Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A duty of care to prevent online exploitation of consumers? Digital dominance and special responsibility in EU competition law

A duty of care to prevent online exploitation of consumers? Digital dominance and special... Abstract Article 102 Treaty on the Functioning of the European Union imposes a special responsibility on undertakings with a dominant position. Digital dominance based on data but also on network effects, platforms, and first user advantage can lead to extreme inequality in bargaining positions with regard to online services, which can result in unfair treatment of competitors, suppliers, and consumers—such as exploitation and discrimination. Online consumers are the focus of this article. In order to tackle competition issues regarding online consumers, I propose interpreting the special responsibility of digitally dominant undertakings as a duty of care in their regard. Digitally dominant undertakings may not just be burdened by but also benefit from this approach, which increases predictability and trust in online markets. It resembles treating online undertakings as information fiduciaries as has been proposed in the USA, but is different because privacy is not the driving concern here and the context is that of antitrust. I look at standards for triggering the duty of care and make proposals for the application of this norm, but both remain to be specified further. I. INTRODUCTION Online markets are posing new challenges to European Union (EU) competition law. The question addressed by this article is if and when this branch of law could be used to protect consumers from unfair behaviour by parties who exploit their dominance based on data, but also on network, platform, and first- mover advantages. I aim to develop the thesis that: (i) there is a need to counter new problems in digital markets,1 such as discrimination and exploitation of consumers including in terms of anticompetitive personalization of online prices and offers2; (ii) hence, there may be an increased need to have recourse to the special responsibility of undertakings with digital dominance3; (iii) that may take the form of a duty of care vis-à-vis their consumers. Based on these three arguments, this articles sets out a prima facie case for purposes of discussion. First, let me further define the terms that I have just used. Special responsibility is a so far loosely defined category from the case law of the Court of Justice of the European Union (CJEU) that refers to the obligation of dominant undertakings to avoid the abuse prohibited by Article 102 Treaty on the Functioning of the European Union (TFEU) (an open-ended category) and to compete exclusively on the merits. Here, this concept is applied primarily to relationships with consumers in the context of increasingly personalized online offers made by undertakings that enjoy data, network, platform, and first mover-based digital dominance. Because I will be dealing primarily with the interests of consumers (and will focus on exploitation and discrimination rather than exclusion), the focus, however, will be on the data dimension.4 In this context, the personalization of online prices and offers means increasing price discrimination between consumers in the digital realm, and the related online personalization of search and advertising.5 This may have both positive and negative effects. I am concerned here with the potential negative effects—such as redlining (or postal code discrimination), as well as exploiting information asymmetries and the lack of outside options (ie alternative choices) of certain (groups of) consumers.6 Online dominance can be strengthened based on access to and control over different categories of data, in particular data that would be difficult to aggregate and replicate by competitors, and that give the owner a significant and persistent degree of leverage over the data subjects.7 This typically includes personal data, including on willingness to pay that could be based on credit ratings, and location as well as search and purchase histories. In addition, network effects and first-mover effects as well as the ability to exploit platforms as two-sided markets also play a role in establishing the digital dominance (as opposed to traditional offline dominance) that I am primarily concerned with here. In many cases, however, undertakings that enjoy digital dominance are also dominant based on more traditional criteria, such as market share, entry barriers, and an absence of countervailing market power. Finally, a duty of care in tort law traditionally means the obligation to prevent an unreasonable loss from a contracting party. As a civil law obligation it has been elaborated by regulation for financial institutions to cover specific requirements to inform and know one’s customer. I will examine whether a duty of care can also be read into the already existing special responsibility of undertakings with dominance. In addition, a notion derived from data protection law, compliance by design and default (that in turn may largely be algorithm based itself) can be used. As argued here, at its core this would involve an obligation not to exploit unfairly the unequal bargaining power of the consumer. The proposed approach involved, thus, builds on a classic EU legal concept: the special responsibility of dominant companies who are prohibited from abusing their position is brought into focus with regard to natural persons. Essentially, this means reading a duty of care for dominant undertakings into this special responsibility. The existence of special responsibility within competition law is not contested, although it remains loosely defined. Two conceptual steps are required. First, the explicit references to the concept so far in relation to exclusion must be extended to include exploitation and discrimination. This is wholly consistent with the overall structure of EU competition law. Secondly, it must be expressed as a duty of care, which is new in this context, although it exists outside competition law. When developing the duty of care argument, I will largely leave aside the implications for other firms, who may likewise be abused by online dominant undertakings such as those providing platforms who may discriminate against and exploit their online rivals or business customers who depend on their services.8 As will be developed further below, there are also consumers who are at the same time undertakings (self-employed persons, professionals but also members of the Gig economy) or consumer producers (prosumers), and therefore frequently are suppliers and competitors. Hence, there can be no neat distinctions here and some overlap is likely. Because special responsibility is a general term, if the concept is to be used at all, a duty of care that is based on it should also make some sense in a broader (non-data) setting in order to preserve the coherence and, therefore, the effectiveness and legitimacy of EU competition law.9 Hence, I will look at the historical background of special responsibility first. Secondly, I will discuss the questions how the duty of care is triggered and what obligations or principles it would entail. Thirdly, I will discuss the question what are the benefits and the drawbacks of linking digital dominance to the duty of care. Finally, I will conclude whether the thesis set out above makes sense. II. SPECIAL RESPONSIBILITY IN EU COMPETITION LAW In a competition context, the special responsibility is limited to instances of dominance. That is to say where undertakings may act as an appreciable extent independently of, competitors, consumers, and suppliers.10 It is with their independence from consumers that I am primarily concerned. Relation to the prohibition as such As a general principle, the special nature of a dominant position is given by the existence of the exceptional category of dominance abuse as such. This is after all a prohibition for behaviour that would not normally constitute a violation of the competition rules, where an undertaking does not enjoy a dominant position. At the same time the intent of a dominant undertaking is not normally a requirement for an Article 102 TFEU abuse, leaving aside exceptions such as the second leg of the AKZO test on predatory pricing as part of a plan to eliminate a competitor,11 and more recently to eliminate as efficient competitors (AECs) in rebate cases.12 Abuse is an objective concept—one of the reasons why in this context object and effect abuses are not formally distinguished,13 even though in practice at least potential significant effects are increasingly required.14 Hence, there is always an implicit obligation not to infringe Article 102 TFEU for dominant undertakings. They must take care not to do so. The concept of special responsibility Special responsibility can be found as a separate category in the case law of the CJEU dating back to the Michelin case involving exclusionary rebates (1983): A finding that an undertaking has a dominant position is not in itself a recrimination but simply means that, irrespective of the reasons for which it has such a dominant position, the undertaking concerned has a special responsibility not to allow its conduct to impair genuine undistorted competition on the common market.15 Therefore, this concept dates back well before the existence of the Internet itself, let alone pervasive online platforms or digital dominance. It is clear, however, that it relates to a measure of self-constraint to be exercised by the undertaking in question, which is not limited to a specific time period. Yet, in spite of its (for an EU competition law concept) long pedigree, the concept of special responsibility and repeated references in competition cases has not been specified and remains open textured—or less fancifully: open to further interpretation. With regard to data issues, this may be interpreted as ‘a feature, not a bug’, or more of an opportunity than a problem. For the purposes of my argument, this open texture allows the claim that the special responsibility does not just concern the structure of competition and, therefore exclusion, but should include discrimination and exploitation. In fact, it would make no sense for undertakings to have a special responsibility vis-à-vis their competitors, but not their consumers. Applying the special responsibility versus consumers is all the more plausible because the EU antitrust regime sees consumer welfare as a core objective and exploitation and discrimination as actionable harms, and especially now that fairness itself is increasingly recognized as a legitimate competition concern. Objections have been raised to reading specific obligations into the special responsibility, notably claims that this might protect competitors and, therefore, obstruct competition on the merits.16 We can leave this objection aside as well as the tricky question what constitutes competition on the merits in the first place. (Broadly but elliptically, competition on the merits is to say: competition within the confines of competition law.17) This is because by competition on the merits by definition concerns exclusion of competitors (a type of problem that we are not engaging with here) and is not directly related to directly protecting consumers from exploitation or discrimination. Harmful versus beneficial discrimination With regard to discrimination in an economic sense it should be added that this is not always harmful, and can even be beneficial from a total and even a consumer welfare perspective because it may increase the number of customers who are able to purchase a good or service.18 This occurs where the wealthier are willing to pay more for the same service and effectively cross-subsidize its purchase for those who can only afford to pay less. However, this optimistic argument suggestive of progressive redistribution appears to break down at least to some extent in an online world. There all the information is on the side of the seller and in fact consumers who have fewest alternatives (and who are likely to be poorer and still less informed) pay the highest prices as a result.19 If a system of perfect price discrimination were to emerge, it would even allow the entire surplus to be appropriated by the producers. Evidently, there are also ethical arguments against such types of price discrimination.20 Distinguishing the three overlapping concepts Having identified and briefly discussed the obligation not to infringe, special responsibility, and duty of care, this now raises the question: (i) how the special responsibility is different from (ii) the obligation not to infringe in general and (iii) from the duty of care postulated as a possible pre-emptive remedy above? Are these three sides of the same coin or manifestations of the same phenomenon, and what would we gain from specifying them? It appears that these are in fact to some extent three distinct categories that overlap in part. Accordingly, some differences in scope are involved. First, the general obligation not to infringe the competition rules that applies to all undertakings generally requires the dominant undertaking to behave as if it is subject to competitive pressure. However, this undertaking must not just refrain from exclusionary behaviour such as the application of a price squeeze—or from exclusionary rebates such as in Michelin. It may also have a positive duty, such as to supply competitors, as for instance in the EU Commission’s Microsoft Decision as upheld by the CJEU in 2007. Here, Microsoft was found to be under an obligation to supply competing providers with its interface documentation enabling these competitors to produce software-based services running on work group servers that used Microsoft’s operating system. This aimed to ensure interoperability of their products, so the market would remain broadly innovative and not just dependent on Microsoft.21 Secondly, the requirement of an active avoidance of competition law infringements that in principle applies to all undertakings is developed further just for dominant undertakings by the special responsibility first cited in the above-mentioned Michelin Case (1983). This occurred in relation to Michelin’s exclusionary rebate scheme concerning new replacement tires for large vehicles (trucks, busses, etc) on the Dutch market, where Michelin held a 57–65 per cent market share compared to 4–8 per cent for its largest competitors. This behaviour was held to have affected negatively the structure of the market.22 The responsibility is special insofar as it only rests on dominant undertakings and involves a requirement to take action in order to protect competition as well as (as argued above) the interest of the consumer. The special responsibility relates to the obligations of a dominant undertaking in a specific case. This is in line with the fact that Article 102 TFEU does not provide a limited list of infringements but a dominant undertaking has the obligation to act as a normal undertaking would if constrained by effective competition, as well as (I would submit) fairly in relation to its consumers.23 Thirdly, the duty of care that flows from the special responsibility of dominant undertakings is not just a negative duty to abstain from infringing but also a duty to take positive steps in order to avoid such an infringement occurring. In relation to consumers, this entails an obligation to act not just as if disciplined by competition, but fairly as defined by norms that are pertinent in the context of exploitation and discrimination—such as privacy in the German Facebook Decision (2019) which found a violation of privacy by a dominant undertaking ipso facto constituted an antitrust infringement.24 Regarding consumers, the duty of care thus requires the undertaking not to act unfairly, as if constrained by the bargaining position of the consumer in a moral sense, and by behavioural norms in antitrust that are grouped under fairness as reflected in the prohibitions on exploitation and discrimination. This serves to guarantee that the unequal/asymmetrical position that in fact exists is not exploited. In other words, the interest of the consumer must be taken into account in a meaningful way. All three of these concepts are expressions of an obligation on dominant undertakings not to infringe Article 102 TFEU. However, there is a sliding scale active here both in terms of specificity and in terms of intrusiveness. In the absence of further practical application of the concepts involved, it does not seem productive to attempt to be more specific at this stage. We now move on to what is the precondition for the duty of care, digital dominance, and its triggers. How does an undertaking know it is dominant? As it is essentially based on prevention, the prohibition on dominance abuse, such as the special responsibility and the duty of care, presupposes not just the existence of a special position but also the awareness on the side of the undertaking concerned of the existence of this dominant position. Yet, how does an undertaking know that it is dominant in order for both the rule and the responsibility to apply? Generally, in Article 102 TFEU cases, market definition can be complex as can be dominance analysis in the sense of showing independent behaviour. This need not be a prohibitive bar to making a dominance assessment, however. For instance, Heike Schweitzer and others have proposed in their 2018 report to the German Economics Ministry that an abuse as such can itself already form decisive proof of the existence of a dominant position and, hence, replace a more formal prior market definition.25 More recently, similar advice has been proffered by Schweitzer, Crémer and de Montjoye to DG Competition of the European Commission (2019).26 In the first place, an undertaking holding a legal or factual monopoly inside or outside the online world should obviously expect that it might well be dominant and carry out due diligence to that effect. The same would hold for an undertaking that was subject to a previous finding of dominance by an authority or ultimately a court. Admittedly, such cases are relatively rare and the duty of care might overlap with obligations imposed or commitments offered by the dominant undertaking. A significant market power (SMP) regime that identified positions of dominance ex ante would play a similar role—if it existed in the digital realm, as it now does for electronic communications network operators in the EU. Secondly, especially in the online world—apart from various types of network, platform, and first-mover effects—could controlling certain data lead to a type of digital dominance which is established more easily or possibly exists per se? Replicability is evidently an issue here. It is also conceivable that a regulatory trigger would be created. Controlling personal data, lack of replicability, network and first-mover effects, and controlling a platform or service might be such criteria. Establishing the relevant market in itself could be an issue here: conceptually, where there is (near) perfect discrimination this might even be the individual consumer.27 As mentioned, an alternative would be for the nature of the behaviour itself to provide proof of dominance, irrespective of market definition. I expect that should this exercise be considered—or by failed attempts proven to be—too difficult in an antitrust context, it could certainly become food for regulation, where it might be applied, for instance, to all undertakings using personal data in their offerings.28 Thirdly, can controlling a program, standard, platform, algorithm, or data lead to a position of super-dominance, as has been variously alleged of Microsoft, Intel, Google, Amazon, Uber, and others? Super-dominance is not an established concept in EU competition law and appears to concern entrenched monopolies. This is illustrated by the fact that the main reference is still Advocate General Fenelly’s Opinion in the 1999 Compagnie Maritime belge Case: Article 86 cannot be interpreted as permitting monopolists or quasi-monopolists to exploit the very significant market power which their superdominance confers so as to preclude the emergence either of a new or additional competitor. Where an undertaking, or group of undertakings whose conduct must be assessed collectively, enjoys a position of such overwhelming dominance verging on monopoly … it would not be consonant with the particularly onerous special obligation affecting such a dominant undertaking not to impair further the structure of the feeble existing competition for them to react, even to aggressive price competition from a new entrant, with a policy of targeted, selective price cuts designed to eliminate that competitor.29 We will leave aside the obvious exclusion context that motivated this quote under the assumption that super-dominance could be equally noxious in relation to other types of abuse (on which more below). Super-dominance should not be required for a duty of care to hold.30 However, extreme market positions may mean that the special responsibility is intensified as the undertaking involved would both enjoy a greater asymmetry of power and a greater ability to make amends. It would presumably also be more aware of the antitrust responsibilities that its position entails. Finally, how would the undertaking know that an infringement that ought to be avoided could take place? This question arises especially with regard to fines, making the instrument of periodic penalty payments perhaps more suitable. It could be solved based on process/criteria like the proposed duty of care (or by compliance by design and by default as in the General Data Protection Regulation31), or compliance by algorithmic design,32 which could also exonerate the undertaking if the relevant criteria had been met. Perhaps on the supply side, there are lessons to be learned from risk management here.33 My argument is that these requirements could be qualified as a direct duty impendent on dominant undertakings as a duty of care obligation in competition law. The reasons for this are set out in more detail in Section IV below. However, as a technical matter the relevant norm could equally well be formulated in soft law,34 or in fact by means of direct intervention as a matter of regulation. Triggering the duty of care Determining the addressees of this duty is obviously linked to the question posed above: how do they self-identify? It appears obvious that there should not be simple quantitative turnover or market share thresholds to trigger the duty of care. Hence, we can also define special responsibility and the duty of care based on what it is not—an approach based on fixed market shares or similar that has been the focus of earlier critiques pleading to ditch the concept, largely based on the straw man argument that a market share approach would be simultaneously unavoidable and inappropriate.35 It is difficult to see how such fixed thresholds could work in any event because it would presuppose an essentially uncontested market definition. Instead, it may be preferable to develop the use of dominance criteria that do not rely on a predefined market such as a demonstrable lack of choice/absence of outside options for consumers or control over data and/or over an essential platform. In other words, it may be possible to define digital dominance in an online setting qualitatively, thereby making quantitative threshold irrelevant while securing the benefits of the special responsibility in a proportionate manner. An alternative is using the fact that certain behaviour is taking place as prima facie evidence that dominant position is in place—much as an abuse can sometimes do that. Such approaches have been advocated in above-mentioned reports by Schweitzer and others on reviewing the German and EU competition rules, respectively the application of these rules.36 Relevant types of abuse A distinction based on the type of abuse concerned may well be useful here, also because new forms of abuse appear to be relevant. In this categorization we should also include whether abuse targets (i) competitors, (ii) suppliers, (iii) purchasers at large, or (iv) consumers (individual end users).37 Arguably, this is a sliding scale, where consumers require most protection and competitors least as is illustrated by the AEC test mentioned above. The steadily increasing group of self-employed in the middle where categorization between undertaking and private individual is largely arbitrary appears to be the most problematic category in this case. As undertakings, they are likely to be dependent on dominant firms and arguably unlikely to meet AEC-type tests. However, as citizens they do vote, creating political sensitivity to their position, and moreover they may be affected directly as consumers too. Notably, they are in large part on the supply side of the market.38 Finally, consumers–producers (prosumers), for instance, in the context of energy are also on the supply side. Although at the present stage this is not a key feature of my argument, in future the supply side issues may have to be tackled too. Below I will distinguish between the four broad categories of abuses of exclusion, exploitation, discrimination, and sui generis abuses. Exclusion of competitors is traditionally favoured as the primary object of public competition enforcement.39 It is also the only area where (in 2009) the Commission has so far provided explicit guidance: similar guidance is lacking with regard to the other three categories of abuse identified below.40 It has also been the focus of sectoral rules on dominance, such as those regarding SMP for electronic communications (based on EU law, and at least under Dutch law, for healthcare providers and postal operators as well).41 A further example is exclusionary rebates by the nationally dominant operator weakening already marginal residual competition in Post Danmark II (2015). Here, the Court’s reasoning appeared to be that because the existence of a position of dominance means only limited scope for competition remains, further restricting it is in violation of the special responsibility of the dominant operator (which moreover was propped up by exclusive rights). This application of special responsibility seems to refer mainly to exclusion.42 Solving the exclusion puzzle after all leads to self-correcting markets, rendering direct protection of consumers superfluous. Recently exploitation has become a more relevant form of abuse now, for instance, excessive pricing cases are being pursued both at EU and national level (albeit with mixed results), for instance, in the area of pharmaceuticals.43 The relevance of special responsibility here is possibly linked to notions of fairness in the context of exploitation. (The vertical—between suppliers and purchasers and horizontal—between competitors—substantive fairness dimensions introduced by Michael Trebilcock and Francesco Ducci may be useful here, specified for consumers.44) This category is important to consumers in general, but perhaps less so in the context of different degrees of personalization. This is because as the test for excessive prices for instance is construed today it may for practical reasons be more effective to tackle the problem as one of discrimination (see below). This raises the issue of what legal basis would be appropriate because tackling such discrimination at least against individual consumers or defined groups of consumers who are not also suppliers or competitors cannot be done under Article 102(c) TFEU on discrimination against undertakings where exclusion must be shown.45 Instead, it would arguably have to be pursued as discrimination under the Article 102(a) TFEU fairness standard as unfair pricing—and preferably not as excessive pricing under the same provision because this involves a forbidding standard and difficulties of proof.46 What exactly would be the norm for this type of discrimination to be found remains to be specified as a matter of law. The category of discrimination, in the first place this covers discrimination of various types that occur notably in the absence of vertical integration (so there is no exclusion) and beneath the threshold of exploitation47—the prices are not excessive, or are not shown to be excessive for practical reasons.48 Secondly, it involves price discrimination—including increasingly personalized prices on online platforms targeted at consumers, as well as increasingly personalized online search and advertising using various layers of filters and tracking behaviour as well as using pre-existing data. Such price discrimination is a most relevant category for the duty of care vis-à-vis consumers. Actual competition cases however are so far in short supply, although studies on personalized pricing are being carried out by some competition authorities.49 Yet as mentioned just above, this type of discrimination could be covered by Article 102(a) TFEU as unfair pricing. Finally, there is the category of sui generis abuses that do not neatly fit the above distinction between exclusion, exploitation, and discrimination is associated with actions that deteriorate the fabric of competition at large and other sui generis type abuses that should also be avoided under the duty of care. The examples appear to be: abuse of procedure—such as in the EU and Italian AstraZeneca Cases (largely about exclusion), which is related to the notion that even the exercise of legal rights can under certain circumstances constitute an abuse50; harming the process of competition—regarding ‘as efficient competitors (AEC)’ (again exclusion); and/or harming the freedom of consumer choice, including variety and quality—arguably to be policed as a form of exploitation, but going beyond efficiency considerations. It is primarily this latter category that is relevant to arguments based on fairness. These four categories of abuses (exclusionary, exploitative, discriminatory, and sui generis) are themselves closely linked to the objectives of EU competition policy that were recently subject to a proposal for realignment by Ariel Ezrachi (in a different form than presented here) in a paperon competition law in the digital economy.51As a general trend greater pluriformity, with an emphasis on fairness, appears to be emerging. This will create more room for the duty of care approach discussed here. III. THE DUTY OF CARE Below we will examine the various issues that are raised by the duty of care as an extrapolation of special responsibility. Origins of the duty of care Because that is where I first happened to come across it, I have originally borrowed the notion of the burden of care from financial services regulation, such as insurance or home owner credits/mortgages. However, there is a fair number of different sources that are relevant. These are listed here in broadly chronological order. The use of the concept of duty of care, not only (i) in torts such as negligence where the first law and economics reasoning originated52—it is more efficient to have the dominant undertaking take precautionary measures than for individual consumers to have to sue for damages, in particular when the dominant undertaking has the data required to make the requisite decisions; but also (ii) as the duty a trustee owed to a beneficiary; and (iii) in contract law generally; as well as (iv) the explicit protection of the weaker party in contracts that aim to tackle structurally unequal bargaining positions such as in consumer law, labour law, and tenancy law; and finally (v) more explicitly in statutes such as those regarding as the above-mentioned financial services. In the common law tort of negligence, the test is whether the harm could have been foreseen, if there was proximity between the parties and if imposing a duty of care is fair, just, and reasonable. This is not fundamentally different from civil law approaches. The private law test may seem like ex post reasoning. However, in professions or activities that would normally risk being subjected to this standard, there will obviously be a prophylactic effect as well when measures are taken to avoid not just liability, but harm itself. In the UK, courts have generally treated a breach of competition law as a tort in the context of compensation. Private antitrust enforcement, especially since the implementation of the 2014 Damages Directive,53 is already relevant in this context, and will be all the more so if the duty of care is developed further. The above-mentioned regulatory approach can be traced to regulated financial services, where the obligations of Markets in Financial Instruments Directive I and II54 for financial institutions to inform and know their customers have developed into a duty of care.55 In the Dutch financial services act, this is spelled out in so many words.56 Recently, the UK Financial Conduct Authority (FCA) has proposed a similarly explicit approach.57 [A] ‘duty of care’ is a positive obligation on a person to ensure that their conduct meets a set standard. In the context of firms dealing with consumers, this normally means an obligation to exercise reasonable care and skill when providing a product or service.58 This development is significant also because under concurrent (parallel) powers with the Consumer and Market Authority (CMA), the FCA is responsible inter alia for applying the general competition rules to the financial sector in order to ensure effective competition there in the interest of consumers. If it adopts the proposed approach, it could therefore easily take the route examined in this article: applying a duty of care under the competition rules. Also in the UK, there is a consultation ongoing on the creation of a new independent regulator for the online sector, replacing self-regulation with public powers to intervene with respect to online harms cited as ranging from terrorist content to cyberbullying. Essentially, the new regulator, if indeed instated following the consultation, will in future balance free speech against a duty of care, which is to be the guiding standard for this approach to illegal and unacceptable content. The proposed duty of care presupposes the adoption of interim standards by government as well as definitive standards and a detailed code of practice by the regulator. Possible sanctions include fines against the undertakings involved and managers may be personally liable.59 In the Netherlands, health insurers also have to meet an explicit duty of care, which is to ensure that all care which is included in the standard insured package (set by Parliament) is actually available to patients.60 This may extend to finding non-standard solutions for bankrupted hospitals and planning cross-border provision of services. It is not the closest comparator to the duty of care intended in this article but it does involve taking the needs of a contracting partner into account both at the individual and at a meta-level. As part of an overall trend towards a more comprehensive type of corporate responsibility vis-à-vis retail clients, it is at least context relevant. Finally, the regulatory approach is one whereby a presumed duty of care with regard to the special responsibility of data dominant undertakings is specified in specific steps that they should take to avoid harm to their customers. This would obviously also be relevant in private antitrust suits. It seems self-evident that effective private enforcement on this point would be facilitated considerably by specific guidance that might be similar to what we have seen for dominant companies and exclusion,61 but also for horizontal and vertical restraints.62 This list could obviously be extended and refined further, but appears sufficient for my present illustrative purposes. However, before moving on to the various requirements of fairness, I will briefly engage in more detail with the US concept of information fiduciaries, which also involves invoking a duty of care. Because I have so far largely reasoned from an EU context, this requires a few more words of introduction and merits a separate section. Information fiduciaries In the USA, a debate is ongoing on the merits of treating online platforms as information fiduciaries by analogy to the position occupied by lawyers, doctors, and accountants who have been entrusted with privileged information by their clients.63 For the professionals mentioned, it is considered a legally acceptable exception to the First Amendment to the US constitution protecting freedom of speech to impose limits on their subsequent use of the information that they collected in the exercise of their duties. As advocated by Jack Balkin, Jonathan Zittrain, and others, the analogous fiduciary role involved regarding online platforms would comprise both a duty of care and a duty of loyalty to the client. The current US academic debate (the concept of information fiduciaries itself appears to be older) is not just about the legal justification of this approach and how it might be implemented in practice.64 It also revolves around the question whether its introduction as part of a ‘grand bargain’ with online platforms would not in effect mean pandering to them with a soft touch that might obstruct attempt to introduce more serious regulatory intervention. This would risk being counterproductive with respect to protecting the interests at which the fiduciary approach is aimed, notably privacy.65 Here, I will merely note several parallels and differences between the duty of care based on the fiduciary approach for online platforms in general in the USA and the duty of care based on the special responsibility of undertakings with digital dominance that I am advocating for the EU. In both cases, a concern over the unfair exploitation of a fundamentally asymmetric data position is addressed by imposing a duty to take the interest of the consumer into account ex ante in a compliance based manner. If successful, both might obviate the need for detailed regulation. A key difference is that in the EU, the General Data Protection Regulation in large part already meets the stricter privacy-related needs that much of the US discussion on the merits of information fiduciaries is in fact concerned with. Free speech as exercised by undertakings is in the EU, in any event typically not accorded the wide berth found in the USA under the First Amendment that for instance protects the sale of personal data by undertakings to third parties—meaning that regulation of such acts may well be unconstitutional.66 Moreover, as critics there point out, an information fiduciary approach cannot be an effective substitute where market power requires antitrust action and may at best be complementary.67 However, in a second best world, this dual solution may not be available because if a grand bargain between online platforms and their political opponents to instate the former as information fiduciaries in fact occurs, it is unlikely that intervention against market power will be on the menu for the foreseeable future. In contrast, because I start from the assertion that antitrust is in fact applicable to unfair treatment of consumers it is precisely my intention to invoke the competition rules and, therefore, address economic and not strictly privacy related concerns. This also includes the full gamut of antitrust remedies that would be available against infringements even if initially undertakings are expected to police themselves as a matter of efficacy. This would self-evidently include prosecution of cases, where undertakings demonstrably violate their own stated policies in this regard. Finally, as mentioned at the outset, in EU competition law, the special responsibility underlying the duty of care advocated here applies more narrowly to dominant undertakings, whereas the information fiduciary status would cover all online platforms entrusted with personal data. Hence, there is a significant difference in scope between the two approaches (and presumably in proving their applicability) even although in both cases a duty of care plays a central role. Fairness requirements As discussed above the duty of care is intrinsically linked to fairness. Ezrachi in his above-mentioned paper on the goals of competition policy and the digital economy has grouped six types of intervention under fairness: (i) counteracting unfair market practice or illegitimate wealth transfers; (ii) intervening against discrimination by dominant undertakings; (iii) counteracting misleading information; (iv) counteracting extreme instances of asymmetric information; (v) fighting unfair handling of data, and violations of data protection and privacy rights; and (vi) imposing a fiduciary duty requiring ethical handling of data.68 Although all these categories have some bearing on special responsibility expressed as a duty of care made here, two are directly relevant at least at the current state of the argument: category (i) on illegitimate wealth transfers and sharp business practices and (ii) on discrimination, of which there are as of yet no known competition cases regarding online consumers.69 A related but distinct is category (v) on privacy that has recently increased in relevance with the Facebook case undertaken by the German Bundeskartellamt.70 As I have briefly mentioned above, the Bundeskartellamt ruled on 6 February 2019 that a violation of data protection rules by Facebook in the context of the terms imposed on its consumers ipso facto also constitutes an exploitative abuse of its dominant position.71 Introducing norms from outside the realm of traditional antitrust, such as privacy and possibly consumer protection, and enforcing them by means of the competition rules presents a novel and potentially effective strategy in countering abuses of digital dominance that are otherwise difficult to classify. However, it remains to be seen how this will pan out on appeal and elsewhere in the EU. Obligations formulated as principles Regarding the obligations involved the proposed approach primarily seeks to reestablish the balance between the contracting parties. This is in line with insight in behavioural economics, which suggests that on the demand side consumers must be protected from themselves and exploitation, for instance, by ensuring that default options are on average beneficial.72 It involves elements such as know your customer and a duty to actively inform about downside risks and to follow appropriate procedures and keep records. In addition, it should be clear that offers which are deliberately exploitative or otherwise predictably unfair are not to be made at all. In this context, a blanket obligation to act reasonably seems insufficient. In order to be effective the norms to be respected must be clear and simple. Yet, because the EU competition rules may not count many words in the TFEU itself—but are complex in practice—something more specific is required. Ideally, the duty of care would be defined by general principles that lead to individualized customer specific obligations—the implementation of which would be possible precisely because of the data available. As a result, the duty of care should become not just a compliance issue but one of compliance by default and design. This is a matter of setting up internal checks and balances and screening out future problems that would promote the consumer interest potentially far more effectively than the system we have today based on analog standards and ex post paper-based enforcement. A modest proposal Principles for a duty of care with regard to the online offerings of digitally dominant undertakings can take various shapes. This goes beyond ‘know your customer’, the requirement found in financial services regulation, which may logically be assumed a given in a data-driven environment. In some (or many) dimensions, the digitally dominant undertaking may know more about the customer/consumer than he/she knows about himself/herself. The question is what action a digitally dominant undertaking would have to take based on this knowledge. Such principles might include preemptive behavioural remedies such as the five listed here: protect your consumer by using the relevant consumer data to screen also for the consumers downside and use this information to the consumer’s benefit; do not engage in negative discrimination by treating a consumer differently in order to exploit their inferior information position or lack of outside options; apply transparency by being clear about the consumer’s risks in the transaction, their outside options, and how the choices offered to them compare to those of others; provide a fair non-choice outcome by using smart default settings that ensure consumers reach a normally low-risk outcome if they do not actively pursue an individual choice; take responsibility for results, meaning that apart from procedural compliance there should be substantive compliance documented by actual performance. These ideas are just intended by way of a modest first proposal and as mentioned draw some inspiration from behavioural economics. It includes elements that would be less appropriate in offline markets because there for instance the offer is generally available, prices are posted physically and publicly, and other customers can similarly be directly observed. More importantly, individual consumer data are not collected or used at a comparable level of intensity. Failure to abide by these or similar principles that flesh out the duty of care as an infringement of the prohibition on dominance abuse would trigger private liability as well as antitrust fines. Enforcement It is early days to write on the competition law enforcement dimension of the duty of care, other than to speculate that it is likely to facilitate enforcement in general terms because it compensates for unequal bargaining power and information asymmetry. Thus, in large part, it will ensure prevention, and ensure that the relevant information is available and generally acted upon by the undertakings involved. Also, there are several clear advantages to a compliance based duty of care for dominant companies in terms of actual competition law enforcement: detection is easier as there would not only be a significant degree of self-policing but possibly self-reporting by the undertakings involved. This means there can be more general process based on systemic checks by the authorities; accompanied only by investigation and sanctions of individual infringements that are more serious. Finally, private enforcement would presumably play a role here because the exploitation norm involved is especially relevant to consumers—not least including the (nominally) self-employed—and prosumers. As mentioned above the duty of care principles, if violated, would trigger liability. This in turn could promote effective enforcement and thereby deterrence. Relationship with other remedies Obligations to share data with competitors—and more radically general transparency of data—may be a solution to some problems with regard to exclusion of competitors.73 However, in both cases, they are so only indirectly for exploitation of consumers in online markets (while evidently raising data protection issues too). These are therefore not alternatives to the duty of care obligations outlined above. A duty of care could exist alongside data protection specific remedies targeting exploitation such as imposed those in the German Facebook case because it is not restricted to privacy.74 Unlike the duty of care all three of these remedies are imposed ex post. The relationship with EU consumer protection rules, notably unfair trading practices, is more complex and remains to be examined further.75 These rules do not allow consumers or their protectors to challenge price levels as such76—although they impose certain transparency requirements on the way a tariff is construed that could capture certain forms of discrimination.77 Finally, the duty of care approach is compatible with largely algorithm-based compliance but also with approaches that require specific problematic outcomes to be justified and access to the algorithms of digitally dominant undertakings as systems of antitrust control. However, this is a topic in its own right that I can do no justice to here. Instead, we will look at why a duty of care may be a balanced approach from an industry perspective as well. IV. WHY IS LINKING THE DUTY OF CARE TO DIGITAL DOMINANCE ATTRACTIVE? At this stage, the principles set out above by way of example do not appear to be too demanding as requirements to be realistic, and we will for the moment assume that the underlying EU law account is sufficiently forceful. So if this is indeed prima facie a credible model, is this model attractive and in particular, could it be so for undertakings as well as consumers? Five reasons the duty of model could be attractive to undertakings Without trying to be exhaustive there appear to be at least five reasons why this is the case. In the first place, the general attraction is that the duty of care locates the burden of proof with the party that controls both the data and thereby directs the consumer, while of course it also controls its own behaviour—largely independently from the other market participants as it is dominant. (This seems in line with general requirements of proximity of proof.78) In addition, it carries an additional (special) responsibility as well as a special liability. Because this party controls the playing field, an additional burden or set of requirements designed to compensate for this advantage may be considered fair. Secondly, from an effective enforcement perspective, the duty of care is also attractive, not just regarding the allocation of proof but also because it is largely preventive rather than corrective in nature. This ex ante prophylactic is generally considered preferable not only as deterrence but also to avoid costly mistakes (including reputational damage), fines and private damages actions ex post. From a business perspective, thirdly, promoting trust between the parties involved, which is to the benefit of both consumers and data dominant undertakings as well as for the successful functioning of online commerce as a whole.79 Similarly, and fourthly, fewer surprises are likely to occur also for the dominant party—this has the advantage of creating a more predictable environment and greater investment certainty, again promoting online commerce. Finally, if successful the duty of care mechanism may obviate the need for potentially equally resource intensive and burdensome sectoral regulation on this count. To some, listing the final point as a possible advantage may seem like a provocation. In the US context, critics see legislation creating online information fiduciaries as increasing the risk that the adoption of more effective regulation will be blocked. As I have stated above the EU context already has more apposite regulation in place—notably the General Data Protection Regulation. Moreover, I am proposing an interpretation of the law as it stands that would imply a significant extension and not a potential reduction of the practical scope and application of antitrust, which in my view is implicit in the concept of special responsibility. This poses more serious constraints notably on economic abuses than the construct of an information fiduciary would, which is more likely to be restricted to privacy. Moreover, as we have seen in the German Facebook case where privacy violations were addressed by antitrust, the two regimes can be mutually reinforcing. There is clearly a balance involved here. Increased trust, more legal certainty, and arguably less need for new sectoral regulation will at least partially offset increased regulatory and compliance (and efficiency) costs. Depending on whether their perceived outside option is zero regulation and no effective competition law enforcement in perpetuity, or some considerable effort in this direction, the sum for undertakings involved in this exercise, with at least in large part algorithm based compliance by design, may well be positive. Hence, this approach appears to involve a reasonable a trade-off between legal certainty to the benefit of the digitally dominant party in exchange for fairer treatment, where the consumer is protected against excessive terms and other unfair practices in cases where the position of the party is highly asymmetrical and (in particular) data based. General applicability It could be argued that for an ex ante remedy or special responsibility (translated into a duty of care as proposed here) to be adopted it should make sense outside the realm of digital dominance as well, because the special responsibility exists there too. However, in my view, it could also be approached as an experiment that is justified in the online world/data realm, and sufficiently specific to warrant special treatment at least initially. This is also the case because outside the online world, given their information position dominant undertakings may not be equally aware of what is good or detrimental for their consumers: this makes for a different setting. At the same time this experiment would be grounded in experience: when operationalizing the duty of care we could learn from the transition of preexisting consumer rights such based on doorstep selling (such the ability to revoke the sale within a specific time period)80 to the inside pocket (smartphone) selling environment. V. Conclusion In this article, I have argued in support of the thesis that the already existing special responsibility of dominant firms in EU law can, and should, be translated into a duty of care for digital dominance. First, this coincides with their private law responsibilities: a digitally dominant undertaking can and should avoid causing predictable harm to its consumers. Secondly, explicit requirements could be set to facilitate compliance. This use of the duty of care would enhance the position of consumers in what are today fundamentally and arguably unjustifiably unequal exchanges while promoting trust between the parties, as well as predictability and investment security for the undertakings involved. If successful it may render sectoral regulatory regimes superfluous, and could fit in with a broader trend towards more comprehensive corporate responsibility towards consumers. There are parallels here with the concept of information fiduciaries found primarily in the US debate. This concept would however on the one hand apply more widely—to all online platforms instead of digitally dominant undertakings—and would on the other hand be limited largely to privacy issues, whereas the duty of care that I have argued for here focuses on a broader abuse of economic power. This also makes the criticism that arises in the US context that applying the concept of information fiduciaries in a grand bargain with online platforms may deflate the drive for more effective regulation less relevant in the EU. Here, the General Data Protection Regulation already covers the most pertinent privacy issues and a duty of care based on the special responsibility of digitally dominant undertakings enforced under antitrust would apply as a complement, not a substitute. It would also increase the bite and effectiveness of EU antitrust. Evidently, both the usefulness of this approach and whether it will be accepted and applied depend on the criteria that are used to determine: (i) whether an undertaking enjoys digital dominance and (ii) what exactly the duty of care involves. Also, (iii) the legal basis requires further clarification. Above, I have made some suggestions on all three counts. However, these important elements remain to be specified further and tested in practice. Currently on part-time secondment to the Free University of Amsterdam (Vrije Universiteit: VU) and extramural fellow of the Tilburg Law and Economics Center (TILEC). He is a former professor of economic law at the universities of Groningen and Tilburg. This article was written in a personal capacity. Comments by David Gabathuler, Inge Graef, Viktor Mayer-Schönberger, Martijn Snoep, Hans Vedder, Rein Wesseling, and from workshop participants in Lund, Nijmegen and Tilburg, as well as two anonymous referees are gratefully acknowledged. An embryonic version of this article appeared in Dutch in (2019) 22 Markt & Mededinging 105. Footnotes 1 See Heike Schweitzer and others, Modernising the Law on Abuse of Market Power: Report for the Federal Ministry for Economic Affairs and Energy (Germany) (17 September 2018) <https://ssrn.com/abstract=3250742> accessed 8 June 2019. 2 See European Commission, Personalised Pricing/Offers in the European Union (2018) <https://ec.europa.eu/info/publications/consumer-market-study-online-market-segmentationthrough-personalised-pricing-offers-european-union_en> accessed 8 June 2019 and recent OECD papers such as Online Personalisation from the Perspective of a Multidisciplinary Market Authority—Note by the Netherlands, 28 November 2018 and Personalised Pricing in the Digital Era – Background Note by the Secretariat, 28 November 2018. 3 I have chosen the term digital dominance over alternatives such as online or data dominance because it is more inclusive. See M Moore and D Tambini (eds), Digital Dominance: The Power of Google, Amazon, Facebook, and Apple (OUP 2018). 4 M Stucke and A Grunes, Big Data and Competition Policy (OUP 2016). 5 R Woodcock, ‘Big Data, Price Discrimination and Antitrust’ 68 (2017) Hastings Law Journal 1371; Christopher Townley, E Morrison and K Yeung, ‘Big Data and Personalised Price Discrimination in EU Competition Law’ 36 (2017) Yearbook of European Law 683; O Bar-Gill, Algorithmic Price Discrimination When Demand Is a Function of Both Preferences and (Mis)Perceptions (Harvard Law School 2018) <https://bit.ly/2vJsx3r> accessed 8 June 2019. On price discrimination more broadly, see OECD Background note DAF/COMP(2016)15, Price Discrimination. 6 C O’Neill, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy (Penguin/Random House 2017). 7 See eg J Prüfer and C Schottmüller, ‘Competing with Big Data’ CentER Discussion Paper, Vol 2017-007. 8 Examples are the Google Cases. See: ‘Antitrust: Commission Fines Google €2.42 Billion for Abusing Dominance as Search Engine by Giving Illegal Advantage to Own Comparison Shopping Service’ (Brussels, 27 June 2017) IP/17/1784. Decision of that date in Case AT 39740 Google Search (Shopping). See ‘Antitrust: Commission Fines Google €1.49 Billion for Abusive Practices in Online Advertising’ (Brussels, 20 March 2019) IP/19/1770. Decision of that date in Case AT 40411 Google Search (AdSense). See also ‘Antitrust: Commission Fines Google €4.34 Billion for Illegal Practices Regarding Android Mobile Devices to Strengthen Dominance of Google's Search Engine’ (Brussels, 18 July 2018) IP/18/4581. Decision of that date in Case AT 40099 Google Android. 9 W Sauter, Coherence in EU Competition Law (OUP 2016). 10 Case 85/76 Hoffmann-La Roche & Co AG v Commission Judgment of 13 February 1979, ECLI:EU:C:1979:36, para 38. 11 Case C–62/86 AKZO Chemie BV v Commission Judgment of 3 July 1991, ECLI:EU:C:1991:286. 12 Case C–413/14 P Intel Corp v Commission Judgment 6 September 2017, ECLI:EU:C:2017:632. 13 Case C–549/10 P Tomra Systems ASA et al. v Commission Judgment of 19 April 2012, ECLI:EU:C:2012:221. 14 Case C–525/16 MEO—Serviços de Comunicações e Multimédia SA v Autoridade da Concorrência Judgment of 19 April 2018, ECLI:EU:C:2018:270; Case C–177/16 Autortiesību un komunicēšanās konsultāciju aģentūra/ Latvijas Autoru apvienība v Konkurences padome (AKKA/LAA) Judgment of 14 September 2017, ECLI:EU:C:2017:689; Case C–177/16 Opinion of AG Wahl, 6 April 2017, ECLI:EU:C:2017:286. 15 Case 322/81 NV Nederlandsche Banden Industrie Michelin v Commission Judgment of 9 November 1983, ECLI:EU:C:1983:313, para 57. See also Case T–228/97 Irish Sugar Plc v Commission Judgment of 7 October 1999, ECLI:EU:T:1999:246, para 112; Case T–203/01 Manufacture française des pneumatiques Michelin v Commission Judgment of 30 September 2003, ECLI:EU:T:2003:250, para 55; Case T–219/99 British Airways plc v Commission Judgment of 17 December 2003, ECLI:EU:T:2003:343, para 242; Case C–202/07 P France Télécom v Commission Judgment of 2 April 2009, ECLI:EU:C:2009:214, para 105; Case C–457/ 10 P AstraZeneca AB and AstraZeneca plc v Commission Judgment of 6 December 2012, ECLI:EU:C:2012:770, para 134; Case C–209/10 Post Danmark A/S v Konkurrencerådet Judgment of 27 March 2012, EU:C:2012:172, para 23; Case C–23/14 Post Danmark A/S v Konkurrencerådet Judgment of 6 October 2015, ECLI:EU:C:2015:651, para 71. 16 M de la Mano, R Nazzini and H Zenger, ‘Article 102’ in J Faull and A Nikpay (eds), The EU Law of Competition (3rd edn, OUP 2014) 329–538, 393–94. 17 OECD policy brief, What Is Competition on the Merits? June 2006 <http://www.oecd.org/competition/mergers/37082099.pdf> accessed 8 June 2019. 18 As an economic concept this is different from the legal concept of discrimination and the EU body of law that intends to counter discrimination and ensure equal treatment. S Fredman, Discrimination Law (2nd edn, OUP 2011) and E Ellis and Ph Watson, EU Anti-Discrimination Law (2nd edn, OUP 2012). 19 Public authorities are beginning to worry about this as well: ‘Government and CMA to Research Targeting of Consumers Through Personalised Pricing. Research Commissioned to Look at the Practice of Retailers Targeting Online Shoppers’ CMA Press Release, 4 November 2018; Note from the Netherlands (n 2). 20 M Sandel, What Money Can’t Buy. The Moral Limits of Markets (Penguin Books 2013). 21 Case T–201/04 Microsoft Corp v Commission Judgment of 17 September 2007, ECLI:EU:T:2007:289. 22 Case 322/81 NV Nederlandsche (n 15) 23 Case 85/76 Hoffmann-La Roche (n 10) para 91. Abuse is not the use of economic power but the nature of that use, and therefore an objective concept relating to the behaviour of the undertaking. 24 Preliminary assessment in Facebook proceeding: Facebook’s collection and use of data from third-party sources is abusive, Bundeskartellamt press release, 19 December 2017; Bundeskartellamt prohibits Facebook from combining user data from different sources, Bundeskartellamt press Release, 7 February 2019; Bundeskartellamt, Case Summary, Facebook, Exploitative business terms pursuant to s 19(1) GWB for inadequate data processing (Decision of 6 February 2019) <https://www.bundeskartellamt.de/SharedDocs/Entscheidung/EN/Fallberichte/Missbrauchsaufsicht/2019/B6-22-16.pdf?__blob=publicationFile&v=4> accessed 8 June 2019. 25 See Schweitzer and others (n 1). The same idea was advocated in 2005 by the Economic advisory group to DG COMP in the report An Economic Approach to Article 82<http://ec.europa.eu/dgs/competition/economist/eagcp_july_21_05.pdf> accessed 8 June 2019. 26 J Crémer, Y-A de Montjoye and H Schweitzer, Competition Policy for the Digital Era (European Commission 2019). <http://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf> accessed 8 June 2019. 27 See E van Damme, I Graef and W Sauter, ‘Exploitative abuse in EU competition Law’ (2019, forthcoming) TILEC discussion paper. 28 An analogy might be call termination and roaming on mobile networks which proved to be intractable problems in antitrust, only to be thoroughly regulated as a result. See eg Regulation (EU) No 531/2012 of the European Parliament and of the Council of 13 June 2012 on roaming on public mobile communications networks within the Union(2012) OJ L172/10. 29 Opinion of AG Fennelly, 29 October 1998, in Joined Cases C–395/96 P and C-396/96 P Compagnie maritime belge transports SA (C-395/96 P), Compagnie maritime belge SA (C-395/96 P), and Dafra-Lines A/S (C-396/96 P) v Commission (1998) ECLI:EU:C:1998:518, para 137. 30 J Goyder and A Albors-LLorens, Goyder’s EC Competition Law (5th edn, OUP 2009) 301. Cf. Case T–201/04 Microsoft Corp v Commission, Judgment of 17 September 2007, ECLI:EU:T:2007:289; Joined cases C395/96 P and C-396/96 P, Compagnie maritime belge transports SA (C-395/96 P), Compagnie maritime belge SA (C-395/96 P) and Dafra-Lines A/S (C-396/96 P) v Commission, Judgment of 16 March 2000, ECLI:EU:C:2000:132. 31 This is a category introduced by Article 25 of the Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) (2016) OJ L119/1. 32 This is a method and/or process of meeting specific compliance requirements by means of software. See eg F Olivieri, ‘Compliance by Design: Synthesis of Business Processes by Declarative Specifications’ (PhD thesis, Brisbane and Verona, May 2014); S Sadiq and G Governatori, ‘Managing Regulatory Compliance in Business Processes’ in J vom Brocke and M Rosemann (eds), Handbook of Business Process Management (2nd edn, Springer 2015) 265–88. 33 C Sunstein, Risk and Reason: Safety, Law and the Environment (CUP 2004); U Beck, Risikogesellschaft: Auf dem Weg in eine andere Moderne (Suhrkamp Verlag 1986). 34 On competition soft law, see eg Z Georgieva, Soft law in EU Competition Law and Its Reception in Member States' Courts an Empirical Study on National Judicial Attitudes to Atypical Legal Instruments in EU Competition Law (PhD Thesis, Tilburg University 2017). An example is the Communication from the Commission: Guidance on its enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (2009) OJ C45/7. 35 RA Corcho, ‘Can We Finally Say Farewell to the “Special Responsibility” of Dominant Companies?’ EUI Robert Schuman Center 2007 EU Competition Law and Policy Workshop/Proceedings 2007 <https://www.eui.eu/Documents/RSCAS/Research/Competition/2007ws/200709-COMPed-Allendesalazar.pdf> accessed on 8 June 2019. 36 See nn (25) and (26) . 37 On problems of purchasing power, see V Daskalova, The Monopsony Paradox: Buyer Power and Enforcement of the EU Antitrust Provisions (Prisma Print, 2016). 38 V Daskalova, ‘Regulating the New Self-Employed in the Uber Economy: What Role for EU Competition Law?’ (Stockholm University 2017) Stockholm Faculty of Law Research Paper No 38. 39 Arguably some types of consumer discrimination are also a sort of exclusion although not centred on a mutually competitive relationship. Borderline cases are consumers–producers and self-employed persons. 40 Communication from the Commission (n 34). 41 Communication from the Commission—Guidelines on market analysis and the assessment of significant market power under the EU regulatory framework for electronic communications networks and services C/2018/2374 (2018) OJ C159/1. A comparison between regulation and competition in D Gabathuler and E Martinez Rivero, ‘Communications: Telecoms and Internet’ in Faull and Nikpay (n 16) 1647–08. 42 Case C–23/14 (n 15)paras 70–72. 43 Such as in Italy and the UK: Aspen Pharmaceuticals (Case A480) AGCM Decision of 29 September 2016; Pfizer (Case CE/9742–13) CMA Decision of 7 December 2016. The Danish competition authority made a similar announcement: ‘CD Pharma has Abused its Dominant Position by a Price Increase of 2000 Percent’ Press release, 31 January 2018. The European Commission opened a 27 Member State investigation into excessive pricing by Aspen as well. 44 M Trebilcock and F Ducci, ‘The Multifaceted Nature of Fairness in Competition Policy’ (CPI Antitrust Chronicle, October 2017) 1–6. A reverse view in A Ayal, Fairness in Antitrust, Protecting the Strong from the Weak (Hart Publishing 2016). 45 Case C–525/16 MEO (n 14). 46 See van Damme, Graef and Sauter (n 27). 47 See Faull and Nikpay (n 16) 525–38, with reference inter alia to Case T–128/98 Aéroports de Paris v Commission, Judgment of 12 December 2000, ECLI:EU:T:2000:290. 48 Case C–525/16 MEO (n 14) . 49 See n 19 on a recently announced CMA investigation and OECD note prepared by the ACM. 50 Case C–457/10 P AstraZeneca AB and AstraZeneca plc v Commission, Judgment of 6 December 2012, ECLI:EU:C:2012:770, para 133. More recently, comparable dominance abuses have been fined in Italy and France. 51 See A Ezrachi, ‘The Goals of EU Competition Policy and the Digital Economy’ (August 2018) BEUC discussion paper. A concrete example is the German Facebook case. See n 24 on the Decision of 5 February 2019 and the earlier preliminary assessment in the Facebook proceedings. 52 The calculus of negligence designed as a balancing test by Judge Learned Hand in United States v Carroll Towing Co (1947) 159 F 2d 169 (2nd Circuit). 53 Directive 2014/104/EU of the European Parliament and of the Council of 26 November 2014 on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union (2014) OJ L349/1. 54 Directive 2004/39/EC of the European Parliament and of the Council of 21 April 2004 on markets in financial instruments amending Council Directives 85/611/EEC and 93/6/EEC and Directive 2000/12/EC of the European Parliament and of the Council and repealing Council Directive 93/22/EEC, OJ L2004 145/1; Directive 2014/65/EU of the European Parliament and of the Council of 15 May 2014 on markets in financial instruments and amending Directive 2002/92/EC and Directive 2011/61/EU Text with EEA relevance (2014) OJ L173/349. 55 See D Busch and C van Dam (eds), A Bank’s Duty of Care (Bloomsbury 2017). 56 Law on financial supervision, 2002 (as amended in 2014), art 4.24a ‘duty of care’ that requires providers of financial services to take the justified interests of their clients into account and to act in this interest as well. 57 FCA, Discussion Paper on a Duty of Care and Potential Alternative Approaches DP18/5 (July 2018). This article provides similar examples with regard to prospective financial services regulation in the USA and Australia. 58 ibid, 35. 59 <https://www.gov.uk/government/consultations/online-harms-white-paper> accessed 8 June 2019; J Naughton, ‘The White Paper on Online Harms Is a Global First. It Has Never Been More Needed’ The Guardian (8 April 2019) <https://www.theguardian.com/commentisfree/2019/apr/14/white-paper-online-harms-global-first-needed-tech-industry-dcms-google-facebook> accessed 8 June 2019. 60 art 11 of the Dutch Health Insurance Act of 16 June 2005. 61 Communication from the Commission (n 34). 62 Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements (2011) OJ C11/1; Guidelines on Vertical Restraints (2000) OJ C291/1. 63 J Balkin, ‘Information Fiduciaries and the First Amendment’ (2016) 49 UC Davis Law Review 1185; J Balkin and J Zittrain, ‘A Grand Bargain to Make Tech Companies Trustworthy: Doctors and Lawyers Are Prohibited from Using Clients’ Information for Their Own Interests, So Why Aren’t Google and Facebook?’ The Atlantic (3 October 2016); J Zittrain, ‘How to Exercise the Power You Didn’t Ask for’ (2018) 96 Harvard Business Review. 64 A Dobkin, ‘Information Fiduciaries in Practice: Data Privacy and User Expectations’ (2018) 33 Berkeley Technology Law Journal 1. 65 L Khan and D Pozen, ‘A Skeptical View of Information Fiduciaries’ (25 May 2019, forthcoming) 133 Harvard Law Review; Columbia Public Law Research Paper No 14-622 <https://ssrn.com/abstract=3341661> accessed 8 June 2019. 66 Sorrell v IMS Health (2011) 564 US 552, 557. 67 Khan and Pozen (n 65). 68 Ezrachi (n 51) 15. 69 Offline there is just one—2000/12/EC: Commission Decision of 20 July 1999 relating to a proceeding under Article 82 of the EC Treaty and Article 54 of the EEA Agreement (Case IV/36.888—1998 Football World Cup), (2000) OJ L5/55. S Weatherill, ‘0033149875354 Fining the Organisers of the 1998 World Cup’ (2000) 21 European Competition Law Review 275. 70 Bundeskartellamt press release (19 December 2017) (n 24); Bundeskartellamt Press Release (7 February 2019). 71 Bundeskartellamt, Case Summary (Decision of 6 February 2019) (n 24). 72 R Thaler and C Sunstein, Nudge; Improving Decisions about Health Wealth and happiness (Yale University Press 2008). 73 I Graef, EU Competition Law, Data Protection and Online Platforms - Data as Essential Facility (Wolters Kluwer 2016); Prüfer and Schottmüller (n 7). 74 See Bundeskartellamt, Case Summary (Decision of 6 February 2019) (n 24). 75 Notably Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contractsCouncil Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts (1993) OJ L95/29; Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC, and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (2005) OJ L149/22; Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights, amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council (2011) OJ L304/65; S Weatherill, EU Consumer Law and Policy (2nd edn, Edward Elgar 2014). 76 Council Directive 93/13/EEC, art 4(2): ‘Assessment of the unfair nature of the terms shall relate neither to the definition of the main subject matter of the contract nor to the adequacy of the price and remuneration, on the one hand, as against the services or goods supplies in exchange, on the other, in so far as these terms are in plain intelligible language.’ 77 Directive 2005/29/EC, art 6(1)d concerning misleading information with regard to ‘the price or the manner in which the price is calculated, or the existence of a specific price advantage’. 78 C Volpin, ‘The Ball Is in Your Court: Evidential Burden of Proof and the Proof-Proximity Principle in EU Antitrust Law’ (2014) 51 Common Market Law Review 1159. 79 R Botsman, Who Can You Trust? How Technology Brought Us Together – and Why It Could Drive Us Apart (Penguin/Random House 2017). F Fukuyama, Trust: The Social Virtues and the Creation of Prosperity (Free Press 1995). 80 As in the transition from Council Directive of 20 December 1985 to protect the consumer in respect of contracts negotiated away from business premises (85/577/EEC) (1985) OJ L372/31 to Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights, amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council (2011) OJ L304/64. © The Author(s) 2019. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contactjournals.permissions@oup.com © The Author(s) 2019. Published by Oxford University Press. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Antitrust Enforcement Oxford University Press

A duty of care to prevent online exploitation of consumers? Digital dominance and special responsibility in EU competition law

Journal of Antitrust Enforcement , Volume 8 (2) – Jul 1, 2020

Loading next page...
 
/lp/oxford-university-press/a-duty-of-care-to-prevent-online-exploitation-of-consumers-digital-Bmg1gwGdnX

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
ISSN
2050-0688
eISSN
2050-0696
DOI
10.1093/jaenfo/jnz023
Publisher site
See Article on Publisher Site

Abstract

Abstract Article 102 Treaty on the Functioning of the European Union imposes a special responsibility on undertakings with a dominant position. Digital dominance based on data but also on network effects, platforms, and first user advantage can lead to extreme inequality in bargaining positions with regard to online services, which can result in unfair treatment of competitors, suppliers, and consumers—such as exploitation and discrimination. Online consumers are the focus of this article. In order to tackle competition issues regarding online consumers, I propose interpreting the special responsibility of digitally dominant undertakings as a duty of care in their regard. Digitally dominant undertakings may not just be burdened by but also benefit from this approach, which increases predictability and trust in online markets. It resembles treating online undertakings as information fiduciaries as has been proposed in the USA, but is different because privacy is not the driving concern here and the context is that of antitrust. I look at standards for triggering the duty of care and make proposals for the application of this norm, but both remain to be specified further. I. INTRODUCTION Online markets are posing new challenges to European Union (EU) competition law. The question addressed by this article is if and when this branch of law could be used to protect consumers from unfair behaviour by parties who exploit their dominance based on data, but also on network, platform, and first- mover advantages. I aim to develop the thesis that: (i) there is a need to counter new problems in digital markets,1 such as discrimination and exploitation of consumers including in terms of anticompetitive personalization of online prices and offers2; (ii) hence, there may be an increased need to have recourse to the special responsibility of undertakings with digital dominance3; (iii) that may take the form of a duty of care vis-à-vis their consumers. Based on these three arguments, this articles sets out a prima facie case for purposes of discussion. First, let me further define the terms that I have just used. Special responsibility is a so far loosely defined category from the case law of the Court of Justice of the European Union (CJEU) that refers to the obligation of dominant undertakings to avoid the abuse prohibited by Article 102 Treaty on the Functioning of the European Union (TFEU) (an open-ended category) and to compete exclusively on the merits. Here, this concept is applied primarily to relationships with consumers in the context of increasingly personalized online offers made by undertakings that enjoy data, network, platform, and first mover-based digital dominance. Because I will be dealing primarily with the interests of consumers (and will focus on exploitation and discrimination rather than exclusion), the focus, however, will be on the data dimension.4 In this context, the personalization of online prices and offers means increasing price discrimination between consumers in the digital realm, and the related online personalization of search and advertising.5 This may have both positive and negative effects. I am concerned here with the potential negative effects—such as redlining (or postal code discrimination), as well as exploiting information asymmetries and the lack of outside options (ie alternative choices) of certain (groups of) consumers.6 Online dominance can be strengthened based on access to and control over different categories of data, in particular data that would be difficult to aggregate and replicate by competitors, and that give the owner a significant and persistent degree of leverage over the data subjects.7 This typically includes personal data, including on willingness to pay that could be based on credit ratings, and location as well as search and purchase histories. In addition, network effects and first-mover effects as well as the ability to exploit platforms as two-sided markets also play a role in establishing the digital dominance (as opposed to traditional offline dominance) that I am primarily concerned with here. In many cases, however, undertakings that enjoy digital dominance are also dominant based on more traditional criteria, such as market share, entry barriers, and an absence of countervailing market power. Finally, a duty of care in tort law traditionally means the obligation to prevent an unreasonable loss from a contracting party. As a civil law obligation it has been elaborated by regulation for financial institutions to cover specific requirements to inform and know one’s customer. I will examine whether a duty of care can also be read into the already existing special responsibility of undertakings with dominance. In addition, a notion derived from data protection law, compliance by design and default (that in turn may largely be algorithm based itself) can be used. As argued here, at its core this would involve an obligation not to exploit unfairly the unequal bargaining power of the consumer. The proposed approach involved, thus, builds on a classic EU legal concept: the special responsibility of dominant companies who are prohibited from abusing their position is brought into focus with regard to natural persons. Essentially, this means reading a duty of care for dominant undertakings into this special responsibility. The existence of special responsibility within competition law is not contested, although it remains loosely defined. Two conceptual steps are required. First, the explicit references to the concept so far in relation to exclusion must be extended to include exploitation and discrimination. This is wholly consistent with the overall structure of EU competition law. Secondly, it must be expressed as a duty of care, which is new in this context, although it exists outside competition law. When developing the duty of care argument, I will largely leave aside the implications for other firms, who may likewise be abused by online dominant undertakings such as those providing platforms who may discriminate against and exploit their online rivals or business customers who depend on their services.8 As will be developed further below, there are also consumers who are at the same time undertakings (self-employed persons, professionals but also members of the Gig economy) or consumer producers (prosumers), and therefore frequently are suppliers and competitors. Hence, there can be no neat distinctions here and some overlap is likely. Because special responsibility is a general term, if the concept is to be used at all, a duty of care that is based on it should also make some sense in a broader (non-data) setting in order to preserve the coherence and, therefore, the effectiveness and legitimacy of EU competition law.9 Hence, I will look at the historical background of special responsibility first. Secondly, I will discuss the questions how the duty of care is triggered and what obligations or principles it would entail. Thirdly, I will discuss the question what are the benefits and the drawbacks of linking digital dominance to the duty of care. Finally, I will conclude whether the thesis set out above makes sense. II. SPECIAL RESPONSIBILITY IN EU COMPETITION LAW In a competition context, the special responsibility is limited to instances of dominance. That is to say where undertakings may act as an appreciable extent independently of, competitors, consumers, and suppliers.10 It is with their independence from consumers that I am primarily concerned. Relation to the prohibition as such As a general principle, the special nature of a dominant position is given by the existence of the exceptional category of dominance abuse as such. This is after all a prohibition for behaviour that would not normally constitute a violation of the competition rules, where an undertaking does not enjoy a dominant position. At the same time the intent of a dominant undertaking is not normally a requirement for an Article 102 TFEU abuse, leaving aside exceptions such as the second leg of the AKZO test on predatory pricing as part of a plan to eliminate a competitor,11 and more recently to eliminate as efficient competitors (AECs) in rebate cases.12 Abuse is an objective concept—one of the reasons why in this context object and effect abuses are not formally distinguished,13 even though in practice at least potential significant effects are increasingly required.14 Hence, there is always an implicit obligation not to infringe Article 102 TFEU for dominant undertakings. They must take care not to do so. The concept of special responsibility Special responsibility can be found as a separate category in the case law of the CJEU dating back to the Michelin case involving exclusionary rebates (1983): A finding that an undertaking has a dominant position is not in itself a recrimination but simply means that, irrespective of the reasons for which it has such a dominant position, the undertaking concerned has a special responsibility not to allow its conduct to impair genuine undistorted competition on the common market.15 Therefore, this concept dates back well before the existence of the Internet itself, let alone pervasive online platforms or digital dominance. It is clear, however, that it relates to a measure of self-constraint to be exercised by the undertaking in question, which is not limited to a specific time period. Yet, in spite of its (for an EU competition law concept) long pedigree, the concept of special responsibility and repeated references in competition cases has not been specified and remains open textured—or less fancifully: open to further interpretation. With regard to data issues, this may be interpreted as ‘a feature, not a bug’, or more of an opportunity than a problem. For the purposes of my argument, this open texture allows the claim that the special responsibility does not just concern the structure of competition and, therefore exclusion, but should include discrimination and exploitation. In fact, it would make no sense for undertakings to have a special responsibility vis-à-vis their competitors, but not their consumers. Applying the special responsibility versus consumers is all the more plausible because the EU antitrust regime sees consumer welfare as a core objective and exploitation and discrimination as actionable harms, and especially now that fairness itself is increasingly recognized as a legitimate competition concern. Objections have been raised to reading specific obligations into the special responsibility, notably claims that this might protect competitors and, therefore, obstruct competition on the merits.16 We can leave this objection aside as well as the tricky question what constitutes competition on the merits in the first place. (Broadly but elliptically, competition on the merits is to say: competition within the confines of competition law.17) This is because by competition on the merits by definition concerns exclusion of competitors (a type of problem that we are not engaging with here) and is not directly related to directly protecting consumers from exploitation or discrimination. Harmful versus beneficial discrimination With regard to discrimination in an economic sense it should be added that this is not always harmful, and can even be beneficial from a total and even a consumer welfare perspective because it may increase the number of customers who are able to purchase a good or service.18 This occurs where the wealthier are willing to pay more for the same service and effectively cross-subsidize its purchase for those who can only afford to pay less. However, this optimistic argument suggestive of progressive redistribution appears to break down at least to some extent in an online world. There all the information is on the side of the seller and in fact consumers who have fewest alternatives (and who are likely to be poorer and still less informed) pay the highest prices as a result.19 If a system of perfect price discrimination were to emerge, it would even allow the entire surplus to be appropriated by the producers. Evidently, there are also ethical arguments against such types of price discrimination.20 Distinguishing the three overlapping concepts Having identified and briefly discussed the obligation not to infringe, special responsibility, and duty of care, this now raises the question: (i) how the special responsibility is different from (ii) the obligation not to infringe in general and (iii) from the duty of care postulated as a possible pre-emptive remedy above? Are these three sides of the same coin or manifestations of the same phenomenon, and what would we gain from specifying them? It appears that these are in fact to some extent three distinct categories that overlap in part. Accordingly, some differences in scope are involved. First, the general obligation not to infringe the competition rules that applies to all undertakings generally requires the dominant undertaking to behave as if it is subject to competitive pressure. However, this undertaking must not just refrain from exclusionary behaviour such as the application of a price squeeze—or from exclusionary rebates such as in Michelin. It may also have a positive duty, such as to supply competitors, as for instance in the EU Commission’s Microsoft Decision as upheld by the CJEU in 2007. Here, Microsoft was found to be under an obligation to supply competing providers with its interface documentation enabling these competitors to produce software-based services running on work group servers that used Microsoft’s operating system. This aimed to ensure interoperability of their products, so the market would remain broadly innovative and not just dependent on Microsoft.21 Secondly, the requirement of an active avoidance of competition law infringements that in principle applies to all undertakings is developed further just for dominant undertakings by the special responsibility first cited in the above-mentioned Michelin Case (1983). This occurred in relation to Michelin’s exclusionary rebate scheme concerning new replacement tires for large vehicles (trucks, busses, etc) on the Dutch market, where Michelin held a 57–65 per cent market share compared to 4–8 per cent for its largest competitors. This behaviour was held to have affected negatively the structure of the market.22 The responsibility is special insofar as it only rests on dominant undertakings and involves a requirement to take action in order to protect competition as well as (as argued above) the interest of the consumer. The special responsibility relates to the obligations of a dominant undertaking in a specific case. This is in line with the fact that Article 102 TFEU does not provide a limited list of infringements but a dominant undertaking has the obligation to act as a normal undertaking would if constrained by effective competition, as well as (I would submit) fairly in relation to its consumers.23 Thirdly, the duty of care that flows from the special responsibility of dominant undertakings is not just a negative duty to abstain from infringing but also a duty to take positive steps in order to avoid such an infringement occurring. In relation to consumers, this entails an obligation to act not just as if disciplined by competition, but fairly as defined by norms that are pertinent in the context of exploitation and discrimination—such as privacy in the German Facebook Decision (2019) which found a violation of privacy by a dominant undertaking ipso facto constituted an antitrust infringement.24 Regarding consumers, the duty of care thus requires the undertaking not to act unfairly, as if constrained by the bargaining position of the consumer in a moral sense, and by behavioural norms in antitrust that are grouped under fairness as reflected in the prohibitions on exploitation and discrimination. This serves to guarantee that the unequal/asymmetrical position that in fact exists is not exploited. In other words, the interest of the consumer must be taken into account in a meaningful way. All three of these concepts are expressions of an obligation on dominant undertakings not to infringe Article 102 TFEU. However, there is a sliding scale active here both in terms of specificity and in terms of intrusiveness. In the absence of further practical application of the concepts involved, it does not seem productive to attempt to be more specific at this stage. We now move on to what is the precondition for the duty of care, digital dominance, and its triggers. How does an undertaking know it is dominant? As it is essentially based on prevention, the prohibition on dominance abuse, such as the special responsibility and the duty of care, presupposes not just the existence of a special position but also the awareness on the side of the undertaking concerned of the existence of this dominant position. Yet, how does an undertaking know that it is dominant in order for both the rule and the responsibility to apply? Generally, in Article 102 TFEU cases, market definition can be complex as can be dominance analysis in the sense of showing independent behaviour. This need not be a prohibitive bar to making a dominance assessment, however. For instance, Heike Schweitzer and others have proposed in their 2018 report to the German Economics Ministry that an abuse as such can itself already form decisive proof of the existence of a dominant position and, hence, replace a more formal prior market definition.25 More recently, similar advice has been proffered by Schweitzer, Crémer and de Montjoye to DG Competition of the European Commission (2019).26 In the first place, an undertaking holding a legal or factual monopoly inside or outside the online world should obviously expect that it might well be dominant and carry out due diligence to that effect. The same would hold for an undertaking that was subject to a previous finding of dominance by an authority or ultimately a court. Admittedly, such cases are relatively rare and the duty of care might overlap with obligations imposed or commitments offered by the dominant undertaking. A significant market power (SMP) regime that identified positions of dominance ex ante would play a similar role—if it existed in the digital realm, as it now does for electronic communications network operators in the EU. Secondly, especially in the online world—apart from various types of network, platform, and first-mover effects—could controlling certain data lead to a type of digital dominance which is established more easily or possibly exists per se? Replicability is evidently an issue here. It is also conceivable that a regulatory trigger would be created. Controlling personal data, lack of replicability, network and first-mover effects, and controlling a platform or service might be such criteria. Establishing the relevant market in itself could be an issue here: conceptually, where there is (near) perfect discrimination this might even be the individual consumer.27 As mentioned, an alternative would be for the nature of the behaviour itself to provide proof of dominance, irrespective of market definition. I expect that should this exercise be considered—or by failed attempts proven to be—too difficult in an antitrust context, it could certainly become food for regulation, where it might be applied, for instance, to all undertakings using personal data in their offerings.28 Thirdly, can controlling a program, standard, platform, algorithm, or data lead to a position of super-dominance, as has been variously alleged of Microsoft, Intel, Google, Amazon, Uber, and others? Super-dominance is not an established concept in EU competition law and appears to concern entrenched monopolies. This is illustrated by the fact that the main reference is still Advocate General Fenelly’s Opinion in the 1999 Compagnie Maritime belge Case: Article 86 cannot be interpreted as permitting monopolists or quasi-monopolists to exploit the very significant market power which their superdominance confers so as to preclude the emergence either of a new or additional competitor. Where an undertaking, or group of undertakings whose conduct must be assessed collectively, enjoys a position of such overwhelming dominance verging on monopoly … it would not be consonant with the particularly onerous special obligation affecting such a dominant undertaking not to impair further the structure of the feeble existing competition for them to react, even to aggressive price competition from a new entrant, with a policy of targeted, selective price cuts designed to eliminate that competitor.29 We will leave aside the obvious exclusion context that motivated this quote under the assumption that super-dominance could be equally noxious in relation to other types of abuse (on which more below). Super-dominance should not be required for a duty of care to hold.30 However, extreme market positions may mean that the special responsibility is intensified as the undertaking involved would both enjoy a greater asymmetry of power and a greater ability to make amends. It would presumably also be more aware of the antitrust responsibilities that its position entails. Finally, how would the undertaking know that an infringement that ought to be avoided could take place? This question arises especially with regard to fines, making the instrument of periodic penalty payments perhaps more suitable. It could be solved based on process/criteria like the proposed duty of care (or by compliance by design and by default as in the General Data Protection Regulation31), or compliance by algorithmic design,32 which could also exonerate the undertaking if the relevant criteria had been met. Perhaps on the supply side, there are lessons to be learned from risk management here.33 My argument is that these requirements could be qualified as a direct duty impendent on dominant undertakings as a duty of care obligation in competition law. The reasons for this are set out in more detail in Section IV below. However, as a technical matter the relevant norm could equally well be formulated in soft law,34 or in fact by means of direct intervention as a matter of regulation. Triggering the duty of care Determining the addressees of this duty is obviously linked to the question posed above: how do they self-identify? It appears obvious that there should not be simple quantitative turnover or market share thresholds to trigger the duty of care. Hence, we can also define special responsibility and the duty of care based on what it is not—an approach based on fixed market shares or similar that has been the focus of earlier critiques pleading to ditch the concept, largely based on the straw man argument that a market share approach would be simultaneously unavoidable and inappropriate.35 It is difficult to see how such fixed thresholds could work in any event because it would presuppose an essentially uncontested market definition. Instead, it may be preferable to develop the use of dominance criteria that do not rely on a predefined market such as a demonstrable lack of choice/absence of outside options for consumers or control over data and/or over an essential platform. In other words, it may be possible to define digital dominance in an online setting qualitatively, thereby making quantitative threshold irrelevant while securing the benefits of the special responsibility in a proportionate manner. An alternative is using the fact that certain behaviour is taking place as prima facie evidence that dominant position is in place—much as an abuse can sometimes do that. Such approaches have been advocated in above-mentioned reports by Schweitzer and others on reviewing the German and EU competition rules, respectively the application of these rules.36 Relevant types of abuse A distinction based on the type of abuse concerned may well be useful here, also because new forms of abuse appear to be relevant. In this categorization we should also include whether abuse targets (i) competitors, (ii) suppliers, (iii) purchasers at large, or (iv) consumers (individual end users).37 Arguably, this is a sliding scale, where consumers require most protection and competitors least as is illustrated by the AEC test mentioned above. The steadily increasing group of self-employed in the middle where categorization between undertaking and private individual is largely arbitrary appears to be the most problematic category in this case. As undertakings, they are likely to be dependent on dominant firms and arguably unlikely to meet AEC-type tests. However, as citizens they do vote, creating political sensitivity to their position, and moreover they may be affected directly as consumers too. Notably, they are in large part on the supply side of the market.38 Finally, consumers–producers (prosumers), for instance, in the context of energy are also on the supply side. Although at the present stage this is not a key feature of my argument, in future the supply side issues may have to be tackled too. Below I will distinguish between the four broad categories of abuses of exclusion, exploitation, discrimination, and sui generis abuses. Exclusion of competitors is traditionally favoured as the primary object of public competition enforcement.39 It is also the only area where (in 2009) the Commission has so far provided explicit guidance: similar guidance is lacking with regard to the other three categories of abuse identified below.40 It has also been the focus of sectoral rules on dominance, such as those regarding SMP for electronic communications (based on EU law, and at least under Dutch law, for healthcare providers and postal operators as well).41 A further example is exclusionary rebates by the nationally dominant operator weakening already marginal residual competition in Post Danmark II (2015). Here, the Court’s reasoning appeared to be that because the existence of a position of dominance means only limited scope for competition remains, further restricting it is in violation of the special responsibility of the dominant operator (which moreover was propped up by exclusive rights). This application of special responsibility seems to refer mainly to exclusion.42 Solving the exclusion puzzle after all leads to self-correcting markets, rendering direct protection of consumers superfluous. Recently exploitation has become a more relevant form of abuse now, for instance, excessive pricing cases are being pursued both at EU and national level (albeit with mixed results), for instance, in the area of pharmaceuticals.43 The relevance of special responsibility here is possibly linked to notions of fairness in the context of exploitation. (The vertical—between suppliers and purchasers and horizontal—between competitors—substantive fairness dimensions introduced by Michael Trebilcock and Francesco Ducci may be useful here, specified for consumers.44) This category is important to consumers in general, but perhaps less so in the context of different degrees of personalization. This is because as the test for excessive prices for instance is construed today it may for practical reasons be more effective to tackle the problem as one of discrimination (see below). This raises the issue of what legal basis would be appropriate because tackling such discrimination at least against individual consumers or defined groups of consumers who are not also suppliers or competitors cannot be done under Article 102(c) TFEU on discrimination against undertakings where exclusion must be shown.45 Instead, it would arguably have to be pursued as discrimination under the Article 102(a) TFEU fairness standard as unfair pricing—and preferably not as excessive pricing under the same provision because this involves a forbidding standard and difficulties of proof.46 What exactly would be the norm for this type of discrimination to be found remains to be specified as a matter of law. The category of discrimination, in the first place this covers discrimination of various types that occur notably in the absence of vertical integration (so there is no exclusion) and beneath the threshold of exploitation47—the prices are not excessive, or are not shown to be excessive for practical reasons.48 Secondly, it involves price discrimination—including increasingly personalized prices on online platforms targeted at consumers, as well as increasingly personalized online search and advertising using various layers of filters and tracking behaviour as well as using pre-existing data. Such price discrimination is a most relevant category for the duty of care vis-à-vis consumers. Actual competition cases however are so far in short supply, although studies on personalized pricing are being carried out by some competition authorities.49 Yet as mentioned just above, this type of discrimination could be covered by Article 102(a) TFEU as unfair pricing. Finally, there is the category of sui generis abuses that do not neatly fit the above distinction between exclusion, exploitation, and discrimination is associated with actions that deteriorate the fabric of competition at large and other sui generis type abuses that should also be avoided under the duty of care. The examples appear to be: abuse of procedure—such as in the EU and Italian AstraZeneca Cases (largely about exclusion), which is related to the notion that even the exercise of legal rights can under certain circumstances constitute an abuse50; harming the process of competition—regarding ‘as efficient competitors (AEC)’ (again exclusion); and/or harming the freedom of consumer choice, including variety and quality—arguably to be policed as a form of exploitation, but going beyond efficiency considerations. It is primarily this latter category that is relevant to arguments based on fairness. These four categories of abuses (exclusionary, exploitative, discriminatory, and sui generis) are themselves closely linked to the objectives of EU competition policy that were recently subject to a proposal for realignment by Ariel Ezrachi (in a different form than presented here) in a paperon competition law in the digital economy.51As a general trend greater pluriformity, with an emphasis on fairness, appears to be emerging. This will create more room for the duty of care approach discussed here. III. THE DUTY OF CARE Below we will examine the various issues that are raised by the duty of care as an extrapolation of special responsibility. Origins of the duty of care Because that is where I first happened to come across it, I have originally borrowed the notion of the burden of care from financial services regulation, such as insurance or home owner credits/mortgages. However, there is a fair number of different sources that are relevant. These are listed here in broadly chronological order. The use of the concept of duty of care, not only (i) in torts such as negligence where the first law and economics reasoning originated52—it is more efficient to have the dominant undertaking take precautionary measures than for individual consumers to have to sue for damages, in particular when the dominant undertaking has the data required to make the requisite decisions; but also (ii) as the duty a trustee owed to a beneficiary; and (iii) in contract law generally; as well as (iv) the explicit protection of the weaker party in contracts that aim to tackle structurally unequal bargaining positions such as in consumer law, labour law, and tenancy law; and finally (v) more explicitly in statutes such as those regarding as the above-mentioned financial services. In the common law tort of negligence, the test is whether the harm could have been foreseen, if there was proximity between the parties and if imposing a duty of care is fair, just, and reasonable. This is not fundamentally different from civil law approaches. The private law test may seem like ex post reasoning. However, in professions or activities that would normally risk being subjected to this standard, there will obviously be a prophylactic effect as well when measures are taken to avoid not just liability, but harm itself. In the UK, courts have generally treated a breach of competition law as a tort in the context of compensation. Private antitrust enforcement, especially since the implementation of the 2014 Damages Directive,53 is already relevant in this context, and will be all the more so if the duty of care is developed further. The above-mentioned regulatory approach can be traced to regulated financial services, where the obligations of Markets in Financial Instruments Directive I and II54 for financial institutions to inform and know their customers have developed into a duty of care.55 In the Dutch financial services act, this is spelled out in so many words.56 Recently, the UK Financial Conduct Authority (FCA) has proposed a similarly explicit approach.57 [A] ‘duty of care’ is a positive obligation on a person to ensure that their conduct meets a set standard. In the context of firms dealing with consumers, this normally means an obligation to exercise reasonable care and skill when providing a product or service.58 This development is significant also because under concurrent (parallel) powers with the Consumer and Market Authority (CMA), the FCA is responsible inter alia for applying the general competition rules to the financial sector in order to ensure effective competition there in the interest of consumers. If it adopts the proposed approach, it could therefore easily take the route examined in this article: applying a duty of care under the competition rules. Also in the UK, there is a consultation ongoing on the creation of a new independent regulator for the online sector, replacing self-regulation with public powers to intervene with respect to online harms cited as ranging from terrorist content to cyberbullying. Essentially, the new regulator, if indeed instated following the consultation, will in future balance free speech against a duty of care, which is to be the guiding standard for this approach to illegal and unacceptable content. The proposed duty of care presupposes the adoption of interim standards by government as well as definitive standards and a detailed code of practice by the regulator. Possible sanctions include fines against the undertakings involved and managers may be personally liable.59 In the Netherlands, health insurers also have to meet an explicit duty of care, which is to ensure that all care which is included in the standard insured package (set by Parliament) is actually available to patients.60 This may extend to finding non-standard solutions for bankrupted hospitals and planning cross-border provision of services. It is not the closest comparator to the duty of care intended in this article but it does involve taking the needs of a contracting partner into account both at the individual and at a meta-level. As part of an overall trend towards a more comprehensive type of corporate responsibility vis-à-vis retail clients, it is at least context relevant. Finally, the regulatory approach is one whereby a presumed duty of care with regard to the special responsibility of data dominant undertakings is specified in specific steps that they should take to avoid harm to their customers. This would obviously also be relevant in private antitrust suits. It seems self-evident that effective private enforcement on this point would be facilitated considerably by specific guidance that might be similar to what we have seen for dominant companies and exclusion,61 but also for horizontal and vertical restraints.62 This list could obviously be extended and refined further, but appears sufficient for my present illustrative purposes. However, before moving on to the various requirements of fairness, I will briefly engage in more detail with the US concept of information fiduciaries, which also involves invoking a duty of care. Because I have so far largely reasoned from an EU context, this requires a few more words of introduction and merits a separate section. Information fiduciaries In the USA, a debate is ongoing on the merits of treating online platforms as information fiduciaries by analogy to the position occupied by lawyers, doctors, and accountants who have been entrusted with privileged information by their clients.63 For the professionals mentioned, it is considered a legally acceptable exception to the First Amendment to the US constitution protecting freedom of speech to impose limits on their subsequent use of the information that they collected in the exercise of their duties. As advocated by Jack Balkin, Jonathan Zittrain, and others, the analogous fiduciary role involved regarding online platforms would comprise both a duty of care and a duty of loyalty to the client. The current US academic debate (the concept of information fiduciaries itself appears to be older) is not just about the legal justification of this approach and how it might be implemented in practice.64 It also revolves around the question whether its introduction as part of a ‘grand bargain’ with online platforms would not in effect mean pandering to them with a soft touch that might obstruct attempt to introduce more serious regulatory intervention. This would risk being counterproductive with respect to protecting the interests at which the fiduciary approach is aimed, notably privacy.65 Here, I will merely note several parallels and differences between the duty of care based on the fiduciary approach for online platforms in general in the USA and the duty of care based on the special responsibility of undertakings with digital dominance that I am advocating for the EU. In both cases, a concern over the unfair exploitation of a fundamentally asymmetric data position is addressed by imposing a duty to take the interest of the consumer into account ex ante in a compliance based manner. If successful, both might obviate the need for detailed regulation. A key difference is that in the EU, the General Data Protection Regulation in large part already meets the stricter privacy-related needs that much of the US discussion on the merits of information fiduciaries is in fact concerned with. Free speech as exercised by undertakings is in the EU, in any event typically not accorded the wide berth found in the USA under the First Amendment that for instance protects the sale of personal data by undertakings to third parties—meaning that regulation of such acts may well be unconstitutional.66 Moreover, as critics there point out, an information fiduciary approach cannot be an effective substitute where market power requires antitrust action and may at best be complementary.67 However, in a second best world, this dual solution may not be available because if a grand bargain between online platforms and their political opponents to instate the former as information fiduciaries in fact occurs, it is unlikely that intervention against market power will be on the menu for the foreseeable future. In contrast, because I start from the assertion that antitrust is in fact applicable to unfair treatment of consumers it is precisely my intention to invoke the competition rules and, therefore, address economic and not strictly privacy related concerns. This also includes the full gamut of antitrust remedies that would be available against infringements even if initially undertakings are expected to police themselves as a matter of efficacy. This would self-evidently include prosecution of cases, where undertakings demonstrably violate their own stated policies in this regard. Finally, as mentioned at the outset, in EU competition law, the special responsibility underlying the duty of care advocated here applies more narrowly to dominant undertakings, whereas the information fiduciary status would cover all online platforms entrusted with personal data. Hence, there is a significant difference in scope between the two approaches (and presumably in proving their applicability) even although in both cases a duty of care plays a central role. Fairness requirements As discussed above the duty of care is intrinsically linked to fairness. Ezrachi in his above-mentioned paper on the goals of competition policy and the digital economy has grouped six types of intervention under fairness: (i) counteracting unfair market practice or illegitimate wealth transfers; (ii) intervening against discrimination by dominant undertakings; (iii) counteracting misleading information; (iv) counteracting extreme instances of asymmetric information; (v) fighting unfair handling of data, and violations of data protection and privacy rights; and (vi) imposing a fiduciary duty requiring ethical handling of data.68 Although all these categories have some bearing on special responsibility expressed as a duty of care made here, two are directly relevant at least at the current state of the argument: category (i) on illegitimate wealth transfers and sharp business practices and (ii) on discrimination, of which there are as of yet no known competition cases regarding online consumers.69 A related but distinct is category (v) on privacy that has recently increased in relevance with the Facebook case undertaken by the German Bundeskartellamt.70 As I have briefly mentioned above, the Bundeskartellamt ruled on 6 February 2019 that a violation of data protection rules by Facebook in the context of the terms imposed on its consumers ipso facto also constitutes an exploitative abuse of its dominant position.71 Introducing norms from outside the realm of traditional antitrust, such as privacy and possibly consumer protection, and enforcing them by means of the competition rules presents a novel and potentially effective strategy in countering abuses of digital dominance that are otherwise difficult to classify. However, it remains to be seen how this will pan out on appeal and elsewhere in the EU. Obligations formulated as principles Regarding the obligations involved the proposed approach primarily seeks to reestablish the balance between the contracting parties. This is in line with insight in behavioural economics, which suggests that on the demand side consumers must be protected from themselves and exploitation, for instance, by ensuring that default options are on average beneficial.72 It involves elements such as know your customer and a duty to actively inform about downside risks and to follow appropriate procedures and keep records. In addition, it should be clear that offers which are deliberately exploitative or otherwise predictably unfair are not to be made at all. In this context, a blanket obligation to act reasonably seems insufficient. In order to be effective the norms to be respected must be clear and simple. Yet, because the EU competition rules may not count many words in the TFEU itself—but are complex in practice—something more specific is required. Ideally, the duty of care would be defined by general principles that lead to individualized customer specific obligations—the implementation of which would be possible precisely because of the data available. As a result, the duty of care should become not just a compliance issue but one of compliance by default and design. This is a matter of setting up internal checks and balances and screening out future problems that would promote the consumer interest potentially far more effectively than the system we have today based on analog standards and ex post paper-based enforcement. A modest proposal Principles for a duty of care with regard to the online offerings of digitally dominant undertakings can take various shapes. This goes beyond ‘know your customer’, the requirement found in financial services regulation, which may logically be assumed a given in a data-driven environment. In some (or many) dimensions, the digitally dominant undertaking may know more about the customer/consumer than he/she knows about himself/herself. The question is what action a digitally dominant undertaking would have to take based on this knowledge. Such principles might include preemptive behavioural remedies such as the five listed here: protect your consumer by using the relevant consumer data to screen also for the consumers downside and use this information to the consumer’s benefit; do not engage in negative discrimination by treating a consumer differently in order to exploit their inferior information position or lack of outside options; apply transparency by being clear about the consumer’s risks in the transaction, their outside options, and how the choices offered to them compare to those of others; provide a fair non-choice outcome by using smart default settings that ensure consumers reach a normally low-risk outcome if they do not actively pursue an individual choice; take responsibility for results, meaning that apart from procedural compliance there should be substantive compliance documented by actual performance. These ideas are just intended by way of a modest first proposal and as mentioned draw some inspiration from behavioural economics. It includes elements that would be less appropriate in offline markets because there for instance the offer is generally available, prices are posted physically and publicly, and other customers can similarly be directly observed. More importantly, individual consumer data are not collected or used at a comparable level of intensity. Failure to abide by these or similar principles that flesh out the duty of care as an infringement of the prohibition on dominance abuse would trigger private liability as well as antitrust fines. Enforcement It is early days to write on the competition law enforcement dimension of the duty of care, other than to speculate that it is likely to facilitate enforcement in general terms because it compensates for unequal bargaining power and information asymmetry. Thus, in large part, it will ensure prevention, and ensure that the relevant information is available and generally acted upon by the undertakings involved. Also, there are several clear advantages to a compliance based duty of care for dominant companies in terms of actual competition law enforcement: detection is easier as there would not only be a significant degree of self-policing but possibly self-reporting by the undertakings involved. This means there can be more general process based on systemic checks by the authorities; accompanied only by investigation and sanctions of individual infringements that are more serious. Finally, private enforcement would presumably play a role here because the exploitation norm involved is especially relevant to consumers—not least including the (nominally) self-employed—and prosumers. As mentioned above the duty of care principles, if violated, would trigger liability. This in turn could promote effective enforcement and thereby deterrence. Relationship with other remedies Obligations to share data with competitors—and more radically general transparency of data—may be a solution to some problems with regard to exclusion of competitors.73 However, in both cases, they are so only indirectly for exploitation of consumers in online markets (while evidently raising data protection issues too). These are therefore not alternatives to the duty of care obligations outlined above. A duty of care could exist alongside data protection specific remedies targeting exploitation such as imposed those in the German Facebook case because it is not restricted to privacy.74 Unlike the duty of care all three of these remedies are imposed ex post. The relationship with EU consumer protection rules, notably unfair trading practices, is more complex and remains to be examined further.75 These rules do not allow consumers or their protectors to challenge price levels as such76—although they impose certain transparency requirements on the way a tariff is construed that could capture certain forms of discrimination.77 Finally, the duty of care approach is compatible with largely algorithm-based compliance but also with approaches that require specific problematic outcomes to be justified and access to the algorithms of digitally dominant undertakings as systems of antitrust control. However, this is a topic in its own right that I can do no justice to here. Instead, we will look at why a duty of care may be a balanced approach from an industry perspective as well. IV. WHY IS LINKING THE DUTY OF CARE TO DIGITAL DOMINANCE ATTRACTIVE? At this stage, the principles set out above by way of example do not appear to be too demanding as requirements to be realistic, and we will for the moment assume that the underlying EU law account is sufficiently forceful. So if this is indeed prima facie a credible model, is this model attractive and in particular, could it be so for undertakings as well as consumers? Five reasons the duty of model could be attractive to undertakings Without trying to be exhaustive there appear to be at least five reasons why this is the case. In the first place, the general attraction is that the duty of care locates the burden of proof with the party that controls both the data and thereby directs the consumer, while of course it also controls its own behaviour—largely independently from the other market participants as it is dominant. (This seems in line with general requirements of proximity of proof.78) In addition, it carries an additional (special) responsibility as well as a special liability. Because this party controls the playing field, an additional burden or set of requirements designed to compensate for this advantage may be considered fair. Secondly, from an effective enforcement perspective, the duty of care is also attractive, not just regarding the allocation of proof but also because it is largely preventive rather than corrective in nature. This ex ante prophylactic is generally considered preferable not only as deterrence but also to avoid costly mistakes (including reputational damage), fines and private damages actions ex post. From a business perspective, thirdly, promoting trust between the parties involved, which is to the benefit of both consumers and data dominant undertakings as well as for the successful functioning of online commerce as a whole.79 Similarly, and fourthly, fewer surprises are likely to occur also for the dominant party—this has the advantage of creating a more predictable environment and greater investment certainty, again promoting online commerce. Finally, if successful the duty of care mechanism may obviate the need for potentially equally resource intensive and burdensome sectoral regulation on this count. To some, listing the final point as a possible advantage may seem like a provocation. In the US context, critics see legislation creating online information fiduciaries as increasing the risk that the adoption of more effective regulation will be blocked. As I have stated above the EU context already has more apposite regulation in place—notably the General Data Protection Regulation. Moreover, I am proposing an interpretation of the law as it stands that would imply a significant extension and not a potential reduction of the practical scope and application of antitrust, which in my view is implicit in the concept of special responsibility. This poses more serious constraints notably on economic abuses than the construct of an information fiduciary would, which is more likely to be restricted to privacy. Moreover, as we have seen in the German Facebook case where privacy violations were addressed by antitrust, the two regimes can be mutually reinforcing. There is clearly a balance involved here. Increased trust, more legal certainty, and arguably less need for new sectoral regulation will at least partially offset increased regulatory and compliance (and efficiency) costs. Depending on whether their perceived outside option is zero regulation and no effective competition law enforcement in perpetuity, or some considerable effort in this direction, the sum for undertakings involved in this exercise, with at least in large part algorithm based compliance by design, may well be positive. Hence, this approach appears to involve a reasonable a trade-off between legal certainty to the benefit of the digitally dominant party in exchange for fairer treatment, where the consumer is protected against excessive terms and other unfair practices in cases where the position of the party is highly asymmetrical and (in particular) data based. General applicability It could be argued that for an ex ante remedy or special responsibility (translated into a duty of care as proposed here) to be adopted it should make sense outside the realm of digital dominance as well, because the special responsibility exists there too. However, in my view, it could also be approached as an experiment that is justified in the online world/data realm, and sufficiently specific to warrant special treatment at least initially. This is also the case because outside the online world, given their information position dominant undertakings may not be equally aware of what is good or detrimental for their consumers: this makes for a different setting. At the same time this experiment would be grounded in experience: when operationalizing the duty of care we could learn from the transition of preexisting consumer rights such based on doorstep selling (such the ability to revoke the sale within a specific time period)80 to the inside pocket (smartphone) selling environment. V. Conclusion In this article, I have argued in support of the thesis that the already existing special responsibility of dominant firms in EU law can, and should, be translated into a duty of care for digital dominance. First, this coincides with their private law responsibilities: a digitally dominant undertaking can and should avoid causing predictable harm to its consumers. Secondly, explicit requirements could be set to facilitate compliance. This use of the duty of care would enhance the position of consumers in what are today fundamentally and arguably unjustifiably unequal exchanges while promoting trust between the parties, as well as predictability and investment security for the undertakings involved. If successful it may render sectoral regulatory regimes superfluous, and could fit in with a broader trend towards more comprehensive corporate responsibility towards consumers. There are parallels here with the concept of information fiduciaries found primarily in the US debate. This concept would however on the one hand apply more widely—to all online platforms instead of digitally dominant undertakings—and would on the other hand be limited largely to privacy issues, whereas the duty of care that I have argued for here focuses on a broader abuse of economic power. This also makes the criticism that arises in the US context that applying the concept of information fiduciaries in a grand bargain with online platforms may deflate the drive for more effective regulation less relevant in the EU. Here, the General Data Protection Regulation already covers the most pertinent privacy issues and a duty of care based on the special responsibility of digitally dominant undertakings enforced under antitrust would apply as a complement, not a substitute. It would also increase the bite and effectiveness of EU antitrust. Evidently, both the usefulness of this approach and whether it will be accepted and applied depend on the criteria that are used to determine: (i) whether an undertaking enjoys digital dominance and (ii) what exactly the duty of care involves. Also, (iii) the legal basis requires further clarification. Above, I have made some suggestions on all three counts. However, these important elements remain to be specified further and tested in practice. Currently on part-time secondment to the Free University of Amsterdam (Vrije Universiteit: VU) and extramural fellow of the Tilburg Law and Economics Center (TILEC). He is a former professor of economic law at the universities of Groningen and Tilburg. This article was written in a personal capacity. Comments by David Gabathuler, Inge Graef, Viktor Mayer-Schönberger, Martijn Snoep, Hans Vedder, Rein Wesseling, and from workshop participants in Lund, Nijmegen and Tilburg, as well as two anonymous referees are gratefully acknowledged. An embryonic version of this article appeared in Dutch in (2019) 22 Markt & Mededinging 105. Footnotes 1 See Heike Schweitzer and others, Modernising the Law on Abuse of Market Power: Report for the Federal Ministry for Economic Affairs and Energy (Germany) (17 September 2018) <https://ssrn.com/abstract=3250742> accessed 8 June 2019. 2 See European Commission, Personalised Pricing/Offers in the European Union (2018) <https://ec.europa.eu/info/publications/consumer-market-study-online-market-segmentationthrough-personalised-pricing-offers-european-union_en> accessed 8 June 2019 and recent OECD papers such as Online Personalisation from the Perspective of a Multidisciplinary Market Authority—Note by the Netherlands, 28 November 2018 and Personalised Pricing in the Digital Era – Background Note by the Secretariat, 28 November 2018. 3 I have chosen the term digital dominance over alternatives such as online or data dominance because it is more inclusive. See M Moore and D Tambini (eds), Digital Dominance: The Power of Google, Amazon, Facebook, and Apple (OUP 2018). 4 M Stucke and A Grunes, Big Data and Competition Policy (OUP 2016). 5 R Woodcock, ‘Big Data, Price Discrimination and Antitrust’ 68 (2017) Hastings Law Journal 1371; Christopher Townley, E Morrison and K Yeung, ‘Big Data and Personalised Price Discrimination in EU Competition Law’ 36 (2017) Yearbook of European Law 683; O Bar-Gill, Algorithmic Price Discrimination When Demand Is a Function of Both Preferences and (Mis)Perceptions (Harvard Law School 2018) <https://bit.ly/2vJsx3r> accessed 8 June 2019. On price discrimination more broadly, see OECD Background note DAF/COMP(2016)15, Price Discrimination. 6 C O’Neill, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy (Penguin/Random House 2017). 7 See eg J Prüfer and C Schottmüller, ‘Competing with Big Data’ CentER Discussion Paper, Vol 2017-007. 8 Examples are the Google Cases. See: ‘Antitrust: Commission Fines Google €2.42 Billion for Abusing Dominance as Search Engine by Giving Illegal Advantage to Own Comparison Shopping Service’ (Brussels, 27 June 2017) IP/17/1784. Decision of that date in Case AT 39740 Google Search (Shopping). See ‘Antitrust: Commission Fines Google €1.49 Billion for Abusive Practices in Online Advertising’ (Brussels, 20 March 2019) IP/19/1770. Decision of that date in Case AT 40411 Google Search (AdSense). See also ‘Antitrust: Commission Fines Google €4.34 Billion for Illegal Practices Regarding Android Mobile Devices to Strengthen Dominance of Google's Search Engine’ (Brussels, 18 July 2018) IP/18/4581. Decision of that date in Case AT 40099 Google Android. 9 W Sauter, Coherence in EU Competition Law (OUP 2016). 10 Case 85/76 Hoffmann-La Roche & Co AG v Commission Judgment of 13 February 1979, ECLI:EU:C:1979:36, para 38. 11 Case C–62/86 AKZO Chemie BV v Commission Judgment of 3 July 1991, ECLI:EU:C:1991:286. 12 Case C–413/14 P Intel Corp v Commission Judgment 6 September 2017, ECLI:EU:C:2017:632. 13 Case C–549/10 P Tomra Systems ASA et al. v Commission Judgment of 19 April 2012, ECLI:EU:C:2012:221. 14 Case C–525/16 MEO—Serviços de Comunicações e Multimédia SA v Autoridade da Concorrência Judgment of 19 April 2018, ECLI:EU:C:2018:270; Case C–177/16 Autortiesību un komunicēšanās konsultāciju aģentūra/ Latvijas Autoru apvienība v Konkurences padome (AKKA/LAA) Judgment of 14 September 2017, ECLI:EU:C:2017:689; Case C–177/16 Opinion of AG Wahl, 6 April 2017, ECLI:EU:C:2017:286. 15 Case 322/81 NV Nederlandsche Banden Industrie Michelin v Commission Judgment of 9 November 1983, ECLI:EU:C:1983:313, para 57. See also Case T–228/97 Irish Sugar Plc v Commission Judgment of 7 October 1999, ECLI:EU:T:1999:246, para 112; Case T–203/01 Manufacture française des pneumatiques Michelin v Commission Judgment of 30 September 2003, ECLI:EU:T:2003:250, para 55; Case T–219/99 British Airways plc v Commission Judgment of 17 December 2003, ECLI:EU:T:2003:343, para 242; Case C–202/07 P France Télécom v Commission Judgment of 2 April 2009, ECLI:EU:C:2009:214, para 105; Case C–457/ 10 P AstraZeneca AB and AstraZeneca plc v Commission Judgment of 6 December 2012, ECLI:EU:C:2012:770, para 134; Case C–209/10 Post Danmark A/S v Konkurrencerådet Judgment of 27 March 2012, EU:C:2012:172, para 23; Case C–23/14 Post Danmark A/S v Konkurrencerådet Judgment of 6 October 2015, ECLI:EU:C:2015:651, para 71. 16 M de la Mano, R Nazzini and H Zenger, ‘Article 102’ in J Faull and A Nikpay (eds), The EU Law of Competition (3rd edn, OUP 2014) 329–538, 393–94. 17 OECD policy brief, What Is Competition on the Merits? June 2006 <http://www.oecd.org/competition/mergers/37082099.pdf> accessed 8 June 2019. 18 As an economic concept this is different from the legal concept of discrimination and the EU body of law that intends to counter discrimination and ensure equal treatment. S Fredman, Discrimination Law (2nd edn, OUP 2011) and E Ellis and Ph Watson, EU Anti-Discrimination Law (2nd edn, OUP 2012). 19 Public authorities are beginning to worry about this as well: ‘Government and CMA to Research Targeting of Consumers Through Personalised Pricing. Research Commissioned to Look at the Practice of Retailers Targeting Online Shoppers’ CMA Press Release, 4 November 2018; Note from the Netherlands (n 2). 20 M Sandel, What Money Can’t Buy. The Moral Limits of Markets (Penguin Books 2013). 21 Case T–201/04 Microsoft Corp v Commission Judgment of 17 September 2007, ECLI:EU:T:2007:289. 22 Case 322/81 NV Nederlandsche (n 15) 23 Case 85/76 Hoffmann-La Roche (n 10) para 91. Abuse is not the use of economic power but the nature of that use, and therefore an objective concept relating to the behaviour of the undertaking. 24 Preliminary assessment in Facebook proceeding: Facebook’s collection and use of data from third-party sources is abusive, Bundeskartellamt press release, 19 December 2017; Bundeskartellamt prohibits Facebook from combining user data from different sources, Bundeskartellamt press Release, 7 February 2019; Bundeskartellamt, Case Summary, Facebook, Exploitative business terms pursuant to s 19(1) GWB for inadequate data processing (Decision of 6 February 2019) <https://www.bundeskartellamt.de/SharedDocs/Entscheidung/EN/Fallberichte/Missbrauchsaufsicht/2019/B6-22-16.pdf?__blob=publicationFile&v=4> accessed 8 June 2019. 25 See Schweitzer and others (n 1). The same idea was advocated in 2005 by the Economic advisory group to DG COMP in the report An Economic Approach to Article 82<http://ec.europa.eu/dgs/competition/economist/eagcp_july_21_05.pdf> accessed 8 June 2019. 26 J Crémer, Y-A de Montjoye and H Schweitzer, Competition Policy for the Digital Era (European Commission 2019). <http://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf> accessed 8 June 2019. 27 See E van Damme, I Graef and W Sauter, ‘Exploitative abuse in EU competition Law’ (2019, forthcoming) TILEC discussion paper. 28 An analogy might be call termination and roaming on mobile networks which proved to be intractable problems in antitrust, only to be thoroughly regulated as a result. See eg Regulation (EU) No 531/2012 of the European Parliament and of the Council of 13 June 2012 on roaming on public mobile communications networks within the Union(2012) OJ L172/10. 29 Opinion of AG Fennelly, 29 October 1998, in Joined Cases C–395/96 P and C-396/96 P Compagnie maritime belge transports SA (C-395/96 P), Compagnie maritime belge SA (C-395/96 P), and Dafra-Lines A/S (C-396/96 P) v Commission (1998) ECLI:EU:C:1998:518, para 137. 30 J Goyder and A Albors-LLorens, Goyder’s EC Competition Law (5th edn, OUP 2009) 301. Cf. Case T–201/04 Microsoft Corp v Commission, Judgment of 17 September 2007, ECLI:EU:T:2007:289; Joined cases C395/96 P and C-396/96 P, Compagnie maritime belge transports SA (C-395/96 P), Compagnie maritime belge SA (C-395/96 P) and Dafra-Lines A/S (C-396/96 P) v Commission, Judgment of 16 March 2000, ECLI:EU:C:2000:132. 31 This is a category introduced by Article 25 of the Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) (2016) OJ L119/1. 32 This is a method and/or process of meeting specific compliance requirements by means of software. See eg F Olivieri, ‘Compliance by Design: Synthesis of Business Processes by Declarative Specifications’ (PhD thesis, Brisbane and Verona, May 2014); S Sadiq and G Governatori, ‘Managing Regulatory Compliance in Business Processes’ in J vom Brocke and M Rosemann (eds), Handbook of Business Process Management (2nd edn, Springer 2015) 265–88. 33 C Sunstein, Risk and Reason: Safety, Law and the Environment (CUP 2004); U Beck, Risikogesellschaft: Auf dem Weg in eine andere Moderne (Suhrkamp Verlag 1986). 34 On competition soft law, see eg Z Georgieva, Soft law in EU Competition Law and Its Reception in Member States' Courts an Empirical Study on National Judicial Attitudes to Atypical Legal Instruments in EU Competition Law (PhD Thesis, Tilburg University 2017). An example is the Communication from the Commission: Guidance on its enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (2009) OJ C45/7. 35 RA Corcho, ‘Can We Finally Say Farewell to the “Special Responsibility” of Dominant Companies?’ EUI Robert Schuman Center 2007 EU Competition Law and Policy Workshop/Proceedings 2007 <https://www.eui.eu/Documents/RSCAS/Research/Competition/2007ws/200709-COMPed-Allendesalazar.pdf> accessed on 8 June 2019. 36 See nn (25) and (26) . 37 On problems of purchasing power, see V Daskalova, The Monopsony Paradox: Buyer Power and Enforcement of the EU Antitrust Provisions (Prisma Print, 2016). 38 V Daskalova, ‘Regulating the New Self-Employed in the Uber Economy: What Role for EU Competition Law?’ (Stockholm University 2017) Stockholm Faculty of Law Research Paper No 38. 39 Arguably some types of consumer discrimination are also a sort of exclusion although not centred on a mutually competitive relationship. Borderline cases are consumers–producers and self-employed persons. 40 Communication from the Commission (n 34). 41 Communication from the Commission—Guidelines on market analysis and the assessment of significant market power under the EU regulatory framework for electronic communications networks and services C/2018/2374 (2018) OJ C159/1. A comparison between regulation and competition in D Gabathuler and E Martinez Rivero, ‘Communications: Telecoms and Internet’ in Faull and Nikpay (n 16) 1647–08. 42 Case C–23/14 (n 15)paras 70–72. 43 Such as in Italy and the UK: Aspen Pharmaceuticals (Case A480) AGCM Decision of 29 September 2016; Pfizer (Case CE/9742–13) CMA Decision of 7 December 2016. The Danish competition authority made a similar announcement: ‘CD Pharma has Abused its Dominant Position by a Price Increase of 2000 Percent’ Press release, 31 January 2018. The European Commission opened a 27 Member State investigation into excessive pricing by Aspen as well. 44 M Trebilcock and F Ducci, ‘The Multifaceted Nature of Fairness in Competition Policy’ (CPI Antitrust Chronicle, October 2017) 1–6. A reverse view in A Ayal, Fairness in Antitrust, Protecting the Strong from the Weak (Hart Publishing 2016). 45 Case C–525/16 MEO (n 14). 46 See van Damme, Graef and Sauter (n 27). 47 See Faull and Nikpay (n 16) 525–38, with reference inter alia to Case T–128/98 Aéroports de Paris v Commission, Judgment of 12 December 2000, ECLI:EU:T:2000:290. 48 Case C–525/16 MEO (n 14) . 49 See n 19 on a recently announced CMA investigation and OECD note prepared by the ACM. 50 Case C–457/10 P AstraZeneca AB and AstraZeneca plc v Commission, Judgment of 6 December 2012, ECLI:EU:C:2012:770, para 133. More recently, comparable dominance abuses have been fined in Italy and France. 51 See A Ezrachi, ‘The Goals of EU Competition Policy and the Digital Economy’ (August 2018) BEUC discussion paper. A concrete example is the German Facebook case. See n 24 on the Decision of 5 February 2019 and the earlier preliminary assessment in the Facebook proceedings. 52 The calculus of negligence designed as a balancing test by Judge Learned Hand in United States v Carroll Towing Co (1947) 159 F 2d 169 (2nd Circuit). 53 Directive 2014/104/EU of the European Parliament and of the Council of 26 November 2014 on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union (2014) OJ L349/1. 54 Directive 2004/39/EC of the European Parliament and of the Council of 21 April 2004 on markets in financial instruments amending Council Directives 85/611/EEC and 93/6/EEC and Directive 2000/12/EC of the European Parliament and of the Council and repealing Council Directive 93/22/EEC, OJ L2004 145/1; Directive 2014/65/EU of the European Parliament and of the Council of 15 May 2014 on markets in financial instruments and amending Directive 2002/92/EC and Directive 2011/61/EU Text with EEA relevance (2014) OJ L173/349. 55 See D Busch and C van Dam (eds), A Bank’s Duty of Care (Bloomsbury 2017). 56 Law on financial supervision, 2002 (as amended in 2014), art 4.24a ‘duty of care’ that requires providers of financial services to take the justified interests of their clients into account and to act in this interest as well. 57 FCA, Discussion Paper on a Duty of Care and Potential Alternative Approaches DP18/5 (July 2018). This article provides similar examples with regard to prospective financial services regulation in the USA and Australia. 58 ibid, 35. 59 <https://www.gov.uk/government/consultations/online-harms-white-paper> accessed 8 June 2019; J Naughton, ‘The White Paper on Online Harms Is a Global First. It Has Never Been More Needed’ The Guardian (8 April 2019) <https://www.theguardian.com/commentisfree/2019/apr/14/white-paper-online-harms-global-first-needed-tech-industry-dcms-google-facebook> accessed 8 June 2019. 60 art 11 of the Dutch Health Insurance Act of 16 June 2005. 61 Communication from the Commission (n 34). 62 Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements (2011) OJ C11/1; Guidelines on Vertical Restraints (2000) OJ C291/1. 63 J Balkin, ‘Information Fiduciaries and the First Amendment’ (2016) 49 UC Davis Law Review 1185; J Balkin and J Zittrain, ‘A Grand Bargain to Make Tech Companies Trustworthy: Doctors and Lawyers Are Prohibited from Using Clients’ Information for Their Own Interests, So Why Aren’t Google and Facebook?’ The Atlantic (3 October 2016); J Zittrain, ‘How to Exercise the Power You Didn’t Ask for’ (2018) 96 Harvard Business Review. 64 A Dobkin, ‘Information Fiduciaries in Practice: Data Privacy and User Expectations’ (2018) 33 Berkeley Technology Law Journal 1. 65 L Khan and D Pozen, ‘A Skeptical View of Information Fiduciaries’ (25 May 2019, forthcoming) 133 Harvard Law Review; Columbia Public Law Research Paper No 14-622 <https://ssrn.com/abstract=3341661> accessed 8 June 2019. 66 Sorrell v IMS Health (2011) 564 US 552, 557. 67 Khan and Pozen (n 65). 68 Ezrachi (n 51) 15. 69 Offline there is just one—2000/12/EC: Commission Decision of 20 July 1999 relating to a proceeding under Article 82 of the EC Treaty and Article 54 of the EEA Agreement (Case IV/36.888—1998 Football World Cup), (2000) OJ L5/55. S Weatherill, ‘0033149875354 Fining the Organisers of the 1998 World Cup’ (2000) 21 European Competition Law Review 275. 70 Bundeskartellamt press release (19 December 2017) (n 24); Bundeskartellamt Press Release (7 February 2019). 71 Bundeskartellamt, Case Summary (Decision of 6 February 2019) (n 24). 72 R Thaler and C Sunstein, Nudge; Improving Decisions about Health Wealth and happiness (Yale University Press 2008). 73 I Graef, EU Competition Law, Data Protection and Online Platforms - Data as Essential Facility (Wolters Kluwer 2016); Prüfer and Schottmüller (n 7). 74 See Bundeskartellamt, Case Summary (Decision of 6 February 2019) (n 24). 75 Notably Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contractsCouncil Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts (1993) OJ L95/29; Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC, and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (2005) OJ L149/22; Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights, amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council (2011) OJ L304/65; S Weatherill, EU Consumer Law and Policy (2nd edn, Edward Elgar 2014). 76 Council Directive 93/13/EEC, art 4(2): ‘Assessment of the unfair nature of the terms shall relate neither to the definition of the main subject matter of the contract nor to the adequacy of the price and remuneration, on the one hand, as against the services or goods supplies in exchange, on the other, in so far as these terms are in plain intelligible language.’ 77 Directive 2005/29/EC, art 6(1)d concerning misleading information with regard to ‘the price or the manner in which the price is calculated, or the existence of a specific price advantage’. 78 C Volpin, ‘The Ball Is in Your Court: Evidential Burden of Proof and the Proof-Proximity Principle in EU Antitrust Law’ (2014) 51 Common Market Law Review 1159. 79 R Botsman, Who Can You Trust? How Technology Brought Us Together – and Why It Could Drive Us Apart (Penguin/Random House 2017). F Fukuyama, Trust: The Social Virtues and the Creation of Prosperity (Free Press 1995). 80 As in the transition from Council Directive of 20 December 1985 to protect the consumer in respect of contracts negotiated away from business premises (85/577/EEC) (1985) OJ L372/31 to Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights, amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council (2011) OJ L304/64. © The Author(s) 2019. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contactjournals.permissions@oup.com © The Author(s) 2019. Published by Oxford University Press.

Journal

Journal of Antitrust EnforcementOxford University Press

Published: Jul 1, 2020

There are no references for this article.