Promoting Fair Private Governance in the Platform Economy: EU Competition and Contract Law Applied to Standard Terms

Abstract

In recent years, a platform economy has emerged that is dominated by undertakings such as Google, Amazon, Facebook, Apple, and Microsoft. They have established a form of private governance vis-à-vis their consumers and customers by means of standard terms that create a risk of exploitation. This trend clashes with the internal market effort of the EU that is predicated on consumer rights and fair competition to address market failures such as market power, information asymmetry, and asymmetrical contractual dependency. In this article we examine how the resulting tensions can be addressed by means of EU competition and contract law. This is based on enforcing fairness by requiring (1) the implementation of proportionality—balancing interests—and (2) respect of the duty of care, in the sense of compliance by design. Jointly this can be seen as an expression of accountability that needs to be made explicit. Apart from pre-existing case law and legislation, we take into account the December 2020 Commission proposals for platform regulation, as well as behavioural insights into consumer behaviour.

I. INTRODUCTION

A. Origins of and Reasons for Standard Terms

The industrial revolution alongside mass production of goods and eventually services also introduced mass contracting. This in turn resulted in pre-drafted contracts in the form of standard terms.¹ This development has gradually changed the perception of contracts and contract law. Freedom of contract or party autonomy is the point of departure of classic contract law: following negotiations, parties enter into a contract that is assumed to serve their mutual interests best. As the outcome of negotiations, the contract was originally presumed to be fair as such.² With respect to the French Code Civil of 1804, this was expressed as qui dit contractuel dit juste.³

In modern contract law however it has been recognised that often contracts are concluded without negotiations, and that even if negotiations take place, they do not always result in a fair outcome.⁴ As a result, in the EU context as elsewhere, in the second half of the twentieth century legislation was adopted to protect weaker contract parties, primarily (but not exclusively) consumers, against unilaterally drafted standard terms.

B. Standard Terms and Digital Platforms

Moreover, the notions of contract and contract law are increasingly under pressure from the platform economy. Digital platforms use contracts and standard terms in a different manner from what is described in textbooks on contract law.⁵ To construct a legal framework for their services, digital platforms enter into contracts with various types of customers in multi-sided markets, where one side of the market may for example represent end users of social media services and the other advertisers, respectively app users versus app developers.

In both markets the terms of these contracts are rarely negotiated: instead, platforms draft standard terms unilaterally. These terms are intended to apply regardless of where a customer lives or is established, although sometimes they are slightly modified if required by applicable law.⁶ They concern, inter alia, exclusions of liability, modification of standard terms, forum choice, choice of law, providing a license to the platform to use the user's content anywhere and anytime, entitling the platform to use and sell the user's data, and charging a commission for use of the platform.⁷

Much of the interpretation and application of these rules is administered by the digital platform itself to enforce compliance with its private regulatory regime. In addition, the platform often controls the content of the contract between the end-user and the business offering the goods or service as well.⁸ For instance, at the outset of the COVID-19 crisis, hospitality platform Airbnb notified guests on this basis that they could freely cancel certain reservations for stays with their individual hosts before a certain date that had been booked using Airbnb.⁹ The degree of control exercised at this level differs from platform to platform.

Consequently, the regimes that platforms establish by means of their standard terms form a type of private governance, in the sense of a largely unilateral regime based on private law.¹⁰ In our view, (1) as is the case for standard terms more generally, this should be treated differently from more balanced contracts between peers that are the result of negotiations, and (2) as we will see below, possibly requires additional controls compared to non-platforms as well because of the greater degree of control over their contracting parties that is associated with the digital platforms due to lock-in, asymmetrical contractual dependence, and information asymmetry.

C. EU Competition and Contract Law

These areas of law both serve to protect consumers, although they do so from a different perspective.¹¹ Whereas competition law looks at market power and the degree of competition in markets as a whole, consumer law in principle concerns contractual power in individual transactions. There are also some commonalities. Both share a concern about possible harm to consumer welfare that may result from imbalances that are due to asymmetrical market positions. However, only competition law has a category of rules (the prohibition on dominance abuse) that are formally based on a specific degree of market power—although enforcement priorities in contract law may take a comparable position in practice.¹² Both are enforced by public authorities as well as, at least in theory, by means of private enforcement (which in EU competition law is largely limited to damages actions). In Europe, authorities such as the Autorità Garante della Concorrenza e il Mercato (‘AGCM’) in Italy, the Consumer and Markets Authority (‘CMA’) in the UK, and the Authority for Consumers and Markets (‘ACM’) in the Netherlands combine the relevant public enforcement functions.

Competition law is most relevant to our topic regarding the prohibition on abuse of dominance that serves to control the exercise of market power ex post. The abuse of dominance test requires establishing a dominant position within a so-called relevant market, and proving a specific abuse, subject to objective justification and efficiency defences that can be raised by the undertaking concerned. It is relevant specifically where, until recently more rarely, it counteracts exploitation directly.¹³ The alternative, combating exclusion of competitors under dominance abuse only indirectly protects consumers by protecting competition, which was the prevalent approach. The most relevant exploitative abuse in this context is imposing unfair contract terms as spelled out in Article 102a of the Treaty on the Functioning of the European Union (‘TFEU’).¹⁴

Such direct control of dominant platforms’ conditions has been relatively rare, albeit on the increase with recent cases in Germany (notably the well-reported Facebook case), France, and the Netherlands, as well as at EU level. These are discussed below in relation to both sides of the market: business using platforms to provide products and services, and consumers procuring these products and services from them via digital platforms.

In EU contract law there is more elaborate legislation that sets out specific protection. These rules are adopted within the context of the establishment and proper functioning of the internal market. So far, private governance through standard terms is primarily policed by national legislation that transposes Directive 93/13 on unfair terms in consumer contracts (‘UCTD’).¹⁵ Under the UCTD, unfair terms in business to consumer (‘B2C’) contracts which parties have not negotiated are non-binding, or in other words void. The European Court of Justice has introduced the rational or average consumer as yardstick for the unfairness test under the UCTD.¹⁶ Although as we will see, the behavioural literature questions whether rational consumers exist at all.¹⁷

We will also look at the online intermediation Regulation 2019/1150, which does not aim to protect consumers but businesses that offer their goods and services to consumers in the European Union through a platform (platform to business, or ‘P2B’).¹⁸ The P2B Regulation aims to achieve a ‘fair, predictable, sustainable and trusted online business environment in the internal market’ through mandatory rules that provide transparency and effective redress mechanisms.¹⁹ Jointly this is the relevant EU contract law, looking at both sides of the market. In this paper we will not take into account consumer and other protection provided by conflict rules if the contract involves a cross-border situation. Moreover, we will not consider the data protection rules.

Below we will examine whether there is any degree of convergence between EU competition and contract law with regard to standard terms. More specifically we examine how public and private law concepts of fairness are applied to standard terms in this context to redress imbalances in bargaining power.

D. Adjacent Regulation

In addition, we will consider how fairness is interpreted in the EU's digital strategy²⁰ notably the European Commission's December 2020 proposals for a Digital Markets Act to control the market behaviour of platforms designated as gatekeepers based on turnover thresholds and a Digital Services Act that imposes certain policing duties on platforms providing access to digital content.²¹ These rules, once adopted, will constitute a separate regulatory framework that applies in parallel to EU competition and contract law. Hence, we will look into the relevance of especially the Digital Markets Act both in terms of its content, as well as in terms of its impact on the enforcement of competition and contract law regarding digital platforms’ standard terms.

E. The Behavioural Dimension

As was mentioned above, a behavioural approach to the interaction between consumer psychology and market forces vis-à-vis standard terms points out that it is not realistic and therefore undesirable to rely on rational consumers. Consumers may discount risks inherent in unbalanced contract terms due to over-optimism and a limited and short term focus (myopia), or due to the availability heuristic (the perceived infrequence of adverse events based on the experience of the subject), and due to status quo and omission biases may fail to switch even where clearly superior alternatives exist.²² By using complex contract and cost deferral (such as low initial payments), businesses make the most of the interaction between consumer psychology and market forces in a manner that increases consumer risk.²³ We will examine their implications for fairness in the application of EU competition and contract law to digital platforms’ standard terms.²⁴

F. Research Questions

A contract that consists of standard terms is therefore a way of construing private governance. Our starting point is that fairness norms of contract law may be used to discipline private governance, just as competition law may be able to. More specifically we examine how public and private law concepts of fairness are applied to standard terms to redress imbalances in bargaining power in this context. The main question we wish to address in this paper is therefore whether there is convergence between the developing fairness norms in EU competition and contract law with regard to the control of private governance by means of standard terms. A secondary issue is how these norms can be enforced effectively, and what behavioural modifications are required from digital platforms.

Below, we will first deal with the relevant rules and cases in competition law and discuss the recent proposals for regulation of digital platforms. Next we look at EU contract law, and finally at the implications of behavioural insights.

II. FAIRNESS, COMPETITION LAW, AND DIGITAL PLATFORMS STANDARD TERMS

Although in law the concept of fairness is used frequently, it has not been uniformly defined. Its meaning ranges from more practical notions of equity—a fair share—to the theoretical foundations of justice itself.²⁵ Unsurprisingly therefore, fairness in competition law is a contested topic in its own right.²⁶ It has been parsed in procedural and substantive dimensions, and the latter in different horizontal and vertical spheres,²⁷ as well as according to various types of intervention.²⁸

Especially the distinction between procedural fairness—due process—and substantive fairness in terms of outcome is significant here. Commentators universally subscribe to the former albeit often focusing on rights of the defence instead while ignoring procedural fairness obligations for undertakings enjoying market power and fairness rights for their victims. The latter, substantive fairness, is considered more problematic because it challenges efficiency as the guiding light of competition law, and by definition includes distributional concerns, even if the consumer tends to remain at centre stage. Competition Commissioner Vestager has made fairness a leading theme of her term in office, focusing on a fair deal for consumers.²⁹ Former director general of DG Competition Laitenberger (now a Court of Justice of the European Union (‘CJEU’) Judge) has concentrated instead on due process.³⁰

This implies that the fairness discourse in EU competition law is gaining traction. Yet the concept of fair competition is not new. It is part of the preamble of the TFEU.³¹ Fairness also appears in both legs of EU competition law: in the provision that provides a legal exception to the cartel prohibition and as part of the prohibition on the abuse of a dominant position.

Article 101(3) TFEU requires a fair share for consumers of the benefits of otherwise anticompetitive agreements before they can benefit from this exception. Most significantly for our current perspective on standard terms of digital platforms, it also features in Article 102a TFEU in relation to the prohibition of both unfair prices and unfair trading conditions for dominant undertakings.

More recently, the 2019 ECN+ Directive that harmonises the powers of competition authorities in the EU states its objective as the creation of fairer and more open competitive markets.³² As we will see below the December 2020 proposal for a Digital Markets Act aims at contestable and fair markets. Both imply that competition alone is not enough, but that fair outcomes matter as well. As far as EU competition policy is concerned there appears to be an increasing emphasis not just on fairness in an abstract sense, but on substantive fairness that is in line with a more general trend toward a plurality of goals of competition law beyond consumer welfare and efficiency.³³

It is hard to read across from general fairness to the specific fairness norms embedded in the TFEU. In this article, we will focus not on various theories of fairness in competition law, although the distinction between procedural and substantive fairness appears to be useful. Instead, we will look at how fairness is and/or can be implemented in positive competition law, more in particular in relation to balancing obligations imposed on dominant undertakings as a result of their special responsibility in relation to standard terms. This means looking at how Article 102 TFEU is applied to standard terms, and which lessons may be derived from that. It is therefore Article 102, more specifically Article 102a TFEU that we will be looking at more closely.

A. The Non-digital Case Law on Contract Terms

First, we will briefly look back at the competition case law on standard terms that precedes the emergence of digital platforms. In summary, contract terms are considered unfair in the sense of Article 102a TFEU here if they fail to meet the twin standards of (1) necessity in relation to a legitimate objective and (2) proportionality to that objective in the sense of establishing an appropriate balance between the interests concerned.³⁴ It is currently an open question whether one should start with the harm suffered before addressing these elements of a fair balance. This approach was initially established in a number of copyright cases, notably BRT II in 1974,³⁵ and confirmed in 2009 in der Grüne Punkt, a case regarding a system for packaging waste recycling.³⁶

Thus, in BRT II the Court stated that

the fact that an undertaking entrusted with the exploitation of copyrights and occupying a dominant position within the meaning of Article 86 imposes on its members obligations which are not absolutely necessary for the attainment of its object and which thus encroach unfairly upon a member's freedom to exercise his copyright can constitute an abuse.³⁷

The Commission in GEMA interpreted this as follows:

It may … be inferred from the BRT II case that, in an examination of a collecting society's statutes in the light of the Treaty competition rules the decisive factor is whether they exceed the limits absolutely necessary for effective protection (indispensability test) and whether they limit the individual copyright holder's freedom to dispose of his work no more than need be (equity).³⁸

In its DSD Decision that was confirmed by the Court in der Grüne Punkt, the Commission stated concisely and unreservedly that:

Unfair commercial terms exist where an undertaking in a dominant position fails to comply with the principle of proportionality.³⁹

The proportionality standard was applied to dominant undertakings in other dominance cases as well. It appeared explicitly in United Brands in the context of a meeting competition defence by a dominant undertaking in relation to its refusal to supply (1978).⁴⁰ More recently, the proportionality standard appeared implicitly in the Kanal 5 (2008)⁴¹ and SABAM/Weareone.World (2020)⁴² cases, where it was held that the interest of purchasers in paying for their actual use of a product should be balanced with those of the dominant undertaking when it set its pricing scheme. The method of calculation should not be disproportionally reliant on charges other than actual use and can only deviate from a more accurate use-based method if necessary for the legitimate goal of the agreement and if such alternative methods are unreasonably expensive to apply.⁴³

At the same time this appears to set a standard of manifest (and not strict) disproportionality precisely because it would still be acceptable to base the remuneration on turnover for variables other than actual use, as long as this is done in a proportional manner. This approach squares with case law like that in Höfner regarding manifest inability to meet demand, and excessive pricing cases where the price is not only high compared to a benchmark, but significantly and persistently higher.⁴⁴

B. The Emerging Digital Platform Case Law

There has long been a debate about the need to tackle digital platforms by means of competition law,⁴⁵ and the confluence of competition and data protection law.⁴⁶ Following a period of hibernation after the case to break up Microsoft effectively stalled and was settled by a consent decree,⁴⁷ digital competition cases are now taking off again in the US.⁴⁸ However since its own 2004 Microsoft Decision,⁴⁹ it has been the EU blazing the trail.⁵⁰ In retrospect, the facile approach of enforcers in both systems, which led to merger clearances in cases such as Facebook/WhatsApp⁵¹ that strengthened dominance while damaging the contestability of digital markets, may now be regretted. Below we will focus on the so far limited subset of cases involving platforms and standard terms.

1. The German Facebook Case

The February 2019 Bundeskartellamt (‘BKA’) ruling against Facebook demonstrates that standard terms imposed on consumers may form an abuse of dominance.⁵² The BKA identified a two-sided relevant market that combined nominally free social media services for consumers with paid advertising services to business users. It found that Facebook's standard terms deny consumers choice over the use of their personal data that are collected and combined not only from Facebook itself, but from other services owned by the same company such as WhatsApp (and Instagram), as well as data gleaned from unrelated internet services.⁵³ The BKA construed this as a form of consumer exploitation, while the resulting scale and scope of data collection led to exclusion of competitors. Both were found to constitute dominance abuse under the German competition rules. (The EU rules were not applied, possibly in order to remain outside the European Commission's coordination powers and as the bar for the BKA based on German law was presumed to be lower.)

Consumers were held not to have a choice other than accept Facebook's detailed and one-sided standard terms because due to its dominant position and network effects there were no alternative social media services that provided them with a real alternative. Under these circumstances a limited form of causality between the dominant position and the behaviour harmful to competition was held to be sufficient—in so far as the latter was considered harmful, that would suffice to establish abuse. This relaxes the requirement of a causal link between dominance and the behaviour in question under German competition law.⁵⁴ It also suggests that the special responsibility of dominant undertakings means types of behaviour that are unobjectionable or in any event not actionable for non-dominant firms may be out of bounds to them.

Hence the BKA prohibited the data processing policy that Facebook imposes on its users and imposed an injunction ordering termination of this conduct. The BKA decision was followed by two contrasting rulings from the Oberlandesgericht Düsseldorf (‘OLG’) and the federal Bundesgerichthof (‘BGH’) under expedited proceedings that (temporarily) left the injunction in place pending Facebook's appeal on substance.⁵⁵ These judicial rulings represent opposed readings of freedom of contract for consumers.

The OLG emphasised how consumers are free to decide autonomously and based on their own preferences whether to contract in accordance with Facebook's standard terms of privacy. The OLG held it to be irrelevant whether consumers in fact rarely examined the relevant clauses for what it assumed were reasons of convenience or indifference. Yet consumers may either be rationally ignorant regarding certain information, where the costs of absorbing it are too high, or rationally apathetic where the likelihood of being able to bargain around or circumventing the terms involved by choosing other providers is too low or non-existent due to market power.

The BGH in contrast invoked consumers’ rational ignorance and/or rational apathy as a condition that was influenced by Facebook's dominance and the resulting lack of alternatives for users of social networks itself. In this context it considered decisive the degree to which data were in fact surrendered in an unequal exchange under the standard terms: the BGH was confident this constituted a breach of the data protection rules. In doing so it confirmed the BKA view that the data protection rules could be invoked in establishing an abuse of dominance.⁵⁶ The BGH also amplified the BKA's finding of a parallel exclusionary abuse. The net result is that so far in Germany Facebook's private governance by means of its standard terms has been found at odds with the public rules on data and competition, although it is strictly speaking only a competition law violation—and was curtailed in fairly general terms under an injunction to cease and desist from this behaviour, without a fine.

Meanwhile as part of the appeal procedure the OLG has nevertheless suspended the BKA's decision pending Facebook's appeal, so the saga continues.

2. The French Google Case

By contrast, in December 2019 the French Autorité de la concurrence (‘ADC’) fined Google €150 million for dominance abuse in the search advertising market under Article 102 TFEU as well as the equivalent provision in Article L 420-2 of the French Code of Commercial Law.⁵⁷ Google was held to have adopted opaque operating rules (‘Rules’) that govern the interactions between internet users and advertisers on the two-sided Google Ads advertising platform, and to have applied them unfairly and randomly to firms that purchased advertising space. Although purportedly intended to protect consumers, the Rules were found to be discriminatory and arbitrary in relation to advertisers, and therefore not fit for purpose. Given Google's extreme degree of dominance with a market share consistently above 90% and significant barriers to entry,⁵⁸ and the absence of alternatives for Google Ads, the ADC held this violated Google's resulting special responsibility vis-à-vis its contracting parties.

In a further contrast to the German Facebook Case, by way of remedy the ADC also imposed not just a fine, but also a range of highly detailed injunctions on Google based on requirements to make its Rules clearly understandable and establish clear procedures for establishing violations. It also required those Rules that carried a sanction of suspension to be necessary and proportionate in relation to their legitimate objective of consumer protection.

The ADC required the criteria for identifying breaches to be objective, transparent, and non- discriminatory.⁵⁹ In addition, a range of specific procedural requirements was introduced, including not only the steps and timing for qualification of a breach, but also enhanced compliance policies and the possibility for consumers to report violations.

Jointly, these requirements amount to fairness criteria for Google's private governance by means of its Rules that do not just target the formal legality of the standard terms involved but their quality and the actual manner in which Google materially applies them to its contract partners.

Regarding the business users of Google's platform these criteria are familiar from, the above-mentioned EU case law on standard terms as well as the exception of objective necessity (necessity and proportionality).⁶⁰ They also recall market power based (sectoral) access regulation such as apply to for electronic communications and for standard essential patent holders who have committed themselves to FRAND license conditions.⁶¹ Moreover, the ADC designed a tailor made procedural regime that does not only impose procedural fairness for business users but is notable also for empowering consumers to directly engage with Google where advertisers were though to violate its Rules. Under this application of the EU competition rules, Google is thus held accountable to its various constituents.

3. The App Store Investigations

Meanwhile in 2019 and 2020 respectively, the Dutch Authority for Consumers and Markets (‘ACM’) and DG Competition initiated investigations of Apple's standard terms under the EU competition rules. In 2021 the CMA likewise launched an investigation into Apple in the UK following complaints that its standard terms for app developers are unfair and anti- competitive, and the ADC rejected a request for interim measures regarding advertising in mobile apps against Apple explicitly based on the BRT II framework discussed above.⁶²

The Commission has meanwhile issued a statement of objections that contests the mandatory use of Apple's proprietary in-app purchase mechanism, which it imposes on music streaming app developers such as Spotify to distribute their apps via Apple's App Store. The Commission is also concerned about the foreclosure of cheaper alternative services as Apple applies restrictions on app developers that prevent them from informing iPhone and iPad users about these. It is focusing on at exclusion of competing apps, where Apple also has its own service—an example is music streaming with Apple Music and complainant Spotify. Jointly with the ACM and CMA investigations, this litigation is likely to cast further light on the approach to abusive standard terms.

C. Fairness, Balancing, and Standards Terms in Competition Law

In sum therefore, we see the emergence of cases applying competition law to digital platforms standard terms that complement the earlier BRT II and der Grüne Punkt case law on standard terms per se. Self-regulation and private governance is thus constrained and conditioned by means of competition law. This involves findings of specific unfair behaviour that is struck down, as in the German Facebook case. More broadly it involves fleshing out the meaning of fair treatment by a dominant undertaking both procedurally and in terms of the practical outcomes that result, as in the French Google case.

We have previously argued that the private law notion of duty of care should apply to digital platforms as a pro-active fairness requirement based on the special responsibility of dominant undertakings in their relationship to consumers more generally.⁶³ This route will be examined in more detail below in relation to behavioural insights into standard terms and the notion of compliance by design. For now, we focus on how the proportionality standard that was previously largely applied in public law has been brought to bear on digital platforms’ private governance.

Based on the above-mentioned case law, in relation to unfair standard terms we believe the control of dominance abuse will take the form of balancing tests of the interests involved according to a manifest (dis)proportionality standard. In addition, in the French Google case we have seen notions of transparent, objective and non-discriminatory application of the relevant private governance rules, and of due process more generally, including consumer empowerment. Substantively in this case, consumers’ freedom of choice was protected by lifting constraining standard terms, not by disclosing them more fully or holding consumers themselves responsible for agreeing to them. At the same time, the dominant firm was not just barred from imposing illegal standard terms, but also required to assuring the quality and even-handed application of the terms that it could impose.

This appears to represent the following three developments:

• – First, attempts to apply to a digital platform an array of criteria that are familiar to courts, notably proportionality, but also transparency, objectivity and non- discrimination. These have not yet gelled into a standard approach to private governance but their outline is clear.

• – Second, these criteria are broadly based on the implications of the special responsibility that flows from deeply asymmetric positions of market power and contractual power. Noblesse oblige, hence dominant firms must accept the consequences on their own accord, and therefore behave well proactively.

• – Third, in line with the above, the findings point toward a more modern understanding of consumer behaviour. This relies less on disclosure and personal responsibility of purchasers and on ex post controls.

We argue that these developments are an expression of substantive fairness, and that proportionality in the sense of balancing of interests is at its heart. Jointly, these trends are likely to prioritise on the one hand direct public intervention and on the other a duty of care requirement for digital platforms relevant to enforcement in civil courts. Having examined the application of EU competition law to standard terms, we now turn to the relevance of the recently proposed regulation of digital markets.

III. THE NEW REGULATORY PROPOSALS AND STANDARD TERMS

In its February 2020 vision document on shaping Europe's digital future, the European Commission presented the green and digital transformations as the two key challenges facing the Union.⁶⁴ It stated that three pillars underpin its proposals for the digital dimension: (1) technology that works for people, regarding initiatives aiming to foster trust in digital markets; (2) a fair and competitive economy, aiming to make competition instruments fit for purpose and promote a digital industrial policy; and (3) an open, democratic and sustainable society, including new rules to deepen trust in the internal market for digital services.

In December 2020, the Commission made joint legislative proposals for the Digital Markets Act that is covered by point (2) above and Digital Services Act under point (3), which concern us here.⁶⁵ These proposals were made on the legal basis for internal market legislation in Article 114 TFEU and are currently before the European Parliament and Council under the normal legislative procedure of co-decision. Albeit subject to unprecedented lobbying by the undertakings concerned, it now appears that they could be adopted by the EU and implemented at national level within approximately two years.

A. Fairness in the Digital Services Act

The focus of the Digital Services Act is creating a single market for digital services. It updates the horizontal rules for online platforms in particular in relation to the e-Commerce Directive that was adopted twenty years ago, before most of today's major platforms even existed.⁶⁶ The core of the Digital Services Act are new rules with regard to issues such as liability for online content and online safety of users as well as due diligence obligations related to the societal risks represented by online services, jointly aiming at promoting trust and respect for human rights. The Digital Services Act covers intermediary services, internet access providers and domain name registrars, hosting services, and online platforms. Special obligations are imposed on very large online platforms, which are defined in principle as those with a monthly reach of more than 45 million EU consumers (or in any event corresponding with 10% of the Union's population). Such very large online platforms have to take special measures with regard to risk assessment, risk mitigation, data access, (advertising) transparency, compliance infrastructure, and scrutiny. As is clarified by Recital 56 of the Digital Services Act, these measures very much address concerns of private governance by very large online platforms:

In the absence of effective regulation and enforcement, they can set the rules of the game, without effectively identifying and mitigating the risks and the societal and economic harm they can cause. Under this Regulation, very large online platforms should therefore assess the systemic risks stemming from the functioning and use of their service, as well as by potential misuses by the recipients of the service, and take appropriate mitigating measures.

In the Digital Services Act, fairness is linked with transparency and accountability, and addressed in relation issues such as content moderation, online advertising, complaint handling, and dispute settlement, and a continued prohibition on general monitoring obligations. Balancing, where it is discussed in the Digital Services Act, relates to fundamental rights, not interests. The terms and conditions of online platforms are dealt with exclusively in relation to such issues, which are essentially non-economic in nature, except in the sense that they can raise barriers to economic transactions and influence innovation.

Hence its relevance to our topic is relatively limited, and in the remainder of this discussion we will focus on the Digital Markets Act. However before doing so, it is worth highlighting that in the Digital Services Act, (1) digital platforms are assigned a special responsibility for policing content that might so far have been thought a public task, and (2) very large online platforms face increased scrutiny and a higher degree of responsibility. It is these aspects with their important parallels in the Digital Markets Act that are germane to fairness and digital platforms’ standard terms.

B. Fairness in the Digital Markets Act

The Digital Markets Act proposal is narrower than its counterpart as it pertains exclusively to specific unfair behaviour (which it lists in detail)⁶⁷ by those digital platforms that are defined as gatekeepers. These are similar to the very large platforms in the Digital Services Act but defined according to different criteria, based primarily on fixed thresholds that combine an annual EEA turnover of €6.5 billion or above and 45 million monthly end users or 100,000 yearly active business users. These are taken as indications that the platform concerned has a significant impact on the internal market, operates a core service (from a closed list of defined services) that forms an important gateway to end users for its business users, and enjoys an entrenched and durable position.

Such gatekeepers are faced with a range of ex ante obligations that prohibit for instance combining personal data sources, using data generated by business users in competition against them, and self-preferencing, as well as being required to provide platform access and data portability and to allow business users alternative ways of reaching their end users. As a last resort, the proposal enables the possibility to impose structural remedies—that is to say forced divestiture of parts of the gatekeeper's business, splitting it up.

Fairness is addressed in the Digital Markets Act in relation to platforms’ business practices alongside the desire for contestable markets: the former deals with conditions and the latter with competitive capabilities. Specifically, fair and non-discriminatory terms must be applied to ranking services, and fair and non-discriminatory general conditions are required for business users’ access to software applications stores (such as Apple's App Store and Google's Play Store on Android). The latter is the most explicit link to our topic, albeit a highly specific one. There are other significant links however, to do with the concept of substantive fairness that is used.

Recital 57 of the Digital Markets Act states the following with regard to fairness and standard terms:

In view of the imbalance in bargaining power between those gatekeepers and business users of their software application stores, those gatekeepers should not be allowed to impose general conditions, including pricing conditions, that would be unfair or lead to unjustified differentiation. Pricing or other general access conditions should be considered unfair if they lead to an imbalance of rights and obligations imposed on business users or confer an advantage on the gatekeeper which is disproportionate to the service provided by the gatekeeper to business users or lead to a disadvantage for business users in providing the same or similar services as the gatekeeper.

This criterion is used in Article 10 of the Digital Markets Act to determine whether by means of delegated acts additional or updated obligations must be imposed on gatekeepers.⁶⁸ An imbalance in rights and obligations or a disproportionate advantage is therefore considered unfair under the Digital Markets Act. This is in line with what we have discussed above with regard to the importance of proportionality as balancing of the relevant interests for a substantive understanding of fairness. Moreover, in the recital referenced above it is formulated specifically with regard to general conditions—in other words standard terms. As we will see below, this also reflects an element of the fairness test under the UCTD to assess whether a standard term in a B2C contract is unfair.

In addition, Recital 58 of the Digital Markets Act spells out an intended requirement of compliance by design:

The gatekeepers should ensure the compliance with this Regulation by design. The necessary measures should therefore be as much as possible and where relevant integrated into the technological design used by the gatekeepers.

So far this does not appear to be linked with specific provisions in the Regulation, although it may be assumed that in complying with Articles 5 and 6 of the Digital Markets Act the gatekeepers will have to (re)design their software in such a manner as to ensure compatibility with their obligations under the Digital Markets Act. Nevertheless, as we will discuss in our section on behavioural insights immediately below, applied to private governance through standard terms, compliance by design is a second element of substantive fairness that should be guaranteed by digital platforms.

The Digital Markets Act requires gatekeepers to submit information regarding intended mergers and submit to market investigations intended to examine systematic non-compliance with their obligations. As is the case under the general competition rules, the Commission may impose fines of up to 10% of total turnover. The Digital Markets Act complements the general competition rules based on the premise that as ex ante regulatory intervention which clearly identifies unfair practices that may in part fall outside current competition rules, it space more effective than more time-consuming individual ex post investigations. At the same time, it does not foreclose recourse to the general competition rules, because it applies in parallel.

C. Implications for the Enforcement of EU Competition and Contract Law

Both proposals remain to be discussed in the legislature and are likely to be amended before adoption. Having already noted the parallel application of special responsibility for very large online platforms and gatekeepers, in this short section we will focus on the Digital Markets Act. In their current form, as a list of obligations for gatekeepers they appear to address specific concerns relating to standard terms for business users, in relation to app stores (Apple and Google) and online advertising services (Google), and in relation to end users concerning combining personal data from platforms and other services (Facebook). However, they do not include a general solution or approach to problems of market power regarding all digital platforms, or to all digital platforms’ standard terms.

This means that the control of such terms is likely to remain a patchwork of ex ante enforcement under the new rules and ex post enforcement under EU competition and contract law. In fact, the proposal for a Digital Markets Act adds a wholly new patch because, as mentioned, it is based only on the need to promote the functioning of the internal market as set out in Article 26 TFEU, under Article 114 TFEU on the approximation of laws, and its legislative basis does not (at least so far) include the competition rules, as arguably it might well have done.⁶⁹

The Digital Markets Act obviously provides remedies against breaches of its terms, including to private parties such as victims of self-preferencing. At the same time, its implications for the public and private enforcement of the competition and consumer law rules against digital platforms will be far-reaching:

• - In the first place the Digital Markets Act incorporates norms on what constitutes substantive fairness that rely on balancing of interests and requires compliance by design;

• - Second, and equally important, the Digital Markets Act effectively designates a set of digital platforms as super-dominant, and consequently subject to special responsibilities with regard to their counterparties, both consumers and business users.

This means that the burden of proof on private parties—as well as in public competition suits—is effectively reversed regarding such gatekeepers. In particular, proving dominance should be facilitated. This appears to lower the threshold for abuse of dominance actions against such parties significantly, and perhaps to the point where not just public but private enforcement may for the first time become an effective decentralised tool for controlling their behaviour.

The same applies by analogy for actions based on EU contract law.

This expectation might be tempered somewhat by the experience with significant market power (‘SMP’) regulation regarding telecommunications,⁷⁰ where such an effect appears to have been limited. However, although SMP was likewise predicated upon pre-identifying dominant operators, a much more comprehensive system of market supervision was created to control them. When the SMP regime was created in 1997, private competition law enforcement was in its infancy. Moreover in the context of digital platforms a broader range of interests and more fundamental concerns are involved, including the need to safeguard democracy and the rule of law, which may give rise to greater and more effective activism.

Jointly, this restatement of the applicable norms and creation of a new basis for legal recourse may therefore be expected to be game changing, on top of the Commission's own enforcement based on the Digital Markets Act which will be a game changer in itself.

IV. EU CONTRACT LAW AND STANDARD TERMS

A. Fairness in Consumer and Customer Contract Regulation and Its Enforcement

Together with property law, contract law provides the legal infrastructure for markets, including the EU digital internal market. Moreover, within the EU internal market, contract law has been deployed in an instrumental way. For instance, the harmonisation of contract law directly aims to achieve internal market objectives such as a high level of consumer protection. The relevant EU acts concern specific topics, for instance, product liability, unfair commercial practices, and unfair terms.⁷¹ EU contract law is different from national civil law systems in that it is not aimed at building a coherent system of contract law, but on regulating subject matters that foster the creation of the internal market.⁷²

Within European contract law, there is a debate whether it should be limited to corrective justice, or also involve social or distributive justice.⁷³ In this context, notions such as fairness, solidarity or altruism are used in favour of redistribution.⁷⁴ An argument against this position is that contract law is not an appropriate means to redistribute wealth, which should be achieved by taxes and social security systems.⁷⁵ Proponents of contract law as more than corrective justice however argue that it endorses a broader interpretation.⁷⁶ We agree that contract law aims at more than just corrective justice because EU contract law is in fact instrumental in achieving other goals than just corrective justice. For instance, the Directive on late payments aims at preventing bankruptcies of companies.⁷⁷

In EU contract law, fairness is deployed as a legal concept. A consumer is, for instance, protected against unfair terms in business to consumer contracts and against unfair business practices.⁷⁸ However, the precise meaning of fairness and commercial practice differs depending on the legal instrument concerned. For instance, fairness under the unfair commercial practices directive (‘UCPD’) differs from fairness under the UCTD and the P2B Regulation. Hereafter, we will discuss these three instruments. First, we will focus on their scope, the fairness test and enforcement. Subsequently, we address the notion of balancing of interests under the UCTD and the P2B Regulation.

B. Policing Standard Terms by European Contract Law: Three Instruments

In addition to competition law, the UCTD, the UCPD, and the P2B Regulation can be used to police a platform's standard terms. Their use depends on the circumstances, since the instruments’ scope, the fairness tests and their implementation into the legal systems of the Member States differ.

1. The Scope of the UCTD, UCPD, and the Online Intermediation Regulation

First, the scope of the UCTD and the UCPD is restricted to business to consumer contracts (‘B2C’), whereas the P2B Regulation concerns contractual relations between platforms, regardless of whether a platform is established within or outside the EU, and businesses that offer their goods and services to consumers (‘P2B’).⁷⁹ However, the businesses involved must be established in the EU and must offer their goods or services through the platforms to consumers within the EU. If the scope of the UCTD and the online intermediation Regulation are taken together, two groups of contracts fall outside the scope of the protection of both instruments.⁸⁰ They are, (1) contracts between the platform and purchasers of online goods or services who are not regarded as consumers, and (2) contracts between a platform and non-professional suppliers or sellers of online goods or services.⁸¹

Second, the UCTD only concerns non-negotiated terms in B2C-contracts.⁸² Although, it is not stated as explicitly, the P2B also concerns standard terms. It does not use these words, but the words ‘terms and conditions’ instead.⁸³A characteristic of such terms and conditions is that the platform sets them unilaterally, which is determined based on an overall assessment. Even if there are some negotiations between the platform and a business, the standard terms may still fall within the definition.

Third, the UCPD differs from the UCTD and the P2B Regulation, in that it does not concern contract terms directly, but commercial practices. This is a broad notion under the UCPD and includes any behaviour or omission of a business ‘directly connected with the promotion, sale or supply of a product to consumers’.⁸⁴

The P2B Regulation employs the notion of commercial practice. Its meaning however differs from a commercial practice under the UCPD. Under the P2B Regulation, this notion concerns standard terms which the platforms impose unilaterally upon the businesses. The wording used is similar to the provision on unfair terms in business-to-business contracts in the Commission's 2011 Draft on a Common European Sales Law (‘CESL’), which was ultimately not adopted.⁸⁵

1. Fairness under Three Instruments of Contract Law

Both directives, the UCPD and the UCTD, include a fairness test that has been interpreted by the CJEU. The notion of fairness under the P2B Regulation seems less important. As it applies to contracts concluded on or after 12 July 2020, at the moment of writing, as far as we are aware there is no case law with respect to this Regulation. Hereafter we will discuss the fairness tests as included in the three instruments and their interpretation by the CJEU.

First, the UCPD polices unfair commercial practices and distinguishes three groups of unfair commercial practices. Misleading and aggressive practices are specific categories that are submitted to separate tests.⁸⁶ In addition, the UCPD provides a blacklist of misleading and aggressive practices, which do not require an individual assessment of the practice at stake and are unfair a priori. Commercial practices that do not fall within the scope of misleading or aggressive practices are submitted to a general test. From the CJEU case law, it follows that only if a practice cannot be considered misleading or aggressive, the general test which concerns a practice ‘contrary to the requirements of professional diligence … materially distorts or is likely to distort the economic behaviour … can be applied.⁸⁷

In this context, Facebook's practice of offering a service for free whereas the consumers in reality pay for this by providing their data can be considered misleading under the UCPD, as the Italian Autorità Garante della Concorrenza e il Mercato did.⁸⁸ In other words, the UCPD can be applied with respect to standard terms, if a business's behaviour or omission is contrary to what is written in the standard terms. This could result in an unfair practice. However, the UCPD does not directly police the content of unfair contract terms.

Second, the UCTD defines a contract term as unfair if, contrary to good faith, it results in a significant imbalance between the rights and obligations of the parties concerned. The CJEU has interpreted this test in its case law. However, terms that concern the main subject matter of the contract, such as price, do not fall within the scope of this test, but are submitted to a transparency test.

Third, in spite of its name, the P2B Regulation does not extensively rely on the notion of fairness. It is mainly concerned with transparency, which is considered a prerequisite of fairness.⁸⁹ Moreover, the essential elements or the main subject matter of the contract between the platform and the business are subject to the general requirements of this Regulation, notably its transparency principle.⁹⁰ Nevertheless, the Regulation aims at balancing the legitimate interests of businesses, the innovative possibilities of the platform economy, competition, and consumer choice.⁹¹ In addition, the P2B Regulation requires platforms to implement an internal complaint-handling system.⁹² This system must be for free for business users and is based on the principles of transparency and equal treatment.⁹³

Thus, in the three instruments the notion of fairness differs and the terms that can be submitted to the fairness test. This has consequences for the enforcement of the three instruments.

2. Enforcement of the Three Instruments

The three instruments of contract law discussed above can be enforced ex ante and ex post as well as privately and publicly. Ex ante enforcement of EU contract law concerns the situation in which the standard terms are drafted, but are not yet incorporated in a given contract. Ex post enforcement concerns the situation in which a contract has been entered into and the standard terms are part of a conflict between the contracting parties.

As to public enforcement, the rules in the Member States differ. For instance, independent consumer authorities in Italy and the Netherlands have the competence to enforce the UCTD and the UCPD, whereas in other Member States such as Germany they do not.⁹⁴ However, this will change under Directive 2019/2161 as regards the better enforcement and modernisation of Union consumer protection rules.⁹⁵ This Directive introduces penalties if businesses do not comply with the rules included in the UCTD,⁹⁶ and sets out how penalties may properly be imposed.⁹⁷

The CJEU decisions on the UCTD and the UCPD concern mainly ex post enforcement of the rules both publicly and privately. With respect to the CJEU decisions on the UCTD a specific body of case law with respect to standard terms of platforms does not emerge yet. However, the standard terms of digital platforms include commonly used terms. Hence, the UCTD provisions and the rules inferred from the general case law can be applied to contracts between platforms and consumers.

In addition, platforms seem to comply voluntarily with the CJEU case law on the UCTD to a certain extent. Following the decision in VKI v Amazon concerning a choice of law clause in standard terms, many platforms changed the choice of law clause in their standard terms to comply with the CJEU ruling.⁹⁸

A recent example of ex ante public enforcement of contract law in the above-mentioned sense concerns the standard terms of AliExpress. A number of consumer organisations filed complaints about AliExpress with their national consumer authorities in 2019.⁹⁹ Among them was the Dutch Consumentenbond that filed a complaint with the Dutch Authority for Consumers and Markets. In cooperation with the European Commission and the European network of national consumer protection authorities (‘CPC’), the ACM confronted AliExpress with its failure to comply with EU consumer law. As a result, in February 2021, AliExpress committed to compliance with the EU rules on consumer protection including those on standard terms in all Member States from 1 May 2021 onward.¹⁰⁰

Another example of ex ante public enforcement under the UCPD is the Italian Facebook case. As explained above, an omission to refer to standard terms or behaviour contrary to standard terms may result in an unfair commercial practice, as the Italian AGCM did.¹⁰¹

The choice between the three different instruments with respect to their enforcement depends on their scope, the tests to be applied and the competences of national consumer authorities. For instance, if the price, an essential element of the contract, is challenged, under the UCTD only the transparency test can be applied, whereas under the UCPD the test of misleading or aggressive practices may be applied. This may explain why in Italy the Facebook case was considered under the UCPD. In Germany it was dealt with under competition law, since in that instance the BKA did not have competence to pursue this matter under the UCPD. Other relevant aspects are the data protection component of the case, which is not a federal responsibility in Germany either, and the fact that the BKA based itself on the German competition rules, which are closer to unfair trading law than their EU equivalent.

In the following sections we will elaborate how the interests of the contracting parties and other interests are balanced under the UCTD and the P2B Regulation by discussing the fairness and transparency tests included in those measures.

C. Balancing Interests under the UCTD

The underlying idea of the UCTD is one of balancing as a consumer needs protection because he is in a weak position vis-à-vis the business,

as regards both his bargaining power and his level of knowledge. This leads to the consumer agreeing to terms drawn up in advance by the [business] without being able to influence the content of those terms.¹⁰²

Thus, the consumer's weaker position follows both from a lack of bargaining power (asymmetrical contractual dependency) and less knowledge with respect to the content of the standard terms (information asymmetry).¹⁰³ Moreover, concerning the non-binding character of an unfair term¹⁰⁴ the CJEU held that this

is a mandatory provision which aims to replace the formal balance which the contract establishes between the rights and obligations of the parties with an effective balance which re-establishes equality between them ….¹⁰⁵

1. The UCTD Case Law on Fairness

The UCTD includes a fairness test to assess whether a standard term in a B2C contract is unfair.¹⁰⁶ From the relevant CJEU case law it can be inferred that the fairness test has three elements and a fair standard term must meet each requirement. The consequence of an unfair term is that it is not binding under the UCTD.¹⁰⁷

The CJEU decisions with respect to the fairness of a standard term concern online and offline situations, but unlike in EU competition law where there is an emerging body of case law with respect to standard terms of digital platforms specifically, no such body of specific case law exist in contract law yet.¹⁰⁸ However, the standard terms of digital platforms include commonly used terms which concern inter alia exclusion of liability, modification of standard terms, forum choice and choice of law. Therefore, the UCTD provisions and the rules inferred from the general case law can be applied to contracts between platforms and consumers.

In Aziz, the CJEU held that the Article 3(1) UCTD unfairness test includes two elements: a good faith requirement and an imbalance-test.¹⁰⁹ A bank had provided Mohammed Aziz with a loan which was secured by a mortgage to buy a house. Aziz stopped paying the loan's instalments and under the standard terms of the loan and/or mortgage contract he had to pay interest for late payment, which he did not pay either. After the bank had sold Aziz's home for 50% of its value, Aziz was evicted from his home. One of the preliminary questions was whether the clause in the standard terms that set the annual default interest if the debtor failed to pay at 18.75%, was an unfair term. To answer that question the CJEU formulated the two above-mentioned tests:

• – A term is contrary to good faith if an average consumer would not have agreed to the term, had the parties negotiated about the term in that situation.

• – A term is unbalanced if in the case of a comparison between the law applicable in the absence of the term in the contract and the term itself, the consumer would be worse off by applying the term included in the standard terms.¹¹⁰

From these two tests, it can be inferred that different interests are balanced. The good faith test reflects classic contract law, since the presumption seems to be that negotiations result in a fair outcome. So, if the parties had negotiated about the term's content and would have agreed upon it, the term can be considered fair. The imbalance test refers to rules of national contract law that would have been applicable if the term had not been agreed upon. In that case, the balancing done by a legislator when she drafted the legislation is considered to be fair based on the democratic process by which such rules are established.

It must be emphasised however that terms that concern the main subject matter of the contract, such as price, do not fall within the scope of these two tests.¹¹¹ However, the price and other ‘main subject matters’ of the contract are submitted to the transparency test, which is the third element of the fairness test.¹¹²

2. The UCTD and Transparency

The transparency principle set out in the UCTD¹¹³ requires a term to be in ‘plain, intelligible’ language.¹¹⁴ Apart the requirement that a given term must be grammatically correct, an average consumer must be able to foresee its economic consequences taking into account the contractual framework.¹¹⁵ The average consumer is someone ‘who is reasonably well-informed and reasonably observant and circumspect’.¹¹⁶ It is this less intrusive test to which price and other main subject matter of the contract is submitted.¹¹⁷

In addition, the CJEU has deduced from the transparency principle and the preamble¹¹⁸ of the UCTD¹¹⁹ that a trader must provide the standard terms to the consumer prior to the conclusion of the contract.¹²⁰ This has two purposes: it serves the aim of informing the consumer; and it enables the consumer to file the standard terms so that he can determine his rights and obligations if something goes wrong.¹²¹

The transparency test is linked to discussion about disclosing information to empower the consumer. Consequently, the information must be comprehensible both from a linguistic and economic perspective. However, the CJEU takes the average consumer as the benchmark, which is problematic from a behavioural perspective, as will be discussed further below.

D. Balancing Interests under the P2B Regulation

The P2B Regulation aims at promoting fairness and transparency in relations between platforms and businesses that offer their goods and services to consumers.¹²² It does so by establishing ‘a fair, predictable, sustainable, and trusted online business environment within the internal market’ by mandatory rules.¹²³ This is considered necessary, because platforms have significantly more bargaining power than ordinary businesses, which they may use to impose practices upon such businesses that are contrary to good commercial practice, good faith and fair dealing.¹²⁴ Moreover, it is considered crucial that platforms behave responsibly in establishing this business environment.¹²⁵ This Regulation is relevant to our discussion not only because it represents one leg of the triangular relationship between digital platforms, businesses and consumers, but also because it deals specifically with platforms and standard terms.

1. Transparency under the Regulation

The Regulation uses the notion of transparency both as a means to prevent unfair behaviour by platforms¹²⁶ and to provide predictability to business users.¹²⁷ Hence, the platforms’ standard terms must meet certain requirements that can all be linked to transparency,¹²⁸ as the relevant provisions concern duties to disclose information in the platform's standard terms in plain and intelligible language.

Article 3 of the Regulation sets out the main transparency requirements. As regards basic transparency, the platform's standard terms must be set out in plain and intelligible language and be easily available at all (pre-)contractual stages. Moreover, this provision requires the platform's standard terms to include specific information. For instance: (1) the standard terms must explain the reasons upon which a platform can suspend, terminate or restrict the service to a business; (2) platforms must inform business of changes in the standard terms within a reasonable time before they will be changed; and (3) the business must have the possibility to terminate the contract before the end of the notice period. Standard terms which do not meet these requirements are void. These effects do not only apply with respect to the contracting parties, but are erga omnes and ex tunc.¹²⁹

Other provisions include information disclosure duties for the platform as to their standard terms.¹³⁰ However, the rationale of those disclosure duties seems to differ. For instance, the disclosure of parameters with respect to ranking aims at predictability for businesses,¹³¹ whereas fair competition and consumer choice are the reasons for transparency with respect to differentiated treatment between the platform's own goods and services and the business’ goods and services.¹³² Yet the Regulation does not include a remedy if the platform fails to disclose the relevant information. It is left to the legal systems of the Member States to impose measures that are effective, proportionate and dissuasive.

The transparency test in the Regulation is elaborated (1) with respect to ranking of services and goods by business users, (2) preferred treatment of the goods and services offered by the platform itself or business controlled by the platform, and (3) the termination, suspension, or restriction of a particular business user on the platform. Its provisions on retroactive changes and termination uses the words good faith and fair dealing, but the meaning of good faith and fair dealing is not made clear: arguably each contract between a platform and a business is governed by these requirements. However, again the consequences are left to the legal systems of the Member States.

1. Remedies under the Regulation

As to the policing of standard terms in the contract between the platform and a business, the Regulation only provides a remedy to those mentioned in Article 3: nullity of the contract term concerned. In this respect a business is protected vis-à-vis a platform. This protection is restricted to transparency, and there is no other fairness test to which the platform's standard terms are submitted. Yet, it is questionable whether all the information that is to be included in the platform's standard terms in fact protects those businesses. First, the remedies are left to the legal systems of the Member States. Second, behavioural studies in the area of consumer transaction show that the more complex a contract is, the less likely it is to be understood by the consumer. As we will discuss this in the section on behavioural studies and standard terms, this probably also applies to small and medium sized businesses.

In spite of its title, therefore the notion of fairness is not developed in this Regulation.¹³³ Instead fairness is so far exclusively filled in by the prerequisite of transparency.

V. COMPARISON BETWEEN EU COMPETITION AND CONTRACT LAW

A. The Balancing Tests

On the one hand, fairness as a concept appears to be in ascendance both within EU competition and contract law. On the other hand, as we have seen, fairness as a form of balancing is so far used as an effective substantive test more in EU competition law and the Digital Markets Act than in contract law. Within EU contract law itself there are also notable differences.

• - The UCTD, the B2C instrument employs a fairness test to assess whether a term is unfair and non-binding. This test has three elements—good faith, significant imbalance, and transparency—and each element represents a different balancing test. The first two elements of the fairness test under the UCTD do not apply to the essential elements of a contract, for instance, the price. However, these are submitted to the transparency test.

• - The P2B Regulation seems so far to be limited to a transparency test. Although balancing is suggested by its language about the superior bargaining power of platforms, transparency appears to be key. Arguably a contract between a platform and a business is subject to good faith and fair dealing requirements, the Regulation does not elaborate what this entails.

In contract law we therefore notice a range of standards that at first sight appear to be fairness -and balancing-based, but are in fact at least partly based primarily on the less restrictive notion of transparency. The difference between businesses and consumers, B2C and P2B, and therefore between the two sides of the multi-sided markets involved, may well go some way in explaining this anomaly. However, this contrasts to comparable recent sectoral regulation on B2B outside the realm of digital platforms such as Directive 2019/633 on agribusiness and food which does include a substantive fairness requirement.¹³⁴ Nor is this distinction found in competition law, where P2B and P2C relationships are both subjected to a similar if not identical fairness standard based on balancing. Moreover, arguably in their relationship with a platform, many businesses are faced with information-asymmetry, asymmetrical contractual dependency and market power and consequently are in a position very similar to that of consumers.

There is so far little formal traction for the idea that in a balancing exercise under contract law the degree of market power should play a distinct role. Yet where market power exists there will always be asymmetric contractual dependence. Hence, enforcement priorities in contract law lead to a comparable approach: the contracting practices of parties with market power are focused upon more often.¹³⁵ As discussed above, it may be modified further by the ex ante designation of gatekeepers in the Digital Markets Act, which will facilitate both public and private enforcement against them in both EU competition and contract law. If this is indeed the case, then we are likely to see further convergence and cross-pollination between EU competition and contract law, partly via regulation.

B. Choice of Instrument between Competition and Contract Law, and Within Contract Law

With respect to the choice of instrument for enforcement, competition law, the UCPD or UCTD, the scope of the tests in the legislative instrument involved and the competences of the consumer authorities are relevant. They often explain why a certain route has been chosen.

• - For example, under German law, the BKA does not have competence to enforce the UPCD, nor is there an independent authority to enforce consumer law. The UCPD's enforcement is left to, inter alia, consumer organisations and chambers of industry and commerce.¹³⁶ This explains why the Facebook under German law was tackled under competition law even although this involves a higher standard of proof, and under Italian law under the arguable lower bar of the UCPD—possibly also due to less strict enforcement of the data protection rules than in Germany.

• - Under the UCTD, the price of a product or a service is not submitted to the general fairness test of good faith and balancing the parties’ interest, but merely to a transparency test. This may explain why the AGCM preferred to deal with payment by data under the UCPD rather than under the UCTD.

Hence, a combination of institutional design and related practical considerations are likely to influence enforcement decisions. Finally, as we have already mentioned above, the designation of gatekeepers by the proposed Digital Markets Act is likely to promote more effective public and private enforcement both in EU competition and contract law.

VI. STANDARD TERMS AND BEHAVIOURAL SCIENCE

A. Behavioural Competition Law and Standard Terms

Although the study of behavioural competition law includes an analysis specifically of standard terms,¹³⁷ it remains an emerging field, also with regard to digital platforms.¹³⁸ The general concern of behavioural competition law is that because consumer biases can weaken competition, firms are likely to exploit such biases. They can do so for instance by tying and bundling services, by individualised price discrimination supported by algorithmic decision making, by competing for active customers while overcharging inactive customers, and by foreclosing and overcharging in aftermarkets after offering low up-front prices. Profit maximising schemes can also deliberately facilitate consumer error by obfuscating their terms through complexity.¹³⁹

In fact, behavioural exploitation cannot readily be solved by competitive process because it tends to become a parameter of competition itself, especially where the gains of educating consumers cannot be appropriated effectively.¹⁴⁰ This leads to a bleak vision where competition by a ‘confusopoly’ may entail results that are suboptimal even to those of a monopoly.¹⁴¹ Such exploitation of biases leads to consumers making suboptimal choices from an individual perspective that not only amount to exploitation. It also promotes customer lock-in that frustrates freedom of choice and excludes competitors at the same time. And it facilitates a transfer of the consumer surplus to the producers. Hence features such as the excessive complexity of consumer contracts hurt not only the consumer, but also the competitive process and consumer welfare in general.¹⁴²

This problem appears to be amplified in a digital platform context, where consumers are isolated from each other, at an extreme disadvantage in terms of the available data, meaningful comparisons are difficult to make, and alternatives are in any event hardly available. Tying and bundling of services may be reinforced by status quo bias, with the resulting lock-in frustrating contestability. The German Facebook Case and possibly the App Store cases involve such a narrative. Fairness may also be compromised as consumers choose based on what is most salient (such as upfront price or quality), and ignore restrictive terms and conditions, transfer of personal data and lock-in effects.

At the same time behavioural competition law still accords value to competition and therefore competition policy, and has implications for possible solutions as well. To some extent competition law will be able to take account of the effects of consumer biases, for instance in its market definition and dominance analysis, as well as in the design of remedies. Bounded rationality in terms of status quo and default bias in relation to digital platforms has been identified as a concern in EU law as early as the 2004 Microsoft browser remedy designed to activate user choice, and again in the remedies in the EU's Google cases where status quo bias was identified as a major driver of the abuse.¹⁴³ Taking the role of information intermediaries into account in competition policy enforcement is another possible avenue.

Regarding the role of regulation, regulators as repeat players with institutional resources, investigating powers at their disposal, are likely to be better positioned and less bias-prone than consumers. In this sense the Digital Markets Act may promote fairness in the area of digital platforms’ standard terms as well. Finally, the law on unfair contract terms may be one way of helping to ensure that competition delivers good consumer outcomes.¹⁴⁴ Although it is one of the key assumptions of this article that this is indeed the case, as we will see below consumer law faces its own struggles in the light of behavioural constraints. Hence it may require more proactive measures on account of digital platforms, such as compliance by design.

B. Behavioural Consumer Law and Standard Terms

Behavioural consumer law mainly focuses on whether regulation and its interpretation by the courts, in particular with respect to transparency, can guide human behaviour and consequently is effective.¹⁴⁵ This research is particularly relevant in the case of ex post control of standard terms under EU contract law. It seems less relevant in the case of ex ante control of such terms by means of collective action.¹⁴⁶ Another relevant aspect of behavioural consumer law is how the law could police the use of consumer biases by digital platforms in their algorithms. From a contractual behavioural perspective there seems to be less research into this.¹⁴⁷

In the literature, the use of the average consumer as a benchmark is widely criticised on the basis of behavioural insights. In the UCTD the concept of the ‘average consumer’ is not codified, but the CJEU introduced this concept as a yardstick with respect to the good faith and the transparency test.¹⁴⁸ The concept of the average consumer resembles the homo economicus and because behavioural science has demonstrated that the average consumer does not in fact exist, using the average consumer as a benchmark is questionable. The seller's or supplier's disclosure duty prior to the conclusion of the contract is based on this notion of an average consumer. Although this is now generally accepted as a myth,¹⁴⁹ the CJEU still presumes that the average consumer reads this information and enters into a contract well- informed.

Empirical studies show that the vast majority of consumers never read standard terms before they enter into a contract.¹⁵⁰ As a result, alternative solutions have been offered, such as the use of rankings or symbols to warn the consumer. Moreover, if consumers would read the standard terms, they will not understand them, since they are often incomprehensible.¹⁵¹ Yet other empirical studies suggest that plain and comprehensible terms do play a role in a later phase, in particular in the case of conflicts between the contracting parties.¹⁵² Firstly, consumers do seem to read the standard terms in this phase, and secondly, plain and intelligible information may empower consumers cognitively. If they understand a text better, consumers are more likely to inquire with the seller or supplier if there is a problem. This also affects the ex post control of standard terms which may be deemed transparent and fair if designed as such.

Fairness by design is a category that is being asserted and developed in modern theories on economic approaches to consumer law.¹⁵³ However, this principle is already found in the EU data protection rules, and, as we have seen, in the proposal for a Digital Markets Act. Thus, Article 25 of the General Data Protection Regulation requires data protection by design and default,¹⁵⁴ and Recital 68 of the proposal for a Digital Markets Act (cited above) likewise spells out a compliance by design requirement.

Jointly, this suggests that private governance should made be fair by design and behaviour- proof by actively avoiding the exploitation of biases and promoting the use of nudges of average consumers in the right direction (including defaults), as well as cognitive empowerment of consumers by informing them in a comprehensible manner. Certain consumer authorities and financial market regulators have already started demanding this more broadly in relation to unfair commercial practices.¹⁵⁵

VI. CONCLUSION

In this article, we have looked both at (1) consumers and (2) business users/purchasers in their relation to (3) digital platforms. Our aim was to establish whether there is convergence in control of private governance by means of standard terms could be based on the developing fairness norms in EU competition and contract law; if so, whether these norms could be enforced effectively, and what behavioural modifications are required from digital platforms.

In this context identifying a workable standard for fairness in private governance was key. The findings that we have presented above, which are necessarily preliminary on further research, are that this can plausibly be done by defining fairness in this context as predicated upon (1) balancing and (2) a duty of care that involves compliance by design. Both are proactive duties that reflect the special responsibility of the undertakings involved, and they can jointly be framed as an expression of accountability. For effective accountability to occur, the application of these requirements should be made explicit, so they can be tested.

A. Fairness and Proportionality: Balancing of Interests

The correction of a structural imbalance between market parties, which forms the basis of both abuse of dominance law and consumer law in this area, is to do with balancing.

Therefore, it is in line with the core of proportionality—the balancing of interests. In this case, its reason is the correction of market power. The dominant parties have to do this balancing themselves and are vetted on that basis.

From the case law we derive rules on digital platforms demonstrating sequentially (1) the necessity of a legitimate objective for the standard terms, as well as (2) the balancing of the interest involved, and (3) to apply due process, in other words a proportional application of their terms. However, these norms are more pronounced in EU competition law than in EU contract law, as is the focus on market power. EU contract law by contrast places more emphasis on the less stringent norm of transparency.

B. Fairness and Behavioural Propriety: The Duty of Care

We believe the implications of the above-mentioned insights from behavioural science relate primarily to the notions of duty of care and compliance by design. Previously, we have already argued in favour of the former.¹⁵⁶ Compliance by design is now being developed in modern theories of economic approaches to consumer law,¹⁵⁷ and more broadly can be found in regulatory frameworks such as the General Data Protection Regulation (‘GDPR’) and the proposal for a Digital Markets Act.

In order to become behaviour proof, private governance rules established as standard terms should encourage: (1) consumer/purchaser biases not being exploited; (2) the use of possible nudges, such as appropriate defaults in the context of standard terms; and (3) cognitive empowerment of consumers/purchasers to overcome biases.

This can be linked to the balancing requirement that we discussed earlier, as it would redress the market power and informational balance between the parties involved. It can also be linked to a duty of care and to compliance by design as positive duties to act in order to avoid damaging the interests of the contracting parties—notably consumers and customers—by digital platforms. As positive duties they are required of the dominant party, whose behaviour and private governance—the standard terms—should be vetted based both on EU competition and contract law on that basis.

C. Practical Tests

We propose giving the concept of fairness a specific meaning in the context of applying competition and contract law to digital platforms’ standard terms that has a very practical dimension. Linking fairness to proportionality and compliance by design introduces requirements of balancing and appropriate proactive behaviour that enable the use of existing legal tests. Greater reliance on the special responsibility of dominant companies leads to testing for illegal conditions based on the standards of necessity for a legitimate objective and adequate balancing of interests, as well as due process. The behavioural insights mentioned imply qualitative requirements of effective form and content. This dovetails with the demand that a duty of care is observed: compliance by design should not just be implemented but actively improved upon.

D. Enforcement

Both EU competition and contract law can be enforced ex post and ex ante, as well as publicly and privately. With respect to public enforcement both the scope of the instruments and the competences of especially national consumer authorities appear to explain the choice of instrument to policy digital platforms’ standard terms. This suggests a need for more coordination between enforcement of anti-trust rules and rules of contract law, or more integration of such enforcement where it is located with a single consumer law and competition authority at national level.

E. The Relevance of Regulation

Finally, there is nevertheless a broad-based view that further asymmetrical controls are needed, which check for market power. The targeted regulation for gatekeepers which the European Commission has recently proposed in the Digital Markets Act is in line with this demand. By reversing the burden of proof in relation to the existence of market power that requires additional checks, this regulation will reinforce both public and private EU competition and contract law enforcement. Jointly this is likely to lead to a further dynamic development of this area of law that may help to establish greater fairness in private governance.