter in specific circumstances.¹⁰⁸

What emerges from the ECJ’s judgment is that while competition authorities can take into account the GDPR in their competitive assessments, they cannot enforce the GDPR as such. Whether in this specific case it was necessary for the BKA to rely on the GDPR to establish the existence of the abuse, or whether it crossed the line into the area of competency of DPAs, will need to be determined by the German court. Since the BKA has tied the breach of GDPR to anticompetitive conduct, however, it does appear that it fulfils the conditions set out by the ECJ.

Enhanced consumer harm

Under this last condition, it must be assessed whether consumers are harmed by the conduct in a way that goes beyond the unfair terms themselves. The BKA argued that Facebook’s conduct impaired competition since the excessive data processing made it more difficult for others to compete against it. According to the BKA, Facebook has ‘gained a competitive edge over its competitors in an unlawful way and increased market entry barriers, which in turn secures Facebook’s market power towards end customers’.¹⁰⁹ By linking the exploitative conduct to exclusionary effects, the BKA showed how consumers can be harmed beyond the unfair terms alone, as reduced competition in the social media market can have wider short- and long-term negative effects on users. This is also how causation in this case was established; rather than arguing that Facebook’s dominance allowed it to impose unfair terms, which is more difficult to prove, the BKA stated that it was enough to show that the conduct had exclusionary effects.

Concluding remarks

The application of the four conditions to the Meta case appears to be an adequate and comprehensive way to assess the desirability of bringing this type of exploitative abuse case. The discussion above has shown that the structured approach can be applied to unfair trading practices, inasmuch as it allows for the assessment of the peculiarities of these types of exploitative behaviour by dominant companies. The approach allows us to reflect on whether the (potential) harm justifies intervention. More specifically, the Meta case has revealed that when it comes to unfair trading practices in digital platform markets, in particular concerning data, the third limb of the approach, that is the absence of sector-specific regulation, may require particular attention.¹¹⁰

The Google News case

In November 2019, news publishers in France lodged a complaint with the French competition authority (Autorité), accusing Google of an abuse of dominance under the French Commercial Code¹¹¹ and Article 102 TFEU, as well as an abuse of a situation of economic dependency. The complaint was directed at Google’s refusal to negotiate licences and its announcement that it would no longer reuse the protected content of news publishers as of the date of entry into force of the Law on Related Rights¹¹² unless the latter gave it permission to do so free of charge. If the news publishers declined to license the use of snippets of protected content without remuneration, Google would solely display a link to the relevant webpage. Even though the right provides publishers with the power to deny the use of snippets, only having a link displayed in search results generates less referral traffic (ie visits to news publishers’ webpages originating from other channels, such as search engines and social media).¹¹³ Such traffic constitutes a substantial part of the visits and is therefore important to generate advertisement revenue for news publishers. The Autorité found that Google’s conduct amounted, among other things, to an imposition of unfair trading conditions.¹¹⁴

The Autorité addressed Google’s alleged imposition of unfair conditions both under Article L. 420-2 of the French Commercial Code and Article 102(a) TFEU. According to the Autorité, by forcing the news publishers to license their content for free, Google imposed unfair trading terms on them.¹¹⁵ Zero remuneration is an unreasonable measure, as Google derived an economic interest from reusing the protected content, and the intention behind the Law on Related Rights was to transfer part of that benefit to news publishers.¹¹⁶

Presence of high and non-transitory barriers to entry

The positive feedback loop of data-driven network effects clearly applies to Google which, as elaborated below, has been dominating the general Internet search market and the market for search-related advertising for more than a decade. Network effects and switching costs appear to be significant barriers to entry. Nonetheless, these economic characteristics do not inherently hinder new firms from displacing the existing incumbent.¹¹⁷ What makes the barriers to entry in Google’s situation (as well as Meta’s)¹¹⁸ less transitory, is that the company not only benefits from network effects on the user side but also on the advertiser side. The more advertisers use the search engine for their online ads, the higher the search engine’s revenue; and this revenue can be reinvested in the maintenance and improvement of the general internet search service so as to attract more users.¹¹⁹

Enhanced dominance

In many national markets around the world, Google has by far the largest market share in the general internet search market and the market for search-related advertising. Google quickly expanded into a number of other markets, but general internet search continues to be its flagship service. Google’s growth is arguably not purely based on competing on the merits but also on the employment of anti-competitive practices. Between 2017 and 2019, the European Commission imposed three prohibition decisions against Google’s parent company Alphabet for abusing its dominant position in various product markets in the EU.¹²⁰ Two of those decisions were upheld by the EU Courts.¹²¹ In a nutshell, Google was found to have engaged in anticompetitive practices that have prevented rivals from becoming more serious competitors. For instance, in Google Android, the Commission argued that apart from weakening innovation in the wider mobile space, Google’s conduct diminished competition and innovation by rival search engine providers and therefore entrenched its dominant position in its core market.¹²² Even more so than mobile operating systems, news referral is closely connected to the general search market.¹²³ Google is thus able to exploit news publishers due to its super-dominant position in the general search market, which it strengthened inter alia by using exclusionary measures in the past.

It could also be argued that Google’s super-dominance stems from barriers that are characteristic of natural monopolies.¹²⁴ As we discussed elsewhere, Google’s super-dominance is comparable to statutory monopolies and undertakings enjoying exclusive rights.¹²⁵ In Google Shopping, the General Court (GC) stated that ‘a system of undistorted competition can be guaranteed only if equality of opportunity is secured as between the various economic operators’.¹²⁶ Monti commented that ‘this passage is normally used to motivate abuse cases where the dominant firm enjoys exclusive rights or has enjoyed a state guaranteed monopoly in the past’.¹²⁷ In Google Shopping, the GC arguably extends this notion to gatekeepers.¹²⁸ Thus, despite initially earning its super-dominance lawfully with its investments and superior technology, Google’s super-dominance, at least partly, results from an unlevel playing field. Based on the above, Google’s dominance is not entirely based on competition on the merits but to some extent also the result of exclusionary conduct or exclusive rights in relation to the general internet search market.

Absence of sector-specific regulation

News publishers and digital platforms find themselves in a complex relationship of competition and interdependence.¹²⁹ The former compete for advertising revenue in a market that is now dominated by Google and Facebook. At the same time, news publishers depend on search engines and social media to reach and interact with audiences. They rely on click-throughs from search engines and social networking platforms.

The granting of a ‘related right’ in Directive 2019/790 on copyright and related rights in the Digital Single Market (CDSM) to press publishers is one of the conceived measures to tackle the impact of digital platforms upon press publishers.¹³⁰ Before the introduction of this right, the position of press publishers in terms of their published content was precarious and non-formalized, since press publishers only enjoyed a derived right stemming from the authors of publications. In light of these difficulties, Article 15 CSDM was intended as a way to formalize and strengthen the rights of press publishers in the digital single market, a tool for press publishers to employ against news aggregators to improve their bargaining position, and a tool to ensure media pluralism and quality.¹³¹

While the copyright legislation in question specifically relates to press publishers, it is important to note that in the Google News case, no media regulator intervened and that the only agency or authority involved in this case was the French competition authority. The fact that there was no media regulator but merely press-specific copyright protection, which Google managed to circumvent, appears to be even more reason for intervention by the NCA.

Enhanced consumer harm

The Autorité found that ‘serious and immediate harm has been caused to the press sector as a result of Google’s behaviour, which, in the context of a major crisis in this sector, deprives publishers and news agencies of a vital resource to ensure the continuity of their activities….’¹³² Google’s unfair trading conditions not only harmed the press sector directly but also indirectly consumers’ access to quality journalism, both online and offline. In terms of the digital readership, the Autorité observed that due to its growth, internet user traffic ‘is of the utmost importance to the activity of news publishers’.¹³³ In particular, smaller local and regional newspapers had been discontinued in the past decade over the lack of resources. As mentioned above, enhanced consumer harm might ensue when excessive prices threaten to prevent the emergence of new goods or services or competition in adjacent markets. Google’s unfair trading conditions may have had a similar effect. They undermined news publishers’ offering of digital content for which there clearly was consumer demand as evidenced by the growing digital readership.

Concluding remarks

The application of the four conditions to the Google News case confirms the suitability of this approach when assessing the merits of an exploitative abuse case. More conspicuous than the high and non-transitory barriers to entry and the super dominant position of Google was the fact that there was an absence of effective regulation in the press sector to stop Google’s exploitation and that the harm to media plurality exceeded ordinary consumer harm.

A structured approach to unfair trading practices on digital platforms

When a digital platform charges a high price or engages in practice to recoup costs for investments and innovations, the benefits it reaps are the reward for its efforts. Competition law intervention in these instances might deter such investments in the future. The characteristics of the digital platform market make it difficult to distinguish pro-competitive from anti-competitive conduct when it comes to trading conditions, which increases the risk of error costs of enforcement not only compared to exclusionary abuses but also compared to exploitative abuse actions in traditional markets. The two case studies above exemplify how the structured approach developed in the context of excessive prices is useful in determining in what situations against unfair trading practices by a digital platform is warranted, thus reducing the risk of over-enforcement. In this section, we suggest that the characteristics of digital platform markets justify minor adjustments to the structured approach.

The first two conditions of the structured approach, high and non-transitory barriers to entry and the existence of an enhanced dominant position, are inextricably linked. This appears to be even more true in digital platform markets that are characterized by competition for the market. As noted by Graef:

[w]hereas old network industries such as electricity, gas and railways depend on physical infrastructures and tangible assets, new network industries including online intermediaries such as search engines, social networks and e-commerce platforms are characterised by the predominance of virtual networks and technologies.¹³⁴

The significance of barriers to entry in those digital platform markets is dependent on whether they constitute a lasting competitive advantage for the incumbents. Network effects constitute the most significant barrier to entry in the platform economy.¹³⁵ Those network effects can potentially result in one company dominating the market and engaging in anticompetitive conduct.¹³⁶ Another common characteristic of digital platform markets is the existence of switching costs. Such costs may be high if consumers are locked-in to a particular platform.¹³⁷ If a dominant platform takes active measures to increase switching costs, for example, in the form of hindering multi-homing, this may be an indication of the abusive nature of the conduct.¹³⁸

With regard to the second condition, enhanced dominance in digital platform markets can be established if the platform operator is a rule setter.¹³⁹ In such a case, it can be argued that the digital platform enjoys an exclusive right similar to statutory monopolies in regulated network industries.¹⁴⁰ As such, the platform operator’s dominance stems more from its exclusive right and no longer from competition on the merits. We can distinguish two situations. First, where a platform operator not only sets the terms and conditions of its platform but also competes against third parties. In other words, the platform has a dual role. A good example is the Google Shopping case, where Google’s own comparative shopping service competed against other comparative shopping services appearing on Google Search.

The second situation is where a platform operator takes advantage of its role as a platform operator vis-à-vis third parties that depend on the platform. In this situation, the platform does not compete against third parties on its own platform. An example of this situation is the Google News case. Third parties’ or consumers’ dependence gives the platform operator superior bargaining power, enabling it to exploit its position. It is important to note that bargaining power (or ‘relative market power’ as recognized in some national competition laws) is not synonymous with dominance within the meaning of Article 102 TFEU. Superior bargaining power only relates to the relationship between trading partners, while dominance or significant market power relates to a firm’s position in the market as a whole.¹⁴¹ As such, it is well possible for the second largest firm in the market to have superior bargaining power over its customers, but it would not have dominance.¹⁴² Despite this difference, for the purpose of the structured approach, we propose to treat a platform’s ability to be a rule setter as a form of enhanced dominance.

In digital platform markets, the application of the third condition, that is the absence of sector-specific regulation, is not as straightforward as in many liberalized industries. Even though the DMA recently entered into force, supplementing competition law with regard to creating a more level playing field in digital platform markets, the DMA is not entirely comparable to sector-specific regulation, such as EU energy regulation. In addition to competition law, other legislation might apply to digital platforms. As seen in the cases above, large digital platform companies like Google and Meta not only distort competition but their significant market power may cause harm outside the scope of competition law, such as the degradation of privacy or the reduction of public interest journalism. Those other types of harm may be better addressed by specialized regulators or authorities. A competition authority should only intervene after checking if no other regime is better suited to address the exploitation.

Under the fourth condition, enhanced consumer harm, a competition authority should be able to show that the abuse is harmful to competition, beyond the imposition of the unfair trading conditions themselves. When it comes to this condition, one particularly relevant factor is whether the conduct is linked to an exclusionary practice.¹⁴³ For instance, in data ecosystems, the boundary between exclusionary and exploitative conduct is often blurred. Personal data collection may be both exploitative, if excessive data are collected, and exclusionary, if the dominant company obtains a competitive advantage through those data, as was the case in the Meta case. In this regard, van de Waerdt emphasizes the need to look at the exploitative component, since ‘focusing only on the foreclosure elements of such a market does not do justice to the complicated interplays of competition for data that characterize it’.¹⁴⁴ In cases where a dominant undertaking’s conduct is both exclusionary and exploitative, a competition authority may decide strategically (eg, in the interest of a higher likelihood of proving an infringement) whether to go after the exclusionary or exploitative conduct.

The European Commission’s App Store Practices case against Apple exemplifies the blurry line between exclusionary and exploitative abuses in digital markets. In its first statement of objections,¹⁴⁵ the Commission outlined two concerns relating to Apple’s agreements with music streaming service providers. First, mandating the use of Apple’s proprietary in-app purchase system for the distribution of paid digital content and charging a 30 per cent commission fee on all subscriptions bought through this system. Secondly, preventing music streaming service providers from informing iOS users about alternative subscription possibilities. These two agreements, according to the Commission, distorted competition in the market for music streaming services and led to higher prices for consumers. In a subsequent statement,¹⁴⁶ the Commission announced that it would only focus on the second conduct, the anti-steering obligations, and eventually fined Apple over €1.8 billion for its abuse of dominance.¹⁴⁷ The Commission treated this case as an exploitative abuse case, describing Apple’s anti-steering provisions as unfair trading conditions. However, this form of abusive conduct appears to be more of an exclusionary type than an exploitative one, since ultimately it was about the competitive advantage obtained by Apple by limiting users’ access to competitors.

5. CONCLUSION

This article has started with the assumption that, in digital platform markets, there is an increased risk of consumer harm through exploitative practices, and argued that this calls for a renewed scrutiny of the possibilities available to competition authorities to address them. The increased risk derives, among other things, from the characteristics of digital platform markets, which are conducive to high barriers to entry. While NCAs have shown willingness to bring exploitative abuse cases in digital platform markets, the European Commission does not envision changing its prioritization of exclusionary over exploitative abuses. This is in part connected to the higher error cost of enforcement compared to exclusionary abuses. Given this high error cost, various scholars have come up with structured approaches to identify in which situation it may be justified for competition authorities to bring exploitative abuse cases.

The article has examined the extent to which the structured approaches developed in the literature in relation to excessive prices can be transposed to cases of unfair trading conditions in digital platform markets. The structured approach consists of the following conditions: (i) the presence of high and non-transitory barriers to entry; (ii) an enhanced dominant position; (iii) the absence of sector-specific regulation; and (iv) enhanced consumer harm. The idea was that such a structured approach may help the Commission and NCAs to identify when to bring exploitative abuse cases against digital platforms. The application of the four conditions to Meta and Google News showed that, with some minor adjustments, the approach is a suitable way to assess the merits of these types of exploitative abuse cases. It allows to reflect on whether the (potential) harm justifies intervention and whether other forms of intervention are preferable. It can, therefore, be used by the Commission and other competition authorities when deciding whether to bring an exploitative abuse case involving unfair terms in a digital market context.

Footnotes