Choose the experimental features you want to try

This document is an excerpt from the EUR-Lex website

Document 52020SC0363

    COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT REPORT Accompanying the document Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on contestable and fair markets in the digital sector (Digital Markets Act)

    SWD/2020/363 final

    Brussels, 15.12.2020

    SWD(2020) 363 final

    COMMISSION STAFF WORKING DOCUMENT

    IMPACT ASSESSMENT REPORT

    Accompanying the document

    Proposal for a

    REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

    on contestable and fair markets in the digital sector (Digital Markets Act)

    {COM(2020) 842 final} - {SEC(2020) 437 final} - {SWD(2020) 364 final}


    GLOSSARY

    Term or acronym

    Meaning or definition

    ACCC

    Australian Competition and Consumer Commission (Australia competition authority)

    B2B

    Business-to-business

    B2C

    Business-to-consumers

    BEUC

    European Consumer Organisation

    CPC

    Consumer protection cooperation

    CMA

    Competition and Markets Authority (United Kingdom competition authority)

    DESI

    Digital Economy and Society Index

    DMU

    Digital Markets Taskforce

    DSA

    Digital Services Act

    ECA

    European Court of Auditors

    EEA

    European Economic Area

    ECD

    e-Commerce Directive

    ECHR

    European Convention on Human Rights

    ECN

    European Competition Network, consisting of NCAs

    FTE

    Full-time equivalent

    GDP

    Gross Domestic Product

    GDPR

    Regulation (EU) 2016/679 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’)

    ICN

    International Competition Network

    ICT

    Information and Communication Technologies

    IMCO

    Internal Market and Consumer Protection

    JURI

    Legal Affairs Committee

    LIBE

    Committee on Civil Liberties, Justice and Home Affairs

    MCAD

    Misleading and Comparative Advertising Directive

    MFN

    Most Favoured Nation

    NCA

    National Competition Authority of the EEA

    NCT

    New Competition Tool

    NFC

    Near-Field-Communication

    NGO

    Non-Governmental Organisation

    OECD

    Organisation for Economic Co-operation and Development

    OPC

    Open Public Consultation

    P2B Regulation

    Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services

    SME

    Small and Medium Enterprise

    TFEU or Treaty

    Treaty on the Functioning of the European Union

    UCPD

    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 (‘Unfair Commercial Practices Directive’)

    UCTD

    Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts (‘Unfair Contract Terms Directive’)

    UNCTAD

    United Nations Conference on Trade and Development



    CONTENTS

    1.    INTRODUCTION: POLITICAL AND LEGAL CONTEXT    

    1.1.    Political context    

    1.2.    Field of intervention    

    1.3.    Relationship of the initiative with other ongoing initiatives    

    2.    PROBLEM DEFINITION    

    2.1.    What are the problems?    

    2.1.1.    Weak contestability of, and competition in, platform markets, or risk thereof    

    2.1.2.    Unfair gatekeeper practices vis-à-vis business users    

    2.1.3.    Legal uncertainty for market players    

    2.2.    What is the size of the problem?    

    2.3.    Problem drivers    

    2.3.1.    Market failures    

    2.3.2.    Fragmented regulation and oversight    

    2.3.3.    Conclusion: problem drivers’ effects    

    2.4.    How will the problem evolve?    

    3.    WHY SHOULD THE EU ACT?    

    3.1.    Legal basis    

    3.2.    Subsidiarity: necessity and added value of EU action    

    4.    OBJECTIVES: WHAT SHOULD BE ACHIEVED?    

    4.1.    General objective    

    4.2.    Specific objectives    

    4.2.1.    Address market failures to ensure contestable and competitive digital markets for increased innovation and consumer choice    

    4.2.2.    Address gatekeepers’ unfair conduct    

    4.2.3.    Enhance coherence and legal certainty to preserve the internal market    

    4.3.    How do the objectives link to the problems identified?    

    5.    WHAT ARE THE AVAILABLE POLICY OPTIONS?    

    5.1.    What is the baseline from which policy options are assessed?    

    5.2.    What are the main parameters that determine the range of available policy options?    

    5.2.1.    Scope: core platform services, gatekeepers, thresholds    

    5.2.2.    Unfair practices    

    5.2.3.    Speed and Flexibility    

    5.2.4.    Enforcement framework    

    5.2.5.    Summary and main trade-offs    

    5.3.    Policy options    

    5.3.1.    Option 1 – Pre-defined list of gatekeepers and immediately applicable obligations    

    5.3.2.    Option 2 – Partially flexible framework of designation and updating of obligations, including regulatory dialogue for implementation of some    

    5.3.3.    Option 3 - Flexible option based exclusively on qualitative scoping thresholds    

    5.4.    Policy options discarded at an earlier stage    

    5.4.1.    A broad scope across platforms    

    5.4.2.    Information obligations without addressing unfair conduct    

    5.4.3.    A broad scope across markets    

    6.    WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?    

    6.1.    Internal market    

    6.2.    Growth and productivity    

    6.3.    Competition and Innovation    

    6.4.    International trade    

    6.5.    Employment    

    6.6.    Businesses    

    6.6.1.    Gatekeepers    

    6.6.2.    SMEs    

    6.7.    Consumers    

    6.8.    Regulatory Authorities    

    7.    HOW DO THE OPTIONS COMPARE?    

    7.1.    Effectiveness    

    7.2.    Efficiency    

    7.3.    Coherence    

    7.3.1.    Coherence with the Digital Strategy    

    7.3.2.    Coherence with the DSA    

    7.3.3.    Coherence with other instruments    

    7.4.    Proportionality    

    7.5.    Subsidiarity    

    7.6.    Conclusion    

    8.    PREFERRED OPTION    

    8.1.    The main building blocks of the preferred option    

    8.2.    The scope of application    

    8.2.1.    Identification of core platform services    

    8.2.2.    Designation of gatekeepers    

    8.2.3.    Obligations applicable to gatekeepers’ core platform services    

    8.3.    Enforcement framework and governance    

    9.    HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?    

    1.INTRODUCTION: POLITICAL AND LEGAL CONTEXT

    1.The digital transformation has profoundly changed the functioning of the global economy and society. The Covid-19 crisis and the increased importance and use of digital services has only further evidenced the importance of ensuring a borderless, fair, and contestable Single Market for digital services where companies can thrive and where citizens have genuine choices and control.

    2.This Impact Assessment examines the possible policy options to ensure a competitive Single Market for digital services and in particular fair and contestable platform markets. It combines the assessment of two initiatives previously presented in separate Inception Impact Assessments: (i) the Digital Services Act (‘DSA’) package: ex ante regulatory instrument of very large online platforms acting as gatekeepers; 1 and (ii) the New Competition Tool. 2 These two initiatives have been subject to two parallel public consultations. 3  

    3.Given the breadth of the topics covered, both Inception Impact Assessments – including their respective consultations – were initially published separately. However, since the outset, both consultations were aimed at complementary solutions by “ensur[ing] a joint analysis of the results”, “with a view to exploring synergies and ensuring consistency on the policy options pursued, in particular as regards possible remedies and enforcement”. 4 The holistic approach presented in this impact assessment is the result of such exercise.

    Political context

    4.Over the last years, a wide range of studies at international level as well as by National Competition Authorities (‘NCAs’) have brought to the fore the acute problems afflicting digital markets in terms of contestability as well as of the fact that a number of large platforms are taking advantage of their position to restrict competition including by means of the imposition of unfair conditions on their trading partners and on consumers.

    5.In this respect, the Commission has initiated a reflection process about the role of competition policy in a fast-changing world, which included commissioning a report from a group of independent Special Advisers to Commissioner Vestager published in April 2019. Among other aspects, the report concluded that “the specificities of competition in the digital world […] make market power “sticky”, and there is legitimate fear that the market power [large platforms] have acquired will be hard to challenge. Furthermore, they have been able to build, on top of their core competencies, entire ecosystems which make it hard for new entrants to compete on the merit and which, many observers feel, face little competitive pressure”. 5  

    6.A subset of issues pertaining to all digital platforms had previously been addressed through regulation in the so-called ‘Platform-to-Business Regulation’ (‘P2B Regulation’), aiming to increase transparency and fairness in platforms that can easily hold asymmetric bargaining power. To analyse further emerging issues addressed in the independent Special Adviser Report, the Commission also established the ‘EU Observatory on the Online Platform Economy’bservatory on the Online Platform Economy supported by an expert group 6 , to support the Commission in monitoring and analysing the developments in the online platform economy. Evidence gathered by this expert group further confirmed the findings of the previous reports.

    7.Similar reflections are taking place in some of the EU’s major trading partners, including the United States of America (‘US’), Japan, the United Kingdom (the ‘UK’), Australia and China. These reflections include calls for a new regulatory framework for platforms with “significant and durable market power” (US House of Representatives Majority Staff report 7 ), “substantial market power” (ACCC report 8 ), strategic market status (Furman report 9 ) and “bottleneck power (Stigler Center report 10 ). The US report notably concludes that each investigated platform now serves as main gateway to consumers and other businesses that each platform uses this role as major gateway to maintain its market power; and that the firms have abused their role as intermediaries to further entrench and expand their dominance.

    8.The need to address these concerns in digital markets was expressed in Commission President von der Leyen’s mission letter for Executive Vice-President Vestager, where she stated that in “striving for digital leadership, we must focus on making markets work better for consumers, business and society. The letter tasked Executive Vice-President Vestager with ensuring “that competition policy and rules are fit for the modern economy”. 11  

    9.This objective was reiterated in the Commission’s Communication Shaping Europe's digital future, as “it is important that the competition rules remain fit for a world that is changing fast, is increasingly digital and must become greener”. 12 In the same Communication, the Commission further stated that “competition policy alone cannot address all the systemic problems that may arise in the platform economy”. Against this background, the Commission also announced that it “will further explore, in the context of the DSA package, ex ante rules to ensure that markets characterised by large platforms with significant network effects acting as gatekeepers, remain fair and contestable for innovators, businesses, and new market entrants”. 13

    10.In the European Parliament, the Committee on the Internal Market and Consumer Protection (‘IMCO’), the Committee on Civil Liberties, Justice and Home Affairs (‘LIBE’) and the Legal Affairs Committee (‘JURI’) published draft reports in April and in May 2020, as legislative own-initiative reports. 14 The final IMCO and LIBE Committees reports were adopted in September 2020 15 and the draft JURI report in October 2020. 16 In parallel to these reports, the European Parliament also adopted a resolution on competition policy on 18 June 2020, where it “calls on the Commission to assess the possibility of imposing ex ante regulatory obligations where competition law is not enough to ensure contestability in these markets”. 17 The Digital Markets Act (‘DMA’), by proposing ex ante rules for certain large platforms and aiming at ensuring fair and contestable digital markets, responds to these calls for action.

    11.The European Council confirmed the need to act in its New Strategic Agenda 2019-2024, by stating that “[w]e will continue to update our European competition framework to new technological and global market developments”. 18 The Council of the European Union (‘Council’) also “supports the Commission’s intention to collect evidence of the issue and further explore ex ante rules to ensure that markets characterised by large platforms with significant network effects, acting as gate-keepers, remain fair and contestable for innovators, businesses and new market entrants”. 19

    12.Furthermore, the Council welcomed the public consultation on a ‘New Competition Tool to address structural competition problems across markets’ and expressed its willingness to discuss the Commission’s proposal for a DSA Package. 20 The Council underlined “that new policy approaches for the Single Market have to be fit for the digital age [and] able to cope with new and agile business models, especially in the digital economy”. 21 Finally, the Council reiterated the importance of swift action on the DSA package in its most recent conclusions, in which it “looks forward to the Commission’s proposal for a Digital Services Act by the end of this year”. 22  

    Field of intervention

    13.The feedback and evidence collected pointed to an urgent need to act in the digital sector, due to the particular features of digital markets. On that basis, the present impact assessment focuses on intervention options with regard to digital markets, with a focus on those markets characterised by the presence of large digital platforms where problems are most prominent, and action appeared most pressingly needed.

    14.In the digital sector, there is a small number of online platforms – often embedded in their own ecosystems – which have come to play a crucial role in the lives of millions – if not, billions – of individuals and companies. They intermediate a significant portion of transactions between consumers and businesses, and have emerged as a key structuring element of today’s digital economy. As such, these platforms have a major impact on, control the access to, and are entrenched in digital markets, leading to extreme dependencies of many businesses on these important platforms. The evidence points to negative effects on effective competition and on the contestability of the markets concerned. Member States in the EU observing these tendencies have begun to take regulatory initiatives to address these effects, potentially fragmenting the Internal Market.

    Online platforms cover presently a wide-ranging set of activities including online advertising platforms, marketplaces, search engines, social media and creative content outlets, application distribution platforms, communications services, payment systems, and platforms for the collaborative economy. They share some important and specific characteristics, in particular:

    ·they have the ability to create and shape new markets, to challenge traditional ones, and to organise new forms of participation or conducting business based on collecting, processing, and editing large amounts of data;

    ·they operate in multi-sided markets but with varying degrees of control over direct interactions between groups of users;

    ·they benefit from network effects;

    ·they rely on information and communications technologies to reach their users, instantly and effortlessly, benefitting from economies of scale and scope; and

    ·they play a key role in digital value creation, notably by capturing significant value (including through data accumulation), facilitating new business ventures, and creating new strategic dependencies.

    15.The scope of this initiative is limited to the digital sector. In fact, the market concentration tendencies and the underlying market dynamics in the digital sector, as well as other characteristics of digital markets, have contributed to several market failures in this area, which are likely to lead to inefficient market outcomes in terms of higher prices, lower quality, less choice and innovation to the detriment of European consumers (see Section 2.2).

    16.Even though some of relevant market features are also observed to some extent in other markets, they are most prevalent in digital markets. They include market features such as extreme scale economies, often resulting from nearly zero marginal costs to add customers and business users –- in contrast to off-line business models where such upscaling would involve major investments – and strong network effects associated to the multi-sidedness of online platforms, as well as data driven-advantages that often fundamentally change the competitive process, leading to sudden and radical decreases in competition (see Section 2.3.1). The presence of large platforms, often vertically or horizontally integrated in large ecosystems, exacerbates the negative effects that these features can trigger, thus making it impossible for the markets to self-correct.

    17.The problems in the digital sector are also most pressing from an internal market perspective. In fact, the OPC and the targeted consultation of NCAs have largely shown that the most salient examples of market failures today stem from the digital sector. 23 Consumer organisations like BEUC have also prominently flagged the particular concerns surrounding digital markets. 24 Likewise, digital markets featured prominently in the expert reports commissioned for the Impact Assessment. 25

    18.On the basis of the available evidence, including the Commission’s regulatory and competition enforcement experience, the Commission has mapped a number of ‘core platform services’ which exhibit these features and where absent regulatory intervention the identified market failures would effectively remain un-addressed.

    19.These ‘core platform services’ are those where the problems identified in Section 2.1 are most evident and prominent and where the presence of a limited number of large online platforms that serve as gateways for business users and customers has led or is likely to lead to weak contestability of markets. On the basis of the evidence collected for this Impact Assessment (see Section 5.2.1), the screening of problems led to the identification of the following core platform services: (i) online intermediation services (including marketplaces and app stores) (ii) online search engines, (iii) social networking (iv) video sharing platform services, (v) number-independent interpersonal electronic communication services, (vi) operating systems, (vii) cloud services and (viii) advertising services, including advertising intermediation services, provided by providers of one or more of the above.

    20.Furthermore, as explained in Section 5.2.1, it should be possible to designate gatekeepers whenever it can be demonstrated that a provider of core platform services: (i) has a significant impact on the internal market, (ii) operates one or more important gateways to customers and (iii) enjoys or is expected to enjoy an entrenched and durable position in its operations.

    21.While this Impact Assessment focuses on issues caused by gatekeepers operating in digital markets, it does also fully recognise the benefits that online platforms bring to the economy and society. The purpose of the DMA initiative is therefore to allow these platforms to unlock their full potential by addressing the most salient incidences of unfair practices and weak contestability so as to allow consumers and business users alike to reap the full benefits of the platform economy. 

    Relationship of the initiative with other ongoing initiatives

    22.In parallel to the present Impact Assessment, the Commission is also presenting an Impact Assessment for the Digital Services Act (‘DSA’), an initiative seeking to address primarily societal risks of digital markets and, more specifically, of very large platforms. The definition of ‘gatekeepers’ in the DMA is different in nature and scope from the definition of ‘very large platforms’ falling within the scope of the asymmetric obligations under the DSA. Whilst a handful of gatekeepers may be subject to both the DSA and the DMA, the risks addressed by the DSA and DMA are, however, very different. The DMA addresses risks to contestability and fairness in digital markets where gatekeepers as defined are present. The DSA addresses risks derived from the fact that very large platforms have become de facto public spaces, playing a systemic role for millions of citizens and businesses, creating a need for more accountability for the content which these providers distribute on their platforms. The different risks that both initiatives seek to tackle also translate in different obligations, the content and applicability of which is clearly distinguishable. 

    23.The impact assessment of the DMA is also being conducted in parallel with a number of ongoing reviews of certain competition rules, most notably the review of the Block Exemption Regulations for horizontal and vertical agreements, including in the motor vehicle sector. 26 These reviews are without prejudice to the Impact Assessment of the DMA, an initiative of a regulatory nature that does not affect – and is not affected by – those reviews. The Block Exemption Regulations pursue a different objective. They apply Article 101(3) of the TFEU by regulation to certain categories of agreements falling within Article 101(1) TFEU, thereby block exempting them from the application of Article 101(1) TFEU.

    24.Also ongoing is the review of the Market Definition Notice, which is without prejudice to the Impact Assessment of the DMA. 27 The Market Definition Notice pursues a different objective since it is a soft law instrument providing guidance as regards how the Commission “applies the concept of relevant product and geographic market” in its enforcement of EU competition law. 28

    2.PROBLEM DEFINITION 

    What are the problems?

    25.Over the past decade, online platforms have established their presence as important economic players and boosting efficiency, as well as spurring innovation and the development of new business models. Online platforms play an important role in many industries, allowing buyers and sellers of goods and services to trade and communicate with each other. They increase consumer choice and convenience, improve efficiency and competitiveness of industry and can enhance civil participation in society. Online platforms are key drivers of innovation in the digital world and their success is closely tied to the success of a range of businesses that use platforms to reach customers. Platforms allow especially smaller businesses to extend their operations beyond their home state, catering for consumers across the entire Single Market. 29  

    26.At the same time, they also raise new issues relating to fairness, transparency and market distortions. According to the evidence collected by the Commission, digital markets are particularly vulnerable to the following three problem clusters.

    27.First, in many digital markets large digital providers have emerged as gatekeepers serving as gateways for their business users and consumers. Some of these gatekeepers exercise control over whole platform ecosystems that are essentially impossible to contest by existing or new market operators, irrespective of how innovative and efficient they may be (‘weak contestability of platform markets’). As a result of the weak competitive pressure experienced by these large players, the likelihood increases that these markets do not function well – or may soon fail to function well – and thus do not deliver the best outcome for consumers in terms of prices, quality, choice, and innovation (weak competition in digital markets, or risk thereof).

    28.Second, many businesses are increasingly dependent on these gatekeepers, which in many cases leads to gross imbalances in bargaining power and, consequently, unfair practices resulting in conditions for business users that would not be achievable under normal circumstances (‘unfair business conditions for business users’). 30 Such imbalances in bargaining power, coupled with the economic dependency of many business users and costumers on gatekeepers, allow the latter to obtain conditions that they would not be able to obtain in case of well-functioning and competitive markets.

    29.Third, digital players typically operate at a global scale and deploy global business models. As a result, different national legislations within the EU 31 may lead to increased regulatory fragmentation and increased compliance costs for these large market players and the business users that rely on them (‘fragmented regulation and oversight’). Smaller players and startups are also negatively affected by this situation, as it impedes them from scaling up easily within the Internal Market in order to grow into contenders vis-à-vis the large, established players in the market.

    30.This section describes each of the three problem clusters above in more detail.

    Weak contestability of, and competition in, platform markets, or risk thereof

    31.The market features of digital markets tend to favour the emergence of a few large firms that have become gatekeepers for many digital products and services. These gatekeepers represent a key segment of the digital economy and play an important role in providing third parties with online access to a large number of European consumers. 32 Where such markets have not yet gravitated towards high concentration, they are at the risk of doing so in the near to medium term.

    32.There is evidence for a trend of growing market concentration (and, relatedly, growing mark-ups) at the industry level, which has been documented both for the US and for the EU. 33 In digital markets in particular, the level of concentration of economic power is unprecedented: the top seven of the large platforms account for 69% of the total EUR 6 trillion valuation of the platform economy, as a result of vertical and horizontal integration. 34 Large online platforms intermediating between businesses and consumers are growing at an exponential pace. They have several hundreds of millions of users (both businesses and citizens/consumers). 35 Total net revenues of some of these platforms (of billions of euros) double and triple over a few years. Moreover, five out of the world’s ten largest companies by market capitalisation are digital conglomerates (see Figure 3 ).

    33.Several respondents, including startups, research institutes and trade associations, point out the positive impact of platforms on startups: by lowering the barriers to entry and extending to companies of all sizes the advantages of cost and speed that can be gained from trading online, they stimulate innovation and the dissemination of new products and technologies. 36 Nevertheless, the large majority of respondents to the OPC 37 and NCAs 38 broadly agreed that “one or few large players on the market (i.e. concentrated market)” constitutes a very important or important source or part of the reasons for market failures. In certain markets, it may be challenging to maintain ‘competition in the market’, notably where having only one network may be the most beneficial outcome for consumers. However, in such a situation it is essential to keep ‘competition for the market’ open. Any successful attempt by a firm to lock in a group of consumers, so that the market is no longer contestable for a new entrant, will prevent such ‘competition for the market’, with possible adverse consequences for prices, quality, choice and innovation. 39  

    34.Large gatekeepers benefit significantly from the entry barriers characterising digital markets. In this context, new market operators that may want to enter or expand in digital markets where a gatekeeper is present may find it extremely difficult to overcome some of the inherent barriers to entry or expansion without access to a sufficiently large user base. 40 For instance, a new entrant must convince a sufficient number of users (due to the importance of network effects) to coordinate their migration to a new service, taking e.g. part of the social network along, or other associated data assets such as purchase or preference histories, or ratings. This lack of contestability due to high barriers to entry is extensively echoed in the academic literature. 41

    35.These gatekeepers therefore have an entrenched market position, which is hard to contest, and which they further expand through the creation of ecosystems. The largest platform companies are active across many different markets, creating extended data-driven ecosystems around their core activities, often cross-subsidising one service with data or revenues from another. In this regard, a large number of respondents identified online intermediation services, search engines, operating systems for smart devices, consumer reviews, network and/or data infrastructure/cloud services, digital identity services, online advertising intermediation services, payment services, fulfilment services and data management platforms as activities that can strengthen the gatekeeper role of such large online platforms when any or all of these are integrated within a single corporate structure. 42

    36.It is sometimes argued that incumbent offer their services often for free and that competition is ‘just one click away’ or that it is vigorous in some segments. This is a too narrow and selective view of the overall dynamics of the digital platform economy. However, the entrenched position of gatekeepers has shown to be lasting and essentially unchallenged by competing platforms, thus leading to weak inter-platform competition.

    Unfair gatekeeper practices vis-à-vis business users

    37.Gatekeepers’ successful business models based on platform economy specificities have allowed them to gain strong market positions and economic power, enabling them to create ecosystems for which they set the rules by which other economic players should abide. If set in an unfair manner, these rules can be detrimental to business users, and limit Small and Medium Enterprises’ (‘SMEs) online visibility and associated sales.

    38.The enforcement experience and input to the OPC show that unfair practices can take on different forms. They could relate to gatekeepers’ size, their capacity to acquire and monopolise data, imposition of contractual conditions or preferential treatment and/or the interplay between these elements. For a full overview of different unfair gatekeepers’ practices and the ongoing and closed investigations and antitrust cases please see Section 5.2.2 and Annex 5.6 to the Impact Assessment.

    39.One of such unfair practices is the imposition on business users of ‘anti-steering’ provisions, by which gatekeepers prevent business users from directing acquired consumers to offers other than those provided on the platform, even though such alternative offers may be cheaper or otherwise potentially more attractive. 43 For instance, an app store that does not allow its business users to advertise alternative subscription options outside its platform to acquire customers. Also, cross-platform parity clauses, 44 i.e. clauses that oblige business users to offer the same or better retail conditions as those offered on other platforms to the contract party, tend to disincentivise competition between platforms. In particular, they prevent business users from ‘rewarding’ other platforms that may provide better or cheaper platform services, by offering better retail prices or conditions on those platforms.

    40.Another example is the imposition of the platform’s ID services, which is a lock-in strategy where the user is required to sign up/register with an email service of the gatekeeper’s core platform services when using another of its products (e.g. an operating system, social network). The US House of Representatives Judiciary Committee also describes lock-in strategies including free tier offerings for cloud services. 45

    41.The broad category of ‘self-preferencing’ refers to practices in which a usually vertically integrated gatekeeper acting in the dual role of providing core platform services to business users and at the same time competing with them when providing ancillary services applies more favourable conditions to its own services compared to the third-party services hosted on the gatekeepers’ platform. Self-preferencing occurs in many situations in the online and offline world (e.g. in supermarkets with own brands). Such behaviour may not be considered generally anti-competitive under the EU competition rules or unfair in all business relationships. However, certain forms of self-preferencing may amount to an unfair business practice. An important concern here is the fair balancing of interests, in this case those of the gatekeeper platforms versus that of their business users. 46 In particular, the special position of gatekeeper platforms that play a dual role and may engage in favouring their own services may lead to the exclusion of alternatives by business users that are largely dependent on these gatekeeper platforms to reach consumers, reducing choice for them, and potentially undermining the quality of service and increasing prices. 47  

    42.One example is an app store, which markets a number of its own popular apps and at the same time maintains a marketplace for competitors, self-preferencing its own marketplace by applying more favourable policies for its own products and selectively drafting rules favouring its own products. Another example is a search engine or marketplace treating more favourably its own products and services in the results displayed to end users.

    43.Feedback to the OPC shows that business users consider self-preferencing to be a very common practice deployed by large, vertically integrated platforms. Responses by business users suggest that search and ranking algorithms often give preference to the platform’s own services, but also that a platform often has an incentive to bias its recommendations towards the content provider charging a lower royalty. 48

    44.Business users are faced with limited or no access to vast amount of data (e.g. app store limiting the information that third-party app providers receive about their subscribers) as well as lack of any or meaningful interoperability to access such data that may be collected by gatekeepers.

    45.The Impact Assessment study and input to the OPC point to practices that prevent both consumers and business users from switching. In the digital sector, being able to port both historical and real-time data is an important precondition for both multi-homing and switching. Business users and consumers alike repeatedly raise the issue of not being able to use any other platform or service because the incumbent refuses to provide an enhanced and continuous real-time ability to port personal and non-personal data in interoperable format. These practices affect contestability, and limit business users’ possibilities to move to or rely on alternative platforms or services. As yet another example, an advertising intermediation services provider collecting multiple datasets from business users’ services which it uses for better targeting and attribution measurement, but does not share them with advertisers.

    46.Another example of a data related practice that could be considered unfair, and has been raised by many stakeholders in the context of the OPC, is the situation where a gatekeeper restricts business users from accessing and using the data that they provide, receive from their customers or generate in the course of their use of the gatekeeper’s platform or service, as is the case as regards an app store limiting the information that third-party app providers receive about their subscribers or an online intermediation service restricting data generated in the curse of the use of its platform by third party sellers and their customers. Feedback to the OPC shows that business users are regularly confronted with the imposition by large platforms of proprietary services and an authentication through the platform even when third party services are used to create a direct link with customers to the detriment of third-party providers. Respondents suggest that gatekeepers exclude business users from access to user data and attempt to remove the direct link between the client and third party suppliers (so-called disintermediation). 49

    47.Other examples include (i) gatekeepers that use certain data that they received from business users for a particular use, for instance advertising services, for other, unrelated purposes, (ii) gatekeepers operating a marketplace benefit from their dual role and ability to evaluate product, sales and customer data generated from the sales of goods provided by third party merchant business users on its marketplace, or (iii) gatekeepers operating a video sharing platform that has access to a rich set of (first party) data about its consumers, data that it can re-use to improve its own products, including in other areas, but restricts the access to this rich set of data to its competitors.

    48.Thanks to their strong market position, gatekeepers, can either limit access to their platform or make such access conditional upon specific requirements. Gatekeepers often impose unfair terms of access to business users, for instance in relation to price for the services they offer or accepting specific bundles which do not allow the mix-and-match by customers (e.g. a provider of cloud services bundling this service with other services or a social network services provider applying terms and conditions, which make the use of its services conditional on the possibility to collect and combine user data from multiple sources).

    49.Another example is gatekeepers limiting the access to or the interoperability of certain of their platform services/functionalities (e.g. operating system) with the services offered by business users, reserving those functionalities to their own services. For example, in certain circumstances third-party providers of payment wallets may require access to near-field-communication (‘NFC’) functionalities in the hardware and the gatekeeper exclusively reserves such functionality to its own services.

    50.The work supporting this Impact Assessment shows that the above problematic practices are most prominent in relation to the following core platform services: e-commerce marketplaces, online search engines, app stores, social networks, video-sharing services, operating systems, cloud, number-independent messaging services, and online advertising. 50 Also the negative effects of the problem drivers are most severe in relation to these core platform services. These services present characteristics that have been identified as driving the problematic practices assessed in this document, i.e. they have a multi-sided character, which allows them to accumulate data on all sides of the market and thus benefit from strong indirect network effects (e.g. social network services, video sharing services); they have an important intermediation function (e.g. marketplaces, app stores) and/or playing important ‘visibility’ role (e.g. online search, operating systems), and are characterised by the presence of big ‘gatekeepers’ (e.g. number-independent messaging services) which often are vertically integrated and operate a large ecosystem (e.g. cloud services).

    51.From the foregoing, it becomes apparent that there is substantive evidence concerning the urgent need to address unfair practices by gatekeepers. The fact that this evidence cuts across not only jurisdictions of both common and civil law, but also across numerous enforcement bodies of different kinds and with diverse mandates is yet another indication of the urgency underlying the intervention. In fact, as outlined by the evidence quoted in this impact assessment and the numerous reports, there are only few other questions currently triggering a similar level of consensus between enforcement authorities, judicial bodies and law makers around the world like the need to tackle problems related to digital gatekeepers.

    Legal uncertainty for market players

    52.While in many areas of the single market, the objective is to ensure further integration, the online platform market is naturally integrated (due to the intrinsically cross-border nature of the platform economy). However, there is an increasing regulatory fragmentation of the online platform space in the EU (see Annex 5.4 and 5.5 to the Impact Assessment). In addition, coordination among national legislators may be insufficient, leading to potentially heterogeneous responses across the EU.

    53.Such fragmentation becomes problematic where it creates increased compliance costs for all market players. This is particularly harmful for smaller platforms and startups, potential entrants and smaller business users since it creates regulatory barriers to entry and limits their ease of scaling up across the Single Market. At the same time, diverging laws may also endanger the benefits stemming from large platforms’ activity; such costs may imply regulatory shopping, ultimately resulting in unequal impacts on EU consumers. If emerging platforms are unable to grow sufficiently in order to compete with gatekeepers, the latter would be able to further gain power, strengthening their ability to establish market rules and (potentially) behave unfairly. This would exacerbate the above described issues of weak market contestability and competition, as well as unfair business practices.

    54.In the OPC, respondents from all categories mention that EU level rules would prevent further legal fragmentation across Member States, considering that several Member States have already started to introduce new rules to address concerns arising from the presence of gatekeepers. Stakeholders generally consider that an effective coordination between EU bodies and the relevant national regulatory authorities is needed, especially in light of the fact that issues related to gatekeepers are likely to have an important cross-border component. Platforms in particular point out the need to minimise fragmentation and allow for a pan-European approach. 51

    What is the size of the problem?

    55.As explained in more detail in Section 2.1.1, the characteristics of digital markets often favour the lack of market contestability and the emergence of strong concentration, which tends to be accompanied by rising mark-ups and weaker competition. The trend of increasing industry concentration has been documented for both digital and non-digital industries alike. For instance, in 2014 the mean European high ‘digital intensity’ industry 52 had 4 percentage point higher sales concentration than in 2000. 53  

    56.As regards trends in mark-ups, empirical studies suggest that company mark-ups have increased by 4% to 6% for the period 2001-2014, on average across country, 54 and that the result is mainly driven by the top of the mark-up distribution in the digital sector. 55 For the top 10% of the firms in the sample, the growth in mark-ups over the period 2001-2014 amounted to 20%, while the remaining firms in the sample exhibit a flat trend, i.e. mark-ups stayed roughly the same. 56 To the extent that this observed trend of increasing market power of this top 10% of firms is a sign of insufficient competitive constraints faced by these firms, increasing competition in these markets could contribute to slowing down the growth trend in mark-ups, decrease prices and increase choice, quality and innovation. For example, a recent study shows that more concentrated industries also feature a more negative relation between markups and investment and innovation. 57  

    57.As regards the size of the problems related to unfair practices by gatekeepers (explained in Section 2.1.2), it is importance to note that the number of merchants and small businesses affected by gatekeepers’ conduct varies depending on the sector, but can be estimated to reach between one and four million. 58  

    58.A good indicator of businesses’ dependence on platforms is turnover from sales and share of revenue via online platforms as a proportion of the company’s total revenue from e-commerce. According to the Observatory’s estimates, around half of enterprises derived more than 25% of their revenues from online platforms. For almost 10% of companies, online platform sales exceed 75% of all revenues; while according to Statista estimates, in 2017, 18% of company revenues across the EU-28 came from e-commerce, the highest proportion being 33%. 59

    59.Another indicator of businesses’ dependence on platforms is the use of platforms to publish online advertising. Of SMEs in the EU that sell online, more than eight in ten rely on search engines as a mean of marketing their products or services. In 2018, an average of 26.2% of enterprises across the EU paid to advertise online. In northern European countries, such as Sweden and Denmark, this figure was over 44%. 60

    60.The degree to which businesses have integrated into and depend on the platform economy is further illustrated by the fact that in some cases more than 50% of goods sold on a marketplace come from third-party sellers. There are over 26.4 million software developers in the world who depend entirely on large platforms providing the infrastructure and setting the rules for the distribution of their apps.

    61.Gatekeepers’ unfair practices affecting businesses do not represent a one-off problem, but are systemic and recurrent. In the OPC on Ex Ante Rules, 88% of the businesses and business users that replied, encountered issues concerning trading conditions on large platforms. 61  According to Cullen International’s database, around 30 antitrust investigations concerning platforms have been formally opened in the EU (by DG Competition or NCAs) since 2015. 62 However, the prevalence of unfair practices by large gatekeeper platforms is evidenced not only in the number of cases that have been investigated by competition authorities, but also from the interviews and case studies run in the context of the support study for the Impact Assessment. 63

    62.As regards the anti-competitive use of third party data, both the Commission and NCAs (Italy, the Netherlands and Germany) are running a number of investigations against four different large online platforms. In case of preferential treatment, except for the Commission cases, also at least two NCAs (Italy and the Netherlands) are running investigations concerning self-preferencing by online platforms. With respect to inadequate or late access to own (business-user related) data and lack of access to key functionality, the investigations were initiated in each case against three different online platforms in three different Member States, while anti-steering and MFN clauses were recently subject to investigations against two different online platforms.

    63.The size of the market and the implications of the issues identified are also an angle of approaching the ‘the size of the problem’ question. The digital economy is estimated to account for between 4.5% to 15.5% of global Gross Domestic Product (‘GDP’) in 2019, depending on the definition. 64 The top 50 online platforms, representing an average of over 60% of traffic share 65 across the Member States, achieved worldwide revenues of almost EUR 276 billion in 2018, and employed almost 600 000 people. 66  Online platforms’ role is constantly increasing due to e-commerce upward trends; it has further strengthened with the widely introduced lockdowns due to the COVID 19 outbreak in 2020; consumers have shifted their habits more towards search engines, social media and online entertaining media. 67

    64.The ongoing fragmentation in the Digital Single Market might also reverse the positive trends in cross-border online trade. Assuming a 10% decrease per year in online cross-border trade, the opportunity cost of the digital market fragmentation would be EUR 1.76 trillion after 10 years (see Annex 5.5 to the Impact Assessment).

    65.The reduced contestability of digital markets in which gatekeepers operate seems to result in suboptimal innovation levels, with notably implications for societal welfare. 68 Relevant data supports the view that many markets are becoming more concentrated and display less competition. Profit margins are widening, with a few firms reaping a significant share. Innovation levels are also sub-optimal. 69  

    66.Gatekeepers bring benefits for consumers in terms of convenience, increased choice of free of charge online products and services. However, there are also important adverse consequences for consumers, namely reduced choice in terms of number of competitive platforms 70 , insufficiently informed choice decisions 71 , and lack of data/privacy-friendly services 72 . Also, due to the distorted intra- and inter-platform competition described in the problem definition, consumers risk experiencing higher prices and/or less quality.

    Problem drivers

    67.This section describes the problem drivers for the problem clusters described in Section 2.1. These problem drivers can be grouped into two overarching categories, namely gatekeeper related market failures (Section 2.3.1), and fragmented regulation and oversight (Section 2.3.2).

    Market failures

    68.From the competition law perspective, the term ‘market failure’ indicates a situation in which a market does not allow consumers to benefit from the results of effective competition in terms of low prices, better quality, as well as more choice and innovation, while firms are able to earn supra-normal profits which are not competed away over time. 73  

    69.While markets typically feature self-correcting mechanisms, there can be obstacles that prevent these mechanisms from operating, leading to non-transitory losses of economic value. 74 For instance, abnormally high profits in a market should in principle not be sustainable in the long run because they would attract new entry into this market. As the new competitors start offering the same or very similar products as the incumbent(s), they will steal market share, and hence profits, from them, until the abnormally high profits will gradually be competed away. However, this self-correcting mechanism may be impaired when there are, for instance, barriers to entry that make it very difficult or even impossible for potential competitors to enter the market and challenge the incumbents. Such barriers to entry are particularly salient in digital markets, because they do not allow entrants to be cost effective (because of scale and scope economies), to replicate the incumbent’s products or services (because of data dependency or vertical integration), or to induce consumers to switch away from the incumbent(s) (because of network effects, switching costs, or asymmetric information). Such barriers to entry therefore allow incumbents to sustain market power, which in turn leads to longer-term societal losses in terms of higher prices and less product variety for consumers, and less dynamic innovation. 75  

    70.It is important to stress that the features of a market include both structural and behavioural ones and that demand-side considerations, in particular the behaviour of customers, play an equally important role in this regard. Therefore, in many cases, there is a combination of those elements leading to or constituting a market failure. 76

    71.This section analyses a list of gatekeeper related market failures, notably: (i) entry barriers to gatekeeper markets (Section 2.3.1.1) and (ii) economic dependence and imbalanced bargaining power (Section 2.3.1.2).

    Entry barriers to gatekeeper markets

    72.Market players in the digital economy face important barriers to entry. This is due to the fact that digital market features can be exploited by gatekeeper platforms to strengthen their market position and prevent market entry.

    73.There has been broad consensus among the NCAs 77 , as well as among the respondents to the OPC 78 that extreme economies of scale and scope, high start-up costs, high fixed operating costs, high degree of vertical integration, single-homing, switching costs, multi-sidedness, network effects, zero-pricing markets, information asymmetry, data dependency access to data, and behavioural bias are important or very important sources for market failures in digital markets. Moreover, according to an International Competition Network (‘ICN’) report, an important proportion of respondents indicated that most of these factors were playing an important role in digital markets’ power assessment in the competition enforcement cases that they have investigated. 79  

    74.Regarding economies of scale, the Commission in Google Shopping, based its dominance assessment for the market for general search services among other things on the existence of barriers to expansion and entry, notably the significant investments in terms of time and resources required to establish a fully-fledged general search engine. 80 Likewise, in its Google Android decision, the Commission found that “developing a smart mobile OS [operating system] is a costly and time-consuming process”. 81 As regards economies of scope, the Commission made particular reference in its Preliminary Assessment in the Amazon e-book MFNs case to “[t]he ability of e-book readers to drive sales and lock-in customers: that with its Kindle e-book reader, Amazon operates a closed "ecosystem" (or "walled garden"). Customers who own a Kindle can use that e-book reader only for ebooks purchased in Amazon's Kindle store”. 82 In the same case, the Commission also found substantial economies of scale for e-book retailing, in particular because of the need to construct a sufficiently large catalogue of available titles (which requires agreements with a large number of E-book Suppliers), and because of the scale and scope of investments needed to set up a viable e-book distribution platform. The Special Advisers Report refers to “the presence of strong economies of scope favouring the development of ecosystems and giving incumbents a strong competitive advantage. Indeed, experience shows that large incumbent digital players are very difficult to dislodge”. 83  

    75.Due to the two-sided nature of platform markets, once a gatekeeper managed to bring both sides of the market on board, it becomes very difficult for a new, emerging platform, to establish itself in the market, as it has to convince both users and developers simultaneously that it is a viable alternative to the already established platform. For instance, in establishing Google’s market power in the Google Android case, the Commission quoted Orange as saying that “[g]iven the two-sided character of this market (attracting enough developers requires having a large user base and users will reciprocally be attracted to shops offering many apps) it is indeed very difficult to offer an app shop in competition with Google Play given (i) its link with Android OS and (ii) its current size”. 84  

    76.The problem of challenging a gatekeeper is often exacerbated in situations where at least one side of the market (typically the final users) is served at zero prices by the incumbent platform - with firms monetising their services through advertising and/or access to consumer data 85 - so that there is no room for the entrant platform to attract final users through aggressive pricing policies. 86  

    77.The zero pricing strategies described above also explain why network effects tend to favour large incumbents preventing smaller rivals from effectively challenging incumbents and stealing market shares from them. The Commission has found in the Microsoft case, for instance, that network effects represented a relevant barrier to entry because “[a] media player would not meet with significant consumer demand if there was no or no significant amount of corresponding digital content which this player could play back”. 87 In Google Android, the Commission found that “network effects arise because, when deciding which licensable smart mobile OS to develop for, app developers consider the revenue potential of that OS and since they ‘earn their profits mainly by app downloads, mobile OSs with a large user base are considered more attractive by app developers”. 88  

    78.Indirect network effects are particularly strong for large-scaled platforms also due to their unlimited capacity to expand data sets, i.e. data-driven network effects. In addition to this network amplification function, data is a major asset in the digital economy. It is particularly important for a business to have access to data related to its consumers and stemming from its activity on a platform since such data allow the business to adapt its market strategy. Business users’ dependency on data could be used to prevent them from competing effectively on the platform. This is particularly problematic when the business user is in direct competition with a gatekeeper who can use data generated by the business user’s activity to its own interest. Data can thus be used by gatekeepers as a barrier to entry, expansion and competition and is therefore an essential element for enabling market contestability. In the Google Shopping case, the Commission identified the availability of data in the form of user search queries, paired with users’ tendency to single-home on Google for their general searches, as an important barrier to entry 89 : “[B]ecause a general search service uses search data to refine the relevance of its general search results pages, it needs to receive a certain volume of queries in order to compete viably. The greater the number of queries a general search service receives, the quicker it is able to detect a change in user behaviour patterns and update and improve its relevance”. 90  

    79.The presence of network effects and the multi-sidedness of certain markets imply that even markets where initially multiple competitors are active are particularly prone to tipping: once a firm has obtained a certain advantage over rivals in terms of market share, its position may become unassailable and the market may gravitate towards a situation of dominance or (quasi)-monopoly. This advantage can be due to its presence in other related services, access to data or simply because it is the first mover into the market. In these cases, markets may not yet have generated a gatekeeper, but show clear signs of increasing market power in the hands of one firm. Respondents to the OPC generally considered that important or very important market features of a tipping market are the following: (i) direct network effects; (ii) indirect network effects; (iii) users predominantly single-home and (iv) economies of scale. Respondents generally considered that tipping is common or to some extent common in digital markets. 91 When asked about the need for the Commission to be able to intervene early in cases of emerging gatekeepers to preserve/improve competition, the large majority of respondents agreed, including the majority of businesses and business associations, civil society organisations (i.e. consumer associations, NGOs and citizens) and public authorities (including NCAs). 92

    80.Behavioural bias is another important feature of digital markets. This feature merits further attention in this section since it contributes to increasing switching costs and keeping users locked into the gatekeeper platform, i.e. leading to user lock-in, thus strengthening entry barriers. Platform companies routinely design services to optimise their users’ experience, often using advanced behavioural profiling and testing techniques, such as A/B testing 93 , or finely targeted personalisation of their service offerings. Gatekeepers use various techniques 94 (e.g. design of choices, misdirection, social pressure, sneaking items into the user’s shopping basket, and inciting a sense of urgency or scarcity) that ‘nudge’ users into certain decisions. A recent search on 11 000 shopping websites identified 1 818 patterns of practices used to incite users doing things they have not intended to do. 95

    81.From the perspective of platform competition, research on the basis of ‘agent-based simulations’ also found evidence of biases that reinforce consumer lock in, such as ‘escalation of commitment’, and ‘availability bias’. 96 In ‘escalation of commitment’, users commit themselves to one platform, even when switching may provide higher user utility. Hence, those users never switch platforms. For instance, a consumer purchasing on a large e-commerce marketplace offering a range of products, would not switch to one or several other platforms even if the latter are specialised in the specific type of goods the consumer is interested in. Convenience and user habits would prevail over the benefit (e.g. higher quality) potentially resulting from the use of a more specialised platform. Users subject to an ‘availability bias’ may make platform choice decisions using a heuristic that relies on vivid or recent data. For example, users may easily recall a platform that has many users, as social media would be mentioning such a platform. Social norming (e.g. follow friends’ behaviour) may play an additional role for user lock-in and increase switching costs. Behavioural bias discourages switching to different alternatives (such as a different browser, different search engine, etc.) whenever certain software products come pre-installed on consumers’ devices, and therefore has similar adverse effects on competition as would limited information about the existence of these alternatives.

    82.For instance, in the Google Android case, the Commission found that “users that find apps pre-installed and presented to them on their smart mobile devices are likely to ‘stick’ to those apps”. 97 In other words, users suffer from ‘default bias’ or ‘status quo bias’, which in turn makes pre-installation of operating systems, app stores, search engines, etc., a powerful tool to lock in users to these specific services: “In 2016, approximately 260 million smartphones were sold in Europe, of which approximately 197 million smartphones or 76% were Google Android devices. Practically all of these Google Android smartphones had the Google Search app pre-installed with the rest of the GMS bundle”. 98

    83.The Commission’s enforcement practice under Article 102 TFEU has shown that that the presence of high switching costs makes it more difficult for entrants to contest the market position of firms that have already acquired a large customer base. For instance, in an internal document, Microsoft itself stated that “The Windows API […] is so deeply embedded in the source code of many Windows apps that there is a huge switching cost to using a different operating system instead”. 99 Switching costs are also relevant where customers are businesses, not final consumers. This is demonstrated by the Google Android case, where the Commission found that “OEMs wishing to switch to other licensable smart mobile OSs face switching costs. […] For example, Sony has estimated that the initial development cost ‘to implement the Android OS on our devices was approximately 50 million Euro, with lead time of 1.5-2 years’”. 100 One implication of high switching costs in the platform context is that either one (or both sides) of the platform tend to single-home for specific purposes, i.e. users only use one platform, rather than using several platforms simultaneously. 101 For instance, the vast majority of smartphone users owns either an iPhone or an Android phone, but not both at the same time, and they tend to be very loyal to their operating system.

    84.The above entry barriers are gradually reinforcing each other due to the ‘winner–take-all’ dynamics in digital markets. The bigger the platform, the stronger the indirect network effects, the larger the amount of data and the higher its quality. This leads to increased insight into user profiles and preferences, allowing gatekeepers to offer them more personalised services and advertisements, thus attracting even more users and reinforcing consumer lock-in, favouring single-homing and rendering switching to alternative platforms more difficult.

    Economic dependence and imbalanced bargaining power

    85.Dependence and imbalanced bargaining power characterise business relations with all platforms 102 including small ones. What distinguishes however relations with gatekeepers, is the particularly strong level of dependency and the important scale of power imbalance, which together with unfair conduct engaged in by these gatekeepers can have serious harmful effects on the business users and customers.

    86.First, gatekeepers have become a strategic business partner; an enterprise not present on these platforms would not reach a very significant number of consumers. Figures in some sectors illustrate well the strong degree of dependence. In 2024, consumers are projected to download 181 billion apps from biggest app stores. 103  Over 80% of social referrals to e-commerce sites come through the most used social platforms some of which having more than 2 billion monthly active users and 7 million active advertisers. 104 This explains the millions business pages and companies using these social media every month 105 to target this large audience.

    87.Gatekeepers’ role as key trading partners is constantly strengthened due to e-commerce trends. The Business-to-consumers (‘B2C’) e-commerce turnover was growing at an average pace of 13% between 2014 and 2019 with turnover forecasted to hit EUR 621 billion in 2019. 106 On average, 16.2% of retail trade in 2020 in Europe takes place on line, almost double in comparison to 2018. The share of online shoppers in Europe making cross-border online purchase has also increased significantly over the past decade, nearing 50% in 2019. Cross-border B2C e-commerce sales in Europe are projected to grow at a double-digit rate at least through 2022. 107

    88.Second, the incomparable economic strength of gatekeeper platforms show the extent to which their commercial relations with business users are imbalanced. In 2019, the biggest app stores generated over USD 83 in revenue. 108 In 2019, one of the most important social networks had over USD 70 billion in revenue. 109

    Fragmented regulation and oversight

    89.Various national rules in the EU are emerging in partial response to the problems identified. In addition, insufficient coordination among different national authorities setting rules vis-à-vis platforms may lead to potentially heterogeneous responses across the EU. Fragmentation already exists with regard to platform-specific regulation, as for example in the cases of transparency obligations and MFN clauses.

    90.Furthermore, fragmentation results also from differing legislation relating to dependency situations in place in various Member States (Belgium, Bulgaria, France, Germany, Hungary and Italy) while in the remaining Member States there is no legislation addressing dependency in place. Regarding MFN clauses there are two different types of fragmentation: fragmentation due to the fact that some Member States imposed legislative bans and some Member States did not. Furthermore, fragmentation is observable due to differences in the MFN-legislations in those Member States where they are in place. For instance, in some Member States (Austria, Belgium, France, Italy) all types of price parity clauses are prohibited, i.e. also narrow MFN clauses, while in the remaining Member States MFN clauses may be challenged under EU competition law only.

    Conclusion: problem drivers’ effects

    91.The above-described problem drivers can lead to a number of issues. The economic dependence of business users on gatekeeper platforms and the imbalance of bargaining power - between these two types of players as well as between gatekeepers and smaller platforms - can result in important economic harm. Business users need to be present on gatekeeper platforms in order to reach consumers as the above analysis show, which allows gatekeepers to set the rules of access and use of their platforms in an unfair way. Gatekeepers can undermine the trading conditions for dependent business users by behaving unfairly, thus limiting (national and cross-border) sales and trust in the platform economy.

    92.Digital markets’ features (see in particular discussion on market failures in Section 2.3.1) have mutually reinforcing effects which in the winner takes it all dynamics of these markets constitute unsurmountable entry barriers. The latter drive a number of issues related to weak market contestability and gatekeepers’ sustained market position leading to longer term societal losses in terms of products’ and services’ prices, consumer choice and suboptimal innovation opportunities (as illustrated in Section 2.2 and further described in Section 6). Gatekeepers control the conditions for innovation and entry by independent firms. An important effect of the exercise of control by gatekeepers is that they can inhibit innovation by potential alternative platforms or by applications providers operating on their platform. Potential competitors, which might offer an alternative route to customers, may find it challenging to gain a foothold in markets with gatekeepers.

    93.The extensive nature of problems associated with gatekeepers has led a number of Member States to take or consider their own measures to address gatekeeper power. However, isolated and uncoordinated national approaches to addressing a problem which concerns cross-border platforms, and is hence pan-European in scope, risk creating different national rules which increase the compliance costs for platforms (and especially entrants or small scale platforms) operating cross border, and for business users, including SMEs, providing services across the EU. 110

    How will the problem evolve?

    94.Concentration and mark-ups in most digital markets have been increasing over the last years, and there is no indication that this trend will be inverted during the next years. In some cases, markets have already stabilised at a high concentration level and do not show any evidence of possible increase in competitiveness in the future. Data is also becoming more and more important, exacerbating the market failures associated with the control of data. 111

    95.The COVID-19 crisis has dramatically increased the importance of e-commerce and trading via digital platforms in the EU’s economy. 112 This has only accelerated the dependency of users and businesses on the services provided by the few gatekeepers – as evidenced indirectly by the increase in stock market valuation of some of the largest platform companies.

    96.The following graph illustrates the stock price development for five major big tech companies from 2014-2018. 113 To the extent that stock prices reflect market expectations of future profitability, this graph can be interpreted as measuring (future) profits of the respective companies. When comparing these figures to the S&P 500 index 114 , which grew by about 60-70% over the same period, this graph shows how the five digital operators – Alphabet (Google’s parent company), Apple, Facebook, Amazon, and Netflix – have consistently outperformed the market average.

    Figure 1 - Stock price development for 5 big tech companies, 2014-2018

    97.Absent any EU intervention, the economic drivers are likely to increase, exacerbating the observed problems. As an illustration, further development and use of voice assistants can also be expected to reinforce gatekeeper platforms’ position. Voice-activated services may create concerns in relation to search for online products/services/information. The provision of a single answer to a search request limits the possibility to access alternative results, thus reducing choice and limiting competition. 115  

    98.Innovation would remain concentrated within a small number of gatekeepers, ultimately limiting consumers’ possibility to access innovation and data-friendly services provided by a larger number of platforms than gatekeepers.

    99.Regarding fragmentation likely to occur in the near future, legislation to address imbalances in the relationships between digital platforms with economic power and their business users is currently in process to be adopted in several Member States (e.g. Germany, France and Romania). Some Member States (Belgium, Luxemburg and the Netherlands) are currently mainly supporting action at EU level. However, they would most likely take legislative action at national level in the absence of action at EU level. Those legislative projects already in the process of adoption and also those likely to be tabled in the absence of action taken at EU level demonstrate the likelihood of further fragmentation. 116

    3.WHY SHOULD THE EU ACT?

    Legal basis

    100.Given the intrinsic cross-border nature of the services provided by gatekeeper platforms and the risk of further regulatory fragmentation regarding functioning of the Single Market for digital services, in particular in relation to gatekeeper platforms as well as functioning of digital markets, Article 114 TFEU is the relevant legal basis for this initiative.

    101.As set out above, the current regulatory approaches at Member States level are a patchwork of existing or proposed regulatory solutions (see detailed description in Annex 5.4 to the Impact Assessment). This creates legal uncertainty for companies operating in the internal market, whether at national or on a pan-European basis and risks creating an appreciable distortion of competition in the internal market and undermine fundamental freedoms protected by the Treaty.

    Subsidiarity: necessity and added value of EU action

    102.The objectives of the intervention cannot be achieved by Member States acting alone, as the problems are of a cross-border nature, and not limited to single Member States or to a subset of Member States. The digital markets at stake (including those featuring gatekeeper platforms) are often of a cross-border nature, as is evidenced by the volume of cross-border trade, and the still untapped potential for future growth, as illustrated by the pattern and volume of cross-border trade intermediated by digital platforms. Almost 24% of total online trade in Europe is cross-border. It is estimated that by 2025 online marketplaces will represent 65% of cross-border online sales in Europe. 117

    103.Even where these digital markets may be geographically defined as national in scope, the problems at stake nevertheless remain of a cross-EU nature for three main reasons. First, the goods and services offered by the market players concerned are typically of a cross-border nature. Second, digital players typically operate across several Member States, if not on an EU-wide basis, which is particularly the case for markets such as online advertising, social media, online retail, cloud services, e-commerce or online search. This is not to say that services such as online advertising and search do not have to be tailored to Member States’ languages - however, the overall business strategy will normally be EU-wide.

    104.Accordingly, market failures in digital markets have Union relevance, as they can arise across borders and affect several Member States, thus not being limited to a specific national market of a Member State. 118  

    105.As regards the particular case of unfair business practices, in the absence of an EU measure, there is a high risk that with national approaches, business users or application developers seeking to serve the internal market will need to understand a range of diverse rule-sets and pursue actions in multiple countries across the EU, which is likely to fragment the Single Market for digital services, create barriers to expansion and compliance costs, especially for start-ups and SMEs. A lack of harmonised rules in this space risks complicating the regulatory landscape faced by platforms operating on a pan-European or indeed global basis. An intervention at the EU level is therefore more efficient, insofar as it introduces a common set of rules across Member States to address in a consistent manner the same unfair business practices carried out by large digital gatekeepers across the Union.

    106.Similarly, intervention by individual Member States or NCAs would be ineffective in tackling gatekeeper related market failures across the Union. Each Member State can only address market failures in its own territory, 119 imposing its own remedies, whereas market failures may affect the territory of several Member States because of the wider geographic scope of the relevant market concerned or the cross-border business activities of the market players concerned. Addressing market failures with a cross-border dimension at national level could also lead to inconsistencies in the remedies imposed, with the ensuing risk of fragmenting the Digital Single Market.

    107.Therefore, by addressing market failures in respect of key digital markets, the functioning of the internal market will be improved through clear behavioural rules that give all stakeholders legal clarity and through an EU-wide intervention framework allowing to address market failures in a timely and effective manner.

    4.OBJECTIVES: WHAT SHOULD BE ACHIEVED?

    General objective

    108.The general objective of this initiative is to ensure the proper functioning of the internal market by promoting effective competition in digital markets, in particular a fair and contestable online platform environment. This objective feeds into the strategic course set out in the Communication ‘Shaping Europe’s digital future’ as shown in Section 1.

    Specific objectives

    Address market failures to ensure contestable and competitive digital markets for increased innovation and consumer choice

    109.As explained in Section 2.3.1.1, certain digital markets may not be functioning well and delivering competitive outcomes due to their particular features, in particular extreme scale (or scope) economies, and a high degree of vertical integration; direct or indirect network effects; multi-sidedness; data dependency; switching costs; asymmetric and limited information, and related biases in consumer behaviour as well as the conduct of gatekeepers. Therefore, a specific policy objective is to allow identifying and addressing such market failures in respect of key digital markets to ensure that these markets remain contestable and competitive. This will contribute to digital markets delivering low prices, better quality, as well as more choice and innovation to the benefit of EU consumers.

    Address gatekeepers’ unfair conduct

    110.As explained in Section 2.3.1.2, gatekeepers’ economic strength, their position of intermediaries between businesses and consumers together with markets dynamics fueling gatekeepers’ growth lead to an imbalance in power between gatekeepers and their business users. This enables gatekeepers to impose unfair commercial conditions on business users, thus hampering competition on the platform. Such unfair behaviour does also have a negative impact on (the emergence of) alternative platforms since it strengthens consumer lock-in thus preventing multi-homing. In light of this, a specific policy objective is to lay out a clearly-defined set of rules addressing identified gatekeepers’ unfair behaviour, thereby facilitating more balanced commercial relationship between gatekeepers and their business users, which would be also expected to create the right incentives for multi-homing.

    Enhance coherence and legal certainty to preserve the internal market

    111.The gatekeeper-related problems identified above are currently not (or not effectively) addressed by Member States in existing regulation. The national legislative initiatives may partially address problems identified but also lead to increased regulatory fragmentation. In addition, tackling issues identified through legislation at national level is suboptimal in light of the cross-border nature of the platform economy. Moreover, the systemic importance of gatekeepers for the internal market deserve a better coordinated and more effective intervention across the EU. As explained in Section 2.3.2, national laws lead to legal fragmentation of the platform space and increase compliance costs for all market players. Therefore a specific policy objective is to improve coherence and effectiveness of oversight and enforcement of measures vis-à-vis gatekeepers, thus contributing to increased legal certainty.

    How do the objectives link to the problems identified?

    112.The figure below shows how different objectives are linked with the problems and the underlying problem drivers. It also shows that the specific objectives, i.e. address market failures to ensure contestable and competitive digital markets for increased innovation and consumer choice, address gatekeepers’ unfair conduct and enhance coherence and legal certainty in the online platform environment for a preserved internal market, contribute to achieving the general objective of ensuring the proper functioning of the internal market (through effective competition in digital markets and through fair and contestable online platform environment).

    Figure 2: Intervention logic tree – problem drivers, problems and objectives

    5.WHAT ARE THE AVAILABLE POLICY OPTIONS? 

    113.This section presents the three policy options retained (Section 5.3), the baseline scenario from which they will be assessed (Section 5.1), and the options that have been considered but discarded (Section 5.4).

    What is the baseline from which policy options are assessed?

    114.Gatekeepers are currently subject to two main sets of laws: first, competition laws which are applicable across all sectors of the economy, and, second, EU rules of a more sector-specific scope or with a less punitive nature including, among others, the P2B regulation, the General Data protection regulation (‘GDPR’) and the EU consumer law.

    115.Broadly speaking, the EU competition policy toolbox includes rules on antitrust, merger control, State aid, and public undertakings and services. Generally, a distinction is made between competition rules allowing for an intervention ex post or ex ante. Antitrust enforcement – under Articles 101 and 102 TFEU and the accompanying implementing regulations – belongs to the first category, as it aims at detecting anti-competitive behaviour by companies that has the actual or likely effect of causing distortions of competition. Merger control and state aid rules aim at preventing anti-competitive outcomes by assessing ex ante whether a merger between undertakings or the granting of State aid would negatively affect competition. Intervention under the existing EU competition rules can therefore only occur if: (i) a company is dominant pursuant to Article 102 TFEU and abuses this position, (ii) there is an anticompetitive agreement or concerted practice between two or more undertakings covered by Article 101 TFEU, (iii) there is a merger/acquisition with EU dimension falling under the EU merger control rules 120 or (iv) a Member State grants aid falling under the EU State aid rules.

    116.Under the baseline scenario, the Commission would not propose any changes to the current competition legal framework. This means that the Commission would continue to vigorously apply and enforce the existing competition law framework, in particular Articles 101 and 102 TFEU, against gatekeepers in digital markets, should the conditions for such intervention be met. Competition enforcement by the Commission would include making full use of the existing tools within this framework. The ongoing reviews of existing legislation (e.g. the Block Exemption Regulations for horizontal and vertical agreements) as well as of the Market Definition Notice would continue as planned.

    117.The majority of the respondents to the OPC indicated in their replies that, while some of the issues connected to gatekeeper powers could potentially be addressed by competition law enforcement through procedural and/or organisational changes, there were restrictions that could not be overcome with competition law enforcement. Respondents argued that the main challenges with regard to the enforcement of Article 102 TFEU relate to situations where dominance does not exist, and the difficulties with remedying a conduct found to be anti-competitive in an appropriate and effective manner, notably once the damage has already occurred. They considered that these challenges also have a negative effect on the duration of antitrust investigations and the ability of the existing competition law framework to ensure the contestability of the markets concerned. Respondents also highlighted the need for a regulatory solution regarding conduct recurrently showing negative effects on competition, as well as the need to pursue more exploitative cases and to take non-economic objectives into account in the competitive assessment. 121 A minority of respondents to the OPCs argued that Articles 101 and 102 TFEU are suitable and sufficiently effective in addressing market failures, and were primarily satisfied with the level of enforcement of the existing competition rules. These respondents pointed towards interim measures, sector inquiries, merger control and deadlines as potential ways to tackle any shortcomings that competition law may have. Some respondents also pointed to the ongoing reviews of competition legislation such as the Market Definition Notice, as well as the rules applicable to vertical and horizontal agreements as other ways to improve the existing competition law framework. Some respondents also argued for a broader use of sector inquiries and a review of the EU merger regulation. 122  

    118.The Commission considers that the current legal framework would not allow it to address the market failures described in Section 2.3.1 for the following reasons.

    119.First, existing EU competition rules cannot conceptually deal with market failures resulting from the behaviour of gatekeepers in the absence of some preconditions, such as the existence of an anticompetitive agreement in the case of Article 101 TFEU or of a dominant position in the case of Article 102 TFEU. In addition, in some instances, existing EU competition rules may be able to prevent or address a market failure, but not in the most effective manner. The Commission’s enforcement of Articles 101 and 102 TFEU rules can only take place ex post, i.e. after a competition problem has emerged. As a recent report by the European Court of Auditors (‘ECA’) also indicates “particularly in the digital economy, this may be too late to tackle a competition problem”. The ECA report also flags that “the Commission has currently no tools in its hands that would allow it to intervene ex ante i.e. before competition problems would occur”. 123 Moreover, – even when using interim measures, explained below – competition law enforcement requires a detailed economic and legal analysis which, jointly with the procedural safeguards, bring the duration of the investigations to at least around two years and usually more than that. In markets characterised by powerful network effects and economies of scope, competition law interventions may mean not only delays in the interventions but also that irreparable effects such as tipping may no longer be reversible.

    120.Second, market failures associated to tipping markets cannot be tackled on the basis of the existing competition rules, notably where market tipping is triggered primarily by the market structure, and not (or only to a lesser extent) by any specific conduct.

    121.Third, the existing EU competition rules do not necessarily capture all unfair business practices by large digital gatekeepers. This is because these practices do not necessarily have an anticompetitive object or effect under Article 101 TFEU, or may not be captured by Article 102 TFEU, if there is no effect on competition on clearly identifiable relevant markets. 124

    122.Fourth, specific competition tools cannot address the gatekeeper related market failures:

    ·Prohibition (fines) and commitments decisions (Articles 7 and 9 of Regulation 1/2003 125 ) are decisions addressed to individual companies for a breach of the EU competition rules, and not suitable for addressing market failures that are not, or not exclusively, caused by such breach of the EU competition rules.

    ·Sector inquiries (Article 17 of Regulation 1/2003) are investigations that the Commission carries out when it suspects possible breaches of the competition rules in specific sectors of the economy. There is, however, no possibility to impose remedies following a sector inquiry.

    ·Interim measures (Article 8 of Regulation 1/2003) are a tool allowing the Commission to intervene in “cases of urgency due to the risk of serious and irreparable damage to competition” where a ‘prima facie’ infringement of the EU competition rules can be shown. Interim measures, however, would not allow the Commission to tackle the problems explained in this Impact Assessment for two main reasons: first, interim measures can only be imposed where a prima facie infringement of Articles 101 or 102 TFEU can be shown, and second, interim measures are founded on a very specific test requiring the finding of ‘urgency’ as well as ‘serious and irreparable damage’. Interim measures have only been used twice in the last nineteen years.

    123.Fifth, the ongoing reviews of existing legislation (e.g. Block Exemption Regulations for horizontal and vertical agreements) as well as of the Market Definition Notice will also not tackle or address the problem drivers:

    ·The ongoing reviews of Block Exemption Regulations cannot tackle or address the problem drivers. The Block Exemption Regulations pursue a different objective than the DMA. They are by their very nature not aimed at addressing specific competition issues and/or market failures, but at block exempting agreements that are on balance efficiency enhancing, thus helping companies to self-assess compliance of their agreements with Article 101 TFEU.

    ·The ongoing evaluation of the Market Definition Notice cannot, by its very nature, tackle or address the problem drivers as it is a soft law document to provide guidance on the definition on the relevant market and on the relevant parameters to be taken into account for it when using Articles 101 and 102 TFEU or the EU Merger Regulation (Regulation (EC) 139/2004). The Market Definition Notice cannot therefore address competition issues and/or market failures in digital markets.

    124.Some respondents to the OPC – across different stakeholder categories – considered that there was no need for the DMA and that the Commission should rather reassess the situation after the P2B Regulation had shown its effects. A minority of respondents, mainly several large platforms and their trade associations, and some research institutes and academics, disagreed with the need for the proposal of new ex ante rules as they consider that the risks posed by gatekeepers can be addressed with existing regulation. Some platforms, trade associations and national authorities emphasised the need to focus the regulatory attention towards specific actions and perceived market failures. 126

    125.Under the baseline scenario, the Commission would continue to apply and enforce the existing more sector-specific EU rules including, among others, the P2B regulation, the GDPR and EU consumer law. The Commission considers however that the current regulations will also not tackle or address the problem drivers described in Section 2.3 for the following reasons:

    ·Regulation (EU) 2019/1150 on fairness and transparency for business users of online intermediation services (the ‘P2B Regulation’) entered into force on 12 July 2020. It is the first EU-level legislation specifically targeted at commercial issues engaged in by online platforms, or online intermediation services, as well as by online search engines. It applies to more than 10 000 platforms in Europe and reflects the fact that a certain dependency of professionals, or business users, is inherent in any successful online platform. This feature of online platforms means that the fairness, transparency and redress rights and obligations that the P2B Regulation provides are necessarily high-level and principles-based. Since this legal framework establishes a general ‘safety net’ for all professionals active in the online platform economy, it does not address issues deriving from the concentration of economic power and unfair business practices of a limited number of very large gatekeeper platforms.

    ·EU data protection legislation 127 specifies the fundamental right to the protection of personal data. It therefore covers business-to-citizen and government-to-citizen interactions, rather than commercial and competition-related issues. Article 20 of the GDPR provides a limited right to data portability 128 , though it is broadly considered that there are still many implementation challenges and that this right is at present insufficient to significantly lower entry barriers and to facilitate the contestability of markets. 129  

    ·EU consumer law does address a range of potentially harmful practices, at EU level notably through the Unfair Commercial Practices Directive (‘UCPD’) 130 and the Unfair Contract Terms Directive (UCTD). 131 While these directives define a number of relevant concepts, such as 'professional diligence' and 'good faith', their scope is explicitly limited to business-to-consumer transactions. Conversely, the Misleading and Comparative Advertising Directive (MCAD) 132 covers certain Business to Business (‘B2B’) relations. However, the provisions set forth in the MCAD are limited to a narrow subset of advertising practices, which are not specific to online platforms or digital markets, and do not deal with the unfair business practices carried out by large gatekeeper platforms.

    126.Finally, in the absence of further EU legislation, and subject to enforcement of the existing legal framework, the legal fragmentation is likely to further increase as Member States are likely to continue to adopt horizontal or sector specific national measures against gatekeepers, as shown in Annex 5.4 to the Impact Assessment.

    What are the main parameters that determine the range of available policy options?

    127.The problems and drivers map on to a set of parameters that characterise the range of available policy options. These parameters include (a) the scope of the intervention; (b) the range of unfair practices at stake; (c) the speed of the instrument and the degree of flexibility offered; and (d) the investigative and enforcement framework available and appropriate. Before presenting the options, this section explains these main parameters and highlights some of the inherent trade-offs.

    Scope: core platform services, gatekeepers, thresholds

    128.As highlighted in Section 2.1, the scope of the intervention is characterised by two particular concepts – the nature of the ‘core platform services’ where problems arise, and the notion of ‘gatekeepers’, i.e. companies that offer one or more core platform services. Core Platform Services and Gatekeepers can be identified robustly by a series of criteria, as set out below.

    Identification of core platform services

    The enforcement experience under EU competition rules both at the EU and national level, numerous expert reports and studies – including the study supporting the present Impact Assessment – and the results of the OPC show that there are number of services that have the following features:

    (a)highly concentrated multi-sided platform services, where usually one or very few large digital platforms set the commercial conditions with considerable autonomy from their competitors, customers or consumers;

    (b)few large digital platforms act as gateways for business users to reach their customers and vice-versa; and

    (c)gatekeeper power of these large digital platforms is often misused by means of unfair behaviour vis-à-vis economically dependent business users and customers.

    133 While these concepts are broad, the work supporting this Impact Assessment shows that unfair practices by gatekeepers are more prominent in some platform services than in others; in identifying the core platform services to which the regulation of the DMA should apply, the Commission was guided by the following principles:

    (a)Clearly defined obligations should apply only to those services and gatekeepers where the identified problems are most prominent and egregious;

    (b)To ensure the highest level of legal certainty for gatekeepers and other market participants alike, it is important to identify services as clearly as possible in the rules themselves.

    134 The Commission identified several services, which meet these criteria and where absent regulatory intervention the identified problems in Section 2 could effectively remain un-addressed. Such core platform services are:

    (a)online intermediation services (including marketplaces and app stores),

    (b)online search engines,

    (c)operating systems, 

    (d)cloud computing services;

    (e)video sharing platform services,

    (f)number-independent interpersonal electronic communication services,

    (g)social networking services and

    (h)advertising services, including advertising networks, advertising exchanges and any other advertising intermediation services, provided by providers of one or more of the above services. 135  

    The mere fact that a given service is identified as a core platform service does not suggest however that any provider of such a service will automatically be considered as a gatekeeper. Determination of these services as core platform services just means that they satisfy the criteria identified above and that therefore any provider of these services, if meeting the conditions for being designated as a gatekeeper would have to comply with the relevant regulatory obligations as set out in the different policy options.

    136 137 Other categories of digital services were also considered for the scope of ‘core platform services’, such as streaming services or B2B industrial platforms. However, these were excluded from the scope of such core platform services at this point either because (a) they lack the multi-sided market characteristics (e.g. for video streaming or video-on-demand services); or (b) they do not exhibit at this point the strong asymmetry in bargaining power that results from the presence of a service provider acting as gateway between consumers and business (e.g. for industrial B2B platforms).

    129.The table below summarises the main features and practices in relation to the core platform services retained. The details about the evidence supporting those elements are provided in Sections 2.3.1 and 5.2.2. There is no consistent publicly available data about the mark-ups in each of these core platform services. In addition, some of these services do not generate direct revenue as prices are set at zero (e.g. online search services, social network services, number-independent messaging services), but are monetised via advertising services on the other side of the platform. Nevertheless, the players active on those services are some of the most profitable companies in the world (see Section 6.6.1).

    130.There is also not much research developed about the impact of concentration in the innovation efforts in relation to these services. According to the Stigler report “Disruptive innovation in markets that are characterized by high concentration levels and network effects is likely to be reduced compared to a competitive market. […] Entrepreneurs may expect a low payoff to developing a free-standing product because of entry barriers and exclusionary conduct by the incumbent platform.” 138 This report further concludes that “The incipient but growing technical research supports a concern for the impact of big tech on innovation” based on some studies on the innovation in social platforms, internet software (e.g. operating systems and other related apps) and internet retail.

    Table Error! No sequence specified.1: List of core platform services

    Core platform service

    Main features and practices

    Online intermediation service (such as for example marketplaces or app stores)

    Main features: online marketplaces and app stores are examples of core platform services that benefit from strong network effects given that the higher the number of users on one side (e.g. buyers) the more valuable is the platform for the other side (e.g. for sellers or app developers) and vice-versa. These intermediation platforms also benefit from data driven advantages (e.g. information about the preference of consumers) and the presence of high switching costs (e.g. resulting from consumer bias or from the fact that they are part of an integrated offer), resulting in many cases in consumers single-homing. Often these services are vertically integrated with the downstream services that they distribute (e.g. app stores and applications). These features result in very concentrated structures. For instance, app stores (e.g. Google Play) generally enjoy quasi monopoly positions in their respective markets. 139

    Online intermediation services have been the subject of several studies and reports that describe the market power of the main platforms offering these services and their weak contestability and predominance of unfair practices. See for instance the study by the Dutch NCA on app stores 140 , the US House of Representatives Majority Staff report (pages 84-87, 93-100, 211-223, 253-302 and 333-372) 141 , the Furman report (paragraphs 1.54-1.59, 1.145, 1.174, and 2.113) 142 and the Stigler Center report (pages 11, 15 and 51) 143 . See also Support Study to the Observatory for the Online Platform Economy: Report on Business user and third party access to data, 144 Report on Platform data access and secondary data sources, 145  Report on Differentiated treatment 146 .

    Practices: Online intermediation services involve many different types of services, and thus the list of practices by gatekeepers of these platforms is also diversified.

    In the particular case of online marketplaces, the types of practices often observed are:

    ·An online marketplace benefiting from its dual role and thus having the incentives to give more prominence to its own products or services as compared to those offered by other sellers directly competing with it. This reduces the possibility for customers to choose third party products or services.

    ·An online marketplace restricting access to data generated in the course of the use of its platform by third party sellers and their customers, thus gaining an unfair advantage vis-à-vis those sellers.

    ·An online marketplace benefiting from its dual role and ability to evaluate product, sales and customer data generated from the sales of products and services provided by third party sellers on its marketplace.

    In the case of app stores, the types of practices often observed are:

    ·An app store requiring sign-in with its provider’s email service thus being able to combine the data from several sources and foreclose other email service providers.

    ·An app store, which markets several of its own popular apps and at the same time maintains a marketplace (dual role), self-preferencing by applying more favourable policies for its own apps and selectively drafting rules favouring its own apps. This reduces the possibility for customers to choose third party apps.

    ·An app store not allowing its business users to advertise alternative subscription options to consumers, thus preventing its customers from benefitting from such an alternative offer.

    ·An app store app limiting the information that third-party app providers receive about their subscribers, limiting their ability to make innovative offers to those subscribers.

    ·An app store charging unfair conditions to distribute third party applications.

    ·An app store preventing the un-installation of its own pre-installed apps or restricting the installation of third party apps (e.g. another app store), thus foreclosing access to an important distribution channel.

    ·An app store reserving for its sister-services or for some providers with whom it has partnership agreements certain functionalities, thus preventing consumer switching to a different internet access provider.

    More in general, other practices are also observed:

    ·An online intermediation service preventing sellers from offering the same products or services (e.g. holiday package, hotel, publishers’ content, applications) to customers through another channel (e.g. direct channel or a third party distribution channel) at prices or conditions that are different from those offered through that intermediation service.

    ·An online intermediation service preventing sellers from promoting their products and services (publishers’ content) and concluding contracts with customers acquired on that platform outside the platform.

    Online search services

    Main features: These are two sided platforms that, on the on hand, allow users to perform searches of ‘all websites’ 147 for free and, on the other hand, provide inventory for advertisers. They are also an important channel of user traffic for businesses. Search engines benefit from strong economies of scale (associated to a high fixed cost and minimal marginal costs) and network effects (the higher the number of users the more valuable is the platform for advertisers) as well as a data driven advantage (in particular in relation to tail queries). A provider of search engine may also benefit from consumer bias when it takes advantage from the pre-installation in certain devices or default positions in certain browsers. These features result in supply of online search services being very concentrated, with one platform having a share of more than 90% in Europe in 2019. 148

    Online search services have been the subject of several studies and reports that describe the market power of the main platform offering these services and their weak contestability and predominance of unfair practices. See for instance the CMA report on advertising (section 3) 149 , the US House of Representatives Majority Staff report (pages 77-84 and 177-206) 150 , the Furman report (paragraphs 1.77 and 2.25) 151 , Stigler Center report (page 11) 152 and ACCC report (section 2.4) 153 .

    Practices: The most usual the types of practices observed in the field of online search services are:

    ·A provider of online search engines preferencing its own vertically integrated services in its search engine results, e.g. shopping or travel services featured on top of the search results page.

    ·A provider of online search services applying terms and conditions which make the use of its services conditional on the possibility to collect and combine user data from multiple sources.

    ·A provider of online search service making use of its a data advantage over competitors to raise barriers to entry as it has access to a vast amount of query data especially on long tail queries

    See also practices related to advertising and operating systems.

    Social network services

    Main features: These services represent an important gateway not only for consumers but also for business users, notably advertisers, to reach consumers. Social network services benefit from strong network effects and data driven advantages given that a greater number of users increases the value of the social network for each user and the ability to improve services and offer personalised advertising. User prices are almost always set at zero. They are also characterised by high switching costs as the user of a social network would not easily switch to a new network since all his/her friends are on the incumbent network. These features result in the supply of social network services being very concentrated, with one platform having a share of close to 80% in Europe in 2019. 154

    Social network services have been the subject of several studies and reports that describe the market power of the main platform offering these services and their weak contestability and predominance of unfair practices. See for instance the CMA report on advertising (section 3) 155 , the US House of Representatives Majority Staff report (pages 88-93 and 134-170) 156 the Furman report (paragraph 1.80) 157 , Stigler Center report (pages 11, 44 and 50-51) 158 and ACCC report (section 2.3) 159 .

    Practices: The most usual the types of practices observed in the field of social network services are:

    ·A provider of social network services ranking its own services more prominently in users´ timelines than those of third-party, thus foreclosing distribution of rival services.

    ·A provider of social network service applying terms and conditions which make the use of its services conditional on the possibility to collect and combine user data from multiple sources.

    See also practices related to advertising.

    Video sharing platform services

    Main features: These services benefit from economies of scale and strong network effects, and can become the default platform to consume and share video content. In this case, it is almost impossible for video content producers not to be present on this video sharing platform. Likewise, its enormous audience facilitates the ability for advertisers to reach a large audience even with very specific targeting parameters. A video sharing platform has access to a rich set of (first party) data about its consumers, data that it can re-use to improve its own products, including in other areas. These features result in the supply of these services being very concentrated, with two main platforms being used by consumers. 160

    Video sharing platform services have been the subject of several studies and reports that describe the market power of the main platform offering these services and their weak contestability and predominance of unfair practices. See for instance the CMA report on advertising (section 3) 161 , the US House of Representatives Majority Staff report (pages 88-93,190 and 211) 162 and the Furman report (paragraph 5.5) 163 .

    Practices: In the case of video sharing platform services, the types of practices often observed are:

    ·A provider video sharing platform restricting the access to their rich set of data to its competitors, thus raising barriers to entry and expansion to those competitors

    ·A provider of video sharing platform restricting the access to its must-have online inventory, access which is of particular importance to compete for the provision of the so-called ad tech products, and providing exclusive access to its own ad tech products, thereby favouring its own products to the detriment of competitors.

    ·A provider of video sharing platform refusing access to detailed performance data to video content providers that prevents them from improving their offerings.

    See also practices related to advertising.

    Number-independent messaging services

    Main features: These services are characterised by strong network effects resulting in ‘bimodal distributions’ of reach, meaning that they achieve an all-or-nothing reach, with market shares at either above 90% or below 10% and a high incidence of tipping. 164 They also benefit from strong economies of scale (associated to a high fixed cost and minimal marginal costs) and consumer lock-in given the high switching costs if all his/her friends stay in the incumbent network and are thus not reachable. Often these services are integrated with other core platform services (e.g. social network services).

    Number-independent messaging services have been the subject of several studies and reports in the context of the assessment of social network services (see above). See also and the Furman report (paragraphs 1.80 and 1.87) 165 and Stigler Center report (pages 44 and 50-51) 166 .

    Practices: In the case of number-independent messaging services, the types of practices often observed are:

    ·Providers of number-independent messaging services imposing on users the possibility to combine the rich set of (first party) data about them with other data sources in order to build a super profile.

    See also practices related to advertising.

    Operating systems

    Main features: Operating systems are very important for the visibility and distribution of most applications. Operating system are characterised by economies of scale (associated to high development costs) and high switching costs (given that a user often has to buy new hardware to change the operating system provider and is generally has a behavioural bias for a given operating system). They also benefit from network effects given that applications need to be coded for a given operating system and the large number of users the more attractive is a platform for developers and vice-versa. These features result in the supply of operating system being very concentrated, with three platforms having a share of close to 90% in Europe in 2019. 167 Each of these operating systems are particularly strong in specific platforms (e.g. mobile or desktop).

    Operating systems have been the subject of several studies and reports that describe the market power of the main platforms offering these services and their weak contestability and predominance of unfair practices. See for instance the US House of Representatives Majority Staff report (pages 100-107, 211-223 and 333-372) 168 , the Furman report (pages 29 and 47-48) 169 and Stigler Center report (pages 11 and 46-47) 170 .

    Practices: The most usual the types of practices observed in the field of operating systems are:

    ·A provider of operating systems giving differentiated conditions of access to its operating systems or device features to both business users and third-party providers of ancillary platform services (e.g. payment, user ID, fulfilment) - as compared to those used by the its own services or ancillary platform services. This prevents third parties from competing in a level playing field.

    ·A provider imposing a lock-in strategy where the user is required to sign up/register with an email service of that provider when using the functionalities of the operating system.

    ·A provider imposing a lock-in strategy requiring the use of its own ancillary services (browsers, search engines…) or preventing its un-installation.

    ·A provider limiting the access to or the interoperability of its operating system and respective functionalities (e.g. NFC) with the services offered by business users, reserving those functionalities to their own services.

    Cloud services

    Main features: These services provide infrastructure to support and enable functionality in services offered by others and at the same time offer a range of products and services across multiple sectors, and mediate many areas of society. Cloud services can reduce barriers to entry for start-ups by providing them with access to technical capabilities that might otherwise be beyond their reach. They benefit from strong economies of scale (associated to a high fixed cost and minimal marginal costs) and high switching costs (associated to the integration of business users in the cloud). The vertical integration of the large cloud services providers and the business model they deploy has contributed to further concentration on the market, where it is very difficult for other less-integrated players, or market actors operating in just one market segment to compete. 171  Consequently, these start-ups are likely to be completely reliant on large online platform companies.

    See Stigler Center report (page 51) 172 .

    Practices: The most usual the types of practices observed in the field of cloud services are:

    ·A provider of cloud services imposing obstacles to interoperability and data portability as well as strengthening lock-in of cloud service providers’ customers, due to high switching costs.

    ·A provider of cloud services bundling several different services, including services where they are a gatekeeper.

    ·A provider of cloud services copying and using a software that other cloud providers have developed and used.

    Online advertising services

    Main features: These services are often provided in connection with some of the services described above, namely online search services, social network services, online intermediation services, video sharing services, etc. They correspond to the side of the platform that is monetised. The intermediation of advertising services, even when not directly related to the services above, also benefits from those as the data collected is of extreme importance for the provision of these services. These services are characterised by data driven advantages as well as network effects. These features result in the supply of online advertising services being very concentrated. 173

    Online search services have been the subject of several studies and reports that describe the market power of the main platform offering these services and their weak contestability and predominance of unfair practices. See for instance the CMA report on advertising (section 5) 174 , the US House of Representatives Majority Staff report (pages 129-133, 170-174 and 206-211) 175 , the Furman report (pages 27-28 and 112-117) 176 , Stigler Center report (pages 22, 25, 38-41 and 50) 177 and ACCC report (section 3) 178 . See also Support Study to the Observatory for the Online Platform Economy, Report on Transparency in online advertising. 179

    Practices: The most usual the types of practices observed in the field of online advertising services are:

    ·The presence of big platforms on both the supply and demand side of the ad supply chain gives rise to conflicts of interest and a possible preferential treatment of one’s own integrated services.

    ·On ‘walled gardens’, major platforms collect multiple datasets from logged-in users and partner services which they use for better targeting and attribution measurement, but do not share user-level data with advertisers, only committing to placing the ad in front of the targeted demographics. As there is currently no uniform standard of metric for online ad effectiveness, advertisers cannot compare the effectiveness of their ad spend across several walled gardens and the open web. This restricts the ability of advertisers and publishers to make informed decisions.

    ·A provider of advertising intermediation services refusing to provide information about the price paid for each of the intermediation services used to deliver the ad to both advertisers and publishers, thus preventing them from comparing with alternative offers.

    ·A provider of advertising services (that could also offer on the other side of the platform search services, social network services, etc) using certain data that it received from business users for other unrelated purposes, including competing against those business users in other markets.

    131.As mentioned above, the identification of core platform services is relevant for both the designation procedure for gatekeepers, and the identification of those services provided by the designated gatekeeper. The evidence gathered during the OPCs indicated two guiding principles.

    132.First, there is a wide agreement that gatekeepers exist and that their role and practices are increasingly determining the conditions of market participation on the consumer side as well as conditions of individual commercial relationships they engage in on the business user side. In circumstances where such practices are unfair, this may undermine balanced commercial relationships and the contestability of platform markets.

    133.Second, while there is no consensus as to a single means to identify gatekeepers in the digital sector, to the extent that different sources refer to qualitative criteria as a means of identifying gatekeepers there is relatively broad consensus that such criteria need to take into account factors such as access to data, network effects and ability of these providers to leverage their economic power to multiple services. Similar qualitative criteria (e.g. access to data, financial resources, level of vertical integration) are also promoted by a recent initiative on the update of the German competition law. The OPC provided mixed views on this, with a majority of stakeholders calling for a combination of quantitative and qualitative criteria to identify gatekeepers and only very small number of respondents promoting use of qualitative criteria only. 180  

    Conditions for the designation of gatekeepers

    A gatekeeper within the meaning of the DMA is a provider of core platform services which:

    (a)has a significant impact on the internal market;

    (b)operates a core platform service which serves as an important gateway for business users to end users; and

    (c)enjoys an entrenched and durable position in its operations or is expected to enjoy such a position in the near future. 181

    The analysis underpinning the selection of these criteria for the identification of gatekeepers under the present Impact Assessment is based on the following principles:

    (a)Conditions should well reflect the identified problems and be supported by the available evidence base;

    (b)Conditions should allow for an objective determination of a gatekeeper status;

    (c)Conditions should guarantee a high level of legal certainty for gatekeepers and other market participants alike; and

    (d)Conditions should be easily identified and measurable.

    134.This designation of gatekeepers in this way leads to another important design parameter, namely on the choice of thresholds. Stakeholders generally support a mix of quantitative and qualitative parameters (see paragraph 133).

    135.Suitable quantitative thresholds can be constructed from indicators for size (such as turnover and presence in various Member States) and for economic dependency (such as the number of business users and end users served on the platform). The weak inter-platform competition that results from such gatekeepers’ services having become entrenched would be captured by measures of persistence (such as the number of core platform services offered by the same group of undertakings, and the number of years this group has held its position). Depending on the precise indicators used and the level at which these are set, the addressable population can obviously be larger or smaller. Importantly, the combination of these quantitative parameters would in all cases only include a very limited number of companies and would exclude a very large set of providers of digital services. It would therefore not cause any undue lack of legal certainty, and the option of using higher or lower thresholds is a legitimate political choice, presenting some of the trade-offs described in this section.

    136.Subject to the overarching criteria of size, dependency and persistence, a wide range of different indicators have also been investigated during the Impact Assessment. 182

    137.A reliable set of specific indicators has been used to identify the effects of fixing the thresholds at higher or lower levels in terms of the number and identity of the undertakings that could potentially be brought within the scope of the different options. This exercise confirmed the accuracy of the methodology.

    138.Alternative methods for designating gatekeepers were analysed but discarded, notably those that are directly inspired by notions of ‘significant market power’ as it is used for example in the regulatory framework for telecommunications services. While such models can serve as inspiration, and do capture notions of concentration, they are difficult to transpose directly as the diversity of platform business models is poorly captured by such a market-based analysis, and the relevant bottleneck power at stake in this Impact Assessment also derives from the tendency of gatekeeper companies to diversity.

    139.The below paragraph specifies the quantitative parameters used to identify gatekeeper companies and their core platform services. 

    140.For this exercise, the following were selected as proxies for the main three criteria:

    141.For Size & internal market impact, the proxy selected is the EEA annual turnover of the group ( > EUR X billion) or the average market capitalisation 183 or the equivalent fair market value of the group ( > EUR X billion) in combination with its presence in more than three EU countries.

    142.A significant turnover in the Union or the market capitalisation and the provision of a core platform service in at least three Member States is a good indicator that the provider of that service has a significant impact on the internal market. In other words, a provider of a core platform service should be deemed to have a significant impact on the internal market where it provides a core platform service in at least three Member States and where either its group turnover realised in the Union is equal to or exceeds a specific, high threshold or the market capitalisation of the group is equal to or exceeds a certain high absolute value. Both parameters are reflective of gatekeepers’ ability to monetise their users and of their financial capacity, including the ability to leverage access to financial markets to reinforce their already strong position.

    143.For Dependency: the proxy selected concerns the number of users ( > X million EU users) in combination with the number of business users ( > X EU business users) across all core platforms services, as this is a proxy for the bottleneck power that characterises the problem.

    144.A very high number of business users that depend on a core platform service to reach end users and a very high number of monthly active end users are indicative of the provider’s role as an important gateway. 184  

    145.For Persistence, the proxy used relates to the number of large core platform services showing dependency, as well as the number of recent years during which the size and dependency criteria have been met.

    146.Likelihood of an entrenched and durable position or the foreseeability of achieving such a position in future is very high where the contestability of the gatekeeper’s position is limited. This is likely to be the case where that provider has provided a core platform service in at least three Member States to a very high number of business users and end users during at least three years.

    147.Other proxies, such as criteria estimating the degree of multi-homing or the rate of innovative entry have been suggested in the literature, but were discarded for the purpose of finding objective quantitative criteria that also have a relationship with the impact of the gatekeepers in the EU.

    148.The above quantitative parameters could be combined in different ways with a view to providing consistent thresholds. 185 Two meaningful thresholds would be the following:

    ·Low threshold: Fixing the EEA annual group turnover threshold at the level of EUR 6.5-7.5 billion and the required number of core platform services showing dependencies at 30-45 million end users and 10 000 business users during a number of years at a single one. This threshold would result in 10 to 15 providers of core platform services.

    ·High threshold: Fixing the threshold at an EEA annual group turnover of EUR 5-6 billion but also including a minimum of two core platform services, with at least one showing dependencies at 30-45 million end users and 10 000 business users would reduce the group of providers captured to an estimated number of five to seven companies.

    149.Other possible combinations of parameters (e.g. market capitalisation, required turnover, number of business users or end users, number of core platform services; EBITDA) could be used for defining plausible quantitative thresholds capturing the type of gatekeepers targeted by the initiative under this assessment, i.e. gatekeepers engaging in the problematic practices described in Section 5.2.2. The above two combinations of parameters for a low and a high threshold have been selected as representative examples for the purpose of providing clarity in assessing impacts and trade-offs while comparing options, without prejudging other plausible use of parameters and their combinations. Small variations around those numbers do not produce significant changes in the number of possible gatekeepers. At the same time, some combinations have been considered but judged inappropriate. For instance, fixing the EEA annual group turnover at EUR 1 billion and the required number of core platform services showing a dependency (during a number of years) at a single one would mean that more than 25 providers of core platform services would be captured. 186 It was concluded that such thresholds would give rise to many false positives.

    150.An alternative to quantitative criteria is to use qualitative criteria. Such a designation based on qualitative criteria would require a market investigation, which would assess whether the provider of core platform services has a significant impact on the internal market, operates a core platform service which serves as an important gateway for business users to customers, and enjoys (or is expected to enjoy) an entrenched and durable position in its operations. This could be the case where the contestability of the core platform service is affected, or risks being affected, on a lasting basis, due to the position of the gatekeeper, entry barriers derived from network effects, in particular in relation to its access to and collection of personal and non-personal data or analytics capabilities, scale and scope effects, customer bias or other structural market characteristics, with the effect of impeding innovation, high quality of digital products and services, fair and competitive prices and choice for business customers and users.

    151.In summary the scope of the intervention is characterised by the main parameters of choice of core platform service where the data-driven advantages and network effects are strong drivers, as well as the criteria that determine the gatekeeper companies in scope. While there is a good consensus that the criteria should focus on size, intermediation power, and entrenched position 187 , there are choices available on the quantitative thresholds, or the use of qualitative thresholds.

    Unfair practices

    152.The public consultation offered strong support for an intervention tackling gatekeepers’ unfair practices. In fact, the large majority of the respondents to the OPCs and to the NCA questionnaire agreed that the Commission should be able to intervene in markets where gatekeepers are present, including a large majority of businesses and businesses associations, all civil society organisations (including consumer associations, NGOs and trade unions) and all public authorities. 188 Those respondents considered that this would both create the right innovation incentives across the market, and contribute to increased consumer choice paving the way for new platforms and innovative and privacy-friendly services.

    153.Those disagreeing refer to the fact that the concept of a gatekeeper is too broad and should instead be assessed on a case-by-case basis and that the Commission can already intervene in the case of gatekeeper’s conduct using Article 102 TFEU. As explained in Section 5.1, the Commission considers that Article 102 is not sufficient to deal with all the problems associated with gatekeepers given that a gatekeeper may not necessarily be a dominant player, and its practices may not be captured by Article 102 TFEU if there is no demonstrable effect on competition. Moreover, Article 102 does not always allow intervening with the speed that is necessary to address these pressing practices in the most timely and thus most effective manner.

    Identification of obligations applicable to gatekeepers’ core platform services

    The different sources supporting the present Impact Assessment refer to a number of alleged or proven unfair practices by gatekeepers in the digital sector. There is a relatively wide consensus that such practices can be grouped into certain categories, such as (i) unfair data driven practices, (ii) unfair self-preferencing, and (iii) unfair access conditions.

    In identifying which of the specific unfair practices may require ex ante regulatory intervention, the Commission followed several principles:

    (a)There should be sufficient experience with the harmful effects of the identified unfair practices;

    (b)Such experience should point to the egregious nature of the unfair practices in question, which would justify the clear identification of obligations related to them;

    (c)To the extent possible, these obligations should be directly applicable; and

    (d)The unfair practices should be identified in a clear and unambiguous manner to provide the necessary legal certainty for gatekeepers who would need to comply with them, as well as for business users or consumers that may avail themselves of the choices provided for them.

    Based on these criteria, several unfair practices have been identified as those that require immediate attention by the Commission, given their likelihood to cause direct harm to business users and to negatively affect the contestability of core platform services. These practices are well-documented and can be relatively easily circumscribed in ex ante regulation, which also addresses an emerging legal fragmentation. These egregiously unfair practices should therefore be subject to clearly defined obligations addressed to gatekeepers in the proposed EU-level framework.

    However, for a number of other, allegedly unfair practices it would not seem justified to intervene at this point in time. For example, it could be disproportionate to require providers of software application stores not to exclusively pre-install their own software application store, especially taking into account that the proposed framework would already tackle specific unfair practices that a provider of such a software application store may engage in.

    154.The table below summarises the list of obligations retained associated to the key unfair practices identifed against the evidence for each practice. A key element for such a list of practices is the applicability to a gatekeeper’s core platform services. In some cases, the practices at stake target specific core platform services.

    155.The evidence included in the table consists of (i) antitrust decisions adopted by the Commission and other competition authorities, (ii) examples of complaints and investigations compiled by an external contractor in the context of the Impact Assessment study, (iii) evidence from the study supporting the Observatory for the Online Platform Economy, (iv) evidence from the reports by the expert group for the Observatory on the Online Platform Economy, complemented with stakeholder input, (v) studies conducted on digital sectors by other public authorities, and (vi) other Impact Assessments and regulations. 189 For some of the practices listed below there is no decision or judgment confirming its effects on the market. Nevertheless, the multiple complaints, investigations and studies raising awareness, and suggesting solutions, to those practices are a strong indication of their relevance and of their negative impact on the internal market.

    156.Beyond the practices identified in the table, a range of other practices were examined. For example, economic regulation of costs and prices was examined and discarded. Other examples of specific practices examined but not retained include generic and broad rules on fair access conditions to any core platform service, or rules which would allow any provider of ancillary service to get full access and interoperability with a gatekeeper’s service. Other practices considered and frequently proposed in the literature – like for example banning the pre-installation of software – were replaced by more proportionate obligations – in this case, the possibility to give customers the possibility to always un-install applications – or dropped completely – for example, a general ban on tying.

    157.Aside specific practices examined and discarded, a category of other obligations was also considered but rejected, namely broadly formulated or generic practices (such as self-preferencing in general) that would require an in-depth competition like analysis to be carried out.

    Table 2: List of obligations with examples and underlying evidence

    Obligations / unfair practices

    Concrete example/evidence

    Gatekeepers shall not be combining personal data originating from different core platform services with personal data from their other services or data from third party services or automatically signing in end users to other services of the gatekeeper in order to combine data without providing an effective possibility to opt-out

    Example: provider of online social network site collecting data from its users obtained through several different services.

    Ability – due to gatekeeper’s size and associated network effects - to accumulate data and use it as a competitive advantage

    German NCA found that Facebook abused its dominance by applying terms and conditions, which made the use of its social network conditional upon Facebook’s possibility to collect and combine user data from multiple sources. 190

    Italian NCA found that WhatsApp forced its users to share their personal data with Facebook. 191

    According to the IA support study , there are some pending antitrust investigations in this area in relation to online intermediation services.

    IA support study , in particular Annex 4, Case 8: Digital ID – Facebook and Google.

    Support Study to the Observatory, Business user and third-party access to online platform data.

    Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , page 44.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 209-211

    CMA report on Online platforms and digital advertising , pages 188-193. 

    EDPB report on social media and impact of profiling on competition , page 7 explains that “the unrivalled insight capabilities provided by the platform may make it an 'unavoidable trading partner' for online marketers”.

    This issue that is also covered by the GDPR but in a narrower sense and based on opt-in.

    Gatekeepers shall not prevent business users from offering the same products or services to customers through third party online intermediation services at prices or conditions that are different from those offered through the online intermediation services of the gatekeeper.

    Example: a provider of online intermediation services does not allow hotels/e-books publishers to offer better prices on different online travel agents/e-books platforms

    Exclusive dealing requirements possible due to gatekeeper size and businesses’ dependence (indispensability to be present on the platform)

    Case AT.40153 E-book MFNs and related matters (Amazon), Commission Decision of 4 May 2017: the Commission considered that MFN clauses included in Amazon's e-books distribution agreements could make it more difficult for other e-book platforms to compete with Amazon by reducing publishers' and competitors' ability and incentives to develop new and innovative e-books and alternative distribution services.

    As demonstrated in the Annex 5. 4 to the Impact Assessment wide parity clauses were removed in a large part of Member States, while in some Member States also narrow MFN clauses were banned via legislative action. This is the case in France, Austria, Italy and Belgium. The laws of those MS prohibit all MFN clauses.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 295-296.

    P2B Impact Assessment on MFNs: “issues have also arisen in the context of so-called 'most-favoured nation' ('MFN') clauses, also known as 'parity' or 'price-parity' clauses. These are common in Online Travel Agents ('OTAs'), but also exist to a more limited extent on e-commerce platforms, app stores or price comparison tools.”

    J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era , pages 55-57.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , page 48.

    Gatekeepers shall not prohibit their business users from promoting and subsequently concluding contracts with their customers acquired on gatekeeper’s platform outside a gatekeeper’s platform.

    Example: a publisher cannot inform a new user through its newspaper app that the subscription is cheaper if concluded via the publisher´s website.

    Gatekeepers shall not prohibit consumers from accessing and consuming, on the gatekeeper’s platform or services, services which have been acquired outside of the gatekeeper’s platform or services.

    Example: a music streaming subscription is concluded through a website, but cannot be subsequently used via the app.

    Ability to set market rules

    According to the IA support study , there are some pending antitrust investigations in this area in relation to app stores.

    Anti-steering provisions are raised by many stakeholders replying to the OPC as a concern (e.g. publishers; media companies). 192

    According to the IA support study , Annex 4: Case 4: Restriction of access and use of business users to data about their customers – Apple App Store there are several antitrust investigations in this area in relation to online intermediation services.

    Support Study to the Observatory , the Significant Market Status.

    Gatekeepers shall not prevent or restrict business users from raising issues with any relevant public authority relating to any behaviour of gatekeepers

    Example: business users would like to complain about unfair practice by gatekeeper, but is effectively prevented doing so due to contractual constraints.

    Ability to impose requirements on other businesses, due to their economic dependence on the gatekeeper

    Some indications in antitrust investigations that complaints to authorities or sharing of information with the authorities.

    Gatekeepers shall not impose their own user ID services on business users when the latter offer service using the core platform service of the gatekeeper.

    Example: an app store operator unilaterally requires all app developers to integrate the app store´s own user ID functionality in their apps and to show this ID functionality to the customers of their apps.

    Ability to impose market rules due to gatekeeper’s market position

    IA support study , in particular Annex 4, Case 8: Digital ID – Facebook and Google.

    Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , pages 31 and 89.

    Gatekeepers shall not require business users or customers of these business users to subscribe to or register with any core platform service other than the core platform service provided by the gatekeeper, as a condition to access, sign up or register to any of their core platform services

    Example: consumer would like to subscribe to social networking service by a gatekeeper, but is effectively prevented from doing so without subscribing to other services of that gatekeeper.

    Ability to impose conditions of access to the platform due to strong market position

    Case AT.40099 Google Android, Commission Decision of 18 July 2018: the Commission considered that Google has ensured that its Google Search app is pre-installed on practically all Android devices sold in the EEA by tying it pre-installation with the pre-installation of the Google Play Store. Google's practice has reduced the incentives of manufacturers to pre-install competing search apps, as well as the incentives of users to download such apps. This reduced the ability of rivals to compete effectively with Google.

    According to the IA support study , Annex 4: Case 3: Unjustified tying and bundling – Microsoft 365 bundling with cloud services and Case study 8: Digital ID – Facebook and Google there are several antitrust investigations in this area in relation to cloud, social networks and search services.

    Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , page 31.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , page 36.

    CMA report on Online platforms and digital advertising page 279.

    J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era , page 37.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 241-242, 286-290, 397.

    Gatekeepers shall provide advertisers and publishers with information concerning the price paid for the impression of a given ad, including for each of the relevant advertising services provided by the gatekeeper

    Example: advertisers and publishers would like to obtain information about all the intermediation fees charged by the adtech services provider, but they are refused so.

    Ability to impose unclear pricing rules due to the gatekeeper’s strategic position of intermediary

    Support Study for the Observatory, Report on Transparency in the Online Advertising Market .

    Observatory Expert Group Report: Market power and transparency issues in open display advertising.

    Lack of transparency in the advertising value chain is raised by many stakeholders in the OPC. 193

    CMA report on Online platforms and digital advertising pages 297-303.

    ACCC report, Digital Platforms Inquiry, Final Report , pages 150-157.

    Gatekeepers should not use data provided by or generated through activities of business users of its core platform services in competition with those business users

    Example: an e-commerce marketplace using commercially sensitive data collected from individual sellers to compete with these sellers on its own online marketplace

    Ability to accumulate third-parties’ generated data and use it as a competitive advantage

    According to the IA support study , there are several pending antitrust investigations in this area.

    IA support study , in particular Annex 4, Case 5: Prohibition of self-preferencing – Amazon Marketplace.

    Japanese Fair Trade Commission Report regarding trading practices on digital platforms , identifies use of sellers´ data as priority issue for continued investigation. 194

    Support Study to the Observatory, Business user and third-party access to online platform data.

    CMA report on Online platforms and digital advertising , page 109-110.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 218.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , page 34, 47.

    Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , page 52.

    Gatekeepers shall not prevent customers from un-installing any pre-installed software applications on its core platform services.

    Example: app stores/operating systems preventing users from un-installing some of the pre-installed apps, in particular where these are not essential for running the hardware.

    Ability to impose market rules due to its intermediary function

    There is a strong consumer bias towards pre-installed software (see Google Android 195 and Microsoft (tying) 196 antitrust decisions.

    The IA support study , Annex 4, Case 2: Unjustified tying and bundling – Google advertising ecosystem.

    CMA report on Online platforms and digital advertising , page 106-109.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 210-217.

    Gatekeeper shall allow the installation and effective use of third party software applications or software application stores using, or interoperating with, operating systems of that gatekeeper

    Example: consumer is prevented from installing an app store of the mobile game provider and the relevant applications directly from its app store.

    Ability to impose market rules due to its intermediary function

    The IA support study , in particular Annex 4, Case 4: Restriction of access and use of business users to data about their customers – Apple App Store, referring to on-going antitrust investigations in this area.

    As demonstrated by the OPC submissions, providing obstacles to interoperability and data portability and lock- in of cloud service providers’ customers, due to high switching costs are the most common practices in the cloud space. 197

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 210-217.

    Gatekeepers shall not treat more favourably in ranking their own services and products compared to similar services or products of third-party business users and shall apply fair conditions to such ranking

    Example: a search engine preferring its own vertically integrated services in its search engine results (e.g. shopping or travel services are featured on top of search results); a social network ranking its own dating service more prominently in users´ timelines than those of third-party dating services.

    Ability to misuse its intermediation position to its own competitive advantage

    Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017: Google leveraged its dominance in general internet search services to the separate comparison shopping service (CSS) market by favouring Google Shopping on its general web search results page.

    Korea FTC imposes fine and remedies on Naver for ranking self-preferencing and anti-steering. 198

    The IA support study , Annex 4, Case 2: Unjustified tying and bundling – Google advertising ecosystem.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 187-193, 282-286, 359.

    CMA report on Online platforms and digital advertising , page 109-110.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , pages 31, 44 and 117.

    P2B evidence. “The favouring of own products or services by online platforms was identified as one of three most commonly experienced problematic trading practices by business respondents to the public consultation on platforms.”

    Gatekeepers shall not technically restrict the ability of end users to switch between and subscribe to different software applications and services to be accessed using the operating system of the gatekeeper

    Example: an app store reserving for some providers with whom it has partnership agreements certain functionalities, thus preventing consumer switching to a different internet access provider.

    Ability to impose conditions of access

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , pages 35-37.

    Gatekeepers shall not prevent business users and providers of ancillary services access to and interoperability with the same operating system, hardware or software features that are available to or used by any ancillary services provided by the gatekeeper.

    Example: provider of financial services online would like to obtain access to certain features available to the payment services of the gatekeeper that are needed to perform certain operations, but is refused access to such features.

    Ability to restrain access

    DE law on access to technical infrastructures supporting payment services. This has already been regulated in Germany to ensure fair access of other service providers to NFC. Necessary to make above obligation function. NFC antenna just one element of a broader hard/software functionality.

    According to IA support study , Annex 4 Case 6: Device Neutrality – Browser Neutrality and Case 9: Slack – Interoperability / API.

    Italian NCA opens investigation into Google Maps, brought by Enel competing mapping service. 199

    Dutch NCA opens investigation into NFC access, following their study Big Tech and the Dutch payment market. 200

    Gatekeepers shall not refuse to provide advertisers and publishers upon their request, with access to the performance measuring tools of the gatekeeper and the information necessary for advertisers and publishers to carry out their own independent verification of the ad inventory.

    Example: advertisers would like to obtain access to performance measuring tools of gatekeepers to assess effectiveness of its advertising campaign, but is refused access to such tools.

    Ability to refuse accountability on the level of service provided, possible due to the gatekeeper’s market position

    Similar reasons as to above for facilitating/ensuring transparency in the advertising value chain. Grounds similar and linked to concerns raised by the business users of advertising services. 201

    CMA report on Online platforms and digital advertising pages 297-303.

    ACCC report, Digital Platforms Inquiry, Final Report , pages 145-150.

    Support Study for the Observatory, Report on Transparency in the Online Advertising Market , Observatory Expert Group Report: Market power and transparency issues in open display advertising.

    Gatekeepers shall provide business users with effective data porting possibilities for data generated on core platform services, subject to GDPR consent requirements as applicable.

    Example: third-party provider of online newspaper may need access to data of the potential user (i.e. subscriber) of its services in real time.

    Ability to refuse access to data on which businesses depend

    According to the IA support study , there are several antitrust investigations in this area.

    Lock-in on the business user-side. Also undermines effectiveness on end user-side (cf. Article 5(1)(i)).

    Support Study to the Observatory, Business user and third-party access to online platform data.

    Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , pages 88-89.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , chapter 2.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , page 20.

    ACCC report, Digital Platforms Inquiry, Final Report , pages 115-116.

    Gatekeepers shall not prevent free of charge, unhindered access to and use of non-aggregated and aggregated data that is provided for, generated in the context of, or inferred from, the use of the relevant core platform services by those business users and the customers acquiring the products or services provided by those business users.

    Example: online newspaper asks the provider of online intermediation service for contacts of the customer who subscribed to its service through software application store of the gatekeeper, but is refused such data on privacy grounds, even if subscriber was never asked for consent, or lack of it, for such data sharing.

    Ability to limit access to data relevant for the business user’s activity

    IA support study , Annex 4, Case 7: Device Neutrality – Apple Wallet/Pay.

    There are many complaints in particular by app developers about so called disintermediation. That is to say that the gatekeeper gathers and controls the customer data and those who are in a business relationship with a specific customer actually lack access to customer data and are therefore disintermediated form their own customers.

    P2B Impact Assessment: “In the Flash Eurobarometer 43960 42% of the respondents said that they usually do not get the data they need about their customers from online marketplaces.”

    Support Study to the Observatory, Business user and third-party access to online platform data and Platform data access and secondary data sources .

    Gatekeepers shall not prevent access, upon request of business users and any third party providers of online search engines, to query, click and view data in relation to free and paid search generated by consumers on the online search engines of the gatekeeper.

    Example: provider of competing online search engine services asks the gatekeeper to provide access to its click-and-query data, which is refused without any consideration or explanation.

    Ability to refuse access to data

    Gatekeepers in search have a data advantage over competitors and business user insofar as they have access to a vast amount of query data especially on long tail queries (see Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017).

    Vertical and horizontal integration reinforced by agreement, for example Apple-Google exclusivity deal targeted by US DoJ, limits contestability of online search engines. 202  

    CMA report on Online platforms and digital advertising , pages 92-96.

    ACCC report, Digital Platforms Inquiry, Final Report , pages 66-67.

    Gatekeepers shall apply fair and non-discriminatory general conditions of access for business users to its software application store.

    Example: provider of an app store charges different commission rates to different business users without clear identification of reasons for such differentiation.

    Ability to apply discriminatory pricing due to its strategic market position

    According to the IA support study , in particular Annex 4, Case study 4: Restriction of access and use of business users to data about their customers – Apple App Store there are several antitrust investigations in this area in relation to app stores.

    The Dutch competition authority found in a study on app stores that: “app providers have only limited options for reaching consumers on their iPhone outside of the App Store. Even though it is technically possible on Android for app providers to reach consumers circumventing the Play Store, this is only a realistic alternative for companies with an already established user base accomplished on other distribution channels. Furthermore, in this market study, ACM concluded that the majority of app providers have limited bargaining power vis-àvis the app store.

    US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , pages 343.

    Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , page 46.

    P2B evidence of unfair contract terms. 203

    158.Finally, it is also worth noting that gatekeepers frequently raise arguments concerning the efficiencies that their practices bring about as a way to counterbalance and justify their potential negative effects. These arguments – raised not only in the OPC but also in numerous past and ongoing investigations (in fields such as antitrust, consumer protection or privacy) – are often one-sided and do not seem to match the evidence underlying this Impact Assessment including the calls for regulation raised by an overwhelming majority of respondents to the OPCs. Such efficiency-related defenses have also been rejected by the Courts as being unfounded. 204

    Speed and Flexibility

    159.Another important design parameter for the available policy options concerns the architecture of the intervention, notably speed of intervention, and the degree of flexibility concerning the main elements of the intervention.

    160.Concerning the speed of application, the main choices are between (a) model of immediately applicable obligations, (b) a model where a degree of appreciation is necessary, notably as regards the implementation of a given obligation, and (c) a fully flexible model, where obligations or remedies are only imposed subsequent to an investigation carried out by an authority.

    161.Such models have regulatory precedents or analogues in other acts of Union law. For instance, the Unfair Commercial Practices Directive (‘UCPD’) 205  and the Directive on Unfair trading practices in the agricultural and food supply chain 206 have models of immediately applicable black-lists of practices that are essentially lists of prohibited conduct.

    162.In contrast, the European Electronics Communication Code offers a more flexible, case-by-case regulatory framework, whereby a regulator can impose a set of remedies following an analysis of problems in the internal market related to the telecommunications sector. 207

    163.For the flexibility element, there are essentially three parameters that could characterise the main choices. First, a flexibility on the designation of a gatekeeper, e.g. by using only qualitative thresholds, or by updating the thresholds in light of market developments. Second, a flexibility on the list of practices that should be subject to a remedy and to their implementation. For instance, the list of practices could be left fully open, subject to an update, or selected after a market investigation or an analysis from a pre-determined list of practices. A third source of flexibility concerns the type of core services in scope, as discussed in the previous section. It is easily conceivable that future technological developments require different core services to be in scope of the instrument.

    164.Stakeholders in the OPC have generally supported a mix of approaches, combining the speed of immediately applicable obligations and a more flexible approach. 208

    Enforcement framework

    165.A final important design parameter is linked to the enforcement powers necessary and available to ensure that the rules concerned are not undermined, ineffective or absent enforcement. This concerns individual cases of non-compliance as well as cases of more systematic non-compliance by gatekeepers. It is relevant to all options, and also linked to the monitoring of the implementation.

    166.In order to ensure effective compliance with the ex ante rules, in any enforcement framework the Commission should have investigative and enforcement powers to allow it to investigate, enforce and monitor the ex ante rules, while at the same time ensuring the respect of the fundamental right to be heard and to have access to the file in the context of the enforcement proceedings. In particular, the Commission should have access to any relevant documents, data and information necessary to open and conduct investigations and to monitor the compliance with the obligations addressed to designated gatekeepers, irrespective of who possesses the documents, data or information in question, and regardless of their form or format, their storage medium, or the place where they are stored.

    167.In order to have sufficient deterrence powers it is necessary that after a due process the Commission shall be able to impose fines and periodic penalties or take the necessary measures to restore compliance. In this respect the Regulation 1/2003 offers a well- known and legally sound model that can be replicated in Options presented in this Impact Assessment.

    168.In the extreme case, where an investigation shows that a gatekeeper has systematically infringed the obligations and has further strengthened or extended its gatekeeper position, the Commission should be able as it is the case of Article 7 209 of Regulation 1/2003 to impose the structural remedies necessary to guarantee that market participants are not irreversibly harmed by this repeated and illicit behaviour. This consideration applies to all categories of intervention in terms of effectiveness.

    169.However, to guarantee proportionality of the intervention the Commission should only impose structural remedies either where there is no equally effective behavioural remedy or where any equally effective behavioural remedy would be more burdensome for the gatekeeper concerned than a structural remedy.

    170.In addition, the notion of systematic non-compliance should be linked with repeated infringements. Gatekeepers shall be deemed to have engaged in a systematic non-compliance in cases where the Commission issued at least two or more non-compliance or fining decisions. The Commission shall have the duty to explain whether and which remedy or remedies it preliminarily considers necessary and proportionate. At any point in time during the proceedings gatekeepers shall be put in condition to offer commitments that if accepted would terminate the infringement.

    171.The above-described system is construed to safeguard proportionality of the intervention. It is based on existing instruments.

    172.Structural measures are to be seen an ultima ratio measure for repeated infringement and to be taken into account where every other possibility has failed. To be noted that in the application of Regulation 1/2003 this circumstance has never occurred.

    Summary and main trade-offs

    173.In summary, the key parameters that determine the choice of options are related to the scope, the set of obligations related to unfair trading practices, the flexibility of the architecture and the scope of enforcement powers. These parameters are linked via a set of trade-offs that are set out below, mainly in terms of considerations of overall proportionality and effectiveness.

    174.The first trade-off is linked to the gatekeepers in scope and the intensity of obligations. A wider scope of gatekeeper platforms corresponds with a lower intensity of the obligations linked to unfair practices and vice-versa, as the degree of harm caused by unfair practices is linked to the strength of the gatekeeper power.

    175.The second trade-off is linked to the catalogue of obligations and the flexibility of the instrument. The more flexible the intervention, the less prescriptive the obligations need to be and vice-versa.

    176.A third trade-off is linked to the speed and flexibility of the intervention. An intervention that requires a detailed analysis for each case will necessarily be slower in effect than an intervention based on a list of immediately applicable prohibitions.

    177.The fourth trade-off is linked to the scope of the remedies in case of systematic non-compliance. An intervention that requires a structural remedy will necessarily be slower since it will require several enforcement steps before the effective compliance will be ensured. However, at the same time, it is an ultima ratio remedy that will only be considered if all other measures do not produce the required result.

    178.While the set of parameters and trade-offs in this section theoretically allow many different combinations, not all of them are credible or meaningful.

    Policy options

    179.Based on the analysis of the main parameters that characterise the problem, and the trade-offs presented in the previous section, three distinct policy options have emerged from the available matrix of combinations as plausible ones and will therefore be fully assessed.

    180.Option 1 is a non-dynamic option with a set of immediately applicable obligations addressing clearly defined unfair practices by gatekeepers designated solely on quantitative thresholds in specific core platform services. This option contains no dynamic elements, but is presented with distinct two sub-options on scope as distinct alternatives, on the basis of different thresholds. Sub-option A is presented as a sub-option with a small number of gatekeeper companies in scope (some 5-7 companies) while sub-option B contains a wider scope of gatekeeper companies (some 10-15 gatekeepers), based on a lower quantitative threshold.

    181.Option 2 is a semi-flexible option, combining a set of immediately applicable obligations with some degree of flexibility, notably through a dialogue on some of the obligations and a mechanism for updating the practices and obligations. It further comprises a mechanism designating gatekeepers based on a combination of quantitative and qualitative thresholds and including the designation of emerging gatekeepers. Again, this semi-flexible option is presented with two sub-options that reflect alternatives on the scope. Sub-options A and B are sub-options on this semi-flexible option, following the same distinction on the quantitative threshold as Option 1.

    182.Option 3 is a fully flexible option providing for a dialogue on all the obligations listed and a dynamic updating mechanism allowing for the inclusion of additional core platform services and of additional obligations where following a market investigation such an inclusion is considered appropriate and justified. Further, the designation of gatekeepers is based only on qualitative thresholds.

    183.These options represent distinct alternatives based on the inherent trade-offs that underline the problem definition. They are distinguished by the architectural element relating to the flexible or dynamic character of the proposed intervention.

    184.The table below gives an overview of these policy options.

    Table 3: Parameters of policy options

    Option

    Flexibility

    Obligations

    Scope

    1

    Non-dynamic

    Immediately applicable obligations

    1.A

    High quantitative thresholds

    1.B

    Low quantitative thresholds

    2

    Semi-dynamic

    Immediately applicable +

    Obligations with regulatory dialogue

    +

    Updating mechanism for new practices

    2.A

    High quantitative thresholds + qualitative designation

    2.B

    Low quantitative thresholds + qualitative designation

    3

    Fully dynamic

    Obligations with regulatory dialogue

    +

    Updating mechanism for new practices and core platform services

    3

    Qualitative criteria only

    Option 1 – Pre-defined list of gatekeepers and immediately applicable obligations

    Summary of option 1

    185.Option 1 would consist of the following elements:

    (a)a closed list of core platform services identified in Section 5.2.1;

    (b)designation of providers of core platform services as gatekeepers based solely on the quantitative thresholds; and

    (c)the whole list of obligations identified in Section 5.2.2 would be immediately applicable without any ability of a regulatory dialogue.

    Identification of core platform services

    186.Option 1 would provide for a new targeted ex ante regulatory framework, which would apply to identified ‘core platform services’ (see Section 5.2.1) provided by designated gatekeepers. These core platform services define the perimeter both for the designation of gatekeepers as such, and for the identification of those individual core platform services provided by the designated gatekeeper which would have to comply with the clearly defined closed list of obligations identified in Section 5.2.2.

    Designation of gatekeepers

    187.Under Option 1, providers of core platform services would be designated as gatekeepers based on pre-defined quantitative thresholds. These quantitative thresholds would provide a high degree of legal certainty for market operators. Conversely, Option 1 would not allow any flexibility to identify gatekeepers on the basis of criteria other than the quantitative ones and would also not enable the providers of core platform services to, in exceptional circumstances, demonstrate based on serious and substantiated arguments that they do not meet the conditions in paragraph 133. This would also exclude any type of case-by-case analysis following an in-depth market investigation.

    188.In view of this, and as explained in paragraph 148, two sub-options could be considered:

    ·Sub-option 1-A, i.e. high threshold implying the designation of five to seven gatekeepers.

    ·Sub-option 1-B, i.e. low threshold implying the designation of 10 to 15 gatekeepers.

    189.Under Option 1, the fixed quantitative criteria would also exclude any possibility to identify and designate those providers of core platform services that are expected to enjoy an entrenched and durable position in the near future.

    Identification of obligations applicable to gatekeepers’ core platform services

    190.Under Option 1, a closed list of obligations that the designated gatekeepers would have to comply with would be defined in the rules themselves. These obligations would be set on the basis of the criteria presented in Section 5.2.2 and are identified in the Table 2 .

    191.The designated gatekeepers would be required to comply with all the obligations laid down in the rules. Under Option 1 the gatekeepers concerned could not engage in a dialogue with the regulator about the measures they intend to take or have taken in order to comply with these obligations (i.e., the whole set of obligations would be immediately applicable).

    Enforcement framework

    192.Option 1 foresees implementation, supervision and enforcement at the EU level by the Commission as the competent regulatory body. Given the pan-European reach of the targeted companies, a decentralised enforcement model does not seem to be a conceivable option, including in light of the fragmentation that the initiative is supposed to address, nor would it be proportionate given the limited number of gatekeepers that would be in scope of the proposed framework. However, to integrate the national expertise in the platform economy, the initiative would envisage that the Commission consults a ‘network of regulators’ before taking decisions that could be considered under Option 1 (e.g. designation of gatekeepers; non-compliance; fines; period penalty payments; remedies decisions in case of systematic non-compliance).

    193.To be able to effectively carry out its work, the Commission would enjoy clearly-defined and circumscribed procedural powers, which would include:

    ·The power to request information from the gatekeepers or third parties to determine compliance with the rules;

    ·The power to conduct on-site inspections to collect any information that may be necessary to establish such a non-compliance;

    ·The ability to adopt interim measures in case of a risk of serious and irreparable damage for business users or end users of gatekeepers, where there are strong indications of a prima facie finding of infringement of obligations addressed to gatekeepers;

    ·The ability to make legally binding voluntary measures that the gatekeepers may offer in the context of the non-compliance procedure to ensure the effective implementation and compliance with their obligations; and

    ·The ability to adopt non-compliance decisions, including fines and period penalty payments where necessary and justified.

    194.In order to ensure the effectiveness and speed of intervention – as well as a way to ensure legal certainty and to replicate the obligatory nature that data-gathering powers would have on gatekeepers – proceedings would be subject to binding legal deadlines. Respondents to the OPCs and NCAs, generally argued in favour of binding deadlines for both the Commission and the businesses concerned in order to ensure expediency and legal certainty. 210 Respondents also added that deadlines would ensure a swifter outcome, which is all the more necessary, in particular in digital sectors, both for a swift resolution of the case and for providing sufficient legal certainty to the market. As regards binding deadlines for the businesses concerned, respondents argued that this would avoid risks of certain businesses slowing down the process with dilatory conducts, and that these deadlines should be coupled with the possibility of imposing fines for non-compliance to ensure speed and effectiveness.

    195.In addition, in order to ensure due process and protection of rights of the parties to the procedure, it is important that any addressee of the decision has the opportunity of being heard on the final decision considered and that all decisions taken are subject to judicial review.

    196.Finally, under this option the Commission would also have the power to ensure that in case of a systematic non-compliance further appropriate and proportionate measures are taken to ensure that objectives of the ex ante rules are not undermined. The exact scope of such behavioural or structural measure should be proportionate to the infringement committed and necessary to ensure compliance with the ex ante rules (see Section 5.2.4).

    197.The enforcement powers required to enforce the prohibitions and obligations under this option are not new creations but on the contrary largely reproduce existing powers that the Commission has under the competition and regulatory frameworks 211 .

    198.As regards the powers inspired from the EU competition acquis, the main source of reference would be Regulation 1/2003, the legal text governing the conduct of competition investigations by the European Commission. In this respect, Regulation 1/2003 contains also tools which largely mimic the five investigative measures referenced in paragraph 193 as being necessary for the enforcement of this option. Some differences between the investigative measures under Regulation 1/2003 and this option would, however, need to be included. Most notably, this would involve the inclusion of an explicit power to access databases, algorithms and other technical elements that are characteristic to the digital economy. While under Regulation 1/2003 the Commission also has the ability to access and conduct searches into these elements, the nature of the tool under the current impact assessment justify the inclusion of an explicit power in this respect. Remedies sanctioning a refusal to access those should also be provided for. Another difference would concern the fact that – at least for the time being – the ability to inspect other premises under Article 21 of Regulation 1/2003 would not seem necessary for the purposes of the current instrument.

    199.In a similar vein, powers to request information from companies or the ability to propose commitments to the regulators are also commonplace in regulatory systems. One such example can be found, for example, in Article 78 of the EU’s telecommunications framework. 212  

    Option 2 – Partially flexible framework of designation and updating of obligations, including regulatory dialogue for implementation of some

    Summary of option 2

    200.Option 2 would consist of the following elements:

    (a)a closed list of core platform services identified in Section 5.2.1;

    (b)a combination of quantitative and qualitative criteria to designate providers of core platform services as gatekeepers;

    (c)the obligations identified in Section 5.2.2 would consist of immediately applicable obligations including some obligations where regulatory dialogue may facilitate their effective implementation; and

    (d)new practices may be added on the basis of a market investigation.

    Identification of core platform services

    201.Option 2 would also provide for a new targeted ex ante regulatory framework, which would apply to identified ‘core platform services’ (see Section 5.2.1) provided by designated gatekeepers. These core platform services would again serve as the perimeter for the designation of gatekeepers as such, and for the identification of those individual core platform services provided by the designated gatekeeper which would have to comply with the clearly defined obligations as set out in Section 5.2.2.

    202.Similarly to Option 1, once this list of core platform services would be identified, there would be no possibility to update such list beyond the revision of the rules themselves and there would be no ability to update them by means of tools provided in the rules themselves.

    Designation of gatekeepers

    203.Under Option 2, providers of core platform services would be designated as gatekeepers based on the combination of pre-defined quantitative thresholds but also following a case-by-case assessment in the context of a market investigation.

    204.Like under Option 1, also under Option 2, two quantitative thresholds could be considered as defined in paragraph 148:

    ·Sub-option 2-A, i.e. high threshold implying the designation of five to seven gatekeepers.

    ·Sub-option 2-B, i.e. low threshold implying the designation of 10 to 15 gatekeepers.

    205.Because of the combination of quantitative and qualitative thresholds, Option 2 would include a certain degree of flexibility, which would allow to capture two important dynamic elements of the platform ecosystem.

    206.First, Option 2 would have the ability to designate gatekeepers not yet enjoying an entrenched and durable position, but which are expected to enjoy such a position in their operations in the near future. Such a designation would prevent core platform services, where these emerging gatekeepers operate, to tip because of weak contestability of the market concerned. The phenomenon of tipping – an irreversible loss of competition in a given market that occurs in a sudden manner – is further explained in Section 2.3.1.1 above.

    207.Second, Option 2 would also foresee that the designation of the gatekeeper should be regularly reviewed where there would be a material change in any of the facts on which the designation decision was based, and where the decision was based on incomplete, incorrect or misleading information provided by the undertakings.

    208.Finally, under Option 2, in view of the dynamic element and combination with the qualitative assessment of the gatekeeper status, the provider of core platform services would be able to present, in exceptional circumstances, serious and substantiated arguments to demonstrate that, in circumstances in which the relevant core platform service operates, it does not fulfil the objective requirements for a gatekeeper (see paragraph 133) and should therefore not be designated directly based on the application of quantitative thresholds, but only subject to a further investigation. The purpose of such an implementation of a legal presumption would not be to demonstrate, on pure economic grounds, efficiencies deriving from a specific type of behaviour by the provider of core platform services since this is not relevant to designation of such a provider as a gatekeeper.

    Identification of obligations applicable to gatekeepers’ core platform services

    209.As under Option 1, a closed list of obligations that the designated gatekeepers would have to comply with would be defined in the rules themselves. These obligations would be set on the basis of the criteria presented in Section 5.2.2 and are identified in the Table 2 .

    210.The designated gatekeepers would be required to comply with all the obligations laid down in the rules. However, for some obligations, Option 2 would provide the gatekeeper with the possibility to discuss with the Commission the measures it intends to take or has taken in order to ensure their effectiveness. This would provide additional flexibility in tailoring the implementing measures by the gatekeepers to the given obligation and circumstances of each gatekeeper.

    211.Option 2 would also include a flexible element by allowing to update the list of obligations whenever new unfair practices would be determined following a market investigation. Consideration was given to the question whether an additional flexibility in relation to the possibility to also update the list of core platform services should not be included in this option. This possibility has not be retained since it would bring Option 2 too close to Option 3; while the options’ design was aimed at ensuring a set of options that are not only plausible but also well distinct from each other as regards the relevant trade-offs, thus providing a wider array of policy choices.

    Market investigation framework

    212.Option 2 would envisage a possibility for the Commission to carry out a market investigation in the following types of situations.

    213.First, the Commission would carry out a market investigation in order to designate on a case-by-case basis a provider of core platform services that meet the conditions referred to Section 5.2.1 as a gatekeeper. In doing so, the Commission would take into account a number of elements, such as the size, operations, the number of business users depending on the core platform service to reach end users and the number of end users, entry barriers derived from network effects and data driven advantages, in particular in relation to provider’s access to and collection of personal and non-personal data or analytics capabilities or scale and scope effects the provider benefits from including with regard to data.

    214.In the context of the market investigation, the Commission could identify as a gatekeeper not only a provider of core platform services that already enjoys an entrenched and durable position, but also those providers that are expected to enjoy such an entrenched and durable position in the near future.

    215.Second, the Commission could initiate a market investigation to identify possible new practices. A report summarising such a market investigation could serve as a basis for a possible revision of the Regulation, either based on a dedicated empowerment for the Commission enabling it to update the obligations in the rules themselves or by means of a full review of the rules. Under a market investigation additional digital services could also be assessed. However, in order to maintain some legal certainty, this could only result in an update of the list of core platform services in the context of the review of the Regulation, which could possibly take place every three years.

    216.Third, the Commission would carry out a market investigation when there is the suspicion that a gatekeeper has systematically infringed the obligations laid down and has further strengthened or extended its gatekeeper position.

    217.The different procedures to be followed by the market investigation framework would be uniquely designed for this option but would nevertheless have some similarities with the frameworks under EU competition law and sector-specific regulation.

    218.In this respect, market investigations – like competition investigations – would be initiated by opening decisions and gatekeepers which are the object of them would have the opportunity to be heard in relation to the allegations raised by the Commission. Unlike in procedures under Regulation 1/2003, the use of the market investigation framework would also in some cases also be subject to timeframes.

    219.In a similar vein, the procedures governing the market investigation framework would also have parallelisms with regulatory systems. One such example would, for example, be Article 63 of the EU’s telecommunications framework allowing for the designation of ‘undertakings with significant market power’. 213

    Enforcement framework

    220.Similarly to Option 1, also Option 2 foresees implementation, supervision and enforcement at the EU level by the Commission as the competent regulatory body for the reasons explained in Section 5.3.1.5.

    221.Furthermore, the new rules under Option 2 would envisage the same set of enforcement powers for the Commission as envisaged under Option 1.

    222.Finally, under this option the Commission would also have the power to ensure that in case of a systematic non-compliance further appropriate and proportionate behavioural or structural measures are taken to ensure that objectives of the ex ante rules are not undermined (see Section 5.2.4).

    Option 3 - Flexible option based exclusively on qualitative scoping thresholds

    Summary of Option 3

    223.Option 3 would consist of the following elements:

    (a)a closed list of core platform services identified in Section 5.2.1;

    (b)designation of providers of core platform services as gatekeepers following a pure qualitative assessment;

    (c)the obligations identified in Section 5.2.2 would all be subject to a regulatory dialogue; and

    (d)new practices and new core services may be added on the basis of a market investigation.

    224.Consideration was given to the question whether some quantifiable elements should not be included in this option. This possibility has not be retained since it would bring Option 3 too close to Option 2; while the options’ design was aimed at ensuring a set of options that are not only plausible but also well distinct from each other as regards the relevant trade-offs, thus providing a wider array of policy choices.

    Identification of core platform services

    225.Option 3 would also provide for a new targeted ex ante regulatory framework, which would apply to identified ‘core platform services’ (see Section 5.2.1) provided by designated gatekeepers. These core platform services would again serve as the perimeter for the designation of gatekeepers as such, and for the identification of those individual core platform services provided by the designated gatekeeper which would have to comply with the clearly defined obligations as set out in Section 5.2.2.

    226.While Option 3 would start by focusing on a list of core platform services, the flexibility of the tool would mean that it should also be able to assess whether new digital services need to be incorporated. Where such a need is proven after a market investigation, the new digital service in question would be added to the scope of the rules.

    Designation of gatekeepers

    227.Option 3 does not include any quantitative thresholds and is solely based on case-by-case qualitative assessments. As such, these qualitative assessments would be the only way to determine which gatekeepers would fall under the scope of the rules and would seek to determine whether the provider of the core platform service has a significant impact on the internal market, operates a core platform service which serves as an important gateway for business users to customers, and enjoys (or is expected to enjoy) an entrenched and durable position in its operations.

    228.Similarly to Option 2, Option 3 would have the ability to designate gatekeepers not yet enjoying an entrenched and durable position, but which are expected to enjoy such a position in their operations in the near future. However, contrary to Option 2, Option 3 would not be bound by any quantitative elements. While the inclusion of quantitative elements as guidance could in principle be implemented while preserving the flexibility of the tool, such elements would also defeat the purpose of Option 3 as they would still constrain its ability to look at any company, regardless of its size or position in the market. The inclusion of such thresholds would also undermine the flexibility of this option as it would require regular reviews of the legislation in order to update the (indicative) thresholds in a manner that does not constrain its flexibility.

    229.As compared to Option 2, however, Option 3 could result in the designation of a higher number of gatekeepers, namely those active in the new digital services added to the scope of the rules after a market investigation.

    230.Option 3 would also foresee that the designation of the gatekeeper should be regularly reviewed where there would be a material change in any of the facts on which the designation decision was based, and where the decision was based on incomplete, incorrect or misleading information provided by the undertakings.

    Identification of obligations applicable to gatekeepers’ core platform services

    231.As under Option 1 and 2, a closed list of obligations that the designated gatekeepers would have to comply with would be defined in the rules themselves. These obligations would be set on the basis of the criteria presented in Section 5.2.2 and are identified in Table 2 . Similarly to Option 2, Option 3 would include a flexible element by allowing to update the list of obligations whenever new unfair practices would be determined following a market investigation. Option 3 could comprise more obligations than Option 2 in case additional core platform services were to be included in the scope and new practices as regards those additional services were determined to be unfair.

    232.Option 3 entails a maximum degree of flexibility amplifying the level of dialogue to the point of allowing potential gatekeepers to present allegations as regards all obligations.

    Market investigation framework

    233.Option 3 would envisage a possibility for the Commission to carry out a market investigation in the following types of situations.

    234.First, by not including any list of quantitative criteria, the Commission would always need to engage in a market investigation to determine which providers of core platforms services should be considered as gatekeepers. In order to do this, the Commission would pay attention to the conditions prevailing in the market as well as to the position of the provider of services.

    235.Second, a market investigation would also be used by the Commission to specify the services in which the provider would be behaving as a gatekeeper. In case the service in question is not pre-defined in the list of core platform services, the Commission could update the list prior to imposing any obligations.

    236.Third, the Commission could also use a market investigation to update the list of practices that are unfair or are contributing to a lessening of the contestability of the market. Once this practice or practices are identified, possible obligations could be added to the list of obligations imposed.

    237.Fourth, the Commission would carry out a market investigation when there is the suspicion that a gatekeeper has systematically infringed the obligations laid down and has further strengthened or extended its gatekeeper position.

    238.The procedures to be followed by the market investigation framework would be similar to the ones described in Option 2.

    Enforcement framework

    239.Similarly to Option 1 and 2, also Option 3 foresees implementation, supervision and enforcement at the EU level by the Commission as the competent regulatory body for the reasons explained in Section 5.3.1.5. Furthermore, the new rules under Option 3 would envisage same set of enforcement powers for the Commission as envisaged under Option 1 and 2.

    240.Finally, under this option the Commission would also have the power to ensure that in case of a systematic non-compliance appropriate and proportionate behavioural or structural measures are taken to ensure that objectives of the ex ante rules are not undermined (see Section 5.2.4).

    Policy options discarded at an earlier stage

    A broad scope across platforms

    241.As indicated in the Inception Impact Assessment, an option of amending the P2B Regulation was considered. Further horizontal rules could be established for all 10 000 online intermediation services and search engines that are currently falling within the scope of the P2B Regulation. This could cover prescriptive rules on different specific practices that are currently addressed by transparency obligations and beyond.

    242.The impact of this option on the internal market would be further harmonisation in a wider range of areas, but without necessarily addressing the issues at stake. As such, the option would include a risk for these issues to be addressed at national level, resulting in no substantial improvement in the functioning of the Single Digital Market.

    243.Imposing stringent measures horizontally would risk being disproportionate and have a negative impact on innovation and competition in the online platform economy. Stricter rules under this option would be intrusive for many of the 10 000 entities currently falling within the scope of the P2B Regulation, but could be especially harmful for smaller platforms, possibly also limiting their growth. The option would be expected to increase innovation for platforms’ business users but the extra burden for smaller platforms would stifle their potential to invest in innovation. In the same way, this option would risk to negate any positive impact on the ability of business users to compete by the negative impact on the ability of smaller platforms to compete. Given the large number of platforms covered, and in order to avoid a disproportionately negative effect on smaller players, there is also a risk that the rules have to be toned down with the result that problems relating to gatekeepers would not be addressed in the most adequate and vigorous manner.

    244.As any stricter rules would apply to all platforms, compliance costs would be more burdensome for smaller platforms with limited resources. As such these rules would have a limited impact on gatekeepers, and possible a negative impact on smaller platforms due to the disproportionate regulatory burden on them. Because of its wider scope encompassing all platforms, this option would also lead to high enforcement and coordination costs for authorities.

    245.The impact on business users and SMEs would be dependent on whether they would qualify as platforms themselves. For non-platforms, the stricter rules would provide benefits across all platforms. For platforms, the impact would be beneficial if the costs they incur for complying with new rules are lower than the benefit from a fairer behaviour by the platform they use and/or compete with (and vice versa).

    246.The impact on consumers would be limited, as the mitigated impact on the platform economy (due to its scope regulating also smaller platforms) would also limit this option’s positive impact on consumers.

    247.Therefore, an option based on the P2B Regulation, targeting not only gatekeepers but all platforms, was discarded as this would constitute a mismatch with the problems and their drivers as identified in this Impact Assessment.

    248.The reversed scenario – i.e. changing the scope of P2B Regulation to gatekeepers only – would not be a conceivable way forward as it would eliminate the beneficial impact of its fairness and transparency rules addressed to non-gatekeeper platforms.

    Information obligations without addressing unfair conduct

    249.Furthermore, at the Inception Impact Assessment stage, an option was considered which would empower a regulatory body to collect information from large online platforms acting as gatekeepers. These data-gathering powers would be supported by enforcement powers in case of refusal to supply this information. The purpose would be to better inform the implementation of the existing legal framework by gaining, for example, further insights into gatekeepers’ business practices and their impact on these platforms’ users and consumers, the scope of gatekeepers’ data gathering, treatment of their own downstream operations compared with those of third parties and indicators of the outcomes resulting from these practices.

    250.The impact of this option on the internal market would consist of a better knowledge of platform ecosystems but it would not lead to any improvement of the internal market functioning as it would not set any behaviour-changing rules. Furthermore, by leaving the problems unaddressed the option would include a risk for the identified issues to be addressed at national level, resulting in legal fragmentation limiting the digital market functioning.

    251.In the same way, the impact of this option on growth, innovation and competition would be limited, as it would only increase regulators’ understanding of gatekeepers’ trading practices and business models but not foresee any regulatory measures. There might be a reputational effect associated with more transparency but this would not affect competition (substantially).

    252.The impact on platforms would be limited and focused on gatekeeper platforms as other platforms would not be subject to legal obligations and information requests would be proportionate. There would be a benefit to smaller platforms in the long run due to regulators’ better understanding of the issues and their effective redress. Enforcement costs would be incurred by the Commission, but this would be in the public interest since it would allow a better understanding of the platform economy and hence increase EU public administration capacity to tackle related issues.

    253.The impact on business users and SME’s would also be limited. Irrespective of the type of business user concerned, i.e. platform or not, the information-gathering mechanism would only allow a better understanding of issues at hand but would not change competing business users’ situation vis-à-vis gatekeepers. While SME platforms and business users could be requested to provide information to enforcement authorities, this would not be a legal obligation for them. The authorities’ increased insight into gatekeepers’ practices would enable better regulation and enforcement and thus have a positive impact on all SMEs. This impact would, however, be delayed in time given the time-frames applicable to any legislative process.

    254.There would not be a concrete impact on consumers in the short term, as this option would only provide further insight in how gatekeepers treat consumers and the role played by consumers in business models.

    255.Therefore, an option limited to improving access to information on the issues at stake was discarded as it would be insufficient to address the problems identified in this Impact Assessment or affect their drivers.

    A broad scope across markets

    256.Finally, as indicated in the Inception Impact Assessment, an option of having a market investigation regime with a horizontal scope (i.e. extending to all markets) as opposed to just a digital scope was considered. Under this option, market investigations would also be applicable to non-digital markets, which are not directly part of the objectives of this initiative. In fact, respondents to the OPC indicated that market failures occur in all sectors and markets and highlighted that no sector is immune to (potential) market failures. 214 At the same time, a high number of respondents who indicated that market failures can occur in all sectors and markets mainly pointed to digital examples in their replies. Respondents indicating that market failures mainly or solely occur in digital sectors/markets argued that the characteristics of the digital sector (e.g. economies of scale and scope, data accumulation and dependency, network effects, lock-in, zero pricing) make digital markets particularly prone to the emergence of quasi-monopolistic market structures.

    257.BEUC pointed out that competition law enforcement in digital markets, though important, has not been effective enough in dealing with all problems in these markets and consequently not been able to remedy, let alone prevent, harm to consumers in a timely manner. 215 Indeed, there is an extensive economic literature and numerous reports as explained in Section 2 describing the growth of digital markets and their particular characteristics that makes them prone to market failures, and where resources would be better focused, at least at the initial stage of any new investigation regime.

    258.This option would not allow an immediate and continuous response to the most pressing instances of gatekeeper related market failures in respect of key digital markets. In fact, it would divert the focus of investigations to other markets which are not in the scope of this initiative and would therefore not be very effective in addressing the market failures listed in Section 2.3.1. In the particular case of the problem driver related to the fragmented regulation of digital markets, this option would not be effective as it would not set common rules to address unfair practices and increase market contestability in digital markets and would therefore not lead to a more coherent regulatory approach across the EU.

    259.In terms of efficiency, this option would result in a large amount of resources being dedicated to other investigations, and thus limit those available to tackle the pressing problems in digital markets. The costs for the regulatory authority would be necessarily higher than all the alternatives considered with no additional benefits in terms of its ability to achieve the objectives listed in Section 4.

    260.Therefore, an option to have a market investigation regime extended to all markets was discarded as it would be out of the scope of this initiative and not focused in addressing the problems identified in this Impact Assessment or affect their drivers.

    6.WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?

    261.This section presents the main impacts of the three options described in Section 5.3 compared to the baseline scenario.

    262.The categories of stakeholders which would be affected, directly or indirectly, by the retained policy options are: platforms (gatekeepers and non-gatekeeper platforms), business users depending on platforms (e.g. hotels, sellers in marketplaces, app developers, banks) and possibly competing with the gatekeeper, competitors (e.g. innovative entrants), consumers, and regulatory authorities. Impacts for these stakeholder categories have been assessed in the following sub-sections covering the internal market (Section 6.1), growth and productivity (Section 6.2), competition and innovation (Section 6.3), international trade (Section 6.4), employment (Section 6.5), businesses – i.e. gatekeepers and SMEs in their role both as competitors and business users – (Section 6.6), consumers (Section 6.7) and regulatory authorities (Section 6.8).

    263.By way of background, among the respondents who replied to the relevant question in the OPC, 91% agree that there is a need to consider dedicated regulatory rules to address negative societal and economic effects of gatekeeper platforms. 216 This view is supported by many targeted submissions by different groups of stakeholders, such as small and medium platforms and their associations, telecom operators and their associations as well by national regulatory authorities in different sectors (e.g. electronic communication services).

    264.For the impacts developed in this section see also Annex 3 to the Impact Assessment, which specifies in detail who would be affected by the preferred option and how.

    265.The problems and their underlying drivers as identified in Section 2 can lead, individually and jointly, to a number of negative outcomes specified where relevant in this section.

    Internal market

    266. Preventing fragmentation of the internal market is one of the most important policy objectives enshrined in the Treaties of the EU, and preserving the cross-border nature of the platform economy contributes to this objective. A 2016 European Commission Communication on the opportunities and challenges of online platforms for the Digital Single Market stressed the pivotal role of online platforms in the European single market. Services and products such as search engines, price comparison websites, online marketplaces and creative content outlets offer strong links to the rest of the economy. 217 A conservative estimate at the time of the Communication put the number of EU companies in Europe ´heavily´ using online platforms to trade goods and services at one million, with more than 50% of these being SMEs.

    267.A study requested by the IMCO committee of the European Parliament concludes that interventions aiming at increasing the contestability of the digital sector would have a significant positive and growing contribution to achieve all of the potential benefits of a Digital Single Market, also resulting in lower prices and greater consumer choice, productivity gains and innovation. 218

    268.Christensen et al (2018) estimated, using the RHOMOLO model 219 , that implementing the third pillar of the Investment Plan for Europe, including efficiency gains from the Digital Single Market, would contribute to a 1.5% increase in GDP per year until 2030 and create between 1 and 1.4 million jobs. 220 In particular, the impact of a more efficient Digital Single Market ranges from 0.44 to 0.82% changes in GDP and between 307 and 561 thousand additional full-time equivalents (‘FTEs’).

    269.As explained in Section 2, national legislations have started appearing or are under consideration in different Member States, which drives fragmentation of the Digital Single Market in the platform space. One of the main objectives of this initiative is therefore precisely to prevent the fragmentation of digital markets. Option 1 would already allow some quick alignment of platform-related rules across the EU through horizontal measures by relying on automatic quantitative criteria to identify gatekeepers and implementing immediately all obligations. However, given its static nature, it would leave scope for some market fragmentation. Under sub-option 1-A, the obligations would be applied to a smaller number of gatekeepers. This could create some fragmentation resulting from the different treatment given to the largest (and thus captured by the quantitative designation process) gatekeepers, as platforms exhibiting similar features and characteristics as the largest gatekeepers would be subject to no obligations. Under sub-option 1-B, this possible difference in treatment between gatekeepers is less likely but there is a risk that some platforms which do not exhibit gatekeepers features and characteristics are forced to comply with the obligations, reducing their ability to compete with the gatekeepers.

    270.Option 2 would capture a broad scope of unfair practices (via a dynamic updating mechanism) and gatekeepers (including those platforms falling only under the qualitative criteria as well as those platforms that are expected to enjoy a gatekeeper position in the near future). Gatekeepers would thus be treated in a harmonised way across the EU. It would still imply some delay in enforcing the obligations for gatekeepers designated on the basis of the qualitative criteria. Under sub-option 2-A, and as compared to sub-option 2-B, these delays would be more predominant, as a higher number of gatekeepers would have to be designated via a market investigation. In any case, the possible fragmentation associated to the unequal treatment of gatekeepers discussed under Option 1 would be less likely and only occur temporarily.

    271.Option 3 would, on the one hand, add to Option 2 by tackling a broader scope of unfair practices by gatekeepers in additional digital services, thus allowing full harmonisation across the EU. On the other hand, this option could result in a staggered implementation of the measures because of the delays associated to the need for the regulator (i) to conduct market investigations to designate every gatekeeper and (ii) to engage in dialogue with each gatekeeper for the implementation of all obligations.

    Growth and productivity

    272.The platform economy contributes heavily to the European economy as revealed by its size and is expected to continue to grow steadily. The digital economy was estimated to account for between 4.5% to 15.5% of global GDP in 2019, depending on the definition. 221 Traffic share is one of the most important proxies of the sector. The top 50 online platforms represent 60% of the traffic share in Europe reaching revenues for about EUR 276 billion in 2018 and employing almost 600 000 people.

    273.The European market of online platforms makes a significant contribution to GDP and the European economy as a whole. Revenues of the sector in Germany for instance reached EUR 33 billion in 2015. Cross-border ecommerce in Europe was worth EUR 143 billion in 2019 (without travel), and 59% of this market, EUR 84 billion, is generated by online marketplaces. Marketplaces with European capital represent 11% of the market, an increase of 17% compared to one year before. 222  

    274.As already explained in the recent P2B Regulation impact assessment, trade intermediated through online platforms is expected to follow an upward trend as most consumers opt for platforms when purchasing goods and services online. 223 B2C e-commerce turnover was growing at an average pace of 13% between 2014 and 2019 with turnover forecasted to hit EUR 621 billion in 2019 and is set to be worth EUR 717 billion in 2020. This sector is expected to increase in value by around 14% per year. 224 In addition, the COVID-19 crisis accelerated the shift to online retail at an unprecedented pace pointing to the importance of the online platform economy. Usage of digital devices has increased significantly during the COVID pandemic, which is likely to increase the relative importance of online platforms compared with the off-line world. Specifically, following the lockdown, one global survey found that consumers spent more time on social media and mobile applications (by 47% and 36% respectively). 225

    275.More generally, several empirical studies confirm that more competition on markets results in higher productivity in affected industries, which translates into economic growth. 226 Other studies also confirm the positive effects of competition on the productive efficiency of companies due to (i) ‘between-firms’ effect, by which better companies succeed while the worst ones fail and leave the market, and (ii) a ‘within-firm’ effect by which companies in competitive environments are better managed. 227

    276.Addressing gatekeepers’ unfair business practices would have a positive impact on the online platform economy in general. The envisaged measures would limit the chilling effects unfair conduct has on sales. Since gatekeepers are such an important channel to reach markets and consumers, business users argue that unfair practices (e.g. pretended privacy considerations, limitation to data access, etc.) would lead to up to 15% loss in their sales. 228 Businesses, especially smaller ones, would be more confident in engaging with gatekeepers if the latter (are obliged to) comply with clear fairness rules. From that perspective, a regulatory action would be expected to result not only in more sales through smaller platform but also to have a positive impact on market growth.

    277.Importantly, competitive entrants contribute to growth in the digital sphere; the obligations considered – e.g. data access or interoperability - can allow entrants to grow and compete effectively. While digital market features exacerbating the problem drivers (such as e.g. network effects) cannot be changed, the rewarding effect on gatekeepers’ further expansion - when that expansion is due to unfair business conduct - can be attenuated by the measures proposed. The measures would thus create fairer and more equitable conditions for all market players, allowing them to take greater advantage of the growth potential of the platform economy.

    278.All three options considered would reinforce trust in the platform business environment. Option 1 would do it in a quick but relatively static way, and could originate some frictions in the business environment by leaving no room for any implementation dialogue. In terms of sub-options, under sub-option 1-A those frictions would only affect the largest gatekeepers, while under sub-option 1-B, there could be a larger number of gatekeepers impacted. Option 2 would be less immediate but particularly effective in that it foresees an adaptable framework, based both on a clear set of immediately applicable obligations and a flexible list of obligations subject to an assessment of the applicability of the conducts to the specific case. It would also allow tackling practices in markets where there is a risk of tipping, and contribute to a more competitive platform ecosystem. Option 3 would additionally allow a dynamic updating of the list of core platform services, thus tackling a potentially larger set of digital services than Options 1 and 2. However, it would have the drawback of delaying (i) the implementation of obligations which could be made immediately aplicable under Option 2 and (ii) the designation of the largest gatekeepers that could be quickly identified on basis of quantitative criteria under both Options 1 and 2. By giving too much discretionarity power to the regulator, Option 3 could also give rise to a risk of a lower level of legal certainty, which would impact negatively the business environment.

    Competition and Innovation

    279.Weak market contestability and lack of competition - driven by gatekeepers’ strong bargaining power and market features leading to entry barriers - are among the problems identified in this Impact Assessment. There are strong links between patterns of innovation and competition. When businesses compete more fairly on their merits, this incentivises them to innovate and offer a better range of higher quality products and services that meet consumers' expectations. Greater competition also drives efficiency in processes, technology and service. According to Federico et al. (2019), a significant amount of innovation is driven by disruptive firms. By making its offer to customers attractive in a new way, a disruptive firm can destroy a great deal of incumbent profit while creating a large amount of consumer surplus. Competition enforcement precisely seeks to protect the competitive process by which disruptive firms challenge the status quo. 229 Several empirical studies confirm that an increase in competition leads to a significant increase in R&D investment by neck-and-neck firms. 230  Conversely, the view according to which market concentration or large firm size is associated with a higher level of innovation is not supported by empirical evidence. 231 Shapiro (2012) highlights the considerable empirical evidence that greater competition spurs innovation. 232

    280.Innovation patterns in the online platform economy are characterised by the following trends. On the one hand, online platforms drive innovation, driven by a competitive strategy. On the other hand, network effects drive higher concentration which may hinder innovation because it remains concentrated among a reduced number of players. At the same time, gatekeepers – due to their impact on the entire ecosystem - are able to set innovation trends for their sector and even beyond (i.e. to non-platform companies). This has the double effect of spreading gatekeepers’ innovative solutions to smaller players but could also limit the emergence of other types of innovation.

    281.Although the online platform sector invests heavily in innovation, smaller companies that depend on gatekeepers are discouraged from innovating so as not to compete with the gatekeeper. 233 Preventing patents or pre-emptive activities, for instance, is one way to gain monopoly power and to increase barriers to entry. If this pattern is dominant, the pace of innovation in the long run slows down. 234 Acquisition of startups is another way for gatekeepers to cement their market power. 235 While acquisitions may have a positive effect for entrepreneurship and innovation, in the long-run they may result in higher market concentration and insufficient diffusion of innovation. 236

    282.The evidence shows the concentration of R&D investment among few dominant firms, and with a sustained trend. The trends in the investment in R&D depicted in our Impact Assessment study suggest a cluster of high volumes of investment among big five companies; and a widening gap across time between large and small companies. The study shows that financial resources that could be invested in R&D are diverted to mergers and acquisitions, which results in higher market concentration instead of increase in the quality and quantity of products and services for consumers. The pattern of innovation dedicated to competing 'for the market' has a detrimental effect on consumer choice and surplus. 237  

    283.Moreover, market concentration results in accumulation of cash-flow that is available for R&D investment and innovation or mergers and acquisitions. The Impact Assessment study illustrates the concentration of liquidity among the top five companies, each of them ranging between 10% up to 30%, while the remaining 17 companies are on average below 1%. Five companies accumulate 90% of total free cash-flow that could be distributed among all 22 companies. This suggests that smaller companies may face some financial constraints, failing to attract venture capital to finance R&D projects, while large firms have enough own funds to embark on innovation.

    284.Furthermore, Carayannis et al (2014) shows that innovation and productivity are important drivers for competitiveness. 238 Autor et al (2020) and Decker et al (2018) show that a growing productivity gap between very big firms and the rest may result in lower business dynamism and lower productivity growth. 239  A more efficient Digital Single Market with the right incentives to innovate should contribute to a more competitive EU digital economy. The measures under consideration are the most effective in increasing market contestability and can be expected to contribute to lower prices for business users due to increased competitive pressure. For instance, promoting switching through e.g. rules against the misuse of data, self-preferencing, or lack of inter-operability can enhance competition and contribute to dynamic patterns of innovation.

    285.Option 1 is expected to have a positive and quick impact on overall innovation and competitiveness since it would immediately create a fairer and more balanced business environment for business users and platforms; the largest gatekeepers’ compliance costs may decrease to certain extent gatekeepers’ innovation ability but given those gatekeepers’ financing capabilities, the regulation would not substantially affect their innovation capacity. The lack of flexibility of Option 1 in relation to the implementation of all the obligations could however have a negative impact on the innovation efforts of those companies. Depending on the sub-option considered, the potential effects would be different: in case of sub-option 1-A, this lack of flexibility would affect only very large platforms (which are nonetheless the ones with the highest financial capabilities), while under sub-option 1-B, it would affect many more platforms (including some that may not exhibit the features and characteristics of a gatekeeper).

    286.Options 2 and 3 would in principle affect more platforms, including those that are expected to enjoy an entrenched gatekeeper position in the near future. In theory this could have a direct negative impact on the innovation incentives of some smaller gatekeepers. However, by more broadly and flexibly addressing the issues encountered by gatekeepers’ business users and creating more competitiveness opportunities, these two options would allow the creation of a healthier business environment for other platforms contributing to restoring and/or installing competitive dynamics in the platform economy. Alternative platforms are currently facing a number of challenges e.g. for developing compelling offers (lack of data and consumers due to strong network effects), for accessing venture capital for competing services, portability, risk of leverage, etc. Also, business users (e.g. e-commerce merchants, service providers and application developers) face issues such as dependency, unfair contractual relations, unequal distribution of revenues/profits and exclusion. In light of this, the expectations for Options 2 and 3 are to spur overall technological innovation in the digital markets (concentrated so far within a limited number of gatekeepers) to other market players, thus creating more competition and innovation to the ultimate benefit of consumers. Both options are in this respect estimated to yield direct benefits of many billions of euros annually, in addition to improved innovation levels and entrepreneurship, which are complex to quantify in precise terms but likely equally if not more important in size and impact. Option 3 could have a broader impact than Option 2 by potentially affecting companies in a larger set of digital markets, but would also give rise to a lower level of legal certainty for gatekeepers and business users as a result of the excessive discretionary powers attributed to the regulator, and thus potentially risk some of these companies’ innovation efforts.

    International trade

    287.The promotion of higher competitiveness of digital markets is of particular importance in increasing trade and investment flows. According to an United Nations Conference on Trade and Development (‘UNCTAD’) report, 240 digitalisation contributes significantly to increasing the scale, scope and speed of trade. 'Information and Communication Technologies (‘ICT’) products are already a significant part of the global trade (in 2017 they are estimated to have reached USD 530 billion, representing 10% of total global trade in services).

    288.All three options are designed in such a way as to target any gatekeeper platform in an objective and non-discriminatory manner (see Section 5.2.1). The objective scoping criteria applicable to all options target EU presence and do not take into account the location of the corporate headquarters of the company in question. In doing so, the options would be future-proof and consistent with the EU’s international obligations, including non-discrimination under the General Agreement on Tariffs and Trade and the World Trade Organisation. The EU is a major market which will remain open for business but competition in the EU should remain fair and markets contestable.

    289.In addition, as mentioned in Section 1.1, an intense debate is on-going about the need to regulate gatekeepers in most jurisdictions around the world such as Japan 241 , Australia 242 , US 243 or China pointing to the global consensus on the need to complement competition policy with ex ante measures (see also Annex 5.3 to the Impact Assessment). This debate has included deep reflection processes in most of the EU’s main trading partners, many of which are considering options similar to the ones presented in this Impact Assessment.

    290.For example in the US, antitrust hearings of Amazon, Apple, Facebook, and Google have taken place in the US Congress House of Representatives and before the Federal Trade Commission. 244 In October 2020, the House of Representatives’ Committee on the Judiciary issued a Majority Staff Report in which a broad range of significant remedies are proposed, following a detailed assessment of the effects of a number of unfair and anticompetitive practices by these platforms, in order to restore competition in digital markets. 245 These remedies notably include structural separation, line of business restrictions as well as non-discrimination rules for dominant platforms including on access and pricing.

    291.Finally, the present initiative would establish a proportionate regulatory framework promoting a fair and contestable online platform environment in the EU, one in which new platforms can emerge and scale-up, to the benefit of users around the globe, not just in the EU.

    Employment

    292.An overview by the OECD of the main literature covering the links between competition and employment confirms that market competition stimulates employment growth in the long term. 246 The aggregate effect mainly results from a positive impact on productivity growth, which increases labour demand, and through aggregate demand, given that more competition lowers prices and therefore tends to increase real wages. This generates a virtuous circle of output and demand growth in the long run. 247  

    293.In the short run, the response to increased competition can lead to an increase in unemployment, e.g. through process innovation that replaces labour intensive machinery with new machines to increase productivity at the cost of labour. However, econometric simulations of the effect of increased competition leading to redundancies in an industry demonstrate a return to a steady growth path with rising employment after two to three years. 248  

    294.According to an UNCTAD report, 249 digital transformation has strongly contributed to job creation across the G20. Between 2006 and 2016, total employment in the G20 grew by 13%, a net gain of almost 127 million jobs with highly digital-intensive sectors contributing with 43% of these net job gains. Jobs in the ICT sector comprised 11.8% of total employment of the G20 countries, in 2017. The Covid-19 crisis called for the adoption of new labour regulations favouring teleworking regimes. Digital services are of extreme importance as tools enabling teleworking regimes. Therefore, making these services more accessible is even more important today for a functional labour market.

    295.One of the studies carried out by the Commission 250 suggests the possible creation of thousands of additional jobs in case of regulatory corrective measures (see Annex 3 to the Impact Assessment). Even under the assumption that no additional jobs would be created, given the millions of people employed in the sector and the millions of SMEs depending on online platforms to reach their customers, taking adequate measures to ensure the proper functioning of the platform economy would safeguard these millions of jobs.

    Businesses 251

    Gatekeepers

    296.First and foremost, it should be stressed that there is broad consensus across various firms of different sizes and business models in the tech community that there is a need for rules addressing the detrimental impact of gatekeeper practices and conduct. 252 Most respondents to the OPC, including businesses and business associations, consumer associations and NGOs, also agreed with the possibility to adopt a combination of policy options to address concerns in digital markets. 253

    297.Second, as explained in Section 5.2.1, this initiative also foresees a mechanism for the designation of gatekeepers subject to the list of obligations that would capture those that are effectively gatekeepers prone to engaging in unfair conduct and/or reducing the contestability and competition in digital markets.

    298.Third, the targeted scope of options imposing rules only on the largest platforms, or on undertakings contributing to a market failure, strongly contributes to the proportionality of any potentially resulting compliance costs.

    299.Evidence shows gatekeepers’ increasing supra-normal profits as well as their ability to (i) obtain conditions that would not be possible under normal market circumstances, and (ii) act independently from competitors, business users and consumers. For instance, the graph below shows that some of the largest companies by market capitalisation included in the S&P 500 index in November 2020 are the companies running some of the most important digital platforms. 254 Also the multiple antitrust investigations on abuse of dominance against many gatekeepers provide relevant evidence about the unfair conditions imposed on business users by those gatekeepers (see Section 5.2.2).

    Figure 3 – Market capitalisation of largest companies included in the S&P 500 index, Nov 2020

    300.Compliance costs under all three options would largely substitute for the already high costs large platforms incur for complying with divergent regulatory measures gradually put in place in different Member States. Such costs would imply some additional legal compliance officers to check company policies against the new rules; some employees to interface with the regulator and respond to requests for information. These would be higher the longer the list of obligations and the broader the digital services in scope. Compliance costs would thus be the highest under Option 3 and Option 2 as compared to Option 1 as it would include potentially more practices as a result of the updating mechanism and the need to reply to more requests for information in the context of market investigations. On the other hand, the fact that Option 1 would not allow for any dialogue for the implementation of the obligations, would give rise to additional compliance costs as compared to the other two options. In terms of the sub-options, sub-option A would result in lower total costs than sub-option B as it would affect a smaller number of gatekeepers, although the cost per gatekeeper should be the same for both sub-options.

    301.Compliance costs under all options would be miniscule as compared to the gatekeepers revenues and could be absorbed by gatekeepers with little incentive for them to pass on costs to business users (e.g. by limiting their access to the gatekeeper platform) or to consumers (see Section 6.1.8).

    302.Indirect (other than compliance) costs may be higher, as proposed measures are expected to have impact on gatekeepers’ business models and potentially reduce their supra-normal profits. The impact of such changes is difficult to quantify. While some loss of revenue for gatekeeper is expected, there are no indications that this would result in significantly higher fees and/or reduced quality for businesses and consumers. Consumers are at the core of platforms’ business strategy and, due to the relevance of indirect network effects and economies of scale, gatekeepers need to attract an important number of consumers in order to be able to (i) attract businesses (and vice versa) thus allowing online matching of offer and demand, and (ii) benefit from the virtual growth cycle characterising the platform economy.

    303.All three options would moreover not be geared towards eliminating legitimate monetisation opportunities. They would aim at eliminating unfair behaviour towards business users and other market failures, thus rather enhancing trust in the platform business model. A set of measures that contribute to a more dynamic online platform economy and more contestable markets would particularly benefit smaller competitors who would face lower barriers when entering the market. It can therefore be expected that an increased market contestability would continue to incentivise gatekeepers to bring innovative products to the market and compete for consumers and business users; this even in case gatekeepers’ business models are impacted by the regulatory measures.

    304.Fourth, all options are designed in a targeted way, taking into account the currently available experience and evidence about the impact of specific unfair practices by gatekeepers on their business users and customers as well as on the contestability of digital markets.

    305.Given that the rules only aim to prevent unfair and harmful conduct, they should not hamper market entry (even) by gatekeepers if the latter is based on fair means of competition. As far as they do not use their market positon in an abusive way their ‘first mover advantage’ could be preserved.

    SMEs

    306.SMEs would not be targeted by the list of obligations as they are very unlikely to qualify as gatekeepers. On the contrary, the adoption of rules levelling the playing field would allow SMEs (including business users competing with gatekeepers) to grow throughout the internal market.

    307.All three options foresee a comprehensive form of regulatory oversight and SMEs would benefit from a more innovative and competitive business environment incentivising them to seize the digital single market opportunities and grow (see Annex 3 to the Impact Assessment).

    308.Competitors and new entrants would benefit from the levelling of the playing field and from enhanced opportunities to scale up and compete with these gatekeepers as a result of the removal of important barriers to entry and expansion. Measures preventing unfair self-preferencing and limitations in interoperability would give them the ability to compete on the merits (e.g. develop their own distribution channels or their own ID services). Data-related rules, including data portability, which would facilitate switching and multi-homing and thereby increase potential user base would allow them to bring innovative solutions to the market. Measures promoting multi-homing and user switching would give competitors and new entrants a real chance to capture a new stream of demand, propose competitive offers and grow. The increase transparency would give them opportunities to compete more equally with the gatekeepers. Rules on MFN clauses would increase incentives for competitors to develop alternative (to the gatekeepers’ ones) distribution channels, since they could expect that better service or lower price can be awarded by the business user with better commercial conditions.

    309.Given that measures envisaged under all three options are aimed at increasing market contestability, it could be expected that they would result in more competition for business users. Business users would have more confidence in selling online, as they would be protected from unfair practices. Measures against data misuse would prevent that their data are exploited for the only benefit of the platform. Access to data generated by business users’ activity on the platform would allow them to adjust their business model to demand and better meet customers’ expectations. Business users would have the ability and incentive to choose among different platforms where to offer their service/product. Business users would have an increased possibility to multi-home and switch thus benefitting from increased choice of services and the ability to combine services according to their actual needs and interests (due to the obligation for gatekeepers not to make the use of the core service conditional upon the use of ancillary services). They could as a result benefit from lower prices for intermediation services and reduced distribution costs.

    310.Finally, access to digital markets allow SMEs to increase their productivity and reduce their costs. According to a study from OECD countries, in 2015 only 20% of SMEs engaged in sales through e-commerce, against 40% of large firms. This digital gap slows productivity growth and widens inequalities. More competitive digital markets resulting in more affordable services would allow SMEs an easier access to digital technologies. Ultimately, given that SMEs are the bulk of many national economies, a massive adoption of digital technologies by them would generate a shift of aggregate productivity and welfare. 255 Since several of the business models of the gatekeepers are extensively benefiting from network effects and thereby large number of business users or end users, it is not expected that the obligations introduced by a new framework would result in gatekeepers terminating provision of services to SMEs that are often dependent on these gatekeepers and their core platform services. Not only would this remove many of the benefits that gatekeepers enjoy due to their unique position, but could further accelerate switching by both business users and end users to alternative providers of same or similar core platform services.

    311.Option 2 and 3 as compared to Option 1 would allow the Commission to address the issues SMEs face in the dynamic digital markets in a more agile way, including issues associated to markets that risk tipping in the absence of an intervention and new unfair practices by gatekeepers. By foreseeing a dialogue between the regulator and gatekeepers for (some of) the obligations, they would allow a more flexible implementation of those obligations that would disrupt less the commercial relationship between gatekeepers and their business users. By foreseeing the possibility to resolve problems stemming from additional core platform services, Option 3 would additionally allow capturing the fast changing character of digital markets, but at the same time would give rise to a risk of a lower level of legal certainty to gatekeepers and business users as a result of the discretionary power given to the regulator. On the other hand, Option 1 could have an immediate effect on SMEs by allowing the automatic identification of all gatekeepers under scope and the immediately implementation of all obligations. In terms of sub-options, sub-option A would benefit more the business users of the largest platforms that would be captured by a high quantitative threshold, while sub-option B would benefit also business users of smaller platforms.

    Consumers

    312.Digital markets are becoming more and more relevant for consumers. According to the Digital Economy and Society Index (‘DESI’) 2020’, internet use has continued to increase year-on-year with 85% of Europeans surfing the internet at least once per week. 256 Using the internet for listening to music, playing games or watching videos is still the most common activity (81% of individuals). Reading news online is the second most popular activity (72% of individuals), followed by e-commerce (71%), bank online (66%) and social networks (65%). According to Eurostat figures, more than six out of 10 consumers from the EU28 made online purchases in 2019, the highest proportion made purchases three to five times in a period of three months and bought goods or services for a total of between EUR 100 to EUR 499. 257 Improving competition enforcement in digital markets is thus particularly relevant for the protection of European consumers.

    313.Even though the digital sector and the companies offering digital services contribute strongly to consumer surplus, the increased market concentration in digital markets does not allow consumers to enjoy the full potential of these dynamic markets. In fact, the high concentration level is detrimental for consumer surplus as it results mainly in lower choice and higher prices/costs. Although data to estimate the loss in consumer surplus is limited, there is some illustrative evidence. For example, if commission fees in large app stores were to be reduced from 30% to 15%, the average prices of apps and digital content acquired through these apps would fall, which would increase consumer surplus by up to EUR 490 million in the EU per year based on Statista data. 258  

    314.The choices for consumers are limited by lock-in effects and lack of innovative alternatives that are restricted by gatekeepers’ unfair business practices and more generally by the market failures in digital markets. In the longer run, consumers risk experiencing lower quality and/or less innovative services and/or higher prices. This initiative aims at addressing these concerns with a view to ensuring optimal and secure consumer experience online.

    315.A regime that protects EU consumers from business practices that keep the prices of goods and services artificially high would ensure that consumers have access to better quality, wider choice and innovative goods and services at affordable prices. Numerous studies confirm the benefits of competitive markets for consumers. 259 More competitive digital markets will allow consumers to multi-home among alternative platforms offering differentiated commercial propositions. In addition, some of the measures considered under the options aim at reducing the search and switching costs associated with mulit-homing, for instance by allowing portability of data, creating conditions for interoperability, increasing transparency in the market, etc.

    316.All options would indirectly contribute to safeguarding value added for consumers and to ensuring greater respect of privacy and consumer interests. 260 This would be achieved by contributing to (i) fairer competition on gatekeeper platforms (intra-platform competition) and among platforms (inter-platform competition), (ii) stronger contestability of the markets where gatekeepers are present, and (iii) better functioning of the internal market through enhanced regulatory oversight at EU level.

    317.It is also important to notice that although interventions foreseen by this initiative may require changes to the existing business models, this does not risk harming consumers. The need for gatekeepers to maintain a large user base in order to optimise indirect network effects and higher level of contestability and competition in which the assessed measures would result, will rather increase gatekeepers’ incentive to innovate and offer lower prices.

    318.It could also be argued that the rules under assessment would lead to curtailing the size of network effects and economies of scale thus reducing associated advantages for consumers. It is important to note that the objectives behind the measures considered aim at allowing also non gatekeeper platforms benefitting from such advantages. This would contribute to a competitive dynamics that benefits consumers who will be able to benefit from larger choice while gatekeepers can continue enjoying important network effects and economies of scale and scope.

    319.It should be stressed that the obligations under this initiative will neither ban specific monetisation models (such as ad-based models) nor prevent the uptake of new services by gatekeepers - they prevent them from acting unfairly in their operations and reduce competition in the markets where they are present. The obligations envisaged would also not prevent an emerging gatekeeper from enjoying network effects and economies of scale and scope, instead they would ensure that other market players can also benefit from those features and would thus be able to compete under fair terms and innovate.

    320.Even in those cases where, due to the multi-sided character of platform markets, there is a cross-subsidisation between the different sides with consumers benefitting from zero prices, additional regulation-compliance costs for gatekeepers cannot be expected to translate into ‘higher’ prices for consumers. This is so since, as explained in Section 6.6, consumers are at the core of platforms’ business strategy of indirect network effects and feedback loops. This is evidenced by the fact that in services like general search and social networks, even smaller platforms offer their services to consumers for free while obtaining their revenues via advertisers. In addition, gatekeepers would not risk losing consumers by setting prices for services which are currently free of charge. Consumers would expect zero-priced services to remain free of charge. Setting a price for gatekeepers’ services that are currently free would be perceived by consumers differently as compared to increases in already existing monetary prices (i.e. not zero). Consequently, any attempt by gatekeepers to make users pay for services that were previously offered for free would imply the risk for them of reducing the attractiveness of their services and of encouraging users to switch to other platforms continuing to offer their services free of charge. 261

    321.Instead of increase in consumer prices, the expected increase in market contestability and competition would increase the diversity of offers available to consumers and would reduce the prices for advertisers, which would then indirectly translate in lower prices charged by those advertisers when selling their products and services to consumers. 262 , 263 Current excessive expenditures on advertising per user are driven by high market concentration, and could be a proxy for consumer detriment. For example, in 2019 a total of EUR 55.4 billion was spent in digital ads in 21 countries of the EU (including the UK) 264 , corresponding to around EUR 110 on advertising per user per year. A more contestable and competitive market would reduce those costs significantly (e.g. a reduction of 10% in ads expenditure would already generate gains of more than EUR 5 billion per year). In addition, for many digital markets where consumers are offered services ‘free of charge’, in practice they receive the service in exchange for their attention and their data, which can then be monetised through digital advertising. In a more contestable and competitive market, it would be clear to consumers what data is collected about them and how it is used and, crucially, consumers would have more control over their data. 265

    322.Moreover, the Impact Assessment study for this initiative estimates that gatekeepers’ financial resources that could be invested in R&D are currently diverted to mergers and acquisitions, which results in higher market concentration instead of increase in the quality and quantity of consumer products and services. This pattern of innovation dedicated to competing 'for the market' has a detrimental effect on consumer choice and surplus. In addition, the positive impact on innovation stemming from higher market contestability is not limited only to diversion of money from mergers and acquisitions to R&D. Other expected indirect effects include an increase in entrepreneurship and creation of new products and solutions meeting consumers' needs rather than focused on exploiting a gatekeeping position. This may have a multiplicative effect increasing the size of the European single market, and hence, GDP and online cross-border trade. All options are estimated to allow to recover to a large extent this opportunity cost. All options would thus have a clearly positive effect on overall welfare.

    323.As shown in Section 6.6.2, all three options would lead to positive implications for business users who would benefit from reduced prices for intermediation services. This in turn would allow business users to lower prices for consumers and offer them higher quality of service. Consequently, consumers would benefit from increased choice of products and services, better tailored to their needs (since they would e.g. be able to have a direct contact with businesses), and offered by different business users possibly through a larger number of platforms. This could lead to higher search costs but consumers would still have the possibility to use the gatekeeper services, if they find it preferable to use a single platform; their choice would however not be limited to offers provided through/on the gatekeeper platform. They would also benefit from lower prices for intermediation services which would be passed down to consumers in the form of lower prices for goods and services, thus generating cost savings to consumers.

    324.Following the above, all three options would generate high benefits for consumers. Option 1 would generate immediate benefits but to a smaller extent given the risk associated the static nature of its application. Option 2 would be more flexible, thus favouring a more effective implementation of obligations and designation of gatekeepers. This would create more competitiveness opportunities. It would also allow tackling practices in markets where there is a risk of tipping as well as new unfair practices by gatekeepers, and thus generate more consumer benefits. Option 3 would on the one hand generate benefits for consumers of a broader range of services, and on the other hand, risk some innovation efforts by businesses (because of the low levels of legal certainty) with the consequent negative impact on consumers. In terms of sub-options, sub-option A would benefit more the consumers of the largest platforms and of the business users of those platforms, while sub-option B would benefit also the consumers of smaller platforms and respective business users. A possible drawback of sub-option B would be the risk of preventing platforms that are wrongly designated as gatekeepers from competing intensively with those gatekeepers, which would thus reduce their ability to innovate and launch new services in the market.

    325.Detailed overview of the implications for gatekeepers, competitors, business users and consumers is presented in Annex 3 to the Impact Assessment.

    Regulatory Authorities

    326.All three options imply enforcement costs to be essentially incurred by the EU Commission, with some administrative burden for national authorities. This includes the costs with preparing and processing information requests as well as the preparation of guidelines, designation of gatekeepers, enforcement of the general obligations, including the specification of some of the obligations. Annex 3 to the Impact Assessment provides a qualitative and quantitative overview of these costs. As compared to Option 1, Option 2 and Option 3 imply additional resource-related costs both for the Commission and for national authorities. However, it can be objectively considered that this higher administrative burden would be largely outbalanced by the benefits of reducing the impact of practices which severely undermine the trading conditions for millions of business users and further entrench gatekeepers’ incontestable positions.

    7.HOW DO THE OPTIONS COMPARE?

    327.This section assesses the effectiveness, efficiency, coherence, proportionality and subsidiarity of the different policy options as compared to the baseline scenario and among each other.

    Effectiveness

    328.Three parameters appear essential for assessing the effectiveness of each option: legal certainty, speed of intervention and flexibility of the approach. The relative importance given to each of these three parameters is specified in the following three paragraphs.

    329.Speed of intervention is essential in digital markets where, due to the market specificities explained in Section 2, the larger the gatekeeper the greater and quasi-automatic its capacity to gain power and strengthen its position, further reinforcing its ability to engage in unfair practices. In particular, in the digital sector it is common to observe markets tipping quickly in favour of one gatekeeper once that gatekeeper has obtained a certain advantage over rivals. The unfair practices identified is this Impact assessment are harmful and action is required in the most efficient manner possible. They affect negatively SME business users and small scale platforms, which may force the latter to exit the market, thus further weakening market contestability and strengthening legal fragmentation issues. Such negative effects of the problems identified could not be easily reversed and should therefore be addressed in a timely manner preventing their further proliferation and irreversibility. Speed is therefore given important consideration in the options’ comparison in this Section and in Table 4 .

    330.Legal certainty is important for meeting expectations of all economic actors interacting in a given ecosystem. Together with regulatory predictability, legal certainty guarantees business trust and creates the right incentives for investment and innovation, for both gatekeepers and SMEs. It is therefore given important weight in Table 4 and in the overall assessment in this section.

    331.Flexibility is an important criterion with a view to guaranteeing that a system is future-proof and agile. In the present case flexibility could be introduced at several levels of the options, namely, in relation to the designation of gatekeepers, the implementation of the obligations, the update of the list of obligations and of the list of core platform services. Some of these elements are more relevant than others and an excess of flexibility may not always be desirable as it creates negative externalities on other parameters. In fact, the optimal level of flexibility needs to strike the right balance between a regulation being agile and providing for a solid regulatory intervention setting a stable and clear framework. Flexibility has therefore been given less weight when comparing options’ effectiveness.

    332.By including a set of obligations on gatekeepers’ behaviour the three options would contribute to both objectives of addressing unfair practices by gatekeepers and facilitating further contestability of the platform markets concerned. Measures related to data portability as well as interoperability and self-preferencing are particularly important for the objective of addressing unfair practices by gatekeepers. Such measures allow business users to benefit from fairer business conditions in relation to gatekeepers’ core platform services, thus also contributing to a level playing field. Gatekeepers would no longer be able to benefit from preferential treatment that derives from unfair behaviour, e.g. in terms of display/ranking or conditions of data access, portability, interoperability, which would also facilitate further contestability of the platform markets concerned. Rules set for anti-steering, side-loading as well as obligations concerning other unfair practices address the issue of unfair platform-to-business practices in specific contexts (i.e. economic dependence of one of the parties; imbalance in commercial relationship) thus contributing to more balanced P2B relations and acting on the imbalance of bargaining power (i.e. one of the drivers behind the fairness concerns). At the same time, such rules allow to address the weak contestability on digital markets since they would contribute to business users’ and consumers’ ability to use alternative services to those offered, or in certain circumstances even imposed by gatekeepers, thus attenuating both entry barriers (driving weak market contestability), and consequently businesses’ economic dependence on gatekeepers (the other driver behind unfairness). Consumer choice – which is closely related to competition and hence to market contestability - could also increase directly, notably requiring consumer portability provisions for gatekeepers, making it easier for consumers to switch (thus acting on the entry barriers driving the weak contestability problem). Indirectly, consumer benefits would also derive from lower prices for gatekeepers, although rules would need to be designed to avoid adverse effects on security and privacy, for instance.

    333.Option 1 would contribute to the objectives of addressing gatekeepers’ unfair conduct and ensuring contestable and competitive digital markets, by allowing to tackle those gatekeepers’ unfair practices on the basis of a list of obligations. Given that all obligations under this option would be immediately applicable, they would have direct quick effects.

    334.However, given that the designation of gatekeepers is based only on quantitative criteria, it could lead to type I errors (false positives) or type II errors (false negatives) depending on whether the threshold would be set at a low level (as per sub-option 1-B) or at a high level (as per sub-option 1-A). In fact, in the case of sub-option 1-A there would be the risk of failing to identify gatekeepers that, similar to the designated gatekeepers, may equally have an important internal market impact, operate an important gateway to end users and have an entrenched position, but which are relatively smaller. This would result in the unfair practices by those gatekeepers not being tackled. In the case of sub-option 1-B, there would be the risk of designating an excessive number of platforms as gatekeepers, including those that are not engaging in unfair practices but that, since they would be above the thresholds, would have to comply with the obligations.

    335.Option 1 would fail to include in scope emerging gatekeepers whose position is likely to become entrenched in the near future as well as any new unfair practices by gatekeepers not part of the initial list of obligations. It would accordingly be a very static approach to deal with the dynamics of digital markets and, where potentially equally harmful gatekeeper behaviour would not be adequately addressed.

    336.With respect to the objective of enhanced coherence and legal certainty, Option 1 allows an immediate aligning of platform-related rules across the EU through horizontal measures by relying on automatic quantitative criteria to identify gatekeepers and implementing immediately all obligations. These rules would preclude Member States from legislating in the areas covered by the new framework. Option 1 provides for effective and coherent EU-wide oversight through the establishment of a single regulator at the EU level, in cooperation with a network of national authorities. By basing the gatekeepers’ designation process on pure quantitative elements, this would provide a high degree of legal certainty through a clear signaling effect to the market. However, given its static nature, Option 1 would leave margin for some market fragmentation to remain. Under sub-option 1-A, only very large gatekeepers would be in scope. This could create some fragmentation resulting from the different treatment given to the largest gatekeepers (and thus captured by the identification process), as platforms exhibiting similar features and characteristics as the largest gatekeepers would not be subject to the obligations. Under sub-option 1-B, this possible difference in treatment between gatekeepers is less likely but there is a risk that some platforms which do not exhibit gatekeepers features and characteristics are forced to comply with the obligations, reducing their ability to compete with the real gatekeepers.

    337.Option 2 would be effective in curtaining a wider range of unfair practices and increasing contestability in digital markets in a flexible way. For some of the practices a dialogue between the competent regulatory body and the gatekeepers concerned may be required to ensure that measures considered or implemented by the gatekeepers better achieve its goals. By introducing the possibility for such a dialogue, Option 2 can be expected to be more effective in addressing unfair practices hampering market contestability and competition. It will, at the same time, be proportionate for the gatekeepers concerned, since they would have certain margin of appreciation in implementing measures that effectively ensure compliance with the identified obligations. Therefore, it would be legitimate to expect that it would both create the right innovation incentives across the market, and contribute to increased consumer choice in terms of number of platforms proposing innovative and privacy-friendly services. By comprising a dynamic updating mechanism, Option 2 would also allow tackling new unfair practices. It would also allow tackling market failures related to gatekeepers that are expected to have an entrenched position in the near future. Option 2 is therefore also more effective in fulfilling the specific objective of addressing weak market contestability and competition than Option 1.

    338.A drawback of Option 2 as compared to Option 1 is the fact that, by being based on market investigations to designate additional gatekeepers and by foreseeing a dialogue between the competent regulatory body and gatekeepers for some of the obligations, it could generate some delays in the implementation of those obligations and for those gatekeepers.

    339.Similarly to Option 1, the sub-options in Option 2 would be subject to a trade-off. In the case of sub-option 2-A, the impact of the intervention would be less immediate given that a higher number of gatekeepers would have to be designated via a market investigation, and thus unfair practices by those gatekeepers would not be tackled for a period of time. In the case of sub-option 2-B, more gatekeepers would be automatically designated by means of the quantitative criteria but this would entail the risk of designating platforms that do not qualify as gatekeepers and that as a result of that type I error, these platforms would have a reduced capacity to compete with the gatekeepers.

    340.As regards the objective of enhanced coherence and legal certainty, similarly to Option 1, Option 2 provides for effective and coherent EU-wide oversight through the establishment of a single regulator at the EU level, in cooperation with a network of national authorities (same regulatory design as under Option 1), thus contributing to legal certainty. The flexibility of tackling new unfair practices by gatekeepers and including gatekeepers that are expected to have an entrenched position in the near future could be expected to further reduce regulatory interventions at national level, thus contributing to the extent possible to preserving the digital single market in the online platform space. Given the combination of quantitative and qualitative criteria, type II errors would be less likely and thus gatekeepers would be treated in a more harmonised way across the EU than under Option 1. The possibility for the provider of core platform services to present, in exceptional circumstances, serious and substantiated arguments to demonstrate that it does not fulfil the objective requirements for a gatekeeper and should therefore not be designated directly based on the application of quantitative thresholds, but only subject to a further investigation, allows to address most of the concerns related to the possible lack of reliability and robustness of the quantitative thresholds set.

    341.Option 2 would however imply some delay in enforcing the obligations for gatekeepers designated on the basis of the qualitative criteria, which could result in a temporary fragmentation of the market. Under Sub-option 2-A, these delays would be more relevant, as a higher number of gatekeepers would have to be designated via a market investigation, but the possible fragmentation associated to the unequal treatment given to gatekeepers discussed under Option 1 would be less likely.

    342.Like under Option 1, to integrate national expertise in the platform economy, this option would also envisage that the Commission consults a ‘network of regulators’ before taking decisions. Option 2 would thus contribute to both addressing legal uncertainty (the problem identified) and to reducing fragmentation of regulatory approaches across the EU (the driver) in relation to a defined list of practices within a closed list of core services.

    343.Option 3 provides for a fully flexible approach in achieving the specific objective of addressing gatekeepers’ unfair conduct and ensuring contestability of digital markets, including the possibility to include in scope new unfair practices by gatekeepers (as in Option 2) and additional digital services. 266 The inclusion of new digital services in scope is however not very likely as the core platform services listed in Section 5.2.1 are precisely the ones for which there is strong evidence that market failures are present and need to be addressed. In addition, it would always be possible under Options 1 and 2 to also include new digital services in scope during the review of the Regulation, which could possibly take place every three years.

    344.Option 3 does however have the drawback of being slow in effectively addressing the problems compared to Options 1 and 2. In fact, given the need for a market investigation to designate all gatekeepers, and the possibility for a dialogue with the latter to determine the implementation of all the obligations, Option 3 would not allow an immediate response to the most pressing instances of gatekeeper related market failures. Given the importance of a timely regulatory response to the issues identified (explained above), this is an important drawback.

    345.Regarding the objective of enhancing coherence and legal certainty, Option 3 provides for an effective and coherent EU-wide oversight through the establishment of a single regulator at the EU level, in cooperation with a network of national authorities. Option 3 would, on the one hand, add to Option 2 by tackling a broader scope of unfair practices by gatekeepers in additional digital services, thus reducing even more the need for intervention at national level and contributing to a more homogeneous approach to the issues at stake. However, the regulatory powers would be less circumscribed and may create a lower level of legal certainty. In addition, the need for the regulator (i) to conduct market investigations to designate all gatekeepers, and (ii) to engage in a dialogue with each gatekeeper for the implementation of all obligations would result in a staggered implementation of the measures. This may in turn create chilling effect and counteract the positive impact of the effectively addressed unfair behaviour on market contestability, which could also negatively affect innovation and consumer choice. Consequently, Option 3 would have a mitigated impact on the objective of ensuring market contestability and competition.

    346.In light of the above, Option 2 – offering a reasonable trade-off between speed, legal certainty and flexibility, appears as meeting the overall general objective of improving the internal market functioning the most effectively since it allows for a more adaptive solution, which is a more appropriate way to tackle issues arising in the fast-changing platform environment.

    347.It is essential to stress that the effectiveness of the measures (including their underlying remedies) under all options depend on their enforcement. Experience and evidence show that sanctions such as fines would not be sufficient to incentivise the gatekeepers targeted in this Impact Assessment to stop unfair practices they are engaging in. This is an important consideration to make when assessing the proportionality of the regulatory rules considered.

    348.Table 5 below compares the effectiveness of the three options in achieving each specific objective pursued taking into consideration their respective speed of intervention, flexibility and legal certainty.

    Table 4: Comparison of options in terms of effectiveness

    Option

    Address unfair practices and market contestability

    Ensure increased consistency across the EU

    Legal certainty

    Flexibility

    Speed

    1-A

    Given that only a limited (5 to 7) number of providers of core platform services would be designated, there would be the risk of failing to identify some gatekeepers (type II error).

    Allows to quickly tackle the largest gatekeepers’ unfair practices on the basis of a list of obligations, leading to positive impact on market contestability, innovation and consumer choice.

    It would not tackle market failures in tipping markets and new unfair practices.

    Some regulatory fragmentation would persist since only a limited number of gatekeepers would be designated, and the lack of flexibility would likely lead to more national regulatory initiatives.

    This would create legal certainty by basing the designation of gatekeepers on pure quantitative criteria.

    The main drawback relates to the immediate application all obligations which would preclude the possibility to exchange with the regulator to specify the application of the certain obligations.

    No flexibility, as this is a fully static option given that (i) all gatekeepers are designated via quantitative criteria, (ii) all obligations are immediately applied, with no dialogue possible, and (iii) no new practices could be added to the list of obligations.

    Immediate effect on the market.

    1-B

    Compared to Option 1-A, a higher number of providers of core platform services (10 to 15) would be designated, with a risk of capturing those that are not engaging in unfair practices but that, since they would be above the thresholds, would have to comply with the gatekeeper obligations (type I error). This could risk some innovation efforts by those platforms.

    Addresses a significant share of emerging regulatory fragmentation as the same obligations are applied to a larger set of gatekeepers.

    2-A

    It automatically captures a number of gatekeepers falling under the quantitative threshold (5-7) as well as a number of gatekeepers designated on the basis of the qualitative criteria.

    Allows to tackle the gatekeepers’ unfair practices on the basis of a list of obligations, leading to positive impact on market contestability, innovation and consumer choice.

    It would additionally allow to tackle market failures in tipping markets and those associated to new forms of unfair conduct.

    Addresses most of the regulatory fragmentation problems as obligations are applied to both gatekeepers designated on the basis of quantitative and qualitative criteria, including emerging gatekeepers.

    It also allows covering new practices.

    Some temporary fragmentation remains as a result of delays in enforcing the obligations for gatekeepers designated on the basis of the qualitative criteria.

    Creates legal certainty for gatekeepers designated both on the basis of quantitative and qualitative criteria.

    Possibility of a dialogue for some of the obligations would also help to create legal certainty.

    Flexible option, by complementing static quantitative designation criteria and immediate implementation of some obligations with flexible elements in the qualitative designation and the implementation dialogue.

    Flexibility also resulting from the possibility to add new unfair practices by gatekeepers to the list of obligations.

    Medium speed as it still requires market investigations to designate some of the gatekeepers and a dialogue to implement some of the obligations.

    2-B

    Similarly to Option 2-A, addresses unfair behaviour, including new unfair practices by gatekeepers and tipping markets, leading to positive impact on market contestability, innovation and consumer choice.

    However, a higher (10 to 15) number of providers of core platform services would be automatically designated, with a risk of including those that are not engaging in unfair practices but that, since they would be above the thresholds, would have to comply with the obligations (type I error).

    Similar to Option 2-A, it addresses most of the regulatory fragmentation.

    It is less likely to result in delays in enforcing the obligations than Option 2-A.

    Creates legal certainty for gatekeepers designated both on the basis of quantitative and qualitative criteria.

    Possibility of a dialogue for some of the obligations would also help to create legal certainty.

    Flexible option, although, to a less extent than Option 2-A given that most of the gatekeepers would be designated via the static quantitative criteria.

    Faster than Option 2-A since most gatekeepers would be designated via the quantitative criteria.

    3

    Addresses unfair behaviour and weak contestability for a relatively high number of gatekeepers designated via qualitative criteria, including emerging gatekeepers.

    It would additionally allow to tackle market failures in tipping markets and those associated to new forms of unfair conduct and services other those in scope.

    Implies several market investigations that could not be carried out in parallel, thus leading to a fragmented/staggered approach to the market.

    Low legal certainty given the discretionary power of the regulator in relation to the designation of gatekeepers, implementation of obligations as well as the possibility to add new unfair practices and services to the scope of regulation.

    Fully flexible option both in terms of designating gatekeepers and the possibility of a dialogue for the tailored implementation of the obligations.

    Full flexibility also resulting from the fact that the regulator could add new unfair practices and additional digital services to the scope of the regulation.

    Very slow impact on the market given the need to conduct a market investigation to designate all gatekeepers and the possibility of a dialogue for all obligations.

    Efficiency

    349.The efficiency comparison is based on a benefit estimate for the preferred option and on a cost comparison between the options (both detailed in Annex 3 to the Impact Assessment). A robust quantitative cost-benefit comparison between the different options proves however difficult for the initiative under consideration given the absence of reliable data, and thus estimates of costs and benefits are only provided for the preferred option, which serves as a reference for the magnitude of the remaining options. The approach taken for estimating costs and benefits under each of the options is as follows: assumptions in relation to costs are overestimated in order to ensure that costs stemming from the measures assessed are not minimized; assumptions in relation to benefits are conservative to guarantee that benefits are not overestimated.

    350.All options imply (a) regulatory costs at EU and national levels, i.e. for the Commission as the competent regulatory body at EU level to ensure implementation, supervision and information gathering, and for the network of national regulators to respond to eventual consultations from the Commission; (b) compliance costs for gatekeepers to deal with new rules and to respond to the regulator’s requests for information; and (c) (minimal) costs for the business users and platforms not subject to the obligations but which may at times need to respond to the regulator’s requests for information. 267

    351.It needs to be noted that while all obligations would be legally applicable to all designated gatekeepers, not all obligations would be relevant for each gatekeeper, since not every gatekeeper would be engaging in all unfair practices targeted by the initiative. If a gatekeeper were engaging in one or more of the unfair practices, this would require changes in its behaviour but would not necessarily translate in direct costs. Compliance costs have been estimated at EUR 1.41 million per year and per platform. This regulatory burden should be weighed in light of the economic power of gatekeepers in scope and against the fact that they would have already existing internal services to comply with other pieces of EU legislation (e.g. EU Merger Regulation; Consumer protection cooperation (‘CPC’) Regulation). This possible synergy in terms of compliance would further reduce the impact of additional costs which is marginal as compared to the enormous revenues earned by gatekeepers. In addition, the regulatory dialogue foreseen under Options 2 and 3 for obligations requiring further specification would reduce the burden on gatekeepers since it would allow tailoring the specific obligation to the particular situation of the gatekeeper concerned, which can be expected to reduce the overall compliance cost. Benefits stemming from the initiative would amount to billions (see following paragraphs) and lead to greater innovation potential amongst smaller businesses as well as improved quality of service, with associated increases in consumer welfare (as specified below).

    352.Under Option 1, five to seven platforms would be covered under sub-option 1-A and 10 to 15 platforms under sub-option 1-B, which implies an overall annual cost for platforms’ compliance with rules ranging between EUR 9.87 million and EUR 21.15 million. This calculation is based on the estimate of the compliance costs per platform, i.e. EUR 1.41 million per year. Based on benchmarks of similar practices within the Commission, networks and national authorities, enforcement cost for the Commission can be estimated at between EUR 6.4 million (under sub-option 1-A) and EUR 10.5 million (under sub-option 1-B) while for national authorities, these are estimated at EUR 4.3 million per year when summing the costs of all 27 Member States.

    353.The estimates on the number of platforms under Option 2 are based on screening of the quantitative criteria and an assumption on the number of additional gatekeepers designated via a market investigation. In relation to the latter, this is very difficult to estimate upfront given that only after a market investigation it would be possible to determine whether a given provider of core platform services meets the criteria. In this context, and for the purpose of these calculations, it is assumed that Option 2 would cover up to a maximum of between 15 gatekeepers (sub-option 1-A) and 20 gatekeepers (sub-option 1-B). 268 This implies an overall compliance cost ranging between EUR 21.15 million and EUR 28.2 million per year. The administrative costs for the EU Commission are estimated at EUR 16.7 million per year. Costs for national authorities (for all 27 Member States) are estimated at EUR 6 million per year.

    354.While a precise assessment of the number of gatekeepers targeted under Option 3 would be misleading (since their designation would be based on qualitative criteria only), it is assumed to be 25 (as an upper bond) for the purpose of quantification and options’ comparison. The number of gatekeepers designated under Option 3 could theoretically be higher than the number of gatekeepers designated under Option 2, given that it could also include gatekeepers active in core platform services that could be added after a market investigation. The above upper bond assumption has been made in order to provide a cost estimate in the worst case scenario, i.e. reflecting the highest possible costs that the measures could generate under this option. Under this assumption, Option 3 would imply a compliance cost per year for gatekeepers of around EUR 35.25 million. The administrative costs for the EU Commission are estimated at EUR 18.2 million per year, and those for national authorities (for all 27 Member States), at EUR 6 million per year.

    355.In terms of benefits, while a quantified comparison of the different options 269 proves difficult to establish, it is an objective qualitative assessment to consider that the impact of putting in place an effective and proportionate regulation addressing dysfunctions in the platform economy in an effective and flexible way (as foreseen under Option 2) would lead to a different profit distribution, i.e. with a greater societal benefit. Concerns about excessive gatekeeper profits that could be extracted based on their grossly imbalanced bargaining power would be effectively addressed and any such profits would be distributed to business users and consumers; hence, the more appropriate (effective but also proportionate) the regulatory measures, the more optimal the re-distribution of profit.

    356.If we assume that the measures foreseen under Option 2 contribute to preserving the internal market in the platform space - thus allowing cross-border trade projections by 2025 to be maintained - this would lead to EUR 92.8 billion benefits. 270 The benefits can be expected to lead to greater innovation potential amongst smaller businesses as well as improved quality of service, with associated increases in consumer welfare. Assuming that interventions foreseen would reduce competitive asymmetries between gatekeepers and other platforms, a consumer surplus of the preferred option could be estimated to increase by EUR 13 billion, i.e. around 6% increase as compared to the baseline. 271 Impact on economic growth is estimated to range between EUR 12 billion and EUR 23 billion. 272 Benefits would be similar under sub-options 2-A and 2-B, with the former having the advantage of not incurring in a type II error, and the latter having the advantage of implementing the obligations on gatekeepers quicker than if they would only be captured after a market investigation under sub-option 2-A.

    357.The benefits under Option 1 would be lower, they would cover a more limited set of market failures given that it would not be possible to add new unfair practices to the scope neither to tackle market failures in tipping markets. The benefits stemming from Option 3 could potentially be higher than under Option 2 given that additional services could be added to the scope of the obligations. However, as explained in Section 7.1, the inclusion of additional services in scope is not very likely to take place given that the core platform services listed in Section 5.2.1 are precisely the ones for which there is strong evidence that market failures are present. 273 The potentially higher benefits under Option 3 could therefore not materialise. Furthermore, Option 3 would also originate a lower level of legal certainty, which would result in a chilling effect and counteract the positive impact sought on market contestability.

    Coherence

    358.An assessment was carried out of the various policy options’ coherence with (i) the Commission’s digital strategy, (ii) the DSA and (iii) other regulatory instruments.

    Coherence with the Digital Strategy

    359.All three options are coherent with the Commission’s digital strategy in their contribution to ensuring a fair and competitive digital economy, one of the three main pillars of the policy orientation and objectives announced in the Communication Shaping Europe's digital future. They would constitute a coherent, effective and proportionate framework to address problems in the digital economy that currently cannot be tackled or cannot be tackled effectively.

    Coherence with the DSA

    360.All three options are coherent with and complementary to the proposal for the update of the e-Commerce Directive (‘ECD’) under the DSA. While the DSA is a horizontal initiative focusing on issues such as liability of online intermediaries for third party content, safety of users online or asymmetric due diligence obligations for different providers of information society services depending on the nature of the societal risks such services represent, the present options are concerned with economic imbalances, unfair business practices by gatekeepers and their negative consequences, such as weakened contestability of platform markets. To the extent that the DSA contemplates an asymmetric approach which may impose stronger due diligence obligations on very large platforms, consistency will be ensured in defining the relevant criteria, while taking into account the different objectives of the initiatives.

    Coherence with other instruments

    361.All three options align with other EU instruments, including with the EU Charter of Fundamental Rights and the European Convention on Human Rights (‘ECHR’), the GDPR, the EU’s consumer law acquis and the P2B regulation.

    362.The definitions to be used under all options are coherent with the definitions used in EU existing legislation, in particular the definitions of ‘online intermediation services’ and ‘online search engines’ used in the P2B Regulation. With their scope targeted to gatekeepers, the options complement well the horizontal obligations for all online platforms under the P2B Regulation. All options also complement existing EU competition law by addressing ex ante unfair practices by gatekeepers that either fall outside the existing EU competition rules, or cannot be addressed in the most effective manner by these rules.

    363.All options complement the data protection laws. Transparency obligations on deep consumer profiling will actually help inform GDPR enforcement, whereas mandatory opt-out for data combination across core platform services goes beyond GDPR protections. Anti-circumvention clauses will clarify that compliance with obligations in this initiative may require consent under GDPR. This is also without prejudice to data minimisation principle, including using anonymised data where possible. The introduction of the dynamic updating of core platform services and gatekeepers’ practices would be subject to full respect of the fundamental rights to fair proceedings and good administration as enshrined in the ECHR, which are binding on the EU institutions. Given that the mechanism for the imposition of remedies is administrative in nature and not criminal or quasi-criminal, the fundamental rights of the Charter enjoyed in the case of criminal proceedings would not apply. 274 However, when acting under the new framework in general and market investigation regime in particular, the Commission’s investigation powers would be counterbalanced by ensuring that undertakings involved enjoy effective fair process rights such as the right to be heard, the right to a reasoned decision and access to judicial review, including the possibility to challenge enforcement measures. These rights apply in case of administrative proceedings. 275 This design to preserve fundamental rights is also consistent with – if not superior to – the safeguards applicable similar investigation regimes elsewhere in the world. 276

    364.All options leverage existing platform regulation, without conflicting with it, while providing for an effective and proportionate enforcement mechanism that matches the need to strictly enforce the targeted obligations vis-a-vis a limited number of very large cross-border providers.

    365.Different from the P2B Regulation 277 , all options foresee EU-level enforcement of a narrow set of very precise unfair practices engaged in by a restricted group of large, cross-border gatekeepers. This EU-level enforcement mechanism is consistent with the enforcement of the P2B Regulation. Gatekeepers are likely to exist in respect of several cross-border core platform services, and a central EU-level regulator with strong investigatory powers is required both to prevent fragmented outcomes as well as to prevent circumvention of the new rules. To this end, the new EU-level regulator can leverage the transparency that each of the online intermediation services and online search engines have to provide under the P2B Regulation on practices that could precisely be illegal under the list of obligations – if engaged in by gatekeepers.

    366.All three options would - while recognising the differences – align with the experiences from the targeted and tailor-made ex ante regulation of specific sectors, including the rules applicable to electronic communication services or short-selling. All options would be coherent with existing initiatives targeting harmful trading practices in the offline world. These existing initiatives are designed to tackle practices relevant to and specific for the offline sector or context in which they arise. 278 They do not overlap with the unfair practices described under Section 2.1.2 which, together with the different business models by which gatekeepers operate, are very different and warrant separate treatment.

    Proportionality

    367.Option 1 would be targeted to gatekeepers above a pure quantitative threshold. This would create legal certainty for gatekeepers. The obligations it would cover would be identified based on objective criteria supported by the evidence gathered. However, the absence of dialogue between the regulator and gatekeepers would preclude the possibility to exchange with the regulator to specify the application of certain obligations.

    368.Option 1-A would leave uncovered some gatekeepers that would be similar to the designated gatekeepers, in the sense that they may equally have an important internal market impact, operate an important gateway to end users and have an entrenched position, but which are relatively smaller. This would call for a potential intervention by the Member States or competition law to tackle those. This would thus leave more room for action at national level but at the same time raise a concern of further regulatory fragmentation. Option 1-B would cover a larger number of platforms, thus reducing the risk of regulatory fragmentation as the obligations are applied to a larger set of platforms. However, it would risk capturing under scope providers of core platform services that would be above the thresholds but that do not present features and characteristics similar to gatekeepers. There is thus a risk of disproportionality in relation to the application of the obligations to those platforms.

    369.Option 2 allows achieving the objectives effectively since it sets a comprehensive ex ante framework providing for a list of obligations that are clearly identified and circumscribed. Option 2-A would also result in a uniform application of obligations to designated gatekeepers (including those that are expected to enjoy an entrenched position in the near future) that although not falling under the high quantitative threshold criteria would nonetheless operate core platform services that exhibit similar features and characteristics to those of gatekeepers corresponding to the quantitative criteria. The main drawback of sub-option 2-A would be the possible temporary fragmentation resulting from having gatekeepers automatically designated and others designated via longer market investigations. Sub-option 2-B would, similarly to the case of sub-option 1-B, provide for a harmonised application of the obligations given that a larger set of platforms would be in scope, but still risking to capture providers of core platform services that would be above the thresholds but do not present features and characteristics similar to gatekeepers. In both cases compliance costs for gatekeepers are reasonable thus allowing to safeguard the benefits they create for the internal market.

    370.Option 2 is proportionate since it would be also addressing the wider possible range of unfair practices at EU-wide level identified on the basis of similar criteria as Option 1, while at the same time providing for a regulatory dialogue in relation to the application of some of the obligations, where necessary and justified. Similarly to Option 1, Option 2 foresees cooperation with NCAs and with sectorial bodies.

    371.Option 3 leaves strong discretionary power to the regulator in terms of both designation of gatekeepers (based on qualitative criteria only assessment) and scope of intervention (given the flexibility left to the regulator to include additional digital services and practices in the scope). Option 3 would from that perspective lead to a lower level of legal certainty, which is essential for a thriving business environment. Option 3 allows however for extensive regulatory dialogues, which possibly makes it less burdensome than regulatory measures not allowing for such dialogues. At the same time the longer procedures which would stem from the market investigation nature of this option limit its potential for addressing the problems identified in a timely manner.

    372.For all options, and as explained in Section 5.2.1, the core platform services in scope are only those where there is strong evidence of (i) high concentration, where usually one or very few large digital platforms set the commercial conditions with considerable autonomy from their competitors, customers or consumers; (ii) few large digital platforms acting as gateways for business users to reach their customers and vice-versa; and (iii) gatekeeper power often misused by means of unfair behaviour vis-à-vis economically dependent business users and customers. Option 3 provides for the possibility of other digital services being added to the list after a market investigation and based on an empowerment given to the Commission. This would allow to cover in a flexible way all the digital services where there is weak contestability and gatekeepers engage in unfair behaviour, but would create lower legal certainty.

    373.As explained in Section 5.2.1, the gatekeepers in scope under Option 1 would only be those that meet the quantitative criteria that serve as a proxy for the features characterising gatekeeper status. This could risk including too few providers of core platform services in scope if thresholds are set at a too high level (false negatives) or too many platforms in case lower thresholds are fixed (false positives). Under Options 2 and 3, gatekeepers in scope are more likely to correspond to those which (i) have a significant impact on the internal market; (ii) operate a core platform service which serves as an important gateway for business users to customers; and (iii) enjoy an entrenched and durable position in their operations or are expected to enjoy such a position in the near future.

    374.Finally, and as explained in Section 5.2.2, the list of obligations foreseen under all options is justified as the obligations have been limited to those practices that (i) are of egregious nature, (ii) can be identified in a clear and unambiguous manner to provide the necessary legal certainty for gatekeepers, and (iii) for which there is sufficient experience with the harmful effects. Option 2 and 3 provide for the possibility of a regular dialogue between the Commission and the gatekeepers concerned, as well as for the addition of other practices if deemed unfair following a market investigation and based on an empowerment by the Commission.

    375.In order to ensure the effectiveness of the tool, all options could include a series of remedies to ensure that designated gatekeepers comply with the obligations. This would include initially fines and penalty payments in case gatekeepers do not comply with the obligations. As a last resort, and in case of systematic failure to comply with the obligations, even after the imposition of fines and penalty payments, other types of measures could be envisaged under specific conditions and circumstances (see Section 5.2.4). These would however be applied in extreme cases, i.e. once all other means to ensure fair behaviour have proven insufficient, and thus not put at risk the proportionality of the measures, under all three options.

    Subsidiarity

    376.All three options respect the subsidiarity principle. The intrinsic cross-border nature of the digital economy and of the provision of core platform services provided by gatekeepers, suggests that the objectives pursued cannot be effectively reached by Member States alone. Rather to the contrary, as shown in Annex 5.4 to the Impact Assessment, regulatory initiatives by Member States lead to divergent regulatory solutions and regulatory fragmentation. EU action would avoid further fragmentation of the single market into different, potentially contradictory frameworks – including the resulting jurisdictional issues. This is expected to decrease gatekeepers’ incentives to develop unfair practices in relation to new core platform services or expand further unfair behaviour related to existing practices.

    377.Furthermore, while all three options foresee enforcement and strong coordination at EU level, they also envisage the involvement of national authorities in the decision making process to ensure that Member States’ expertise is taken into account. The new ex ante framework would harmonise rules in the areas targeted by these rules, without prejudice to the right of Member States to consider further measures in order to improve contestability of markets or to fight against acts of unfair competition that are unrelated to the presence of gatekeepers within the meaning of the new framework and where such obligations would be compatible with EU law.

    Conclusion

    Table 5: Policy options - comparison

    Option

    Effectiveness

    Efficiency

    Coherence

    Proportionality

    Subsidiarity

    Benefits

    Costs

    1-A

    +

    +

    Small

    ++

    +

    ++

    1-B

    +

    +

    Medium

    ++

    +

    ++

    2-A

    ++

    ++

    Medium

    ++

    ++

    ++

    2-B

    ++

    ++

    Medium

    ++

    ++

    ++

    3

    +

    ++

    High

    ++

    ++

    ++

    378.Effectiveness. Compared to Option 1, Options 2 and 3 would allow tackling new unfair practices and market failures related to gatekeepers that are expected to have an entrenched position in the near future. Option 1 would be particularly effective in quickly dealing with the market failures in digital markets. The combination of immediately applicable obligations and of the possibility for a regulatory dialogue with gatekeepers for some of the obligations would make Option 2 more effective than Options 1 and 3. The latter two would be: too static (in the case of Option 1) or too flexible (in the case of Option 3). Option 1-A could give rise to type II errors (false negatives) while Option 1-B could result in type I errors (false positives). Under Option 2-A, the risk of false negatives would be minimised by the possibility of designating gatekeepers also on the basis of qualitative criteria. In that case, the drawback in comparison with sub-option 2-B would be the delays associated to the need of conducting market investigations to designate some of the gatekeepers. 279 Option 3 is future-proof and would reduce the risk of type I or type II errors by basing designation on a pure qualitative test. At the same time, it does not allow addressing the problems in a timely manner, implies less legal certainty than the other two options and would result in a staggered approach to the market.

    379.Efficiency. Compared to Option 1, Options 2 and 3 would generate higher benefits because they would allow tackling a higher number of market failures. Option 3 could in theory target even more market failures given the possibility to add other digital services to the scope of the regulation, following a market investigation. This is though unlikely to take place. However, as a consequence of discretionary power given to the regulator, Option 3 would also create a lower level of legal certainty which would result in a chilling effect, thus counteracting the positive impact (that measures effectively addressing unfair behaviour would have) on market contestability. As regards sub-options A and B, magnitude of benefits would be similar, with the former having the advantage of not incurring in type II errors and the latter of implementing quicker the obligations to gatekeepers (that would otherwise be captured after a market investigation under sub-option 2-A). In relation to compliance costs, they would be miniscule as compared to the profits of gatekeepers. They would also be insignificant as compared to the range of benefits resulting from tackling unfair practices in digital markets. Nevertheless, Option 1 would result in lower compliance costs as gatekeepers would not be subject to any market investigation and less practices would be tackled. Option 3 would imply the highest compliance costs given the need to conduct a market investigation for any designation and the possibility to designate gatekeepers from services other than the current core services. As for sub-options 2-A and 2-B, the former would result in lower compliance costs since it would imply designating a lower number of gatekeepers, although the higher number of market investigations associated to sub-option 2-A could imply higher costs of replying to requests for information.

    380.Coherence: All three options are coherent with other EU instruments and EU international commitments.

    381.Proportionality. Option 1 would be too static, in the designation process and the implementation of the obligations. This could originate a disproportional implementation of obligations. Options 2 and 3, by allowing a more flexible approach would be more proportionate. Sub-option 2-A could lead to temporary fragmentation resulting from some gatekeepers being automatically designated while others via longer market investigations; under sub-option 2-B there would be the risk of capturing platforms that do not present features and characteristics similar to gatekeepers’ ones. Option 3 would lead to a lower level of legal certainty which is essential for a thriving business environment.

    382.All three options respect the subsidiarity principle.

    383.In light of the above, political choice is needed on whether Option 2-A or Option 2-B would better address the issues at stake and achieve the policy objectives pursued. Option 2 allows for timely intervention for the most egregious practices and more gradual approach for measures needing further tailoring and specification. The comparison between Options 2-A and 2-B leads to a trade-off to be made between the speed of the regulatory intervention and the scope of the target population. As specified above, in the case of sub-option 2-A, the impact of the intervention would be less immediate (given that a higher number of gatekeepers would have to be designated via a market investigation), and thus the unfair practices by those gatekeepers would not be tackled for a period of time. In the case of sub-option 2-B, more gatekeepers would be automatically designated but with the risk of designating platforms that do not qualify as gatekeepers. Both sub-options (i) address unfair behaviour, including new unfair practices and tipping markets, leading to positive impact on market contestability, innovation and consumer choice, (ii) address most of the regulatory fragmentation problems as obligations are applied to both gatekeepers designated on the basis of quantitative and qualitative criteria, including emerging gatekeepers, and (iii) create legal certainty for those.

    8.PREFERRED OPTION

    The main building blocks of the preferred option

    384.The core substantive elements of the preferred option are explained in Section 5.3.2. The preferred option provides for a new ex ante regulatory framework built around the following elements:

    (a)First, the ex ante framework would only apply to clearly identified and closed list of core platform services (i.e. a numerous clausus of core platform services) that are most broadly used by business users and end users and where, based on the evidence collected and presented in this Impact Assessment, more apparent and urgent concerns about weak contestability and unfair practices by gatekeepers arise;

    (b)Second, only providers of core platform services that meet the specific conditions analysed could be designated as gatekeepers. Such designation would take place by applying mix of quantitative and qualitative thresholds;

    (c)Third, designated gatekeepers would be required to comply with the set of clearly defined obligations in order to address a negative impact of unfair practices discussed in Section 5.2.2 on fairness in commercial relationship between these gatekeepers and their business users and contestability of platform markets. Such obligations would encompass (i) immediately applicable obligations and (ii) obligations where a degree of appreciation would be required in view of the implementation of a given obligation.

    The scope of application

    Identification of core platform services

    385.The preferred option foresees up-front a set of clearly identified core platform services that feature number of specific characteristics discussed in Section 5.2.1.

    386.The analysis underpinning the Impact Assessment shows that there are number of core platform services that meet these characteristics, notably:

    (a)Online intermediation services, such as online marketplaces and software application stores which enable business users to reach and contact end users, to provide or offer services or products to the latter. They can become a key access point for business users to reach end users.

    (b)Online search engines that can significantly affect the commercial success of business users and therefore unfair practices carried out by a gatekeeper providing such online search engine services have the capacity to affect a large number of end users and businesses alike.

    (c)Operating systems are at the heart of devices and ecosystems and are characterised in particular by economies of scale and high switching costs, and benefit from network effects. A gatekeeper can use its control over the operating system to engage in unfair practices limiting the contestability of the services concerned.

    (d)Online social networking services are characterised in particular by strong network effects, data driven advantages and high switching costs. When controlled by a gatekeeper, online social networking services represent an important gateway not only for end users but also increasingly for business users.

    (e)Video-sharing platform services that can become the default or at least a preeminent platform to consume and to share video content online. When operated by a gatekeeper, they are thus a very important access point for video content providers and offer significant audiences for advertisers.

    (f)Number-independent interpersonal communication services are services for which network effects are particularly strong when they are run by a gatekeeper, the risk of unfair business practices and a lack of contestability is particularly strong.

    (g)Cloud computing services provide infrastructure to support and enable functionality in digital services offered by others and at the same time offer a range of products. The vertical integration by a gatekeeper of a large cloud computing services provider can lead to unfair business conditions, for instance unjustified limitations to interoperability and data portability.

    (h)Online advertising services are often related to other core platform services, such as online search engines and online social networking services. Gatekeepers operating the latter types of core platform services very often also provide online advertising services and may engage in unfair practices, which goes in particular to the detriment of their business users that is, advertisers and publishers and also further limit contestability.

    Designation of gatekeepers

    387.The addressees of the preferred option would be those providers of core platform services that meet the following conditions: (i) have a significant impact on the single market, (ii) operate one or more important gateways for business users to reach end users, and (iii) they enjoy or are expected to enjoy an entrenched and durable position in their operations.

    388.Under the preferred option, these providers of core platform services would be designated based on combined application of quantitative and qualitative criteria. Where a provider of core platform services would meet cumulatively a set of quantitative thresholds established in the regulation it would be automatically designated as a gatekeeper by the Commission. This allows a fast and effective protection of the interest of all the business users affected by the gatekeepers’ unfair behaviour. Two specific combinations of parameters, for a low and a high threshold respectively, have been selected for the purpose of providing clarity in assessing impacts and trade-offs for the options’ comparison. As explained in Section 5.2.1, other plausible policy options exist as to the use of the economic parameters or their possible combinations. These thresholds act as quantifiable proxies of the qualitative criteria in terms of size and reach of the business, number of business users and duration in time of the market position. Such a quantitative threshold may be set at a high level (sub-option 2-A) or low level (sub-option 2-B). In choosing between sub-options 2-A or 2-B it is preferable that these thresholds should be set at a sufficiently high level. This is to ensure that only very large systemic players with a significant internal market presence and which are gateways to a large number of end users, clearly holding an entrenched and durable position, should be deemed to be a gatekeeper on the basis of quantitative criteria. Such undertakings should be subject to a fast designation process which is limited to verifying whether the quantitative criteria are met. 

    389.While considering the high probative value of the considered quantitative threshold, it cannot be completely excluded that in very exceptional circumstances a provider of core platform services that meets these quantitative thresholds nonetheless does not act as a gateway for its business users and end users. To ensure necessary proportionality in such exceptional circumstances, the provider of core platform services should have the opportunity to present serious and substantiated arguments in order to demonstrate that, in the circumstances in which the relevant core platform service operates, and taking into account other relevant elements 280 , the provider does not meet the conditions discussed in paragraph 387. It is important however to note that the purpose of the possibility to rebut the legal presumption is not to demonstrate, on pure economic grounds, efficiencies deriving from a specific type of behaviour by the provider of core platform services since this is not relevant to designation of such a provider as a gatekeeper.

    390.Furthermore, even if a provider of core platform services does not meet the quantitative thresholds that does not in itself mean that it may not constitute a gatekeeper. In fact, the preferred option would envisage a possibility to designate the provider of core platform services as a gatekeeper following a market investigation, which would have to show that the provider of core platform services meets the conditions discussed in paragraph 139.

    391.This would also ensure the regulatory playing field and regulatory symmetry of the obligations laid down in the rules. This also represents the optimal trade-off in terms of necessity to assess the market conditions and time that would take to designate gatekeepers and thereby time within which the problems identified would be effectively addressed. In this respect it is to be noted that quantitative thresholds set at a high level would take comparatively more time and resources than a more straightforward designation based on lower quantitative thresholds. Conversely, while relatively low quantitative threshold level would allow immediately capturing the majority of gatekeepers, it would entail the risk of extending disproportionately gatekeeping obligations to a large number of platforms.

    392.The designation of a provider of core platform services as a gatekeeper following a mechanism that combines quantitative and qualitative indicators features strong support by stakeholders and is considered as an appropriate mix of ensuring flexibility, speed and legal certainty.

    393.Under the preferred option, the qualitative criteria would allow designating not only providers that are already enjoying an entrenched and durable position in their operations, but also those for which this is not yet the case, but which are rapidly acquiring market strength and building towards becoming gateway due to specific market features and their capacity to put competitors at a disadvantage in the market, i.e. emerging gatekeepers.

    394.The preferred option would provide for regular review of the gatekeepers status, a possibility which seems particularly important in such a dynamic market environment. Such a regular review of the gatekeeper status would in principle have to be carried out on regular intervals of two years. 

    395.The designation decision addressed to providers of core platform services that meet the conditions would be subject to judicial review and would, beyond the regular review, also foresee reassessment at the request of the affected firm in case of material changes concerning the designation conditions.

    Obligations applicable to gatekeepers’ core platform services

    396.Under the preferred option, once a provider of core platform services is designated as a gatekeeper, all its core platform services that individually meet the conditions of being an important gateway for business users to reach end users would have to comply with a clearly defined set of obligations relating to a clearly identified set of unfair practices.

    397.The obligations under the preferred option would address the unfair practices by gatekeepers that weaken market contestability (Section 2.1.1) and undermine the fairness of commercial relationship of gatekeepers towards their business users or in some cases towards third parties (Section 2.1.2).

    398.The obligations under the preferred option would be either immediately applicable or would in certain cases envisage the possibility of a regulatory dialogue between the Commission and the gatekeeper concerned in view of ensuring that the measures gatekeepers intend to implement ensure effective compliance with the obligations. The set of obligations that would be included in the preferred option is explained in details in Section 5.2.2.

    399.The distinction between immediately applicable obligations and obligations subject to dialogue is based on the analysis of the measures in question. The obligations relative to transparency and non-discrimination are self-evident. Obligations which require evaluation of interoperability conditions or customisation considering the specific nature of the core platform service offered are subject to a dialogue. Such a dialogue can be launched by the Commission, either upon request of the gatekeeper concerned, or where the Commission finds on its own initiative that any measures that the gatekeeper has already implemented or still intends to implement are likely to fall short of what is required to ensure compliance with the obligations concerned. This possibility of a regulatory dialogue should facilitate compliance by gatekeepers and allow them to signal any circumstances.

    400.As an additional element to ensure proportionality, gatekeepers should be given an opportunity to request the suspension of a specific obligation in exceptional circumstances that lie beyond the control of the gatekeeper. Where compliance with a specific obligation is shown by the gatekeeper to endanger the broader economic viability of the EU operations of the gatekeeper concerned, for example because an unforeseen external shock has temporarily eliminated a significant part of end user demand for the relevant core platform service, it would ultimately harm innovation and welfare if that core platform service were unable to continue its operations once the exceptional circumstances would cease to apply. Similarly, in exceptional circumstances solely justified on the limited grounds of public morality, public health or public security, and based on a reasoned request by the gatekeeper, the Commission could decide that the obligation concerned does not apply to a specific core platform service.

    401.The combination of a regulatory dialogue to facilitate compliance with limited exemption possibilities will ensure the proportionality of the obligations without undermining the intended ex ante effects on fairness and contestability.

    Enforcement framework and governance

    402.Under the preferred option implementation, supervision and enforcement would be carried out at the EU level by the Commission as the competent regulatory body.

    403.The preferred option is built on the clearly identified behavioural measures, which will be laid down in the new rules. These rules will be based on a legal presumption that, to meet the objectives of safeguarding contestability of core platform services and fairness of their commercial relationships, gatekeepers need to comply with such regulatory behavioural measures, i.e. obligations. This will ensure the necessary legal certainty and predictability of the rules as well as ensure that the rules apply only where this is necessary and justified, i.e. are proportionate to the objective sought.

    404.The preferred option would also lay down adequate and proportionate enforcement powers of the Commission with clearly defined procedural enforcement framework and clearly set deadlines that the Commission would need to respect. The enforcement powers and processes applicable would be unique to the preferred option. However, as explained in Sections 5.3.1.5 and 5.3.2.5 they would nevertheless be able to learn in some of their aspects from existing regulatory and competition law powers. In addition, the procedural framework would also lay down clear rules on redress available to gatekeepers or other concerned parties, including access to judicial remedies.

    405.The ex ante rules under the preferred option will be complemented by the possibility for the Commission to launch a market investigation in a limited and well identified number of cases:

    (a)to designate gatekeepers that meet the conditions laid down in Section 5.2.1, or may meet them in near future;

    (b)to update the list of unfair practices and corresponding obligations laid down in the rules; and

    (c)to serve as a basis for further remedial action if the behavioural measures clearly prescribed by the rules are systematically infringed by the designated gatekeepers.

    406.It is worth recalling that the possibility of updating the core platform services by means of empowerment by the Commission following a market investigation has been excluded from the preferred option. The reason is linked to legal limits to the market investigation powers under the chosen ex ante legal instrument that cannot include implicit powers to adapt the scope. Instead, the Commission should propose the necessary legislative adaptations by including in the evaluation of the effectiveness of the regulation the regular review of the list of core platform services in view of ensuring that digital markets across the EU are contestable and fair.

    407.Under the preferred option, further remedies (see Section 5.2.4) would be envisaged for the purpose of ensuring effective remedies against systematic non-compliance.

    408.Such remedies should be modelled on the well-established precedent of Regulation 1/2003, and offer a graduated, step-by-step process of increasing sanctions, with due process rights at each step. This enforcement framework reflects the potentially limited dissuasive power of monetary fines alone, and may, as explained in Section 5.2.4., contain as ultima ratio behavioural or structural remedies after all other avenues have been exhausted, noting that even in established legislation such a tool has never been used.

    409.In the preferred option the Commission will ensure close cooperation with and between the competent independent authorities of the Member States, with a view to informing its implementation and to building out the Union´s expertise in tackling fairness and contestability issues in the digital sector. In this context, the Commission will establish an information exchange and consultation network consisting of relevant independent authorities of the Member States, which shall also deliver opinions on the individual decisions of the Commission.

    410.As regards the possible integration of the new Commission powers under the Digital Markets Act and responsibilities of the Board envisaged under the DSA it is important to note the very different objectives of the two sets of rules and corresponding expertise and competences that may be required from the competent enforcement bodies to ensure compliance with the respective rules.

    411.The Board under the DSA, including the participation of the national Digital Services Coordinators, enhances the cooperation system, particularly necessary for ensuring the supervised risk management approach for regulating the due diligence of very large platforms. This system ensures in particular that primarily systemic societal, and not economic, concerns brought by those platforms with an EU-wide impact are appropriately addressed through cooperation at the EU level supported by the activities of the Board, thereby ensuring sufficient expertise and appropriate competencies. However, the main regulatory compliance activities continue to be carried out by the competent national Digital Services Coordinators.

    412.Furthermore, contrary to the decentralised approach under the DSA, where the focus of regulatory compliance activities is on the competent national Digital Services Coordinators, the implementation and enforcement of harmonised rules under the DMA is to be ensured at the EU level by the Commission who has the necessary means and expertise, without any decentralised competences.

    413.In view of this, it could not be considered under the preferred option that any of the investigation and enforcement competences and powers could effectively be carried out by the Board whose tasks relate to facilitating implementation, cooperation and enforcement of very different rules as those envisaged by the DMA.

    9.HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?

    414.Given the dynamic nature of online platforms, monitoring and evaluation of impacts needs to constitute an important part of the proposal. It also responds to explicit demands by stakeholders, including Member States (e.g. France), for a dedicated monitoring function, and reflects the self-standing monitoring option considered in the Inception Impact Assessment. The monitoring therefore will be divided into two parts: (i) continuous monitoring which will report on the latest developments in the market every second year potentially involving the EU Observatory of the Online Platform Economy, and (ii) operational objectives and specific indicators to measure them.

    415.Regular and continuous monitoring will cover the following main aspects:

    a)Monitoring scope-related issues (e.g. indicators for the designation of gatekeepers, range of designated gatekeepers and its evolution, use of the margin of appreciation in the designation);

    b)Monitoring unfair practices (compliance, enforcement patterns, evolution); and 

    c)Monitoring as a trigger for launch of a market investigation.

    416.The following indicators would be potentially used:

    Table 6: Measuring indicators

    Specific objective

    Operational objectives

    Potential Measuring indicators

    Enhance coherence and legal certainty in the online platform environment in the internal market

    Limit the diverging national regulatory interventions

    Ensure coherent interpretation of obligations

    Number of regulatory interventions at the national level

    Number of clarification requests per year

    Address gatekeeper platforms' unfair conduct

    Preventing identified unfair self-preferencing practices

    Number of compliance interventions by the Commission per gatekeeper platform/per year

    Number of sanction decisions per gatekeeper platform/per year

    Address market failures to ensure contestable and competitive digital markets for increased innovation and consumer choice

    Preventing unfair practices concerning access to gatekeeper platforms’ services and platforms

    Preventing unfair data related practices and ensuring the compliance with obligations

    Share of users multi-homing with different platforms or services

    Share of users switching between different platforms and services

    417.The monitoring will also take due account of the conceptual work of the Expert Group of the Online Platform economy under its work stream on Measurement and Economic Indicators. 281

    418.Of particular importance in the monitoring framework is the evolution of the market, in terms of new unfair practices, additional core platform services, and new gatekeepers. The monitoring framework, including through the dedicated Observatory of the Expert Group on the Online Platform Economy as well as through the market investigation part of the preferred option, will continuously monitor the evolution of these factors. This is a core part of keeping the regulation future proof.

    419.To this end, the preferred option contains a specific obligation on the Commission to review whether designated gatekeepers continue to meet the scope of the obligation. This monitoring function is essentially part of the preferred option to keep the rules in line with market developments.

    420.Furthermore, specifically the legislation proposed should be reviewed at least every three years, to ensure that other elements, notably other scope related issues (such as new services) require adjustments.

    421.Finally, the monitoring framework also needs to monitor compliance with the regulation, and the effectiveness of the enforcement framework, including to which extent the range of available remedies were actually used, the effectiveness of the implementation dialogues, and the responsiveness of the companies in scope to the obligations.

    (1)

          Inception Impact Assessment for the Digital Services Act package .

    (2)

          Inception Impact Assessment of the New Competition Tool .

    (3)

          Open Public Consultation on Single Market - new complementary tool to strengthen competition enforcement and Open Public Consultation on Digital Services Act package – ex ante regulatory instrument of very large online platforms acting as gatekeepers .

    (4)

          Inception Impact Assessment of the New Competition Tool , at page 3; and Inception Impact Assessment for the Digital Services Act package , at page 4.

    (5)

         J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era , page 70.

    (6)

          EU Observatory on the Online Platform Economy .

    (7)

             US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (8)

             ACCC report, Digital Platforms Inquiry, Final Report , June 2019.

    (9)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (10)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (11)

          Mission letter to Executive Vice-President Vestager , 10 September 2019.

    (12)

         European Commission Communication, Shaping Europe’s digital future , 19 February 2020, at page 5.

    (13)

         Ibid, at page 5.

    (14)

         Pursuant to Article 225 of the Treaty on the Functioning of the European Union (‘TFEU’).

    (15)

          https://emeeting.europarl.europa.eu/emeeting/committee/en/agenda/202009/IMCO .

    (16)

         In its draft report, the IMCO considered that “by reducing barriers to market entry and by regulating large platforms, an internal market instrument imposing ex-ante regulatory remedies on these large platforms has the potential to open up markets to new entrants, including SMEs and start-ups, thereby promoting consumer choice and driving innovation beyond what can be achieved by competition law enforcement alone”. The draft report of JURI considers that “the acquisition of significant market power by dominant platforms has led to a situation in which “the winner takes it all”, and the market is composed of a small number of players each exerting market dominance over their competitors and imposing their business practices on users”. It further “calls on the Commission to assess the possibility of defining fair contractual conditions to facilitate data sharing with the aim of addressing imbalances in market power; suggests, to this end, to explore options to facilitate the interoperability and portability of data”.

    (17)

          European Parliament resolution of 18 June 2020 on competition policy – annual report 2019 (2019/2131(INI)) , 18 June 2020.

    (18)

         European Council, A new strategic agenda 2019-2024 , 20 June 2019. See also European Council Conclusions of 22 March 2019 .

    (19)

          Council of the European Union Conclusions of 9 June 2020 , responding to the Commission’s Communication Shaping Europe’s Digital Future.

    (20)

         Council of the European Union, Conclusions on a deepened Single Market for a strong recovery and a competitive, sustainable Europe , 11 September 2020.

    (21)

         Ibid. 

    (22)

         Council of the European Union, Special Meeting of the European Council– conclusions , 2 October 2020.

    (23)

          Summary of the Stakeholder Consultation on the New Competition Tool and Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (24)

         For example, BEUC’s reply to the OPC states that the “challenges posed in particular by large players in digital markets require new instruments in addition to traditional competition law enforcement in order to protect consumers’ interests in an effective and timely manner.

    (25)

         See Annex 5.3 to the Impact Assessment.

    (26)

         See the dedicated webpages on DG Competition’s website at: https://ec.europa.eu/competition/consultations/2018_vber/index_en.html (VBER), https://ec.europa.eu/competition/consultations/2019_hbers/index_en.html (HBERs), https://ec.europa.eu/competition/sectors/motor_vehicles/legislation/mvber_review.html (MVBER).

    (27)

         See the dedicated webpages on DG Competition’s website at: https://ec.europa.eu/competition/consultations/2020_market_definition_notice/index_en.html .

    (28)

          Commission Notice on the definition of relevant market for the purposes of Community competition law , OJ 97/C 372/03, at point 2.

    (29)

         See brochure How do online platforms shape our lives and businesses?  

    (30)

         In this Impact Assessment the notion of unfair business practices refers to both terms and conditions as well as the actual business practices of gatekeepers.

    (31)

         Where appropriate in this Impact Assessment, references to EU should be understood as comprising the EEA.

    (32)

         In the OPC on Ex Ante Rules a majority of stakeholders note that “certain platforms and their ecosystems have become unavoidable to access a large variety of contents and services on the internet. Those structuring platforms have become gatekeepers not only within their services, but for the internet at large.” See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (33)

         M. Bajgar, G. Berlingieri, S. Calligaris, C. Criscuolo & J. Timmisn (2019), Industry Concentration in Europe and North America , OECD Productivity Working Paper; G. Grullon, Y. Larkin, & R. Michaely (2019), Are US Industries Becoming More Concentrated?, Review of Finance, Volume 23(4), pages 697–743; Market Concentration - Note by Jason Furman, Hearing on Market Concentration , 7 June 2018; G. Gutiérrez & Th. Philippon (2017), Declining Competition and Investment in the U.S., ; G. Gutiérrez & Th. Philippon (2018), How European Markets Became Free: A Study of Institutional Drift .

    (34)

         Source: R. Fijneman, K. Kuperus, J. Pasman (2018), Unlocking the value of the platform economy , KPMG report for the Dutch Transformation Forum 

    (35)

          Observatory expert group Progress report on the Measurement of the Online Platform Economy .

    (36)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (37)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (38)

         See Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (39)

         See M. Motta & M. Peitz (2020), Intervention trigger and underlying theories of harm - Expert advice for the Impact Assessment of a New Competition Tool , Chapter 2, and G. S. Crawford, P. Rey, & M. Schnitzer (2020), An Economic Evaluation of the EC’s Proposed “New Competition Tool” , Section V.C.

    (40)

         In the OPC on Ex Ante Rules, respondents in general consider that unfair practices by gatekeepers have a concerning impact on competition, innovation and consumer choice. Competition is hampered when gatekeepers create barriers for new market operators to enter the market, thereby resulting in reduction of investments and innovation and consumer choice stifling. Unfair practices are considered to be the means by which digital platforms increase the cost of switching or multi-homing for users, thereby limiting market contestability and preserving their market power. See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (41)

         G. Biglaiser, E. Calvano & J. Crémer (2019), Incumbency advantage and its value, Journal of Economics and Management Strategy, volume 28(1), pages 41-48; A. Afilipoaie, K. Donders & P. Ballon (2019), What Are the Pro- and Anti-Competitive Claims Driving the European Commission’s Platform Policies? A Case Study Based Analysis of the European Commission’s Take on Platform Cases .

    (42)

         In the OPC on Ex Ante Rules, several hundreds of respondents identified each of these activities. See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (43)

         Several stakeholders, such as media publishers or game developers, raised concerns about this specific issue in the OPC on Ex Ante Rules as well as through different legal actions taken both in the Europe and the US.

    (44)

         Often called Wide Most Favoured Nation (‘MFN’) clauses. These clauses generally also apply to the business user’s direct sales channels, however this element of such clauses would not be affected by the options presented in this impact assessment.

    (45)

          https://judiciary.house.gov/calendar/eventsingle.aspx?EventID=3113 .

    (46)

         See I. Graef (2018), Algorithms and Fairness: What Role for Competition Law in Targeting Price Discrimination towards End Consumers?, Columbia Journal of European Law, volume 541, pages 546–8. and I. Graef (2019), Differentiated Treatment in Platform-to-Business Relations: EU Competition Law and Economic Dependence, Yearbook of European Law, Volume 38, pages 448–499.

    (47)

          Observatory expert group report on differentiated treatment .  

    (48)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (49)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules. 

    (50)

         For a more detailed overview see IA support study .

    (51)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (52)

         This notion of ‘digital intensity’ is rather broad and encompasses all firms with relatively high exposure to ‘Information and Communication Technologies’ (in terms of their investments, or input purchases), as well as firms reporting online sales. For the definition of ‘digital intensity’, see F Calvino, C Criscuolo, L Marcolin, & M Squicciarini (2018), A taxonomy of digital intensive sectors, OECD Science, Technology and Industry Working Papers 2018/14.

    (53)

         M. Bajgar, G. Berlingieri, S. Calligaris, C. Criscuolo & J. Timmis, (2019), Industry Concentration in Europe and North America, OECD Productivity Working Papers, 2019-18, OECD Publishing, Paris.

    (54)

         The study used firm-level data sourced from the commercial dataset Orbis® by Bureau van Dijk (BVD). It provides information on firms' localisation, annual balance sheet and income statements, although the number of observations per country can vary significantly. It covers the period 2001-2014 for 26 countries: Australia, Austria, Belgium, Bulgaria, Denmark, Estonia, France, Finland, Hungary, Germany, Indonesia, India, Ireland, Italy, Japan, Republic of Korea, Luxembourg, the Netherlands, Portugal, Romania, Slovenia, Spain, Sweden, Turkey, the UK, US. See also J. Federico, D. Leigh & S. Tambunlertchai (2018), Global Market Power and its Macroeconomic Implications, IMF Working Paper WP/18/137.

    (55)

         See S. Calligaris, C. Criscuolo, & L. Marcolin (2018), Mark-ups in the Digital Era, OECD Science, Technology and Industry Working Papers 2018/10. See also https://www.oecd-ilibrary.org/docserver/4efe2d25-en.pdf?expires=1601160653&id J. De Loecker, & J. Eeckhout (2017), The Rise of Market Power and the Macroeconomic Implications .

    (56)

         Ibid.

    (57)

         See J. Federico, D. Leigh & S. Tambunlertchai (2018), Global Market Power and its Macroeconomic Implications, IMF Working Paper WP/18/137.

    (58)

         According respectively to estimates from the P2B Impact Assessment and the Online Platform Economy Observatory.

    (59)

          https://platformobservatory.eu/state-of-play/power-over-users/ .

    (60)

          https://platformobservatory.eu/state-of-play/power-over-users/ .  

    (61)

         In general, most of the issues presented by the users are due to a perceived imbalance in bargaining power between platforms and business users, which hampers competition, fosters uncertainty in relation to contractual terms and also results in lock-in of consumers. See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (62)

         See IA support study .

    (63)

         See IA support study .

    (64)

         UNCTAD (2019), Digital economy report .

    (65)

         According to the Online Platform Economy Observatory, traffic share is the most revealing indicator of the economic significance of online platforms.

    (66)

          https://platformobservatory.eu/news/covid-19-and-online-platform-economy/ .  

    (67)

          Ibid . 

    (68)

         A. Ezrachi & M. Stucke, Digitalisation and its impact on innovation, R&I Paper Series, Working Paper 2020/07, European Union 2020, page 34.

    (69)

         Ibid., page 22.

    (70)

         The choice for consumers is limited by lock-in effects and lack of innovative alternatives blocked out of the gatekeeper platform(s).

    (71)

         The Online Observatory report on differentiated treatment shows that they may be subject to search diversion, i.e. platforms may have incentives to a biased order of products/services presentation, which would divert consumers from products/services they initially intended to buy, pushing them to purchase more and/or more expensive products/services.

    (72)

         Gatekeepers’ extraction of information leads to consumer profiling, unwanted advertisement targeting and privacy concerns.

    (73)

         See for instance J. De Loecker, J. Eeckhout & G. Unger (2020), The Rise of Market Power and the Macroeconomic Implications, The Quarterly Journal of Economics, Volume 135(2), Pages 561–644; J. De Loecker & J. Eeckhout (2018), Global Market Power, NBER Working Paper No. 24768; S. Barkai (2020), Declining Labor and Capital Shares, Journal of Finance, Volume 75, Issue 5; S. Calligaris, C. Criscuolo, & L. Marcolin (2018), Mark-ups in the Digital Era, OECD Science, Technology and Industry Working Papers 2018/10, on trends in firm-level mark-ups across 26 countries for the period 2001-14. They find that average mark-ups are higher and have grown more in ‘digital intensive’ sectors than in less ‘digital intensive’ sectors over the 2001-2014 period.

    (74)

         See M. Motta & M. Peitz (2020), Intervention trigger and underlying theories of harm - Expert advice for the Impact Assessment of a New Competition Tool , Chapter 2.

    (75)

         See M. Motta (2014), Competition Policy -Theory and Practice, Cambridge University Press.

    (76)

         See R. Whish (2020), The New Competition Tool: Legal comparative study of existing competition tools aimed at addressing structural competition problems, with a particular focus on the UK’s market investigation tool .

    (77)

         See Summary of the contributions of the NCAs to the impact assessment of the new competition tool . Some NCAs indicated that some of the questions in the questionnaire did not apply to them, because they did not have come across this particular feature or scenario in their recent case-work. When reporting on the views expressed by NCAs on particular issues, this Impact Assessment only reflects the views of those NCAs that did in fact express such a view.

    (78)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (79)

         77% for network effects, 51% for economies of scale, 49% for data, 44% for consumer bias 41% for switching costs. See ICN ‘Report on results of the ICN survey dominance/substantial market power in digital markets’ (‘ ICN Report on digital markets’ ).

    (80)

         Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017, paragraph 272.

    (81)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 462.

    (82)

         Case AT.40153 E-book MFNs and related matters (Amazon), Article 9 Decision of 4 May 2017, paragraph 65.

    (83)

         J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era , page 70.

    (84)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 600.

    (85)

         See M. Motta & M. Peitz (2020), Intervention trigger and underlying theories of harm - Expert advice for the Impact Assessment of a New Competition Tool , Chapter 2.

    (86)

         See OECD Policy Roundtable on Two-Sided Markets , 2009, in particular the contribution by the European Commission, pages 157-186.

    (87)

         Case COMP/C-3/37.792 Microsoft, Commission Decision of 24 March 2004, paragraph 420.

    (88)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 464.

    (89)

         In addition, in Apple/Shazam (Case M. M.8788 Apple/Shazam, Commission Decision of 6 September 2018, paragraphs 221 ff.), the Commission found that the merger would give Apple access to Shazam’s consumer data, which would give it the “[a]bility to use the Customer Information to put competitors at a competitive disadvantage”, while the evidence on Apple’s incentives to do so was mixed.

    (90)

         Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017, paragraph 287.

    (91)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (92)

         See Summary of the Stakeholder Consultation on the New Competition Tool and Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (93)

         A/B testing (also known as split testing) is a process of showing two variants of the same web page to different segments of visitors at the same time and comparing which variant drives more conversions. A/B testing is one of the most important ways to optimise a website's funnel in digital marketing.

    (94)

         A recent JRC report - Technology and Democracy: Understanding the influence of online technologies on political behaviour and decision-making - describes such techniques as “design choices that benefit an online service by coercing, steering, or deceiving users into making unintended and potentially harmful decisions”. There are patterns used in websites and apps that make users do things that they didn't mean to, like buying or signing up for something (see https://darkpatterns.org/ ). To explain such coercive and manipulative techniques, the JRC report refers to the “roach motel” example, i.e. it is easy for users to get into a certain situation, but difficult to get out. For instance, creating an account would require just a few clicks, but deleting it would involve more than 10 steps that are difficult to achieve without instructions.

    (95)

         A. Mathur, G. Acar, M. J. Friedman, E. Lucherini, J. Mayer, M. Chetty, & A. Narayanan (2019), Dark patterns at scale, Proceedings of the ACM on Human-Computer Interaction, volume 3, pages1–32.

    (96)

         This section quotes E. Katsamakas & H. Madany (2019), Effects of user cognitive biases on platform competition, Journal of Decision Systems, volume 28(2), pages 138-161.

    (97)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 781.

    (98)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 783.

    (99)

         Case COMP/C-3/37.792 Microsoft, Commission Decision of 24 March 2004, paragraph 463.

    (100)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 470.

    (101)

         See Support study to the Observatory for the Online Platform Economy, Report on the main obstacles and opportunities for multihoming, https://platformobservatory.eu/research/ .  

    (102)

         See Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services .

    (103)

         https://www.statista.com/statistics/1010716/apple-app-store-google-play-app-downloads-forecast/  Statista

    (104)

          https://www.businessofapps.com/data/facebook-statistics/ .

    (105)

          https://www.businessofapps.com/data/facebook-statistics/ .

    (106)

          Ecommerce report 2019 .

    (107)

         Report European Cross Border B2C E-Commerce Market 2020 Double Digit Growth Expected after 2020.

    (108)

         Source: Statista .

    (109)

          https://sproutsocial.com/insights/facebook-stats-for-marketers/ .

    (110)

    See IA support study .

    (111)

    The amount of data created each year in the digital economy is growing at an exponential rate. In 2020, it is estimated to reach 47 zettabytes at worldwide level compared to 12 zettabytes in 2015. Forecasts point to 142 zettabytes in 2035. Source: Statista (2019), Digital Economy Compass .

    (112)

          https://platformobservatory.eu/news/covid-19-and-online-platform-economy/ .  

    (113)

         Stock prices of each company are normalised to 100 in 2014, i.e. they are expressed relative to their respective value in 2014. This graph therefore allows to compare the development of stock prices across different companies, but not their absolute level.

    (114)

         The S&P 500 is a stock market index that measures the stock performance of 500 large companies listed on stock exchanges in the United States. It is one of the most commonly followed equity indices.

    (115)

    Competition in the voice-assistant markets will become more and more difficult as the algorithms underpinning the assistants benefit from the concentration of access to customers’ and users’ accumulated data. Incumbent platforms benefitting from large volume and variety of datasets will be able to provide more refined search results through their own assistants. This is an important competitive advantage vis-à-vis smaller and/or start-up platforms.

    (116)

    The likelihood of forthcoming legislative action is further supported by numerous reports of influential national authorities.

    (117)

         See IA support study .

    (118)

    The replies of citizens and stakeholders to the Commission’s OPC and the feedback of the NCAs replying to the Commission’s questionnaire indicate that market failures appear to be widespread across the Union, in particular in digital markets of cross-border nature. See Summary of the Stakeholder Consultation on the New Competition Tool and Summary of the contributions of the NCAs to the impact assessment of the new competition tool . While respondents indicated that market failures may occur in all industry sectors, several respondents emphasised that they are particularly prominent in the digital sphere.

    (119)

         In addition, Member States may not have the means to adopt appropriate measures to tackle market failures. Only some NCAs of Member States have instruments that enable them to tackle, to a certain extent, market failures, such as Greece and Romania. It can be expected that further Member States will adopt such national tools. Eight NCAs signalled that the competition rules applicable in their respective Member States have been amended in order to deal with market failures or that there are plans for doing so, namely Belgium, Bulgaria, Austria, Romania, Lithuania, Iceland, Germany and Greece. See Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (120)

         A merger or acquisition will be of an ‘EU dimension’ where the aggregate turnover of the undertakings concerned exceeds given thresholds; irrespective of whether or not the undertakings effecting the concentration have their seat or their principal fields of activity in the EU, provided they have substantial operations there. Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (the EC Merger Regulation ), OJ L 24, 29.1.2004, pages 1–22, Article 1(2).

    (121)

    See Summary of the Stakeholder Consultation on the New Competition Tool and Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (122)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (123)

    European Court of Auditors, Special Report 24/2020: EU audit report: merger control and antitrust proceedings , 19 November 2020, at paragraph 59.

    (124)

         While certain forms of unfair business practices can be abusive under Article 102(a) TFEU, finding such an abuse not only requires a dominant undertaking but generally also an effect on competition. If an undertaking imposes on its trading partners or obtains from them terms and conditions that are unjustified, disproportionate or without consideration but without affecting competition on the market, competition law generally does not apply (See recital 9 of Regulation (EC) No 1/2003. Some national competition laws also prohibit the abuse of economic dependence). Such behaviour resulting from imbalances in bargaining power that do not affect competition is usually the domain of unfair trading laws.

    (125)

      Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty , OJ L 001 4.1.2003, page 1.

    (126)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (127)

         Regulation (EU) 2016/679 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 ), OJ/L 119/1 (2016).

    (128)

         While some voluntary efforts for data portability by some platforms have been underway since 2017 in the ‘Data Transfer Project’, the project described itself still as ‘early stage’ and activity peaked in 2018 on the project. It should not be underestimated that this ‘Data Transfer Project’ is at present limited to only several large online platforms, which means that actual or potential competitors do not (yet) benefit from this project.

    (129)

         See for example O. de Hert, V. Papakonstantinou & G. Malgieri (2018), The right to data portability in the GDPR: Towards user-centric interoperability of digital services, Computer Law & Security Review, volume 34(2), pages 193-203.

    (130)

          https://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1601555056590&uri=CELEX:32005L0029 .  

    (131)

          https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A31993L0013 .  

    (132)

          https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32006L0114 .  

    (133)

         Sources supporting the assessment include the existing enforcement experience under competition rules, both within the EU and beyond, and other areas of law (e.g. protection of personal data); the numerous expert studies and reports both within the EU as well as internationally; complaints from business users and customers of gatekeepers as well as several on-going regulatory (e.g. Australia; Japan) or enforcement interventions (e.g. US); the reports and support studies for the Observatory on the Online Platform Economy drawn up by the independent Observatory Expert Group and external contractors respectively; the IA support study, which provided both quantitative (e.g. data analysis; case studies) and qualitative input; and the broad consultation across stakeholder groups.

    (134)

    The activity undertaken by the firm has been also considered in a recent advice by the CMA on the Digital Markets Unit. CMA is recommending Digital Markets Taskforce (‘DMU’) initially prioritising digital firms active in particular activities (e.g. online marketplaces, app stores, social networks, web browsers, online search engines, operating systems and cloud computing services).

    (135)

         The IA support study contains an analysis of business areas including mobile operating systems, app stores, desktop operating systems, search, social media, advertising (incl. search, display & video), e-commerce and cloud services. See also Annex 5.6 to the Impact Assessment.

    (136)

    Video streaming or video-on-demand services are currently characterised by less pronounced network effects (given the high costs of producing the distributed content) and switching costs are also not particularly high as subscriptions can be easily cancelled. In addition, there is no evidence of the presence of the problems described in Section 2.1 in video streaming or video-on-demand services.

    (137)

    In industrial B2B platforms the clients tend to be big, sophisticated companies which are not easily swayed by the platforms choice of ranking. They do not exhibit a similar dependency of the provider-side of the market at present. Switching costs are significantly lower: if products or services are delisted, corporate clients can insist that they be reinstated. Corporate clients have more leverage over platform decision. Support study to the Observatory for the Online Platform Economy, Developments concerning B2B platforms and emerging issues.

    (138)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (139)

    See for instance, Case AT.40099 Google Android, Commission Decision of 18 July 2018, section 9.4.

    (140)

    ACM, Market study into mobile app stores , 11 April 2019.

    (141)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (142)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (143)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (144)

          https://platformobservatory.eu/app/uploads/2020/09/Analytical-Paper-5-Business-user-and-third-party-access-to-data_final.pdf .

    (145)

      https://platformobservatory.eu/app/uploads/2020/09/Analytical-paper-1-Platform-data-access-and-secondary-data-sources_final.pdf .

    (146)

          https://platformobservatory.eu/app/uploads/2020/09/Analytical-Paper-2-Differentiated-treatment_final.pdf .

    (147)

       As defined in Directive (EU) 2016/1148.

    (148)

         Source: Statcounter .

    (149)

         CMA report on Online platforms and digital advertising .

    (150)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (151)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (152)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (153)

         ACCC report, Digital Platforms Inquiry, Final Report , June 2019.

    (154)

         Source: Statcounter .

    (155)

         CMA report on Online platforms and digital advertising .

    (156)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (157)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (158)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (159)

         ACCC report, Digital Platforms Inquiry, Final Report , June 2019.

    (160)

         See CMA report on Online platforms and digital advertising , page 119-123.

    (161)

         CMA report on Online platforms and digital advertising .

    (162)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (163)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (164)

         See for instance data at worldwide level (including platforms mostly active in China) in Statista .

    (165)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (166)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (167)

    Source: Statcounter .

    (168)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (169)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (170)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (171)

         See for instance shares reported by Statista at a worldwide level.

    (172)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (173)

         See CMA report on Online platforms and digital advertising , section 5.

    (174)

         CMA report on Online platforms and digital advertising .

    (175)

         US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets , October 2020.

    (176)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (177)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (178)

         ACCC report, Digital Platforms Inquiry, Final Report , June 2019.

    (179)

          https://platformobservatory.eu/research/ .

    (180)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (181)

         Very similar criteria have been considered by the CMA in its recent Advice on the DMU (see in particular point 4.19 of the Advice; available here: https://assets.publishing.service.gov.uk/media/5fce7567e90e07562f98286c/Digital_Taskforce_-_Advice_--.pdf ).

    (182)

         See IA support study , indicator and cluster analysis.

    (183)

         Market capitalisation is the aggregate valuation of the company based on its current share price and the total number of outstanding stocks. It is calculated by multiplying the current market price of the company's share with the total outstanding shares of the company and reflects the relatively significant potential to monetise these users in the near future.

    (184)

    The respective relevant levels should be set at a level reflecting a substantive percentage of the entire Union population when it comes to end users and of the entire population of businesses using platforms to determine the threshold for business users.

    (185)

         Very similar parameters pointing to digital firms with the likely Strategic Market Status (‘SMS’) have recently been considered and proposed by the CMA in its Advice of the DMU. For example, CMA proposes the use of the firm’s revenue as a criteria for the focus of the DMU’s enforcement activities, which should be on firms with annual UK revenue in excess of GBP 1 billion, and particularly those which also have annual global revenue in excess of GBP 25 billion.

    (186)

         This estimate as well as other estimates as to the number of gatekeepers to be covered under different options is largely based on publicly available data, supplemented with data on end user numbers sourced from different external providers.

    (187)

    See notably CERRE Report, Digital Markets Act: Making economic regulation of platforms fit for the digital age , 24 November 2020.

    (188)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules, Summary of the Stakeholder Consultation on the New Competition Tool and Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (189)

    See also Annexes 5.3 and 5.6 to the Impact Assessment.

    (190)

      https://www.bundeskartellamt.de/SharedDocs/Entscheidung/DE/Entscheidungen/Missbrauchsaufsicht/2019/B6-22-16.pdf?__blob=publicationFile&v=8 .

    (191)

      https://en.agcm.it/en/media/press-releases/2017/5/alias-2380 .

    (192)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (193)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (194)

      https://www.jftc.go.jp/en/pressreleases/yearly-2019/October/191031.html .

    (195)

         Case AT.40099 Google Android, Commission Decision of 18 July 2018.

    (196)

         Case AT. 39530 Microsoft (Tying), Commission Decision of 16 December 2009.

    (197)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (198)

          https://www.ftc.go.kr/solution/skin/doc.html?fn=508d97db636c2f7f0961bf6361cfd44f09977d1a7a06f4dd5603f17c11d61013&rs=/fileupload/data/result/BBSMSTR_ 000000002402/ .

    (199)

      https://www.agcm.it/dettaglio?db=41256297003874BD&uid=C66CBC6B87379163C125840000581AE0&view=vw0301&title=A529-GOOGLE/COMPATIBILITÀ%20APP%20ENEL%20X%20ITALIA%20CON%20SISTEMA%20ANDROID%20AUTO&fs=%20%2082_CE/102_CE-Abuso%20di%20posizione%20dominante .

    (200)

      https://www.acm.nl/en/publications/big-tech-and-dutch-payment-market-tightening-rules-needed-maintain-level-playing-field .  

    (201)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (202)

      https://www.justice.gov/opa/pr/justice-department-sues-monopolist-google-violating-antitrust-laws .

    (203)

      https://ec.europa.eu/digital-single-market/en/news/study-contractual-relations-between-online-platforms-and-their-professional-users .

    (204)

    Csse T-201/04 Microsoft v Commission, judgment of 17 September 2007, at paragraphs 1091 ff.

    (205)

      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 .    

    (206)

      Directive (EU) 2019/633 of the European Parliament and of the Council of 17 April 2019 on unfair trading practices in business-to-business relationships in the agricultural and food supply chain .

    (207)

      Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code (Recast) .

    (208)

         See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (209)

    Article 77 of the Code establishes that where “the national regulatory authority concludes that the appropriate obligations […] have failed to achieve effective competition and that there are important and persisting competition problems or market failures identified in relation to the wholesale provision of certain access product markets, it may, on an exceptional basis […] impose an obligation on vertically integrated undertakings to place activities related to the wholesale provision of relevant access products in a business entity operating independently”. Article 77 as well as its accompanying recitals establish a series of proportionality criteria that must be fulfilled prior to the imposition of such a requirement.

    (210)

         See Summary of the Stakeholder Consultation on the New Competition Tool and Summary of the contributions of the NCAs to the impact assessment of the new competition tool .

    (211)

         See for example also Article 9 of Regulation (EU) 2017/2394 of the European Parliament and of the Council of 12 December 2017 on cooperation between national authorities responsible for the enforcement of consumer protection laws and repealing Regulation (EC) No 2006/2004 (Text with EEA relevance), OJ L 345, 27.12.2017, page 1.

    (212)

      Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code (Recast) .

    (213)

      Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code (Recast) .

    (214)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (215)

    See workshop with BEUC members on the Impact Assessment for a possible New Competition Tool, 1 October 2020. See also BEUC’s response to the OPC on the NCT: “challenges posed in particular by large players in digital markets require new instruments in addition to traditional competition law enforcement in order to protect consumers’ interests in an effective and timely manner.

    (216)

    See Annex 2.1: Synopsis Report Open Public Consultation Ex Ante Rules.

    (217)

          https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52016DC0288&from=EN .

    (218)

          Contribution to Growth: European Digital Single Market: Delivering economic benefits to citizens and businesses (2019) .

    (219)

         For more information see: https://ec.europa.eu/jrc/en/rhomolo .

    (220)

         M. Christensen, A. Conte, F. Di Pietro, P. Lecca, G. Mandras, & S. Salotti (2018), The third pillar of the Investment Plan for Europe: An impact assessment using the RHOMOLO model (No. 02/2018). JRC Working Papers on Territorial Modelling and Analysis.

    (221)

         UNCTAD (2019), Digital economy report .

    (222)

          Cross-border Europe, annual analysis of the best global cross-border platforms operating in Europe , EU 28.

    (223)

          Impact Assessment Annexes, SWD(2018) 138 final : 71% of consumers would have preferred platforms for their purchases. This figure is an underestimate given the COVID epidemics but provides already an idea of the important use of platforms by consumers.

    (224)

          European Ecommerce Report 2017 . While the causal link between GDP growth and the economy of online platforms is difficult to demonstrate, considering these figures, it is reasonable to expect a relatively significant positive and growing contribution of the platform economy to the digital internal market and economic growth.

    (225)

        Hootsuite Digital 2020 global statshot report .

    (226)

         Ahn (2002) similarly concluded that “[a] large number of empirical studies confirm that the link between product market competition and productivity growth is positive and robust. […] Empirical findings from various kinds of policy changes […] also confirm that competition brings about productivity gains, consumers’ welfare gains and long-run economic growth". S. Ahn (2002), Competition, Innovation and Productivity Growth: A Review of Theory and Evidence, OECD Economics Working Paper No. 317. See also S. Nickell (1996), Competition and Corporate Performance, Journal of Political Economy, volume 104(4), pages 724-746, which found that the most competitive firms experienced productivity growth rates 3.8-4.6% higher than the least competitive. See also R. Disney, J. Haskel & Y. Heden (2003), Restructuring and productivity growth in UK manufacturing, The Economic Journal, volume 113(489), pages 666-694; R. Blundell, R. Griffith & J. van Reenen (1999), Market Share, Market Value and Innovation in a Panel of British Manufacturing Firms, Review of Economic Studies, volume 66(3), pages 529-54; S.I. Januszewski, J. Köke & J.K. Winter (2002), Product market competition, corporate governance and firm performance: an empirical analysis for Germany, Research in Economics, volume 56(3), pages 299-332. 

    (227)

         J.M. Arnold, G. Nicoletti & S. Scarpetta (2011), Regulation, Resource Reallocation and Productivity Growth, European Investment Bank Papers, volume 16(1), pages 90-115; OECD’s project of 2013 on Supporting Investment in Knowledge Capital, Growth and Innovation .

    (228)

    Commission services’ meetings with stakeholders.

    (229)

         G. Federico, F. Scott Morton & C. Shapiro (2019), Antitrust and Innovation: Welcoming and Protecting Disruption, NBER Working Paper No. 26005.

    (230)

         P. Aghion, N. Bloom, R. Blundell, R. Griffith & P. Howitt (2005), Competition and Innovation: an Inverted-U Relationship, The Quarterly Journal of Economics, volume 120(2), pages 701-728. On empirical work, see P. Aghion, S. Bechtold, L. Cassar & H. Herz (2014), The causal effects of competition on innovation: Experimental evidence, Journal of Law, Economics, and Organization, volume 34 (2), pages 162-195.

    (231)

         S. Ahn (2002), Competition, Innovation and Productivity Growth: A Review of Theory and Evidence, OECD Economics Working Paper No. 317; R. Gilbert (2007), Competition and innovation, Competition Policy Centre, UC Berkeley.

    (232)

         C. Shapiro (2012), Competition and innovation. Did Arrow hit the bull’s eye?, chapter 7 of Josh Lerner and Scott Stern (eds.), The Rate and Direction of Inventive Activity Revisited, pages 361-404.

    (233)

          https://ec.europa.eu/digital-single-market/en/news/impact-assessment-proposal--promoting-fairness-transparency-online-platforms .

    (234)

         R. Gilbert & D. Newbery (1982), Preemptive patenting and the persistence of monopoly. The American Economic Review, volume 74(1), pages 514-526.

    (235)

     The magnitude of online firms' acquisition is on the rise as highlighted by the Furman report, Unlocking digital competition, Report of the Digital Competition Expert Panel , reporting that the top 5 larger online platforms have carried out more 400 acquisitions worldwide in the last 10 years.

    (236)

    K.A. Bryan & E. Hovenkamp (2020), Antitrust limits on startup acquisitions, Review of Industrial Organization, volume 56, pages 615–636.

    (237)

         See IA support study .

    (238)

    E. Carayannis & E. Grigoroudis (2014), Linking innovation, productivity, and competitiveness: implications for policy and practice, The Journal of Technology Transfer, volume 39(2), pages 199-218.

    (239)

         D. Autor, D. Dorn, L.F. Katz, C. Patterson, & J. Van Reenen (2020), The fall of the labor share and the rise of superstar firms, The Quarterly Journal of Economics, volume 135(2), pages 645-709 and R.A. Decker, J.C. Haltiwanger, R.S. Jarmin & J. Miranda (2018), Changing business dynamism and productivity: Shocks vs. responsiveness. Technical report, NBER.

    (240)

         UNCTAD (2019), Digital economy report .

    (241)

      Japanese Fair Trade Commission Report regarding trading practices on digital platforms , October 2019.

    (242)

         ACCC report, Digital Platforms Inquiry, Final Report , June 2019; Japanese Fair Trade Commission Report regarding trading practices on digital platforms , October 2019.

    (243)

         Stigler Center report, Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee Report , July 2019.

    (244)

          https://judiciary.house.gov/calendar/eventsingle.aspx?EventID=3113 ; https://on.ft.com/33Bnq6T ; https://www.ftc.gov/policy/hearings-competition-consumer-protection

    (245)

         US House of Representatives Majority Staff Report, , October 2020 .

    (246)

         OECD (2015), Does competition kill or create jobs?, OECD Global Forum on Competition, DAF/COMP/GF(2015)9.

    (247)

         See also A. Dierx, J. Heikkonen, F. Ilzkovitz, B. Pataracchia, M. Ratto, A. Thum-Thysen & J. Varga (2015), Distributional macroeconomic effects of EU competition policy – A general equilibrium analysis, paper to be published in a World Bank-OECD publication on Competition Policy, Shared Prosperity and Inclusive Growth, who estimate that enforcement of the EU competition rules by the European Commission has a sizeable impact on the creation of new jobs (they estimate around 650 000 after 10 years).

    (248)

         OECD (2015), Does competition kill or create jobs?, OECD Global Forum on Competition, DAF/COMP/GF(2015)9, paragraph 78.

    (249)

         UNCTAD (2019), Digital economy report .

    (250)

    See IA support study .

    (251)

    Detailed overview of the preferred option’s implications for gatekeepers (but also for competitors, business users and consumers) is presented in Annex 3 to the Impact Assessment.

    (252)

    OPC, direct submissions.

    (253)

         See Summary of the Stakeholder Consultation on the New Competition Tool .

    (254)

         Source: Statista .

    (255)

         See https://www.oecd.org/going-digital/sme/resources/D4SME-Brochure.pdf .

    (256)

         See https://ec.europa.eu/digital-single-market/en/use-internet .

    (257)

         See https://ec.europa.eu/eurostat/statistics-explained/pdfscache/46776.pdf .

    (258)

    See IA support study .

    (259)

         See for instance S. Ahn (2002), Competition, Innovation and Productivity Growth: A Review of Theory and Evidence, OECD Economics Working Paper No. 317. See also for example, a study by the European Commission (2015) on The Economic Impact of enforcement of competition policies in the functioning of EU energy markets , which found that the Commission's decision finding an abuse of dominance by E.ON lead to a reduction in prices for both wholesalers and retailers to the benefit of consumers. See also the Note by the UNCTAD Secretariat (2014), The benefits of competition policy for consumers .

    (260)

    Online platforms benefit from asymmetry of information (they dispose of large data sets compared to consumers). Platforms’ analytical capacity gives them the possibility to use advanced algorithms and machine learning techniques to facilitate targeting, discriminatory practices and behavioural manipulation. BEUC considers that such practices can have  an impact on demand and distribution of wealth  – “the most vulnerable consumers might end up paying higher prices than under a competitive price scenario (when personalisation is combined with commercial practices seeking to increase the individual consumer’s willingness to pay). They may also be used to target biases and reinforce existing or desired viewpoints with the aim of keeping users engaged with the firm’s platform so as to generate advertising revenues .” BEUC (2019), The Role of Competition Policy in Protecting Consumers’ Well-being in the Digital Era .

    (261)

    Cf. General Court in Cisco and Messagenet.

    (262)

    In fact, higher advertising prices represent increased costs to the companies producing goods and services which are purchased by consumers. These costs are expected to be passed through to consumers in terms of higher prices for goods and services, even if the downstream market is highly competitive.

    (263)

         See Section 6 of CMA report on Online platforms and digital advertising .

    (264)

    See IAB Europe AdEx Benchmark 2019 Report .

    (265)

    See Section 6 of CMA report on Online platforms and digital advertising .

    (266)

    As explained in Sections 5.3.3.3 and 5.3.3.4, the inclusion of additional core platform services in the scope could also imply the inclusion of additional obligations that would relate to those services and additional gatekeepers that would be active in those services.

    (267)

         See detailed explanation in Annex 3 to the Impact Assessment.

    (268)

         For sub-option 2-A, five to seven out of the 15 gatekeepers would be identified on the basis of the quantitative threshold; the corresponding number would be 10 to 15 for sub-option 1-B.

    (269)

    Annex 3 to the Impact Assessment aims at quantifying benefits for the preferred option. It seems difficult to distinguish however which benefits could be attributed to each of the options. It appears therefore that a quantitative comparison of benefits would not be sufficiently reliable.

    (270)

         Cross-border e-commerce in Europe was worth EUR 143 billion in 2019, with 59% of this market being generated by online marketplaces. This is projected to increase to 65% in 2025 ( Ecommerce News Europe (2020 )).

    (271)

         In line with the impact assessment requirements an attempt was made by JRC (see Annex 4.2 to the Impact Assessment) to quantify consumer surplus which would stem if the measures foreseen under the preferred option were to be adopted and implemented. It is important to stress however, that such quantification remains a highly theoretical exercise; this is the reason why the qualitative assessment of implications for consumers should be considered as a more reliable analysis of the impacts of a regulatory intervention.

    (272)

     Higher investment in R&D in the ICT sector in EU27 leads to an overall increase in the EU27 income between 0.09% to 0.17% of 2014 EU GDP, this is between EUR 12 billion and EUR 23 billion; input-output micro-econometric modelling, See Annex 3 to the Impact Assessment.

    (273)

    In addition, it would always be possible under Option 2 to also include new digital services in scope during the review of the Regulation, which could possibly take place every three years.

    (274)

         See H. Schweitzer (2020), The New Competition Tool: Its institutional set up and procedural design , Chapter II.

    (275)

         Ibid, at Chapter V and Chapter X.

    (276)

    See R. Whish (2020), The New Competition Tool: Legal comparative study of existing competition tools aimed at addressing structural competition problems, with a particular focus on the UK’s market investigation tool , at Chapter 4 and Chapter 7.

    (277)

    The P2B Regulation applies to all online intermediation services and all online search engines, regardless of their size. Given the relatively wide scope of the P2B Regulation, which includes many services that are provided locally, the regulation is enforced at the level of Member States. Given that the Regulation mainly requires the relevant providers to engage in one-off actions that simultaneously benefit their entire user bases (e.g. providing transparency in general terms and conditions, or putting in place an internal complaint-handling mechanism), the regulation’s dual private and public enforcement mechanism is geared towards finding ‘systemic’ breaches of the regulation, which generally will not require in-depth investigations and economic analyses by regulators. For example, a regulator or commercial court will likely only find that an online intermediation services provider breached the obligation to provide internal complaint-handling if they can establish that it is entirely absent or that a pattern exists of unfairly rejected complaints. 

    (278)

    See for example Directive (EU) 2019/633 of the European Parliament and of the Council of 17 April 2019 on unfair trading practices in business-to-business relationships in the agricultural and food supply chain . To improve farmers’ and small and medium sized businesses’ position in the food supply chain, the EU adopted this legislation banning certain unfair trading practices. These include (but are not limited to): late payments for perishable food products, last minute order cancellations, unilateral changes to contracts, refusal to enter into a written contract, returning unsold or wasted products or payment for buyer’s marketing.

    (279)

         The possibility of false positives under Option 2 is partially addressed by the fact that the provider of core platform services would be able to present, in exceptional circumstances, serious and substantiated arguments to demonstrate that it does not fulfil the objective requirements for a gatekeeper.

    (280)

         Such relevant elements would include: (i) the size, including turnover and market capitalisation, operations and position of the provider of core platform services; (ii) the number of business users depending on the core platform service to reach end users and the number of end users; (iii) entry barriers derived from network effects and data driven advantages, in particular in relation to the provider’s access to and collection of personal and non-personal data or analytics capabilities; (iv) scale and scope effects the provider benefits from, including with regard to data; (v) business user or end user lock-in; and (vi) other structural market characteristics.

    (281)

          https://platformobservatory.eu/app/uploads/2020/07/ProgressReport_Workstream_on_Measurement_and_ Economic_Indicators_2020.pdf .

    Top

    Brussels, 15.12.2020

    SWD(2020) 363 final

    COMMISSION STAFF WORKING DOCUMENT

    IMPACT ASSESSMENT REPORT

    ANNEXES

    Accompanying the document

    Proposal for a

    REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

    on contestable and fair markets in digital sector (Digital Markets Act)

    {COM(2020) 842 final} - {SEC(2020) 437 final} - {SWD(2020) 364 final}


    Annex 1: Procedural information

    1.Lead DG, Decide Planning/CWP references

    Three Directorates-General are in the lead for this impact assessment. These are the Directorate-General for Competition (DG Competition), the Directorate-General for Communications Networks, Content and Technology (DG Connect) and the Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs (DG Grow).

    The impact assessment compiles information on two projects:

    -Initiative for a New Competition Tool, led by DG Competition and registered in Decide as PLAN/2020/7913; and

    -Initiative for a Digital Services Act package: ex ante regulatory instrument of very large online platforms acting as gatekeepers, led by DG CNECT and DG GROW and registered in Decide as PLAN/2020/7452.

    2.Organisation and timing

    The inception impact assessments for both initiatives were published on 2 June 2020. These inception impact assessments set out the background of the initiatives as well as their purpose and scope. The inception impact assessments also presented the consultation activities that would be conducted by the Commission (notably a public consultation, external support studies, exchanges with dedicated stakeholders and, for the New Competition Tool, a targeted consultation of the national competition authorities). The inception impact assessments also explained the data collection methodology that would be followed to gather relevant information for the purpose of the impact assessment.

    The impact assessment was carried out in close cooperation with other interested Commission services. The inter-service steering group (‘ISSG’) set up for that purpose comprises representatives of the Directorates-General FPI, JRC, HOME, ENV, FISMA, AGRI, JUST, EAC, TRADE, RTD, TAXUD, ENER, MARE, SANTE, EMPL, MOVE, and ECFIN, the EEAS, as well as the Secretariat-General and the Legal Service, which are associated by default to any such initiative.

    The impact assessment for the New Competition Tool, was carried out in close cooperation with the NCAs, which were consulted on the milestones for the evaluation study and the study on consumer purchasing behaviour. The different milestones of the evaluation phase are reflected in the table below:

    Timing

    Step

    2 June 2020

    Launch of the initiative in the Commission’s Decide

    2 June 2020

    Publication of the Inception Impact Assessments (4-week comment period) and launch of the open public consultation (2 June until 8 September 2020)

    3 July 2020

    Upstream Meeting with the Regulatory Scrutiny Board on the Digital Services Act

    10 September 2020

    Upstream Meeting with the Regulatory Scrutiny Board on the New Competition Tool

    6 October 2020

    ISSG Meeting to consult on the draft Impact Assessment

    8 October 2020

    Publication of the following documents concerning the NCT pillar:

    -Summary report of the public consultation

    -Summary of the NCA consultation

    -External support studies

    30 October 2020

    Quality check-list

    4 November 2020

    Consultation of the Regulatory Scrutiny Board

    6 November 2020

    First (negative) Opinion by the Regulatory Scrutiny Board

    10 December 2020

    Second (positive) Opinion by the Regulatory Scrutiny Board

    3.External support studies

    3.1.External support studies conducted in the context of the Digital Services Act (‘DSA’) Package: ex ante regulatory instrument of very large online platforms acting as gatekeepers (‘gatekeeper instrument’)

    1.

    2.

    3.

    3.1.

    3.1.1.Impact Assessment Support study

    DG CNECT commissioned a support study for the an impact assessment, Platforms with Significant Network Effects Acting as Gatekeeper, run by Consortium composed of ICF (lead), WiK and CEPS, with a budget of EUR 597 850 [VIGIE 2020-00630]. The study had three objectives:

    1.Providing a structured analysis of (i) the issues raised by digital platforms with strong data-driven network effects and (ii) analysis of the ability of current regulation (e.g. competition law; P2B regulation) to address these issues (regulatory failures).

    2.Scoping the parameters of intervention which match the problem analysis (identify economic players in scope of the initiative, and criteria relevant to identify these players).

    3.In agreement and cooperation with Commission services, help the identification of possible policy options, and provide evidence in analysing their impact.

    3.1.2.Support study to the Observatory for the Online Platform Economy

    DG CNNECT and DG GROW commissioned a support study to the Observatory for the Online Platform Economy run by a consortium composed of PPMI (lead) with Open Evidence, IW and Rand Europe (SMART 2018/0034), with a budget of EUR 830 000.

    The contractor produced the following analytical papers (AP):

    ·AP1:Differentiated treatment (IW)

    ·AP2:Platform data access and secondary data sources (PPMI)

    ·AP3:Transparency in the business-to business commercial relations in the online advertising market (Open Evidence)

    ·AP4: Significant Market Status (RAND) 

    ·AP5: Business user and third-party access to digital platform data (PPMI)

    ·AP6: The main obstacles and opportunities for multihoming (PPMI)

    ·AP7: The structure of the online platform economy post COVID-19 outbreak (Open Evidence)

    ·AP8: Developments concerning B2B platforms and emerging issues (RAND)

    3.2.External support studies conducted in the context of the New Competition Tool

    DG COMP commissioned expert advice reports by renowned academics to inform the most appropriate set-up of the NCT, including:

    a.A study by Massimo Motta and Martin Peitz on structural competition problems in digital and other markets, as well as a possible intervention trigger for the NCT based on the commonalities between the scenarios identified; 1

    b.A study by Alexandre De Streel and Pierre Larouche on the interplay of the NCTs and sector-specific regulation, as well as possible ways to ensure complementarity between both; 2

    c.A study by Heike Schweitzer on the institutional and procedural set-up of the NCT, with the aim of ensuring effective and timely intervention, while safeguarding the right to be heard and judicial review; 3 and

    d.A comparative study by Richard Whish of existing market investigation tools, with a particular focus on the UK Competition and Markets Authority’s market investigation reference tool. 4

    DG COMP also contacted three members of the Economic Advisory Group on Competition Policy (EAGCP), namely Gregory Crawford, Patrick Rey and Monika Schnitzer, who prepared an economic evaluation of the NCT. 5

    These reports are referenced in Annex 5.1.

    4.Consultation of the RSB

    The meeting of the Regulatory Scrutiny Board (‘RSB’) took place on 4 November 2020. The outcome was a negative opinion, issued on 6 November. Following a substantial rethinking of the document in light of the comments of the RSB, the text was resubmitted for a second time. The RSB delivered its second positive opinion with reservations on 10 December 2020.

    The following table provides information on how the comments made by the RSB in its first negative opinion were addressed in this Staff Working Document:

    RSB comments

    Actions taken

    (1) The impact assessment is unfinished. Work on integrating the two pillars of the initiative is incomplete.

    We acknowledge that the first submitted impact assessment was unfinished, and have since then fundamentally reworked the approach, notably on the basis of the feedback from the RSB. In particular, the original two-pillar structure of the impact assessment was abandoned and integrated into a ‘single track’ approach to the problem definition and options’ assessment. The impact assessment further substantiates the internal market nature of the issues at stake and of the measures considered to address those.

    The entire Impact Assessment has been thoroughly overhauled, and no longer contains any distinctions between any pillars present in the previous version of the Impact Assessment.

    As a result of the substantial rethinking of the problem definition – no longer based on a distinction between pillars – genuinely unified policy options ex novo have been created for the impact assessment. This new structure has allowed an objective comparison of the three options presented, in line with the Better Regulation requirements.

    (2) The report does not sufficiently justify the restriction of its scope to digital markets. It does not justify the selection of platform services within the digital sector nor does it clarify the concept of gatekeeper platforms.

    A new Section 1.2 has been included in the impact assessment in order to explain the focus of the initiative not only on digital markets, but more specifically on (i) selected core platform services of (ii) certain gatekeepers engaging in (iii) certain behaviour. Moreover, the evidence on the special incidence of market failures on digital markets has been further added in the text.

    An entire new section 5 explains the main parameters that set out the options space, including explicitly a discussion of the different parameters that set the scope. This includes a clarification and a rationale of the notion of gatekeeper platforms, and provides an explicit overview of which services are in scope and which companies might qualify depending on the choice of criteria.

    (3) The report does not provide an integrated problem definition for the initiative. It does not appropriately describe the shortcomings the initiative intends to address and does not provide a proper evidence base for them.

    The problem definition has been redone from scratch in order to present an integrated problem definition for the initiative.

    In addition to doing away with the Pillar I and Pillar II classification, the distinction between problem drivers and problems as well as their interlinkage have been clarified. Moreover, for the purpose of clarity, the problem definition part no longer distinguishes between existing and emerging market failures.

    Following this new problem definition, the impact assessment now presents a single, more coherent intervention logic, reflecting problems, their underlying drivers and policy objectives pursued.

    The evidence base for the problems identified – in particular as regards the specific unfair practices – is now explicit, notably in the tables in Section 5, and in updated Annexes (notably 5.6); concrete examples of these problems and of their underlying drivers and evidence have been included.

    (4) The report does not provide policymakers with real choices on the different policy options. It does not provide a full range of options and it does not develop these in sufficient detail. It therefore cannot assess their impacts on different stakeholders.

    The new Impact Assessment provides a new structure for the Options design. It sets out upfront the main parameters that determine the options range, and their trade-offs. A completely new set of options (with sub-options as alternatives) is now presented that provide genuine alternatives within the parameters of the problem definition.

    The new Impact Assessment also explains in greater detail the discarded options and why they have been discarded.

    1.Option 1 is a non-dynamic option with a set of self-executing obligations addressing clearly defined unfair practices by gatekeepers designated solely on quantitative thresholds in specific core platform services. This option contains no dynamic elements, but is presented with distinct two sub-options on scope as distinct alternatives, on the basis of different thresholds. Sub-option 1-A is presented as a sub-option with a small perimeter of gatekeeper companies in scope (some 5-7 companies in scope) while sub-option 1-B contains a wider scope of gatekeeper companies (some 10-15 gatekeeper companies), based on a lower quantitative threshold.

    2.Option 2 is a semi-flexible option, combining a set of self-executing obligations with some degree of flexibility, notably through a dialogue on some of the obligations, through a mechanism for updating the practices and obligations, and a mechanism designating gatekeepers based on a combination of quantitative and qualitative thresholds and including the designation of emerging gatekeepers. Again, this semi-flexible option is presented with two sub-options that reflect alternatives on the scope platforms. Sub-options 2-A and 2-B are sub-options on this semi-flexible option, following the same distinction on the quantitative threshold as Option 1.

    3.Option 3 is a fully flexible option providing for a dynamic updating mechanism allowing for the inclusion of additional core platform services and of additional obligations where following a market investigation such an inclusion is considered appropriate and justified, and where the designation of gatekeepers is based only on qualitative (not quantitative) thresholds.

    The impact section is completely updated and revised, and now includes more detailed assessments and comparisons of each option against the baseline and against each other.

    (5) The report fails to assess all risks and trade-offs of the policy options. It does not clarify the extent to which the preferred option, and in particular the interaction between the regulatory measures and the market investigation regime, is coherent and futureproof.

    Section 5 now outlines the trade-offs upfront that motivate the choice and design of options. Section 7 now explicitly compares the new options against each other in terms of trade-offs, and motivates a preferred option.

    The substantial changes in the design of the Impact Assessment – including the disappearance of the pillars – has led to a unified set of measures no longer distinguishing between regulatory measures and market investigation.

    A fresh quantitative and qualitative assessment has been carried out for the preferred option. Annex 3 also includes a qualified and quantified overview of costs associated with each of the three options. Special attention has been given to the monitoring framework of the intervention under the preferred option.

    The following table provides information on how the comments made by the RSB in its second positive opinion with reservations were addressed in this Staff Working Document:

    RSB comments

    Actions taken

    (1) The report should make clearer how the problem drivers may lead to the identified negative outcomes. It should consider the negative consequences of curtailing the size advantages following from network economies and economies of scale for consumers. It should better distinguish problems relating to size advantages from the monopolisation of data and the imposition of market rules like exclusive dealings.

    A conclusion was added on drivers’ effects (Section 2.3.3) explaining in more detail how the problem drivers lead to negative outcomes. Section 6 has been updated, where relevant, to better reflect the link between problem drivers and identified impacts.

    Section 6.7 is now considering the point of curtailing gatekeepers’ size advantages and is assessing the impact on consumers.

    (2) The report should better justify the identification and selection of the core platform services. It should present evidence of what determines persistent misuse of gatekeepers’ power vis-à-vis dependent business users and customers. It should more convincingly demonstrate for each of the selected core platform services that the identified weak contestability has negative effects in terms of higher mark-ups, lower quality of service, or reduced innovation. The report should better justify why other platform services, such as content streaming providers, would not meet the selection criteria.

    Table 2 (in Section 5.2.2) illustrates all practices resulting from misuse of gatekeepers’ power vis-à-vis dependent business users and customers. The Table specifies now under each example of unfair practice the type of behavior concerned (e.g. data-related, size-related, dependence-related, etc.). The evidence provided for each practice was also strengthened.

    Table 1 has been included in Section 5.2.1 explaining the main features of each of the eight core platform services in scope. Reference to several points of evidence describing the unfair practices and weak contestability in these services was also added, as well as a list of the most common practices for each services.

    As an introduction to Table 1 it is also explained the limitations as regards a granular assessment of mark-ups and innovation for each of these services. Nevertheless, some additional information was added in relation to these variables.

    Finally, some justifications are provided on the question why video streaming content services and industrial B2B platforms do not meet the criteria.

    (3) The report should better define and justify the measures covered under the options. It should demonstrate why the proposed set of cumulative quantitative thresholds (under the ‘non-dynamic’ and ‘semi-flexible’ options) can be considered as a robust and reliable trigger across all selected core platform services for the (quasi-automatic) designation of gatekeepers and the imposition of obligations. It should better explain why a market investigation is not deemed necessary or proportionate in these situations.

    In Section 5.2.1 more detail is now provided about the reliability of each quantitative proxy for the determination of the status of gatekeeper.

    Some additional data was collected and analysed during the period of review by the RSB. This resulted in a slight change in the value of the thresholds considered, without changing the number of gatekeepers under each sub-option. This also shows that small variations in the turnover and number of users do not have a significant impact in the designation of gatekeepers.

    The description of Option 2 in Section 5.3.2 (as well as in Section 8) now explains in more detail a flexible element that minimises the risks associated to the possible lack of robustness and reliability of the quantitative triggers. Also circumstances where a market investigation can take place when there is a doubt about the application of the quantitative thresholds are explained. This is also reflected in the comparison about effectiveness.

    (4) From a future proofing perspective, the report should explain why the possibility of updating the list of core platform services following a market investigation was discarded for the ‘semi-flexible’option, while maintained as a key element for the ‘fully flexible’ option. As regards the ‘fully flexible’ option, it is not clear why certain beneficial guidance elements (including indicative quantitative thresholds), which could have provided further legal clarity, have not been considered in the design of this option.

    Section 5.3.2.5 now includes an explanation about the possibility under Option 2 to update the list of core platform services, namely in the context of the review of the Regulation. Any additional flexibility level would defeat the legal certainty created by a fixed scope of core platform services.

    Section 5.3.3.3 now includes an additional explanation as to why even guiding thresholds would defeat the purpose of Option 3 by undermining its inherent flexibility.

    (5) The report should clarify the distinction between the ‘semi-flexible’ and ‘fully flexible’ options in terms of the obligations that can be added following a market investigation. It should also explain, where the market investigation powers and process deviate from the envisaged model and rules under Regulation 1/2003.

    Section 5.3.3.4 now clarifies the distinction between Option 2 and Option 3 in terms of the obligations that can be added.

    Sections 5.3.2 and 5.3.3 now include several paragraphs comparing the enforcement powers and processes of Regulation 1/2003 with those of the instrument that is the object of this impact assessment.

    (6) The report should improve the comparison of options in terms of effectiveness and benefits (including in summary table 5) given that the ‘fully flexible’ option seems to score best in minimising false negatives/positives and future proofing. The report should clarify the relative weight given to the different assessment criteria (e.g. legal certainty vs. flexibility vs. speed). It should better substantiate the assumption that the ‘fully flexible’ option would lead to a higher number of large platforms being covered, and why the decisions taken under this option would be ‘arbitrary’ (given that they would be based on market investigation).

    The relative importance of the tree parameters (speed, legal certainty and flexibility) has been specified in Section 7.1, and reflected throughout the options’ comparison carried out in Section 7.

    The narrative has been further specified and substantiates better the scoring in Table 5.

    The effectiveness and efficiency criteria have been reviewed, and adjusted whenever needed.

    The different assumptions made have been further explained in Sections 5.3 and 7.

    (7) The report should better explain the limitations of the methodology used. When presenting evidence the report should differentiate more clearly between cases which are still being investigated or pending and the established case law. The Board notes the estimated costs and benefits of the preferred options in this initiative, as summarised in the attached quantification tables.

    An explanation of the sources of evidence and their limitations was included before Table 2.

    Table 2 was updated making clear which evidence comes from established case law, which comes from cases being investigated and which one comes from public authorities’ reports.

    5.Other evidence, sources and quality

    Reports by the expert group for the Observatory on the Online Platform Economy 6

    §Measurement of the Online Platform Economy

    §Differentiated treatment

    §Data in the Online Platform Economy

    Published for feedback on 9 July. 7  

    Studies supporting the P2B initiative with relevant input for this proposal

    §ECORYS, Business-to-Business relations in the online platform environment FWC ENTR/300/PP/2013/FC-WIFO, 2017 (commissioned by DG GROW & DG CNECT ).

    §ERNST&YOUNG, Contractual Relationships between Online Platforms and Their Professional Users, SMART 2017/0041 (commissioned by DG CNECT).

    §VVA, Data in platform-to-business relations, November 2017 (commissioned by DG GROW).

    §GfK et al., Behavioural study on advertising and marketing practices in online social media, June 2018 (commissioned by DG JUST).

    Research conducted by the Joint Research Centre

    §B. Martens (2020), An economic perspective on data and platform market power, JRC Digital Economy Working Paper 2020-09. 8

    §B. Martens, M. Sobolewski, & N. Duch-Brown (2020), Market power in app stores, JRC Digital Economy working paper 2020-10JRC. 9  

    §B. Martens & N. Duch-Brown (2020), From platforms to ecosystems: The role of data in linking seemingly separate markets, JRC Digital Economy Working Paper 2020-11. 10  

    §N. Duch-Brown (2020), Entry and contestability in online platform markets, JRC Digital Economy working paper 2020-12. 11

    §JRC, Quality discrimination in online multi-sided markets, 2017. 12

    §JRC, Platform to business relations in online platform ecosystems, 2017. 13

    §JRC, The Competitive landscape of online platforms, 2017. 14  

    §JRC, An Economic Policy Perspective on Online Platforms, 2016. 15

    Other data sources

    §Dealroom economic report, Global platforms and marketplaces custom policy intelligence, April 2020.

    §Report on the Monitoring Exercise Carried out in the Online Hotel Booking Sector by EU Competition Authorities in 2016.

    CERRE reports and events:

    §The role of data for digital markets contestability, September 2020. 16  

    §Seminar of 4 March 2020, How should Europe address gatekeeping platforms. 17  

    § Market Definition and Market Power in the Platform Economy , May 2019. 18

    §Implementing effective remedies for anti-competitive intermediation bias on vertically integrated platforms, October 2019. 19

    §Big data and competition policy, February 2017. 20

    §Internet Platforms and Non-Discrimination, December 2017. 21

    Sources from the Member States

    §Austrian position Paper, Digitalisation and Competition Law, June 2020.

    §Dutch Competition Authority (ACM), Market Study into mobile app stores, April 2019. 22

    §Digital gatekeepers - Assessing exclusionary conduct – a study by e-Conomics commissioned by the Dutch government, October 2019. 23  

    §Dutch position, Future-proofing of competition policy in regard to online platforms, May 2019. 24

    §Seminar on the regulatory challenges posed by ‘structuring platforms’ organised on 24 February 2020 in Paris by the French government.

    §Non-paper by the French Ministry of Economy and Finance, Regulating structuring digital platforms in favour of competition and innovation in the digital economy. 25

    §French paper, Regulation of structuring platforms: the case of operating systems and app stores, ‘gatekeepers’ of our devices.

    §Position of French Competition Authority. 26  

    §ARCEP’s working paper on the structuring platforms, December 2019.

    §The Report by the German Competition Commission, 2020. 27  

    §German report, A New Competition Framework for the Digital Economy, 9 September 2019. 28

    §German Ministry for Economic Affairs and Energy Study, Modernising the law on abuse of market power (2018). 29

    §Position paper by German telecom and competition authorities on monitoring digital platforms, May 2020. 30  

    §DE Monopolies Commission Policy Brief 4/2020, 10th amendment to the Competition Act – meeting challenges in digital and regional markets!. 

    §Italian AGCM/AGCOM/DPA Report on big data and policy recommendations – 20 Feb 2020. 31

    §Joint memorandum of the Belgian, Dutch and Luxembourg competition authorities on challenges faced by competition authorities in a digital world (2 October 2019).

    §Economic Affairs Ministries of DE, FR, PL: Modernising EU Competition Policy.

    §Spanish National Commission on Markets and Competition (CNMC) contribution to conference Shaping competition policy in the era of digitisation. 32

    Sources from non-EU states and international organisations

    §US House of Representatives Majority Staff report, Investigation of Competition in Digital Markets, October 2020. 33

    §Furman report, Unlocking digital competition, Report of the Digital Competition Expert Panel (the UK), March 2019. 34

    §OFCOM, Online market failures and harms, An economic perspective on the challenges and opportunities in regulating online services, October 2019. 35

    §CMA, Online platforms and online advertising – Market study final report, July 2020. 36

    §Stigler Center Report, George J. Stigler Center for the Study of the Economy and the State The University of Chicago Booth School of Business, July 2019. 37

    §ACCC Report, Digital Platforms Inquiry, June 2019, 38 and Interim Report of September 2020. 39

    §Japanese Fair Trade Commission Report regarding trading practices on digital platforms, October 2019, 40 and the Interim Report on the Evaluation of Competition in the Digital Advertising Market. 41

    §BEUC, The role of competition policy in protecting consumers’ well-being in the Digital Era, October 2019. 42

    §OECD, Rethinking antitrust tools in multisided markets, 2018. 43

    External expertise

    The European Commission sought external expertise before drafting this Impact Assessment. Views of the experts have contributed to the problem framing and evidence collection strategy. Consultation of experts listed here below does not imply automatic endorsement on their side of the Impact Assessment report.

    §JRC expert panel

    At the request of DG CNECT, the Joint Research Centre (JRC) of the European Commission established a high-level Panel of Economic Experts on Platform issues with a mandate to produce a report with an economic opinion on the proposed ex-ante regulatory tool, based on existing economic research and evidence. The members of the Panel are well-known economists with a strong track-record in economic research on digital platforms and competition policy. They include: Luis Cabral, Justus Haucap, Geoffrey Parker, Georgios Petropoulos, Marshall Van Alstyne and Tommaso Valletti. Panel members are independent and contribute pro bono to the report. 44

    §Philip Marsden, workshop of 29 January 2020.

    §Paul Belleflamme, workshop of 29 January 2020.

    §Stephen Adshead, workshop of 29 January 2020.

    §Francesco Decarolis, workshop of 29 January 2020

    §Daniel Knapp, workshop of 29 January 2020.

    §Wolfgang Kerber, Updating Competition Policy for the Digital Economy? An Analysis of Recent Reports in Germany, UK, EU, and Australia, September 2019. 45

    §Alexandre de Streel and Peter Alexiadis, Designing an EU Intervention Standard for Digital Platforms, Robert Schuman Centre for Advanced Studies, Research Paper No. 2020/14 . 46

    §Experts for the Observatory on the Online Platform Economy. 47

    Annex 2: Stakeholder consultation

    1.The stakeholders engagement strategy

    This annex presents the results of the consultation activities performed in the context of the Inception Impact Assessments: (i) the Digital Services Act (‘DSA’) package: ex ante regulatory instrument of very large online platforms acting as gatekeepers; 48 and (ii) the New Competition Tool (‘NCT’). 49  

    Given the breadth of the questions asked, both consultations were conducted separately and the results are presented separately below. However, since the outset, both consultations were aimed at complementary solutions by “ensur[ing] a joint analysis of the results”, “with a view to exploring synergies and ensuring consistency on the policy options pursued, in particular as regards possible remedies and enforcement. 50

    As presented in the initiatives’ Inception Impact Assessments, the objective of both consultations was to consult as widely as possible through various means in order to deliver an in-depth impact assessment of the different policy options and their perceived impact on the Commission’s ability to improve effective competition in digital markets.

    As will be presented below, this objective was largely met. To illustrate this, it is worth noting that a total of 3051 respondents participated in both open public consultations. These respondents represented all possible categories of stakeholders.

    In developing the stakeholder engagement strategy for the both initiatives, the merged stakeholder mapping included:

    1.Businesses and their associations, including digital players (online intermediaries, other digital players, third parties involved in the ecosystem around digital services);

    2.Trade associations and labour unions;

    3.Consumers, including users of digital services;

    4.Civil society and consumer organisations;

    5.National authorities including law enforcement, competition, data protection and consumer protection authorities, and other relevant regulatory bodies in Member States and, to the extent possible, in regions and municipalities;

    6.Academics from the technical, legal and social science communities;

    7.International organisations; and

    8.General public, in particular through the open public consultations.

    2.Consultation activities in the context of the Digital Services Act (‘DSA’) package: ex ante regulatory instrument of very large online platforms acting as gatekeepers (‘Gatekeeper Instrument’)

    2.1Consultation on the Inception Impact Assessment

    The Inception Impact Assessment was published on 2 June 2020 with the deadline for comments running until 30 June 2020. During this period, 85 formal submissions were received from a variety of stakeholders (e.g. online platforms; business associations; telecom operators; media publishers; civil society; consumers).

    The largest group of respondents were from the private sector, amounting to more than half of all respondents. Among the private sector, online platforms constituted the largest group of respondents (one third of all respondents).

    Overall, a two-third majority of stakeholders expressed its (general) support of Gatekeeper Instrument, with a one-fifth minority explicitly opposing its introduction. Although most replies were of a preliminary nature, many focused on ‘option 3’ with mixed support for and opposition to blacklisted practices and/or a case-by-case approach.

    Online platforms are split on the issue, with the majority of large online platforms and/or their representative associations questioning the need for a Gatekeeper Instrument. On the other side, many small and medium sized platforms, in particular those that are business users of large online platforms, expressed their support for a Gatekeeper Instrument.

    Market operators from some specific sectors (e.g. telecoms; financial services) have expressed equally strong support for a Gatekeeper Instruments and were specifically referring to the ineffectiveness of ex post competition rules in addressing some of the emerging issues. Having said that, some telecom operators referred to the relatively static nature of a blacklist/whitelist approach, which they therefore consider to not always be an appropriate and effective solution in a very dynamic online platform environment.

    National Authorities expressed their support of a Gatekeeper Instrument and the need for an approach on an EU level to avoid regulatory fragmentation, whilst emphasizing the importance of involving the responsible national government representatives in the legislative project in advance.

    Civil society and media publishers also strongly supported a Gatekeeper Instrument. Both called for an adequate degree of transparency in the market as well as the guarantee of a certain degree of media diversity and the respect of consumers' autonomy and choice.

    2.2Consultation on the Impact Assessment 51

    The open public consultation on the DSA, including the Gatekeeper Instrument was launched on 2 June 2020 and open for feedback until 8 September 2020. During this period, a total of 2863 contributions were received, of which 2128 citizens, 621 organisations and 59 administrations represented stakeholders from across all Member States.

    In terms of geographical distribution of respondents, the majority of answers came from respondents from Germany (28%) followed by the United Kingdom (21%) and France (14%). Other Member States represented in higher proportions are Belgium (9%), Netherlands (4%) and Austria (3%). Member States contributed with response rates lower than 3%. Among respondents originating outside the EU, the highest share comes from respondents from the United States of America (3%).

    Among respondents, the vast majority fully agree (71%) and agree to a certain extent (20%) that there is a need to consider dedicated regulatory rules to address negative societal and economic effects of gatekeeper power of large platforms. The majority of stakeholders considers that, while some of the issues connected to gatekeeper powers can potentially be addressed by improving the efficiency of competition law enforcement through procedural and/or organisational changes, there are restrictions that cannot be overcome with competition law enforcement.

    The vast majority of respondents (85% of those who replied to the relevant question) considers that dedicated rules on platforms should include prohibitions and obligations for gatekeeper platforms. Most of the stakeholders suggest that, rather than having certain practices categorically prohibited, the Commission should scrutinise certain practices and prohibit them on a case-by-case basis in circumstances when they are most likely to have detrimental effects. It is also suggested that remedies could be more procedural in nature rather than prescribing a given course of conduct.

    According to the vast majority of stakeholders, the proposed list of problematic practices, or ‘blacklist’, should be targeted to clearly unfair and harmful practices of gatekeeper platforms; specific enough to avoid confusion of what is and is not permitted; adaptable to a dynamic, fast moving sector; and specific to certain gatekeepers as they would otherwise risk hurting smaller players trying to compete with them.

    The unfair practices listed by the respondents cover exclusionary conducts, exploitative conducts and transparency-related problems, such as: self-preferencing; lack of data sharing and accumulation of data; limited data portability and data access due to lack of interoperability; imbalance on how the revenues are split between platforms and right owners in relation to user generated content; imposition of unfair and unilateral terms and conditions; imposition of exclusionary terms and conditions for attaining and/or retaining access; cross-financing and cross-subsidising of otherwise unprofitable subsidiary companies; and default settings which adversely impact customer choice.

    The respondents consider all the characteristics mentioned in the questionnaire (large user base, wide geographical coverage, large share of total market revenue, impact on a certain sector, exploitation of strong network effects, leverage of assets to enter new areas of activity, raising of barriers to entry, accumulation of valuable and diverse data and information, lack of alternative services, lock-in of users) are relevant in determining the gatekeeper role of large online platforms.

    Respondents among platforms show diverse views on what would define a gatekeeping position. Some platforms argue that incorporating different services into a platform’s offering says little about the strength of a platform, as it is also the case with the ability to leverage assets from one market to another. It is suggested that gatekeeper designations should be business model agnostic, gatekeeper assessments should be reviewed periodically, gatekeeper designations should apply to identified activities in specific markets, and some rules ought to apply on a sector-wide basis.

    In general, stakeholders of all categories point out the need to ensure a high level of coherence and legal certainty, the criteria used should be transparent, objective and easily measurable. At the same time, stakeholders also state that a one-size-fits-all approach might be unfeasible, and that a merely cumulative approach might not be sufficient. Users mostly refer to a combination of both quantitative and qualitative criteria.

    2.3Summary of the targeted consultation of Member States

    The e-Commerce Expert group was set up in 2005 to coordinate with Member States and exchange views on issues relating to electronic commerce and related services, facilitate the exchange of information, experiences and good practices in the area of electronic commerce in order to advise and assist the Commission in the preparation of legislative proposals and policy initiatives.

    During the 21st meeting of the Expert group on 26 May 2020, the preparation of DSA package was presented in detail and discussed with the Member States. The Commission provided a presentation on the context and thinking behind the Gatekeeper Instrument, outlining possible options that might be elaborated in the Impact Assessment and emphasising that the final options will need to be looked at very carefully.

    Throughout the impact assessment, the Commission also met bilaterally with stakeholders that requested this, primarily in the context of the public consultation and the feedback period for the inception impact assessment. These meetings were requested by the parties concerned and aimed primarily at discussing the submissions made by stakeholders, either in the context of the public consultation or outside of it

    During the following Questions & Answers session, Member States welcomed the details provided by the Commission, mentioned ongoing national initiatives and discussions, and asked complementary questions on the possible scope of the proposed tool and evidence base.

    2.4Summary of targeted stakeholder workshops

    1.

    2.

    2.1.

    2.2.

    2.3.

    2.4.

    2.4.1.EU Observatory for the Online Platform Economy

    Workshop, January 2020

    On 29 January 2020, the Commission organised a closed Workshop to support its policy making in the area of online platform economy. The participants included the experts from the expert group for the Observatory on the Online Platform Economy, the Commission Observatory staff from DG CNECT, GROW, COMP and JUST and invited external experts: both from academia (Paul Belleflamme, Francesco Decarolis); industry (Daniel Knapp, Stephen Adshead ) and regulatory authorities ( Philip Marsden).

    The Workshop was devoted to two main topics: market power and transparency in online advertising. The presentation and debate that followed fed into the reports prepared by the expert group and evidence supporting this Impact Assessment.

    2.4.2.Expert Panel – Support Study for the IA

    On 28 July 2020 and 10 September 2020 ICF, WIK-Consult GmbH, Cullen International, and CEPS organised a high-level academic expert panels to support the Commission in the preparation of the Impact Assessment of platforms with significant network effects acting as gatekeeper. The members of the academic panel were selected in consultation with the Commission by virtue of their in-depth experience in issues relevant to the governance of digital platforms and markets. The panel included the following experts: Martin Kenney, Jan Krämer, Marshall Van Alstyne, William E. Kovacic, Pierre Larouche, Giorgio Monti.

    The expert discussed a wide range of issues concerning the platforms with significant network effects acting as gatekeepers, including among others objectives of the future regulatory framework, problems definition, thresholds for intervention, remedies and institutional design.

    2.4.3.Stakeholder Consultation

    The progress reports by the expert group on: (i) Measurement of the Online Platform Economy; (ii) Differentiated treatment; and (iii) Data in the Online Platform Economy, were published for feedback on 9 July. The Commission received nine contributions from citizens, industry associations, platforms and regulatory authorities.

    2.5Other consultation activities

    In addition to the above-mentioned consultations and targeted stakeholder exchanges, the Commission received a number of spontaneous submissions from stakeholders. Some of these contributions were submitted by stakeholders that had participated in the public consultation and were therefore intended to supplement their views with additional evidence. Other submissions were received from EU government bodies and business associations that had not participated in the public consultation. These submissions largely echoed the issues already raised in the different consultation activities.

    Throughout the impact assessment, the Commission also met bilaterally with stakeholders that requested this, primarily in the context of the public consultation and the feedback period for the inception impact assessment. These meetings were requested by the parties concerned and aimed primarily at discussing the submissions made by stakeholders, either in the context of the public consultation or outside of it.

    3.Consultation activities in the context of the New Competition Tool (‘NCT’)

    3

    3.1Consultation on the Inception Impact Assessment

    The Inception Impact Assessment was published on 2 June 2020 with the deadline for comments running until 30 June 2020. During this period, 73 formal submissions were received. The largest group of respondents were businesses and business associations, amounting to more than half of all respondents. Among businesses, technological companies constituted the largest group of respondents.

    Respondents generally agreed that there are structural competition problems that cannot be addressed under the existing competition rules, with some expressing explicit support for an NCT proposal. Respondents expressed different opinions as to whether competition problems should be tackled with competition-based or regulatory tools. Consumer associations pointed out that there is a need for the NCT to complement the current EU toolbox.

    Regarding possible problematic sectors, most views referred to issues relating to digital markets. Most respondents argued that it was less clear which were the structural competition problems outside the digital area that could not be addressed by Articles 101 and 102 Treaty on the Functioning of the European Union (‘TFEU’).

    Given that most respondents did not appear familiar with the investigative processes of similar tools, they questioned how such a tool would work at EU level. Respondents expressing support emphasised that any new intervention tool would require a careful design to ensure legal certainty and procedural safeguards.

    3.2Consultation on the Impact Assessment

    The open public consultation on the NCT was launched on 2 June 2020 and open for feedback until 8 September 2020. During this period, a total of 188 contributions were received, with 154 respondents representing stakeholders from 18 Member States.

    Businesses (68) and their associations (54) represented more than 2/3 of respondents. Other respondents included NGOs, consumer organisations and academic/research institutions. Nineteen contributions were received outside the open public consultation, which largely echoed the issues raised in the contributions to the public consultation. The figures in this summary are based only on contributions to the public consultation submitted through the online questionnaire.

    Respondents generally agreed that there are structural competition problems that Articles 101/102 TFEU cannot tackle or address in the most effective manner. Respondents also generally agreed that an NCT could help address the limits of the existing competition rules.

    More specifically, respondents confirmed that certain market features may lead to structural competition problems. Respondents also confirmed that the examples of structural competition problems set out in the questionnaire, in particular leveraging and monopolisation strategies, as well gatekeepers scenarios and tipping markets, may raise competition concerns that Articles 101/102 TFEU are not suitable or sufficiently effective to address, and that the Commission should be able to intervene in such scenarios. Respondents considered that such structural competition concerns commonly occur in digital markets, while pointing out that there are indications that they are not limited to digital markets.

    As regards the intervention trigger for the NCT, the majority of respondents that expressed a view in this regard considered that such a tool should focus on structural competition problems, thus being applicable to all companies in a market, rather than only to dominant companies or gatekeepers or digital platforms. As regards the scope of application, the majority of respondents considered that such a tool should be applicable to all markets. A majority of respondents that expressed a view also indicated that the tool should not be limited to only markets/sectors affected by digitisation. However, a large number of those respondents who indicated that the tool should apply in all sectors and markets nevertheless provided explanations that mainly highlighted how the tool would be especially beneficial if applied to the problems found in digital markets.

    As regards the interplay with other instruments and policy options, such as those included in the DSA package, there is general support for ex ante rules consisting of obligations and prohibitions for digital gatekeepers in order to address issues in digital markets raised by gatekeeper platforms. Most respondents emphasised that, in order to effectively address contestability issues in digital markets, there is a need for a combined approach, consisting of more than one policy solution. In those respondents’ view, this should include ex ante rules and an enforcement tool applicable to digital markets.

    A more detailed summary of the replies received in the context of the open public consultation on the NCT can be found on DG Competition's website. 52 A full list of supporting materials available on that website is also attached as Annex 5.1 to this document.

    3.3Summary of the targeted consultation of National Competition Authorities

    In the context of the European Competition Network – a network bringing together the Commission, the EFTA Surveillance Authority and all the National Competition Authorities (‘NCAs’) of the EEA – the Commission submitted a questionnaire to gather the views on the NCT within the Network.

    NCAs generally agreed that there exist certain features that may lead to structural competition problems that Articles 101 and 102 TFEU cannot tackle conceptually or cannot address in the most effective manner. The consultation showed a consensus among NCAs with relevant experience that there was a need for a new competition tool to deal with these structural competition problems. More specifically, NCAs pointed out that such a tool should enable the Commission to conduct investigations in markets with structural problems since a case-by-case enforcement against abuses of dominance is not sufficient in the increasingly fast-paced and interconnected economy.

    NCAs with relevant experience were split as to the question in which sectors structural competition problems can occur. According to half of the respondents, structural competition problems may occur in all sectors/markets, whereas others argued that structural competition problems may occur in some specific sectors/markets, including but not limited to digital sectors/markets. NCAs, however, suggested that digital markets were more prominently affected by structural competition problems than other markets.

    NCAs with relevant experience also indicated that a new competition tool to tackle such structural competition problems would only be effective if it were accompanied with adequate and proportionate investigative powers, but also by soft and hard powers to deal with structural competition problems, including possibly imposing structural remedies (e.g. divestitures or granting access to key infrastructure or inputs) where duly justified.

    NCAs with relevant experience generally considered that not adapting existing competition law tools would be at most ‘somewhat effective’, meaning that an ex-ante regulation would in itself not be sufficient to address structural competition problems.

    A more detailed summary of the replies received from NCAs in the context of their consultation on the NCT can be found on DG Competition's website. 53 A full list of supporting materials available on that website is also attached as Annex 5.1 to this document.

    3.4Summary of targeted stakeholder workshops

    Consultation activities have also included the participation of the project team in a number of exchanges with stakeholders across various sectors. Given the particular circumstances of the Covid-19 crisis, all these exchanges took place in a virtual environment.

    First, as is the standard practice concerning pan-European competition policy matters, the Commission organised two meetings in the context of the European Competition Network (‘ECN’) in order to gather the views of NCAs as regards the NCT. These meetings were complemented by a series of questionnaires, whose replies are summarised in Section 2.3 above.

    Second, bilateral calls were organised with a number of national government bodies and NCAs who requested additional information on the ongoing impact assessment. Upon their request, the Commission also introduced the impact assessment to working groups within the European Parliament and the Council of the European Union.

    Third, the Commission also held extensive discussions with all EEA competition authorities (i.e. Greece, Romania) and non-EEA competition authorities (Mexico’s COFECE, South Africa’s Competition Commission and the United Kingdom’s Competition and Markets Authority) having similar tools.

    Fourth, a virtual meeting was also held with the Body of European Regulators for Electronic Communications.

    Fifth, exchanges were also organised, at their request, with consumer organisations (through BEUC), as well as with a number of private sector stakeholders in the context of events organised by trade associations (e.g. European Round Table for Industry).

    3.5Other consultation activities

    In addition to the above-mentioned consultations and targeted stakeholder exchanges, the Commission received a number of spontaneous submissions from stakeholders. Some of these contributions were submitted by stakeholders that had participated in the public consultation and were therefore intended to supplement their views with additional evidence. Other submissions were received from EU government bodies and business associations that had not participated in the public consultation. These submissions largely echoed the issues already raised in the different consultation activities.

    All such submissions are published on the dedicated webpage on DG Competition's website, 54 except for a few submissions which stakeholders had asked the Commission not to publish for confidentiality reasons. The Commission used the latter to enhance its understanding of a particular stakeholder position and to complement its views on the issues subject to consultation.

    Throughout the impact assessment, the Commission also met bilaterally with stakeholders that requested this, primarily in the context of the public consultation and the feedback period for the inception impact assessment. These meetings were requested by the parties concerned and aimed primarily at discussing the submissions made by stakeholders, either in the context of the public consultation or outside of it.

    Annex 2.1: Synopsis Report of the results of the open public consultation on the DSA package - Ex Ante Regulatory Instrument for large online platforms acting as gatekeepers

    I. OUTLINE

    The Commission has undertaken an Open Public Consultation (“OPC”) on the Digital Services Act package (“DSA Package”) following the Commission's Communication “Shaping Europe's Digital Future” of 19 February 2020.

    The purpose of the OPC is to collect views and evidence from respondents as regards the experience with the application of the existing regulatory framework, in particular e-commerce Directive and collect input about the possible future digital services rulebook. To this end, an online questionnaire ran from 2 June 2020 to 8 September 2020, available in all official EU languages. Responses to the questionnaire were submitted online.

    This synopsis report summarises and analyses the views of respondents on the scopes, the specific perceived problems and the implications, definitions and parameters for addressing possible issues deriving from the economic power of large, gatekeeper platforms. 55  

    II. PARTICIPANTS AND METHODOLOGY

    1.Participants - general

    In total, 2863 responses to the OPC on the DSA Package have been received. Additionally, around 300 position papers were received in the context of the OPC. Respondents were asked to categorize themselves into different groups namely academic/research institutions, business associations, companies/business organisations, consumer organisations, environmental organisations, EU-citizens, non-EU-citizens, non-governmental organizations (NGOs), public authorities, trade unions and others.

    By far most feedback was received from EU-citizens (66.2%) and non-EU-citizens (8.2%), companies/business organisations (7.4%), business associations (6.3%) and NGOs (5.6%). This was followed by public authorities (2.2%), others (1.9%), academic/research institutions (1.2%), trade unions (0.9%), as well as consumer and environmental organisations (0.4%) and several international organisations.

    Figure 1: Type of respondent

    In terms of geographical distribution of respondents (Figure 1), most of the respondents are located in the EU, the majority of respondents are from Germany (27.8%, i.e. 797 respondents) followed by the United Kingdom (20.6%, i.e. 591 respondents), France (14.3%, i.e. 410 respondents). Other Member States that are represented in a slightly higher proportion are Belgium (9.3%, i.e. 266 respondents), the Netherlands (4%, i.e. 104 respondents) and Austria (3%, i.e. 92 respondents).

    The respondents from the remaining Member States contributed to a smaller extent, with responses rates lower than 3%. Among respondents originating outside the EU, the highest share comes from respondents from the United States of America (3%, i.e. 79 respondents).

    Figure 2: Country of Origin of Respondents

    2.Participants - Companies/Businesses organizations and business associations

    Of the 211 participating companies/business organizations, 80.1% specified that they were established in the EU and 11.4% indicated that they were established outside of the EU. 26.5% described themselves as a conglomerate, offering a wide range of services online. 21.3% identified as a scale-up and 6.6% as a start-up.

    In terms of annual turnover, more than half of the participating companies/business organizations indicated a turnover of over EUR 50 million per year. 13.3% make an annual turnover of smaller than or equal to EUR 2 million, 3.8% of the respondent revealed an annual turnover of smaller than or equal to EUR 10 Mio, whereas 6.2% specified an annual turnover of smaller than or equal to EUR 50 Mio.

    28.4% of the responding companies/business organizations were online intermediaries, 24.6% were other types of digital services. 12.3% indicated that they were an association, representing the interest of the types of businesses named prior. Of the 180 participating business associations, 15% indicated that they were representing online intermediaries, 19.4% specified that they are working on behalf of digital service providers other than online intermediaries, and 40% indicated that they represented the interests of other businesses.

    3.Participants - NGOs

    Of the 159 participating NGOs, almost half (49.7%) stated, that they represented fundamental rights in the digital environment. 22.6% dealt with flagging illegal activities or information to online intermediaries for removal, and 22% represented consumer rights in the digital environment. Furthermore, 18.9% specified that they were fact checking and/or cooperating with online platforms for tackling harmful, (but not illegal) behaviours and 13.2% represented the rights of victims of illegal activities online. 10.7% represented interests of providers of services intermediated by online platforms, including trade unions, and 10.7% gave no answer. 30.8% of the responding NGOs indicated “other”.

    4.Participants - Public authorities

    59 public authorities participated in the open public consultation, of which 43 representing authorities at national level (72.9%), 8 at regional level (13.6%), 6 at international level (10.2%), and 2 at local level (3.4%). Among EU Member States, authorities replied from Austria, Belgium, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Luxembourg, and Poland. About half of the responding public authorities were governments, administrative or other public authorities other than law enforcement in a member state of the EU (49.2%). 15.3% indicated that they were a law enforcement authority in a Member State of the EU and 15.3% specified that they were another independent authority in a member state of the EU. These replies are complemented by a targeted consultation ran by the Commission with Member States.

    Figure 3: Type of responding public authority

    5.Participants – Subcategories

    For this report, respondents were categorized in 18 different categories to take into account different perspectives, as presented in Figure 4 below. These categories are: telecoms and mere-conduit; caching services; hosting services, other than platforms; large platforms; scale up and startup platforms; creative industry and publishers; brand owners and other businesses selling through platforms; other services; business associations representing the interests of platforms; business associations representing the interests of actors other than platforms; civil society organisations; trade unions; national authorities; European authorities; International Organisations; academia and think-tanks; technical community; and the general public.

    Figure 4: Distribution of Respondents per Category 56

    6.Methodology

    The responses to the OPC were subjected to quantitative and qualitative analysis. In the quantitative analysis, the responses will be analysed in the following main steps:

    ·Ordering the respondents into the main categories considered relevant for the analysis;

    ·Analysis and mapping of respondents/stakeholders. In the mapping of stakeholders, the principles of Better regulation Toolbox were applied 57 ;

    ·Correlating main categories of respondents/stakeholders to their answers to identify patterns as regards “who thinks what?”.

    A methodological challenge of quantitative analysis is that the low numerical frequency of organisations makes it difficult to make statistical generalisations as to what type of organisation thinks what. In contrast, it is less challenging to make generalisations as regards the citizens because their number is much higher.

    In the qualitative analysis the open, qualitative answers were analysed in the following main steps:

    ·The lists of open answers were grouped per type of issue they bring up. In other words, the answers are put in the same group if they bring up the same type of issues, although with different wording.

    ·The content of the group of answers was summarised in bullet points, condensing the main messages of the respondents. Where relevant, some examples of authentic responses are quoted.

    ·The bullet points, each condensing the issue and main messages, are listed. The most frequently mentioned issues are listed first, followed by the less frequently mentioned.

    ·In addition to the grouping of the most frequent issues and views, unique responses have been analysed and presented. Especially, if these were extensive, well-informed, and reflective.

    In order to reach this thorough analysis, quantitative text analysis techniques have been used to group answers by the uniqueness of information and double-check if any relevant input was missing. We used:

    ·Deduplication methodology: the responses were duplicated by comparing their lexical similarity. For doing so, all the answers to each question were transformed into a document term matrix. Each response was processed by tokenizing and turning it into a vector of stemmed words weighted by their Term Frequency-Inverse Document Frequency (TF-IDF). 58 Next, for each question, the cosine similarity of each pair of responses was computed. If a pair achieved a cosine similarity higher or equal than 0.85, only one would be kept for the analysis.

    ·Automatic summarization methodology: Some of the larger submitted reports were automatically summarized. Use was made of the python module “gensim.summarization” which implements the “TextRank” algorithm, a graph based sentence ranking methodology. For each document, a proportion of the sentences with the highest rank would be selected conditional on the length of the document.

    III. SURVEY OUTCOME

    1.Unfair practices by gatekeeper platforms

    There is a wide-ranging majority across all stakeholder groups that there is a need for rules addressing the negative impact of gatekeepers’ practices and conduct. 59 Among businesses and business users who replied to the relevant question, 88% encountered issues concerning allegedly unfair trading conditions on large platforms (Table 1).

    Table 1: Business users of large platforms encountering issues

    with trading conditions on large online platforms

    In general, the vast majority of respondents across stakeholder groups – in particular among businesses and business users, civil society organisations and telecom operators - report that the experienced issues are due to a perceived imbalance in bargaining power between large platforms and business users, which they consider hampers competition, fosters uncertainty in relation to contractual terms and also results in lock-in of consumers. These respondents also consider that unfair practices by gatekeeper platforms have a concerning impact on competition, innovation and consumer choice. On the other hand, a few platforms, academic institutions and representatives of startups emphasised the positive impact of gatekeeper platforms on innovation and consumer choice. Among national authorities, both points of view were expressed. Generally, respondents, in their replies, consider unfair practices to be the means, legal or illegal, by which digital platforms with a gatekeeper role limit market contestability and preserve their position of power.

    Among the 1715 stakeholders who replied to the question, 58% somewhat disagree or fully disagree with the statement that consumers have sufficient choices and alternatives to the offering of online platforms, while 27% somewhat agree or fully agree with this statement (Table 2). The distribution of responses is homogeneous between each stakeholder group as well as within groups, with the majority of respondents somewhat disagreeing or fully disagreeing with this statement. 60 The only stakeholder group for which the majority of respondents consider that consumers have sufficient choices and alternatives are business associations representing the interests of platforms.

    Table 2: View of the respondents on whether consumers have sufficient choices and alternatives to the offering from online platforms

    The unfair practices listed by respondents cover exclusionary conducts, exploitative conducts and transparency-related problems. Practices mentioned most often include:

    1.Self-preferencing is considered to be very common by large platforms when services are vertically integrated, where they often favour their own services or products in detriment of third providers that rely on the large platform’s infrastructure or to favour paid-for content by certain content providers or advertisers. It is for instance perceived that search and ranking algorithms give preference to the platform’s own services or when the platform has an incentive to bias its recommendations toward the content provider charging a lower royalty.

    2.Lack of data sharing and accumulation of data, also linked to the imposition of proprietary services and an authentication through the platform even when third party services/products are used to create a direct link with customers to the detriment of third-party providers. It is considered that gatekeeper platforms incentivize disintermediation as they preserve monopoly access to user data and attempt to remove the direct link between the client and third-party suppliers, creating therefore privileged relation with the client.

    3.Limited data portability and data access due to lack of interoperability (e.g., APIs, limits to sharing customer data, restrictions to access key components, software or hardware), which creates obstacles for emerging competitors and also favours consumers lock-in. Several stakeholders refer to “walled-gardens”, which allow to determine who can access the data uploaded by their end-users and on which terms and conditions.

    4.Imbalance on how the revenues are split between platforms and right-owners in relation to User Generated Content (UGC).

    5.The imposition of unfair and unilateral terms and conditions, which cover pricing, non-price terms, most favoured nation (“MFN”) clauses, restrictions/abusive conditions on data sharing and use, exclusivity clauses, obligations regarding IPR, exclusivity or illegal restrictions, unfair access fees, among others.

    6.The imposition of exclusionary terms and conditions for attaining and/or retaining access such as unfair delisting, and unreasonable performance targets.

    7.Cross-financing and cross-subsidizing of otherwise unprofitable subsidiary companies as a strategy to gain market power in adjacent markets are considered to negatively impact competition.

    8.Default settings which adversely impact customer choice. Similarly, the way privacy settings are presented to users by platforms are considered to potentially lead to manipulation of users into “consenting” to contractual terms of service. 

    9.Bundling of services with ‘must have’ services and apps, which makes the use of a certain service dependent on the use of further services of the company.

    10.General lack of transparency on business practices on platforms (e.g. use of algorithms, content prioritization, lack of clarity in the terms of use, etc.).

    Regarding app stores, the issues, raised by stakeholder groups that make use of app stores, include high commission fees, unreasonable transfers of liability to the app developer without mutual liability being accepted by the platform operator, and the lack of notice given for technical changes in the app stores, which then requires apps to be amended in some cases resulting in lack of functionality.

    Regarding the travel sector, on one hand, it is reported by a business association that 56% of hoteliers feel pressured by online travel agents (“OTAs”) to accept platforms terms and conditions (e.g. regarding cancellation policy, special discounts) that hotels would otherwise voluntarily not offer. On the other hand, OTAs report that a large platform acts as a gatekeeper and diverts traffic away from OTAs and metasearch search engines (“MSEs”) to its own vertical search products, for accommodation, flights and vacation rentals.

    Book publishers consider that the balance of power between gatekeeper platforms and book publishers is uneven and publishers often have to bear unfair terms and conditions, lacking sufficient bargaining power and/or the possibility to switch to another business partner. Unfair practices cited by book publishers relate to shortage claims, price-related claims, special agreement on freight costs, high number of returns, delivery rules, lack of communication access, cancellations, tight delivery slots, hard rejects/wrong codes, among other issues. 61

    Among financial services providers there are concerns related to the fact that large digital platforms provide technical infrastructure and related functionalities that are increasingly relevant for the provision of digital financial services. This infrastructure includes devices and their associated functionalities, such as biometric authentication or communication protocols like Bluetooth and near field communication (“NFC”) and app stores and pre-installed apps on devices. It is reported that this infrastructure is not always available on an equal basis to all market participants, with elements controlled by some market players and/or technical providers. Financial services providers also raise the issue that, under the Revised Payment Services Directive (“PSD II”), banks have to offer application programming interfaces (“APIs”) for competitors as to certain payment services, while there is no such obligation for platform providers. This is considered to place financial service firms at a direct disadvantage, as financial data can be shared easily with platforms (who can then combine it with non-financial data in order to generate insights that may be relevant for the provision of financial services, other products or advertising), while data held in those platforms cannot be shared with financial services providers on the same terms.

    In particular digital rights’ associations pointed to the lack of data access and meaningful interoperability as important barriers to entry and called for measures that would address them. In addition, telecom operators recognised the right of data portability in Article 20 of the General Data Protection Regulation, but referred to the fact that its scope is limited to specific cases and subject to specific legal bases for processing. In particular, this right does not foresee continued and far-reaching access possibilities to different categories of data but is limited to receive the data ‘provided’ by the user, to avoid lock-in effects for individuals.

    Regarding the use of gatekeeper platforms by minors, specific issues have been raised which include: overly complex terms and conditions; lack of transparency and redress; lack of standardised frameworks for age-appropriate terms and conditions; significant amount of data gathereing without the knowledge of minors; and behavioural advertising.

    2.Gatekeepers and media plurality

    Respondents representing the media sector and publishers consider in their replies that social media and search engines have a strong impact on the consumption, distribution and production of news. They impact the way in which information is accessed (demand side) as online platforms act as access points to information, as well as the way revenues are distributed (offer side).

    Respondents consider that traditional business models, mostly based on revenues from advertisements and subscriptions to print media, are particularly challenged by digitalisation. The respondents from the media and publishing sector consider that, on the one hand, gatekeeper platforms have left media without direct contact to their readers, losing access to full audience data and at the will of any changes in company policy and ranking algorithms, with consequent perceived threats on media plurality. On the other hand, it is perceived that the role and importance of online platforms in capturing the lion's share of online advertising spend has had a dramatic impact on newspapers and newswires that rely on advertising to finance their operations. In fact, press publishers consider that platforms exploit and monetise publishers’ content and take advantage of their primary and direct access to users and users’ data.

    It is stated that, without access to sufficient advertising revenues, many media companies struggle to continue to offer valuable content. News publishers also consider that there is a lack of contractual transparency, especially with regards to the value created by leveraging on the content created by the publishers, which raises concerns that the current model is not sustainable and represents an existential threat to quality press media. Press publishers are worried that gatekeeper platforms not only can determine the success or failure of journalistic and editorial publications, but also ultimately have the power to steer political and cultural opinion as well as economic prosperity in the EU.

    3.Platforms and startups

    Generally, all categories of stakeholders consider that startups and small companies are more and more dependent on large platforms for reaching their customers, in particular, 3 of 4 scaleups and 4 of 5 startups answered this question accordingly. Their reliance on platforms is considered to vary heavily according to their fields and/or business models.

    It is argued that startups often end up in a dependent relationship with these mega-platforms from the very beginning. In addition, all categories of stakeholders appear concerned with the practice of acquiring startups and scaleups, since it is considered detrimental for competition and can raise serious concerns related to the accumulation of data (‘killer acquisitions’). It is also pointed out that, by acquiring startups while they are still a niche product, gatekeepers avoid the impression of impeding competition but in practice they prevent the emergence of competition in a targeted market.

    However, some respondents, including several startups, research institutes and business associations, also point out the positive impact of platforms for startups: by lowering the barriers to entry and extending to companies of all sizes the advantages of cost and speed that can be gained from trading online.

    4.Scope and criteria for large gatekeepers

    The respondents consider all the characteristics mentioned in the questionnaire 62 are relevant in determining the gatekeeper role of large online platforms. In particular, the most relevant characteristics for respondents are: the accumulation of valuable and diverse data and information (74% of 1304 stakeholders who replied to the relevant question consider this characteristic very relevant), large user base (70% of the 1300 replies received), and the fact that large platforms raise barriers to entry for competition (68% of 1293 replies received) (Figure 5).

    Figure 5: Most relevant characteristics in determining a gatekeeper platform 63

    For the definition of gatekeeper platforms, some stakeholders suggest to consider the criteria of ‘economic dependence’ on certain platforms which makes them unavoidable trading partners and make it more difficult for consumers to avoid dealing with them. Other relevant criteria to define gatekeeper platforms suggested by news publishing are the ability to utilise a platform which can direct consumer attention combined with the power to choose which information is displayed to consumers or otherwise set the terms on how information is displayed. The Body of European Regulators for Electronic Communications (“BEREC”) proposes to identify digital platforms with significant intermediation power based on a combination of structural and specific criteria in different Areas of Business (“AoBs”). An AoB could be e.g. e-commerce, app stores, online search, OS, voice assistants etc., and would be characterised by features such as strong direct and indirect network effects, significant economies of scale and scope, significant barriers to entry and expansion relating to technical and/or legal aspects, high switching costs and/or consumer inertia.

    Respondents among platforms show diverse views on what would define a gatekeeping position. Some platforms argue that incorporating different services into a platform’s offering says little about the strength of a platform, as it is also the case with the ability to leverage assets from one market to another. In addition, one platform stated that utilizing a company’s own assets to enter new markets is actually welfare enhancing. It is suggested that gatekeeper designations should be business model agnostic, gatekeeper assessments should be reviewed periodically, gatekeeper designations should apply to identified activities in specific markets, and some rules ought to apply on a sector-wide basis. Some large platforms, foreign business associations and a limited number of national authorities are worried that the regulatory proposal would focus on defining the companies it wants to regulate rather than focussing on determining the market and consider that using generic criteria would not be appropriate. Several platforms and foreign business associations highlight the need to ensure that any requirements to define ‘gatekeepers’ are non-discriminatory regarding the national origin of those companies and agnostic to different business models approaches.

    In general, stakeholders of all categories point out the need to ensure a high level of coherence and legal certainty, the criteria used should be transparent, objective and easily measurable, and that a merely cumulative approach might not be sufficient. At the same time, some respondents from different stakeholder categories (including platforms, business associations and telecom operators) state that a one-size-fits-all approach might be unfeasible, while others (mainly from business associations) state that the new legislation should be general in nature, so that it may be applicable regardless of industry, sector, technology or business-model. Some respondents from the telecom sector argue that there should be dynamic methodology with a case-by-case assessment of the companies that should be subject to ex ante regulation. Respondents from the general public mostly refer to a combination of both quantitative or qualitative criteria.

    Several respondents have referred to the operation of the EU telecoms ex ante regulatory regime, that provides valuable lessons as to how different criteria could be assessed to determine the extent of competition, consolidation of market power and the potential of consumer harm. This would ensure proportionality and legal certainty. In addition, several respondents, mainly telecom providers, argue that the new sectorial ex ante instrument for platforms should exclude from its scope those services that are already subject to sectorial ex ante rules that promote competition, as it is the case for electronic communication services and networks.

    5.Need for a regulatory framework 

    Of the 1476 respondents who replied to the relevant question, the vast majority fully agree (70%) or agree to a certain extent (20%) that there is a need to consider dedicated regulatory rules to address negative societal and economic effects of gatekeeper power of large platforms (Table 3). This majority holds also within each group of stakeholders, including platforms, with 73% of the respondents agreeing or fully agreeing with the need for dedicated regulatory rules. 64

    Table 3: Respondents’ view on the need to consider dedicated regulatory rules for gatekeeper platforms

    In general, there is a shared understanding among stakeholders that there are structural competition issues that EU competition rules cannot address or cannot deal with effectively. The majority of stakeholders consider that, while some of the issues connected to gatekeeper powers can potentially be addressed by improving the efficiency of competition law enforcement through procedural and/or organisational changes, there are restrictions that cannot be overcome with competition law enforcement. These include the fact that ex post enforcement is not always best suited to tackle anti-competitive practices in fast-moving digital markets so that by the time an investigation has been concluded, the market may have irreversibly tipped in favour of the dominant firm and it may be very hard to restore competition. Moreover, competition investigations are ad hoc, limited to the narrow facts of the particular case, and may do little to address the same issues arising in different contexts, and the remedies imposed may do little to reinvigorate competition.

    While telecom operators generally argue for the need to ex ante rules for gatekeepers, they also mention the need to carefully assess the results of the implementation of the Platform to Business Regulation (“P2B Regulation”), which only came into effect in July 2020 and the New Deal for Consumers, before suggesting any ex ante regulation. In fact, they consider that these policy instruments address the same market failures that the gatekeeper regulation seeks to address. Similarly, respondents from different stakeholders’ categories consider the need to reassess the situation after the P2B Regulation has shown its effects.

    A minority of respondents, mainly several large platforms and their business associations and some research institutes and academics, disagree with the proposal for new ex ante rules. These stakeholders consider that the risks posed by gatekeeper platforms can be sufficiently addressed with existing competition and antitrust law and tools (enforcement of Articles 101 and 102 TFEU) and other existing regulation. Some platforms, business associations and national authorities emphasized the need to focus the regulatory attention at specific actions and perceived market failures.

    Regarding the form of the new ex ante rules, the majority agree that the rules should include prohibitions of certain practices considered particularly harmful for users and consumers (Table B.3) 65  as well as specific obligations (Table B.4) 66 , as presented below in section 7. However, respondents have diverse views on what would be the best design for such prohibitions and obligations. Some stakeholders’ groups, in particular among the general public, civil society organizations and the representatives of the creative and publishing industry, argue for strict prohibitions and obligations to all gatekeepers, while some respondents among business associations, academia and platforms caution against applying the same restrictions to all gatekeepers. Some respondents have highlighted the need to tailor the new rules to the different markets in which gatekeepers operate. As shown in the next section, the majority of respondents also argue that a case-by-case assessment and tailor-made development of remedies would be necessary to ensure proportionality and satisfactory policy outcome (Table B.5). 67

    It is argued by many stakeholders, especially among platforms, business associations, civil society organisations and academia, that new EU level rules would prevent further legal fragmentation across Member States, considering that several Member States have already started to introduce new regulation to address concerns arising from the presence of gatekeeper platforms. When considering the introduction of ex ante rules for online gatekeeper platforms, stakeholders across all categories have pointed out the need for these rules to be flexible enough to take into account the wide diversity of business models and future-proof against the evolution of these business models and technology.

    6.Case-by-case regulation of gatekeeper platforms 

    80% of the 1216 respondents who replied to the relevant question, considers that there is the need for regulatory intervention on a case-by-case basis (Figure 6). This majority holds all stakeholder groups and, in particular, 100% of European authorities and International Organisations as well as hosting services other than platforms and telecom operators agree with the need for case-by-case remedies against specific large online platform companies with gatekeeper role (Table B.5).

    It is argued that regulatory intervention should consider the high variety of online platforms’ business models and digital markets they operate in, as well as the specific harms that should be addressed. Following the identification of large online platform companies with a gatekeeper role, competent authorities should be empowered to monitor markets, select the remedies needed, attune them to the competition concerns of each particular case and enforce compliance.

    In general, the majority of respondents consider that case-by-case remedies to respond to specific behaviours should go hand in hand with ex ante rules that apply horizontally to all gatekeepers. It is argued that the digital world is evolving rapidly, so a list of ‘dos’ and ‘don’ts’ might not capture all types of detrimental practices. Several stakeholders refer to the approach taken by the European Commission on Standards Essential Patents (SEPs) and on the Fair Reasonable and non-discriminatory (FRAND) commitment, as they consider that it might prove very instructive in the digital platforms’ context.

    Among platforms, some argue that case-by-case intervention following a market assessment would be more effective and efficient than blanket prohibitions in targeting specific market failures. One consumer association instead argues that ex ante rules should focus on a list of obligations and prohibited practices, while case-by-case investigations are better left to competition law.

    7.Practices by gatekeeper platforms that should be prohibited or obliged

    The vast majority of respondents (85% of 1274 respondents to the relevant question) consider that dedicated rules on platforms should include prohibitions and obligations for gatekeeper platforms (Figure 7). 68

    According to the majority of respondents, the proposed list of problematic practices, or ‘blacklist’, should be targeted to clearly unfair and harmful practices of gatekeeper platforms; specific enough to avoid confusion of what is and is not permitted; adaptable to a dynamic, fast moving sector; and specific to certain gatekeepers as they would otherwise risk hurting smaller players trying to compete with them. It is also suggested that remedies could be more procedural in nature rather than prescribing a given course of conduct.

    Most of the respondents suggest that, rather than having certain practices categorically prohibited for all gatekeepers, the Commission should scrutinize certain practices and prohibit them for some gatekeepers in circumstances when they are most likely to have detrimental effects.

    Some respondents from the telecom sector argue that a list of prohibited unfair practices and specific obligations should be introduced to prevent the most frequent and harmful abusive behaviours, while the case-by-case approach would allow to apply specific remedies that reflect the gravity of specific threats to competition and to contestability in a targeted market.

    Some respondents among business associations, research institutes and platforms consider that a ‘blacklist’ of prohibited practices should require careful consideration in relation to a dynamic industry that has multiple business models, types of users, and types of business partners, and it should result from an assessment of market failures to be resolved. In addition, several platforms are of the opinion that blanket banning of market behaviours risks being inefficient, negatively impacting consumers and actually risk worsening competition by limiting the ways in which entrants can innovatively challenge the incumbents.

    The suggested obligations and prohibitions cover mainly issues of transparency, interoperability, portability, and non-discrimination. The suggestions are linked to the unfair practices which are reported by the respondents 69 and include, among a wide variety of others:

    ·the prohibition of discrimination through self-preferencing;

    ·elimination of certain clauses in the terms and conditions in contracts considered unfair, such as obligations to use platform’s ancillary services, unilateral liability issues, contract modifications with retroactive effect;

    ·longer notification times for major changes on business-to-business contracts and market practices; the obligation of interoperability of datasets and APIs;

    ·transparency obligations in relation to interconnection, access, ranking of services and suspension of accounts, practices of micro-targeting;

    ·the prohibition of bundling when it results in restricting consumer choice;

    ·the prohibition of combination of data collected across different services when it results in unfair competitive advantage over rivals (‘data unbundling’);

    ·the prohibition of excessive intermediation fees;

    ·the obligation to connect businesses intermediated directly with their customers and to provide these businesses with more data of the transactions;

    ·a general prohibition on discrimination in access to infrastructure;

    ·restrictions on the use of pre-installation and default settings (in particular for browsers and search engines) and of ‘nudging techniques’;

    ·the provision of consumers control over data use, sharing and mobility, including additional obligations to facilitate the portability of both personal and non-personal data between service providers;

    ·algorithmic accountability and transparency audits (including to verify whether in practice the platforms are operating a preferential treatment for their own services);

    ·the provision of access to individualized data to businesses about their operations or customers/consumers;

    ·imposing accounting, structural or functional separation or firewalls between different businesses under common ownership of a gatekeeper platform.

    Issues relating to data sharing are considered especially important by national authorities. Regarding the issue of ‘killer acquisitions’, it is proposed by national authorities to extend the time for the notification of mergers, so that the competent authorities can carry out in-depth investigations on the impact of a certain acquisition/merger on competition and innovation in the internal market. National authorities also consider highly urgent the issue of algorithmic transparency and accountability.

    Some platforms have offered detailed responses on some of the potential obligations and prohibitions. In particular, they have pointed out that mandating interoperability should only be considered on the basis of a careful analysis on its effects on the market, especially when multi-sided. Similarly, it is considered that unconditional and misguided data access obligations would undermine the platforms’ ability to innovate. More generally, they argue that the remedies should be proportional and that there should be a clear link between a specific conduct considered anticompetitive and consumer harm.

    Regarding IP infringement, some respondents from telecom sector, brand owners and representatives of the creative industry have argued that gatekeeper platforms should be liable for facilitating infringements taking place on the platforms to the extent that the financial benefits have to be refunded to the IP owner.

    In relation to media pluralism, representatives of this sector as well as some civil society organizations is suggested to impose specific media requirements to ensure adequate remuneration of press publishers for their content used by or uploaded on large online platforms. Other suggestions specific to the media sector include: the requirement for platforms to share data with news media businesses about the interaction of users with the content of the publisher, early notification of changes to the ranking or display of news content, the obligation that the digital platform actions to not impede news media businesses’ opportunities to monetise their content appropriately on the digital platform’s sites or apps, and the requirement for the digital platforms to “fairly negotiate with” news media businesses as to how direct and indirect revenue should be shared. It is also pointed out that consumers must always have the choice to choose the relevance of content/media and never be left in the dark on why they see certain articles while not being served others. More generally regarding content diversity, some digital rights’ associations consider that platforms should take steps to ensure that users are exposed to sufficiently diverse content and balanced coverage of issues of public interest by default. Yet, platforms and some civil society organisations consider that a regulatory intervention would not be the most appropriate way to ensure media pluralism.

    One specific suggestion put forward by one digital rights’ association to respond to excessive concentration in social media market is the decentralisation of content moderation. The association suggests a combination of data portability, interoperability and unbundling of hosting and content curation activities, consistent with data protection laws.

    Specific suggestions have also been put forward in relation to the protection of minors, including the requirement to implement a privacy impact assessments to determine how products and services affect children’s privacy, the minimization of the information being collected on children and applying enhanced security measures to protect any personal data that is collected, the provision and communication of child-friendly terms and conditions, making children’s online profiles private by default, and offering simplified, accessible reporting and complaints mechanisms for minors.

    8.Regulatory authority for gatekeeper platforms

    70% of the 1215 respondents who replied to the relevant question consider that there is the need for a specific regulatory authority to enforce the new prohibitions and obligations that might be imposed to gatekeeper platforms, while 80% of the respondents who replied to the relevant question consider that there is the need for a specific regulatory authority to enforce the case-by-case remedies that might be imposed to gatekeeper platforms.

    These stakeholders agree that there should be only one authority overseeing these issues and they point out the need to avoid overlapping with the competence of other authorities. Respondents from the media sector, consumers’ associations and some business associations are among those who advocate for such a solution as they consider that only a specialised interdisciplinary regulatory authority can grasp the complexities of the digital ecosystem. Some national authorities and research institutes instead consider the need to build on existing structures in other to avoid creating parallel structures. These stakeholders suggest to rely on cooperation among Member States authorities as gatekeeper platforms are usually based in only one Member State.

    In addition, several respondents, including some business associations and some national authorities, consider that only competition authorities should deal with potential harm to competition and point out the need to utilise as much as possible the current structure to enforce the rules. These stakeholders consider that there should be very special and weighty reasons for establishing a new regulatory authority for monitoring the online platform market.

    The respondents generally consider that an effective coordination between EU bodies and the relevant national regulatory authorities is needed, especially in the light of the fact that issues related to gatekeepers are likely to have an important cross-border component. Platforms in particular point out the need to minimise fragmentation and allow for a pan-European approach. The majority of respondents consider that regulatory oversight should be both at the EU and national level (63%), while 27% consider that oversight should be only at the EU level and 6% consider that oversight should be only at the national level (Figure 9). 70  

    Figure 9. Platforms regulatory oversight level 71

    In particular, the majority of large platforms (70%) and of scale-up/startup platforms (57%) consider that the oversight should be at the EU level. 72 Among business associations representing the interest of platforms, 50% consider the oversight should be at the EU level and 50% consider that it should be done both at the EU and national level. On the other hand, the majority of national authorities (68%) consider that the oversight should be done both at the EU and national level and 7% of them consider that the oversight should be done at the national level only. Among business associations, 54% consider that the oversight should be done both at the EU and national level, while 40% consider that the oversight should be at the EU level. Among civil society organizations, 49% consider that the supervision should be done both at the EU and the national level, while 47% of consider that the supervision should be done at the EU level only.

    In addition, respondents across stakeholder groups consider that the regulatory authority overseeing the gatekeeper platforms should: rely on an institutional cooperation with other authorities addressing related sectors; support swift and effective cross-border assistance across Member States; have a high level of technical capabilities, including data processing and auditing capacities; have a pan-European scope and cooperate with extra-EU jurisdictions. Several respondents have referred to National Regulatory Authorities (“NRAs”, i.e. telecom regulatory bodies dealing with ex ante regulation for the telecoms sector) to be involved in the monitoring of gatekeeper platforms. Some also referred to the mechanism in place for the enforcement of EU competition law according to Regulation 1/2003, which creates a system of parallel competences between the European Commission and national authorities.

    There is a general agreement between all the stakeholders that the ex ante rules for gatekeeping platforms should complement the current sector-specific framework with a view to granting simplification, harmonization and consistency with the acquis.#

    IV TABLES

    Table B.1. Degree of agreement: Consumers have sufficient choices and alternatives to the offerings from online platforms (rating per category). 73

    Table B.2. Need for dedicated regulatory rules to address any negative societal and economic effects of the gatekeeper role that large online platform companies exercise over whole platform ecosystems (rating per category). 74

    Table B.3. Need for dedicated rules that prohibit certain practices by large online platform companies with gatekeeper role that are considered particularly harmful for users and consumers of these large online platforms (rating per category). 75

    Table B.4. Need for dedicated rules that impose obligations on large online platform companies with gatekeeper role that are considered particularly harmful for users and consumers of these large online platforms (rating per category). 76

    Table B.5. Need for case-by-case remedies against specific large online platform companies with gatekeeper role when necessary (rating per category). 77

    Table B.6. Platforms regulatory oversight level (rating per category). 78

    Annex 3: Who is affected and how?

    1.Who is affected by the Digital Markets Act?

    The Digital Markets Act will have an impact on businesses, including gatekeeper platforms subject to the regulation, competing platforms, business users and SMEs, as well as on consumers and regulatory authorities.

    1.1.Businesses

    The Digital Markets Act would benefit businesses in many different aspects. First, this initiate would allow the Commission to address gatekeepers’ unfair conduct and weak contestability of, and competition in, platform markets, or risk thereof. More open and competitive markets where companies compete on their merits enable wealth and job creation.

    79 Second, and in accordance with the results of numerous empirical studies, an improvement of market competition would result in higher productivity, which would translate into higher economic growth. These effects are expected to be particular relevant in digital markets where structural market features may lead or contribute to market failures, preventing healthy competition between market players.

    80 81 Third, more open and competitive markets would provide more incentives for companies to innovate and offer a better range of high quality products and services. The economic literature shows that firms facing more competition from rivals innovate more than monopolies. Greater competition also drives efficiency in processes, technology and service and creates the conditions to make European's markets more attractive to investors. 

    A set of measures that contribute to a more dynamic online platform ecosystem and more contestable market would particularly benefit SMEs who would face lower barriers when entering the market. It can therefore be expected that an increased contestability of the markets would, even with some changes to their business model due to the regulatory intervention, continue to incentivise gatekeeper platforms to bring innovative products to the market and compete for consumers and business users.

    Fourth, it is worth mentioning that the measures envisaged would limit the chilling effects unfair conduct has on sales. While gatekeepers are an important channel to reach markets and consumers, business users argue that unfair practices would lead to up to 15% loss in their sales. Businesses, especially smaller ones, would be more confident in engaging with gatekeepers if the latter (are obliged to) comply with clear fairness rules.

    A detailed assessment of the way different categories of enterprises are affected is presented in Section 6 as well as in the following parts of this Annex.

    1.2.Consumers

    82 A regime that protects EU consumers from business practices that keep the prices of goods and services artificially high would ensure that consumers have access to better quality, wider choice and innovative goods and services at affordable prices. Numerous studies confirm the benefits of competitive markets for consumers. 

    A more competitive digital market will allow consumers to multi-home among alternative platforms offering differentiated commercial proposition. The fact that consumers can multi-home, altough resulting in possible search and switching costs, would generate a net benefit for consumers as they would only multi-home if the benefits of using alternative platforms compensate for those possible costs. In addition, some of the measures considered under the options aim at reducing those search and switching costs, for instance by allowing portability of data, creating conditions for interoperability, increasing transparency in the market, etc.

    83 This initiative would indirectly contribute to safeguard value added for consumers and would contribute to ensuring greater respect of privacy and consumers’ interests. This would be achieved by contributing to (i) fairer competition on gatekeeper’s platform (intra-platform competition) and among platforms (inter-platform competition), (ii) stronger contestability of the markets where gatekeepers are present and (iii) better functioning of the internal market through enhanced regulatory oversight at EU level.

    It is also important to notice that although some interventions under this initiative may require changes to the existing business models, the evidence from the supra-normal profits that gatekeepers are accruing, their ability to obtain conditions that would not be possible under normal market circumstances as well as their ability to act independently from competitors, business users and consumer indicates that in the long term business users and consumers will not be harmed. This is especially the case since gatekeepers in particular will continue to depend on a large user base and since they will be subject to higher contestability and competition, their incentive to innovate and offer low prices will rather increase and not reduce due to the intervention.

    84 85 86 It should be stressed that the foressen interventions will neither ban specific monetisation models (such as ad-based models) nor prevent the uptake of new services by gatekeepers - they prevent them from acting unfairly in their operations and reduce competition in the markets where they are present. Even in those cases where, due to the multi-sidedness of the market, there is a cross-subsidisation between the different sides and consumers already benefit from zero prices, more contestability and competition would not change the business model. On the contrary, more contestability and competition would increase the diversity of offers available to consumers and would reduce the prices to advertisers which would then indirectly reflect in lower prices charged by those advertisers when selling their products and services to consumers., In addition, for many digital markets when consumers are offered ‘free of charge’ services, in practice they are receiving the service in exchange for their attention and their data, which can then be monetised through digital advertising. In a more contestable and competitive market, it would be clear to consumers what data is collected about them and how it is used and, crucially, the consumer would have more control of the data.

    Section 6 further assesses the impacts on consumers; the table of impacts per practice in this annex specifies expected impacts associated with each of the measures foreseen.

    1.3.Regulatory authorities

    This initiative will allow the Commission to tackle gatekeepers’ unfair practices and existing and emerging market failures in digital markets. The burden that would ensue from giving the Commission this ability is low (mainly redeployment of existing job positions) compared to the benefits for the economy. Section 7 of the Impact Assessment and the sections below qualify and quantify these costs. National authorities would have to bear some administrative costs specified below.

     

    2.How would the ex ante rules foreseen affect stakeholders?

    The table below contains an assessment of impacts for individual practices that could be addressed by the various obligations included in the options considered in the Impact Assessment. It is important to note that the three options considered foresee a combination of obligations, which should resort an overall positive impact as set out in Sections 6, 7 and 8 of the Impact Assessment. Some of these individual practices would also act in concert, mutually reinforcing one another in the most proportionate way.

    Table of impacts per considered obligation

    Unfair practice

    Impacts on gatekeepers

    Impact on competitors and new entrants

    Impact on business users

    Impact on consumers

    Gatekeepers shall not be combining personal data originating from different core platform services with personal data from their other services or data from third party services or automatically signing in end users to other services of the gatekeeper in order to combine data without providing an effective possibility to opt-out

    Example: provider of online social network site collecting data from its users obtained through several different services.

    Gatekeepers would be constrained in combining data collected from all their core platform services which would reduce their data advantage, thereby reducing barriers to entry.

    Gatekeepers would be required to compete on merits or face a risk of losing revenue.

    In view of increased choice, downward pressure on prices and therefore revenue.

    Removal of barriers to entry and ability to enter/expand.

    Ability to compete on merits and higher incentives to innovate.

    Consumers could possibly benefit from lower prices and/or higher quality/price ratio for online services offered on other platforms.

    Gatekeepers shall not prevent business users from offering the same products or services to customers through third party online intermediation services at prices or conditions that are different from those offered through the online intermediation services of the gatekeeper.

    Example: a provider of online intermediation services does not allow hotels/e-books publishers to offer better prices on different online travel agents/e-books platforms

    Gatekeepers would be limited in disincetivising its business users from switching or multi-homing.

    Gatekeepers would be required to compete on merits or face a risk of losing revenue.

    In view of increased choice, downward pressure on prices and therefore revenue.

    Smaller and start-up platforms would have the possibility to offer better conditions to business users and thereby incentivise them to switch or multi-home.

    Increased choice likely to increase innovation incentive for all market players.

    Business users would have the ability and incentive to choose among different platforms where to offer their service/product.

    Lower prices for intermediation services would lead to passing of the cost savings to consumers.

    Consumers’ choice would not be limited to offers provided through/on the gatekeeper platform.

    Consumers could possibly benefit from lower prices and/or higher quality/price ratio for online services offered on other platforms.

    Opportunity costs of comparing different options may increase.

    Gatekeepers shall not prohibit their business users from promoting and subsequently concluding contracts with their customers acquired on gatekeeper’s platform outside a gatekeeper’s platform.

    Example: a publisher cannot inform a new user through its newspaper app that the subscription is cheaper if concluded via the publisher´s website.

    Gatekeepers shall not prohibit consumers from accessing and consuming, on the gatekeeper’s platform or services, services which have been acquired outside of the gatekeeper’s platform or services.

    Example: a music streaming subscription is concluded through a website, but cannot be subsequently used via the app.

    Gatekeepers facing increased choice from alternative sources. Such increased choice would put pressure on the level of prices and conditions of the service offered.

    Incentive to innovate in order to prevent outside-gatekeeper platform/service offers.

    Gatekeeper would likely face lower revenue in view of lower number of transactions.

    Increased incentive to develop an alternative distribution channel.

    Ability to reach customers directly and offer them targeted and tailored offers.

    Ability to reduce costs of distribution, which could be passed through to consumers in form of lower prices and investments in more innovative products/services.

    Increased choice for consumers.

    Lower prices and higher quality.

     

    Gatekeepers shall not prevent or restrict business users from raising issues with any relevant public authority relating to any behaviour of gatekeepers

    Example: business users would like to complain about unfair practice by gatekeeper, but is effectively prevented doing so due to contractual constraints.

    Gatekeepers would not be able to exercise their imbalanced power preventing business users from complaining.

    Business users competing with the gatekeeper would be able to obtain redress and thus compete with the gatekeeper on an equal footing.

    Business users – especially smaller ones - would have a chance to find solution to the issue they face with the platform.

    Consumers would ultimately benefit from an increased number of business users on the gatekeeper platform, since even smaller businesses would dare using the gatekeeper platform’s services.

    Gatekeepers shall not impose their own user ID services on business users when the latter offer service using the core platform service of the gatekeeper.

    Example: an app store operator unilaterally requires all app developers to integrate the app store´s own user ID functionality in their apps and to show this ID functionality to the customers of their apps.

    Gatekeepers required to compete on merits.

    Incentive to innovate in order to compete.

    Gatekeeper would likely face lower revenue in view of lover use of their use ID services.

    Enabling market entry by competitors offering alternative user ID services.

    Ensuring a level playing in the online platform economy.

    Incentives to innovate and compete.

    Business users benefiting from more choice.

    Possible cost savings in use of alternative user ID services.

    Increased choice for consumers.

    Benefiting from higher quality services.

    Gatekeepers shall not require business users or customers of these business users to subscribe to or register with any core platform service other than the core platform service provided by the gatekeeper, as a condition to access, sign up or register to any of their core platform services

    Example: consumer would like to subscribe to social networking service by a gatekeeper, but is effectively prevented from doing so without subscribing to other services of that gatekeeper.

    Gatekeepers would be limited in combining different services together.

    Gatekeepers would have to compete on the merits of quality and characteristics of each individual service.

    Gatekeepers would be faced with lower demand for one or more of its services.

    Gatekeeper would face potential loss of revenue.

    Incentive to compete on merits and innovate.

    Removal of barriers to entry and ability to enter/expand.

    Ability to compete on merits and higher incentives to innovate.

    Facing increased choice, which in turn could lead to lower prices and higher quality of service.

    Ability to combine services according to the actual needs and interests.

    Ability to multi-home and switch increased.

    Consumers facing more choice from different business users.

    More choice enables also better choice to the actual needs of consumers (e.g. one service; combinations of services).

    Gatekeepers shall provide advertisers and publishers with information concerning the price paid for the impression of a given ad, including for each of the relevant advertising services provided by the gatekeeper

    Example: advertisers and publishers would like to obtain information about all the intermediation fees charged by the adtech services provider, but they are refused so.

    Gatekeepers would provide for increased transparency of its advertising system.

    Increased transparency is likely to put downward pressure on pricing, which is considered opaque.

    This may lead to some lost advertising revenue.

    Increased transparency may facilitate entry or expansion of competing service providers.

    Success of competitors largely dependent on the actual switch in demand from business users, with at least part of their demand (reach of gatekeeper platform still being an important barrier to entry).

    This would lead to improved transparency of the advertising value chain.

    Some demand may switch to alternative operators, but that will largely also depend on their reach to consumers.

    Increased choice and competition likely to lead to lower prices paid for advertising, which should subsequently be passed on in final prices for consumers.

    Gatekeepers should not use data provided by or generated through activities of business users of its core platform services in competition with those business users

    Example: an e-commerce marketplace using commercially sensitive data collected from individual sellers to compete with these sellers on its own online marketplace

    Gatekeepers would be constrained in expanding their power to neighbouring markets through unfair practices since they would not be able to gain unfair commercial advantage.

    Gatekeepers would need to establish market trends as any other market operator when determining supply and demand patterns.

    Gatekeepers would be further incentivised to innovate and compete on merits.

    Gatekeepers might grow at a lower pace as when engaging in unfair practices. If they can preserve their ‘first mover advantage’ in terms of number of business users and data accumulation but would need to compete on the basis of fair data- use practices.

    Increased level playing field.

    Ability to compete on merits.

    Ability to bring innovative solutions to the market.

    Business users would have more confidence in selling online, as they would be protected from practices that exploit their data for the benefits of the platforms alone.

    Consumers would continue benefitting from products/services provided through multiple channels (e.g. retailers) thus benefitting from more choice, lower prices and innovative products.

    Gatekeepers shall not prevent customers from un-installing any pre-installed software applications on its core platform services.

    Example: app stores/operating systems preventing users from un-installing some of the pre-installed apps, in particular where these are not essential for running the hardware.

    Gatekeepers would be limited in practices preventing switching or multi-homing.

    Gatekeepers may be facing more competition concerning specific proprietary software applications and therefore limited growth.

    Gatekeepers may be losing some revenue in case of exclusivity arrangements with business users.

    Ensuring a level playing field in the online platform economy.

    Ability to bring innovative products and services on the market in view of more open switching or multi-homing possibilities of consumers.

    In case of pre-installation arrangements with gatekeeper, increased incentive to compete to remain relevant and competitive.

    Increased competition between different developers of software applications.

    Increased consumer freedom of choice with a positive impact on prices and quality of service.

    Gatekeeper shall allow the installation and effective use of third party software applications or software application stores using, or interoperating with, operating systems of that gatekeeper

    Example: consumer is prevented from installing an app store of the mobile game provider and the relevant applications directly from its app store.

    Gatekeepers would be effectively prevented from restricting switching or multi-homing.

    This may lead to decrease in demand in absence of innovative competing products in comparison to competitors.

    Increased switching and/or multihoming likely to lead to some loss in revenues.

    Increased ability to switch shall serve as incentive to target consumers with high quality products or services.

    Switching should incentivise increased innovation in a variety of markets.

    Increased overall competition and contestability of the markets.

    Incentive to switch would lead to a more volatile markets due to competitive choice.

    This would in turn lead to more innovation, lower prices and higher quality products and services.

    Increased choice for consumers.

    Lower prices and higher quality of service.

    Gatekeepers shall not treat more favourably in ranking their own services and products compared to similar services or products of third-party business users and shall apply fair conditions to such ranking

    Example: a search engine preferring its own vertically integrated services in its search engine results (e.g. shopping or travel services are featured on top of search results); a social network ranking its own dating service more prominently in users´ timelines than those of third-party dating services.

    Gatekeepers’ power stemming from ability and incentive to engage in self-preferencing due to the vertical integration will be limited.

    Gatekeepers adjacent services growth may be limited or stalled absent change to compete on merits.

    Competitors would have the ability to compete on equal footing and merits.

    Increased incentive to innovate and invest in view of level playing field.

    Business users offering high quality and/or competitive product/services would benefit from greater visibility in view of the increased competitiveness of their service and therefore relevance in ranking.

    Better informed and more impartial choice for consumers.

    Demand driven by the competitiveness of the offer and not by unfair practices.

    Gatekeepers shall not technically restrict the ability of end users to switch between and subscribe to different software applications and services to be accessed using the operating system of the gatekeeper

    Example: an app store reserving for some providers with whom it has partnership agreements certain functionalities, thus preventing consumer switching to a different internet access provider.

    Gatekeepers would be effectively prevented from restricting switching.

    This may lead to decrease in demand in absence of innovative competing products in comparison to competitors.

    Increased switching likely to lead to some loss in revenues.

    Increased ability to switch shall serve as incentive to target consumers with high quality products or services.

    Switching should incentivise increased innovation in a variety of markets.

    Increased overall competition and contestability of the markets.

    Incentive to switch would lead to a more volatile markets due to competitive choice.

    This would in turn lead to more innovation, lower prices and higher quality products and services.

    Increased choice for consumers.

    Lower prices and higher quality of service.

    Gatekeepers shall not prevent business users and providers of ancillary services access to and interoperability with the same operating system, hardware or software features that are available to or used by any ancillary services provided by the gatekeeper.

    Example: provider of financial services online would like to obtain access to certain features available to the payment services of the gatekeeper that are needed to perform certain operations, but is refused access to such features.

    Gatekeepers would not be able to discriminate against business users ancillary services providers in terms of access to their operating system, hardware or software features.

    Increased level playing field

    Business users would benefit from systems and features needed for their activity.

    Consumers would benefit from higher quality and more diversified choice.

    Gatekeepers shall not refuse to provide advertisers and publishers upon their request, with access to the performance measuring tools of the gatekeeper and the information necessary for advertisers and publishers to carry out their own independent verification of the ad inventory.

    Example: advertisers would like to obtain access to performance measuring tools of gatekeepers to assess effectiveness of its advertising campaign, but is refused access to such tools.

    Gatekeepers would provide for increased transparency of its advertising system.

    Increased transparency is likely to put downward pressure on pricing, which is considered opaque.

    This may lead to some lost advertising revenue.

    Increased transparency may facilitate entry or expansion of competing service providers.

    Success of competitors largely dependent on the actual switch in demand from business users, with at least part of their demand (reach of gatekeeper platform still being an important barrier to entry).

    This would lead to improved transparency of the advertising value chain and better targeting of the service needed.

    Some demand may switch to alternative operators, but that will largely also depend on their reach to consumers.

    Increased choice and competition likely to lead to lower prices paid for advertising, which should subsequently be passed on in final prices for consumers.

    Gatekeepers shall provide business users with effective data porting possibilities for data generated on core platform services, subject to GDPR consent requirements as applicable.

    Example: third-party provider of online newspaper may need access to data of the potential user (i.e. subscriber) of its services in real time.

    Gatekeepers would be required to provide for effective means of data portability, practice which does not seem to function at the moment.

    Gatekeepers would not benefit from an important barrier to entry and expansion faced by existing competitors.

    Interoperability measures required would raise compliance costs for gatekeepers.

    Competitors would benefit from incentive of business users who would have an effective mean for switching or multi-homing.

    Incentive and ability to switch would lead to incentive to enter or expand in the market and provide innovative services.

    Increased ability of business users to switch or multi-homing.

    Incentive to switch or multi-home could lead to more innovative services and choice.

    Increased choice for consumers.

    Lower prices and higher quality of service.

    Gatekeepers shall not prevent free of charge, unhindered access to and use of non-aggregated and aggregated data that is provided for, generated in the context of, or inferred from, the use of the relevant core platform services by those business users and the customers acquiring the products or services provided by those business users.

    Example: online newspaper asks the provider of online intermediation service for contacts of the customer who subscribed to its service through software application store of the gatekeeper, but is refused such data on privacy grounds, even if subscriber was never asked for consent, or lack of it, for such data sharing.

    Gatekeepers will not be able to prevent direct contacts between business users and consumers.

    Gatekeepers might lose some consumers who may switch to business users or competitors and thereby lose some revenue.

    Level playing field among competitors in the online platform environment.

    Competitors would benefit from business users being able to obtain their customer data and possible use similar services from competitors of gatekeepers.

    Business users would be able to better understand behaviour of their customers.

    Would enable better targeting of the offers as well as addressing possible shortcoming in the service (e.g. reasons for termination of a contract).

    Consumers would benefit from a more direct contact with the business that is providing demanded service.

    More targeted supply could lead to more competitive offer in terms of lower price and higher quality.

    Gatekeepers shall not prevent access, upon request of business users and any third party providers of online search engines, to query, click and view data in relation to free and paid search generated by consumers on the online search engines of the gatekeeper.

    Example: provider of competing online search engine services asks the gatekeeper to provide access to its click-and-query data, which is refused without any consideration or explanation.

    Providing access to click-and-query data is likely to lead to increased competition in online intermediation services and online search engines.

    Increased competition expected to lead to incentive to innovate and provide higher quality product than competition.

    By increasing level playing field the competitors obtain an opportunity to differentiate themselves from gatekeepers and show that quality of their product or service.

    Increased competition overall, likely to lead to more innovation and more tailored products.

    Increased competition likely to lead to increased choice and thereby lower prices and higher quality products and services.

    Gatekeepers shall apply fair and non-discriminatory general conditions of access for business users to its software application store.

    Example: provider of an app store charges different commission rates to different business users without clear identification of reasons for such differentiation.

    Gatekeepers would not be able to treat differently third parties on their respective platforms, they would not be allowed to prevent the latter from benefitting from the same access conditions.

    Gatekeepers would however not be prevented from determining the substance of the access conditions for their core platform services – these would merely be subjected to the fairness and non-discrimination check.

    Business users directly competing would benefit from better level playing field vis-à-vis gatekeepers.

    Business users could benefit from better quality of service, i.e. better access conditions, which would also allow them to offer increased quality of services to consumers.

    Consumers would benefit from greater quality products/services by business users.

    3.Summary of costs and benefits

    The costs and benefits associated to this initiative are specified below, comparing where possible the outcome of different options assessed.

    3.1.Overview of benefits for the preferred option

    I. Overview of Benefits – Preferred Option 2

    Description

    Amount

    Comments

    Internal market fragmentation (see also Annex 5.5 on cost of non-Europe)

    EUR 92.8 billion

    It is expected that here will be a substantial decrease in internal market fragmentation, as EU Member States will not need to introduce national legislations. The effect of market contestability on the internal single market is proxied by an increase in online cross-border trade and the indirect/spill-over effect in terms of employment, economic growth, innovation and consumer surplus (see below). If we assume that by preserving the internal market in the platform space cross-border trade projections by 2025 could be maintained, this would lead to EUR 92.8 billion. 87

    Impact on economic growth

    EUR 12 billion - EUR 23 billion

    Input-output micro-econometric modelling: Higher investment in R&D in the ICT sector in EU27 leads to an overall increase in the EU27 income between 0.09% to 0.17% of 2014 EU GDP, this is between EUR 12 billion and EUR 23 billion.

    Both impacts on growth and employment (below) are very conservative estimates because they result exclusively from an increase in R&D investment. However, market contestability and more fair competition are expected to produce important spillover effects that result in higher innovation, increase in market size, increase of entrepreneurship within and beyond the platform economy and growth in other traditional sectors. Online cross-border trade is expected to be highly impacted by this virtuous dynamic. Therefore, this estimation is not taking into account further rounds of direct and indirect effects with positive loops in the long-term.

    Employment

    600 000 jobs preserved (conservative scenario) – b/n 136,387 and 294,236 jobs created (optimistic scenario)

    The preferred option would either preserve the current level of employment in the sector or lead to its increase 88 thanks to the increase in R&D spending (input-output microeconomic modelling)

    Innovation

    EUR 221 billion and EUR 323 billion over 10 years

    Financial resources that could be invested in R&D are diverted to mergers and acquisitions (M&A), which results in higher market concentration instead of improvements in the quality and quantity of products and services for consumers. This pattern of innovation dedicated to competing 'for the market' has a detrimental effect on consumer choice and surplus.

    In addition, the positive impact on innovation stemming from higher market contestability is not limited only to diversion of money from M&A to R&D. Other expected indirect effects include an increase in entrepreneurship and creation of new products and solutions meeting consumers' needs rather than focused on exploiting a gatekeeping position. This may have a multiplicative effect increasing the size of the European single market, and hence, GDP and online cross-border trade (see other impacts in this table).

    Investment in R&D

    EUR 12 billion– EUR 23 billion

    Higher investment in R&D in the ICT sector in EU27 leads to an overall increase in the EU27 income between 0.09% to 0.17% of 2014 EU GDP, 89 i.e. between EUR 12 billion and EUR 23 billion (input-output modelling).

    Competition

    Fall in HHI index 0.25 (user shares) and 0.11 (revenue shares)

    It is expected that competition will improve substantially due among other to a substantial decrease in barriers to entry. Conservative estimate is no increase in the HHI Index, while upper bound means a fall in HHI index on for the user shares by 0.25 points and 0.11 for the revenue shares.

    Online cross-border trade

    EUR 450 billion to EUR 1.76 trillion after 10 years

    Assuming the internal market fragmentation is fully addressed, the online cross-border trade would increase between EUR 450 billion to EUR 1.76 trillion after 10 years.

    Although it is hard to forecast with precision the increase in online cross-border trade, the impacts have been proxied by similar trends in offline cross-border trade resulting from market integration.

    The opportunity costs estimated here are very conservative as the assumed trends were linear and conservative growth rates. The fast change in the platform economy and interlinks with the rest of the economy suggests that online cross-border trade could see an important exponential growth if enhanced by market contestability, fair competition and virtuous patterns of innovation.

    Consumer surplus

    EUR 13 billion

    The higher level of competition may result in lower prices as companies could decrease spending on advertising and lower costs; such savings could be passed onto consumers (especially where (price) competition increases). Consumer surplus of EUR 13 billion is based on the assumption that competitive asymmetry between gatekeepers and alternative platforms would be addressed (see Annex 4).

    3.2.Overview of costs

    The Table below presents a cost comparison between Options 1, 2 and 3 (including underlying assumptions for cost estimates). In relation to the number of platforms captured by the qualitative criteria, it is very difficult to estimate upfront given that only after a market investigation it would be possible to determine whether a given provider of core platform services meets the criteria. The calculations are thus provided on the basis of illustrative numbers.

    It is important to note that the costs for the EU Commission in terms of FTEs (full time equivalents) refer mostly to the internal redeployment of already existing job positions.

    Cost comparison

    Option 1

    Option 2

    Option 3

    Carrier

    Cost qualification

    Cost quantification

    Cost qualification

    Cost quantification

    Cost qualification

    Cost quantification

    European Commission

    Regulatory costs of implementation, supervision, information gathering.

    Associated burden is estimated based on experience from other sectors where regulation requires the preparation of guidelines, designation of actors with market power and enforcement of conditions aimed at supporting contestability and avoiding foreclosure, i.e. telecoms regulation and competition law.

    It is assumed that the Commission would engage in preparing and processing information requests as well as the preparation of guidelines, designation of gatekeepers and enforcement of the obligations

    Annual costs: between €6.4m (sub-option 1-A) and €10.5m (sub-option 1-B).

    This is based on 30 FTEs in case of sub-option 1-A (with a cost of €3.9m) and 50 FTEs in case of sub-option 1-B (with a cost of €6.5m).

    Additional costs (between €2.5m and €4m) are necessary in relation to the support of experts, provision of training, development of required IT systems, expenditure with missions and organisation of meetings.

    In addition to costs identified under Option 1, further data requests, implementation, assessment and enforcement/supervision costs are to be foreseen.

    Further implementation costs would stem from the regulator specifying the obligations imposed to a given gatekeeper.

    Further assessment costs would stem from the need to conduct market investigations to designate gatekeepers and assess new practices.

    Annual costs: €16.7m.

    This is based on 80 FTEs under both sub-options (€10.3m).

    Additional costs (i.e. around €6.4m) are necessary in relation to the support of experts, provision of training, development of required IT systems, expenditure with missions and organisation of meetings.

    In addition to costs identified under Option 2, further costs would be incurred in similar tasks in relation to other digital services, including implementation, assessment, enforcement/supervision costs, and assessments of fairness.

    Annual costs: €18.2m

    This is based on 90 FTEs (€11.7m).

    Additional costs (i.e. around €6.5m) are necessary in relation to the support of experts, provision of training, development of required IT systems, expenditure with missions and organisation of meetings.

    National authorities

    Responses to consultations held by the EU regulator to integrate national expertise before taking a decision (e.g. on guidelines non-compliance, fines).

    Annual costs: €4.3m based on 2.5 FTE for 27 Member States

    In addition to costs under Option 1, Option 2 would imply costs for national regulators to study Commission’s proposed draft decisions on further tailoring of obligations.

    Annual costs: €6m based on 3.5 FTE for 27 Member States

    In addition to Option 2, Option 3 would not imply any additional costs for national regulators.

    Annual costs: €6m based on 3.5 FTE for 27 Member States

    Gatekeepers

    Compliance costs incurred in order to prepare for compliance with rules, set compliance officers, and respond to requests for information.

    Number of information requests would depend on the complexity of the case. Estimate assumes that 20 FTEs are involved in data gathering, monitoring and enforcement activities per gatekeeper platform.

    This scenario does not consider possible synergies with already existing internal organisation/service for complying with other legislation, e.g. service ensuring COMP law compliance.

    Annual costs: between €9.87m and €21.15m for a total number of gatekeepers in scope between 7 (under sub-option 1-A) and 15 (under sub-option 1-B)

    Similar compliance costs per platform as per Option 1.

    On the one hand, the possibility of a dialogue would reduce the compliance costs. On the other hand, the need to reply to request for information in the context of market investigations would imply some extra costs.

    Annual costs: between €21.15m and €28.2m for a total number of gatekeepers in scope between 15 (under sub-option 2-A) and 20 (under sub-option 2-B)

    Similar compliance costs per platform as per Option 1.

    On the one hand, the possibility of a dialogue would reduce the compliance costs. On the other hand, the need to reply to request for information in the context of market investigations would imply some extra costs.

    Annual costs: around €35.25m based on 25 gatekeepers.

    Competitors, start-ups, business users

    Monitoring of unfair conduct as well as new rules’ implementation and supervision of compliance would imply some burden in the form of e.g. responses to information requests.

    However, in order to ensure proportionality information requests would take into consideration the size of the enterprise to which they are sent.

    The resources devoted to these requests might be counteracted by reductions in legal resource required to address unfair contractual conditions, with a substantial portion of the burden previously taken by small firms in this area now addressed through tailored action at EU level.

    Net additional resource requirements likely to be very limited

    Monitoring new forms of unfair practices would create additional costs for market players as compared with Option 1. However, in order to ensure proportionality information requests would take into consideration the size of the enterprise to which they are sent.

    The resources devoted to these requests might be counteracted by reductions in legal resource required to address unfair contractual conditions, with a substantial portion of the burden previously taken by small firms in this area now addressed through tailored action at EU level.

    Net additional resource requirements likely to be very limited

    Monitoring new digital markets would create additional costs for market players as compared with Option 1. However, in order to ensure proportionality information requests would take into consideration the size of the enterprise to which they are sent.

    The resources devoted to these requests might be counteracted by reductions in legal resource required to address unfair contractual conditions, with a substantial portion of the burden previously taken by small firms in this area now addressed through tailored action at EU level.

    Net additional resource requirements likely to be very limited

    Consumers

    Responses to public consultations - questions targeting consumers would be less complex and shorter. Possibly higher search costs

    Additional information gathering from consumers may be needed to inform specification/tailoring of remedies.

    Higher search costs

     

    Additional information gathering from consumers may be needed to inform about other digital services.

    Higher search costs

    Total costs:

    EUR 20.57m – 35.95m

    EUR 43.85m – 50.9m

    EUR 59.45m



    Annex 4 – Analytical methods

    The teams at DG CNECT, GROW and COMP, as well as the contractor of the study supporting the Impact Assessment and JRC conducted calculations to estimate the impact of the unfair practices employed by platform and market failures.

    The quantitative assessments relied on estimates available in empirical studies quoted in the Impact Assessment, correlation analysis was based on data from Statista and an Input-Output macro-modelling. The assessment was guided by the EU Better Regulation Guidelines.

    1.INPUT-OUTPUT MODEL

    1.1Introduction

    The input-output (I-O) model is the name given to a modelling approach developed by Professor Wassily Leontief in the late 1930s 90 . As its name suggests, the I-O model assumes that there is a matrix that links transactions or flows recording payments to and from a sector within a year. Besides, the framework works on double-entry bookkeeping so that total gross output must equal gross input. 0 below illustrates the model.

    The row total represents the total produced (supplied) by a sector while the total column represents the total used (demanded) by such sector. Hence, any element in each cell is what sector j use from sector i.

    Input-output transaction matrix

    Source: Miller and Blair (2009) 91

    The model is built using observed economic data from national account statistics to show the flows of products going from each industrial sector seen as a producer to sectors seen as consumers. The grey area in 0 above is the interindustry trade to which must be added the final demand columns and the value-added rows.

    National account data will populate the matrix which will be used to estimate impacts out of exogenous shocks. For example, each in the matrix below (0) will be constructed from official statistics. Such matrix will be used to find a matrix with the multiplier effects to estimate how exogenous changes in one specific sector of the economic impacts in the other sectors, value-added, final demand and lastly in GDP.

    Example of a two-sector economy

    Source: Miller and Blair (2009)

    The next section describes the implementation of the I-O model to this impact assessment.

    1.2Implementation of the I-O model for the impact assessment

    In this Implementation of the I-O model for the impact assessment analysis, data was taken from the sources below:

    ·The 2014 world input-output table (WIOT) publicly available from the World Input-Output Database (WIOD, www.wiod.org),

    ·Employment (number of persons engaged) and compensation of employees obtained from the Socio-Economic Accounts (SEAs) of WIOD, and

    ·Private R&D investments in information and communication (and its subitems represented by NACE Rev.2’s Section J’s divisions and/or groups), obtained from Eurostat 92 .

    The most recent data were available for 2014 (WIOD Release 2016), which explains the choice of the year in our impact assessments. The WIOTs and SEAs cover 43 countries and the rest of the world region, each detailed by 56 industries according to the International Standard Industrial Classification Rev. 4. All tables adhere to the latest version (2008) of the System of National Accounts.

    The incorporate the impact of market contestability and fairer competition in GDP and employment into the I-O model, we needed to assume that such market dynamic would result in higher investment in R&D in the platform economy, impacting in GDP and job creation. However, as the platform economy is still relatively new to the national account system there is not an exact code for such sector and we had to take some sub-sectors from the ICT sector as a proxy 93 .

    The results suggest that private investments in ICT sectors account only for roughly 0.10% of the EU GDP. The I-O modelling exercises show that these investments imply:

    ·An overall EU income increase from 0.09% to 0.17% (of 2014 EU GDP) and EU employment increase from 0.07% to 0.15% (of 2014 EU employment);

    ·At the EU level, most of the impacts are driven by one ICT subsector, consisting of Computer programming, consultancy and related activities and Data processing, hosting and related activities, web portal;

    ·The impacts are, however, heterogenous across the individual EU countries.

    1.3Limitations

    One of the main limitations is the lack of exact code to identify the platform economy which may be underestimating the actual size of the sector and hence the contribution and links to the overall economy.

    A second limitation is that it only incorporates the increase of R&D but there might be other exogenous shocks resulting from market contestability and fairer competition, including higher market size and higher online-cross-border trade. As it is difficult to know a priori the increase in market size and across which sector, incorporating this into the model proves challenging.

    Other direct and indirect effects such as entrepreneurship, quantitative and qualitative changes in the patterns of innovation as well as lower prices to consumers resulting from market contestability are not included in the model for the same reasons as failing to incorporate change in market size. Therefore, the estimations must be taken as conservative and lower bound.

    2.Methodology used for calculating consumer benefits stemming from a more competitive platform economy in the EU

    2

    2.1Presentation

    This note summarises the method used to assess the economic impact of the DMA in the EU and the (preliminary) results obtained.

    The quantitative methodology adopted can be conceptualised as a partial-equilibrium structural approach. The econometric model is grounded in a partial-equilibrium framework since it uses very detailed data to identify with high precision consumers' substitution patterns for a large set of digital services providers in 19 EU Member States and the UK. It is structural in the sense that economic theory is used to develop statements about how a set of observable endogenous variables are related to another set of observable explanatory variables, and sometimes also to a set of unobservable variables. However, economic theory alone cannot provide enough information for the estimation of the model. For this reason, there is a need to add statistical assumptions about its observed and unobserved variables. A key reason to use economic theory, beyond the specification of the relationship between the variables, is to clarify how institutional and economic conditions affect these relationships. This specificity is essential to make causal statements about the estimated relationships, or use them to perform counterfactuals, i.e., scenarios that have not been implemented but that can represent the likely outcomes of policy interventions.

    The methodological approach is framed in the tradition of structural estimation in empirical industrial organisation in the economics profession. This approach uses discrete choice models for the estimation of demand and adds a simulated supply side to compute the industry equilibrium given by the observed data. Adding a simulated supply side to account for firms' strategic behaviour, the observed market equilibrium can be found. Moreover, by changing supply or demand conditions, the framework allows for the design of counterfactuals that simulate policy changes.

    The model used here is a modified version of Duch-Brown et al. (2015), and was developed by researchers from the JRC. The model is a partial equilibrium approach using detailed data for usage of a large set of digital services, allowing estimating with a high level of accuracy demand substitutability and market equilibrium.

    From a market analysis perspective, there are three potential competitive constraints: demand substitution, supply substitutability, and potential competition. Demand substitution constitutes the most immediate and effective disciplinary force on suppliers, and in particular to their strategic decisions. Supply substitutability and potential competition are relevant in the medium to long terms, since they imply the need of adjustments through tangible or intangible assets, additional investments or strategic decisions, all of which would imply significant changes in the markets under consideration 94 . Hence, a precise estimation of demand substitutability is essential to the analysis of the effects of changes in the institutional setting of a given sector, and this is the basis for the approach taken here.

    We consider the demand for several categories of digital services. Consumers can choose among a large variety of websites that are differentiated in quality. Furthermore, consumers can also decide not to visit a website at all, in which case they can spend their time on other (offline) services goods. To model the substitution patterns, a two-level nested logit model is used which allows for market segmentation according to two discrete dimensions: i) category; and ii) sub-category. This model is useful since the nesting parameters enable one to assess to which extent consumers view the options in the same distribution channel and/or quality category as closer substitutes.

    Assuming that consumers choose the product with the highest utility, one can obtain the choice probabilities for every product in every country, including the probability of selecting the outside good (McFadden, 1978). At the aggregate level, these choice probabilities can be equated to the market shares, relative to a hypothesised potential market, defined here as representing twice as much as the observed website visits. 95 The demand model can be used to compute consumer surplus (McFadden, 1978 or Anderson et al. 1992. If the model conforms to the basic principles of consumer theory, the model translates preference correlations into aggregate substitution patterns. Products in the same subgroup will have higher substitutability than products in a different subgroup.

    An oligopolistic supply side is added to the model to infer marginal costs and current economic profits; as well as to define the observed market equilibrium. The model assumes that firms maximize profits, and that they compete in a differentiated products setting (Bertrand competition). As shown by Berry (1994) and Berry, Levinsohn and Pakes (1995), the profit maximising conditions can be used to compute the current marginal costs. Furthermore, this system can be used to perform policy counterfactuals, and in particular the effects of introducing more competition in this particular setting. The model also calculates consumer welfare (consumer surplus) changes, by computing the welfare measures in the different counterfactuals and in the observed market equilibrium.

    The results with respect to the estimated consumer surplus are:

    Country

    Original

    Counterfactual

    Difference

    AT

    3.47

    3.78

    0.31

    BE

    4.77

    5.54

    0.77

    BG

    1.93

    1.86

    -0.06

    CZ

    6.00

    5.93

    -0.07

    DE

    34.11

    37.17

    3.06

    ES

    20.27

    20.72

    0.45

    FI

    3.43

    3.86

    0.43

    FR

    29.41

    31.88

    2.47

    GB

    36.85

    40.08

    3.23

    GR

    3.83

    3.91

    0.09

    HR

    1.62

    1.61

    -0.01

    HU

    4.01

    3.95

    -0.06

    IE

    2.65

    2.95

    0.30

    IT

    20.75

    21.49

    0.74

    NL

    10.31

    11.59

    1.28

    PL

    18.67

    18.38

    -0.29

    PT

    3.75

    3.81

    0.06

    RO

    3.74

    3.63

    -0.10

    SE

    5.15

    5.71

    0.56

    SK

    2.00

    1.99

    -0.01

    Total

    216.72

    229.87

    13.15

    2.2Limitations

    The methodology suffers several limitations. First, the results come from a simulated counterfactual scenario, based on a series of assumptions, which may not necessarily be true. For instance, the assumption that digital service providers compete according to the Bertrand behaviour is questionable, but practical from an empirical point of view. Second, the data used covers just one month, and there may be singularities in that particular point in time that are different from a more wide perspective (ie, several months, or even years). Third, the results are based on a hypothesised consumer behaviour model, which may not correspond to real choices. Finally, the results refer to alternative platforms only (not considering business users). Consequently, the results account for consumer surplus which would stem from increased inter-platform competition but do not reflect positive effects on intra-platform competition (that can be legitimately expected as a result from the application of the preferred option).

    3.References

    Berry, S., J. Levinsohn, and A. Pakes (1995). Automobile Prices in Market Equilibrium, Econometrica, 63, pp.841-890.

    Berry, S.T. (1994). Estimating Discrete-Choice Models of Product Differentiation, RAND Journal of Economics, 25(2), pp.242-262.

    Duch-Brown, N. and B. Martens (2016). The economic impact of removing GB restrictions in the EU Digital Single Market. JRC/IPTS Digital Economy Working Paper 2016-02.

    Duch-Brown, N., L. Grzybowski and F. Verboven (2015). The Impact of Online Sales on Consumers and Firms: Evidence from Household Appliances. JRC/IPTS Digital Economy Working Paper 2015-15.

    McFadden, D. (1978). Modelling the Choice of Residential Location, in Spatial Interaction Theory and Planning Models. Karlquist, A., L. Lundquist, F. Snickers, and J. Weibull (eds.). Amsterdam: North Holland.



    Annex 5.1 – Overview of consultations and expert advice reports conducted in the context of the New Competition Tool

    96 The following consultations and expert advice reports have been conducted in relation to the market investigation regime. All of these summaries and reports can also be found on DG Competition's dedicated website.

    Inception Impact Assessment of the New Competition Tool:

    ·Feedback received: https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12416-New-competition-tool/feedback?p_id=7937377  

    Open Public Consultation on the New Competition Tool:

    ·Summary of contributions received in the context of the Open Public Consultation on the New Competition Tool: https://ec.europa.eu/competition/consultations/2020_new_comp_tool/summary_stakeholder_consultation.pdf

    ·Contributions outside EU Survey: https://ec.europa.eu/competition/consultations/2020_new_comp_tool/contributions_outside_eu_survey.zip  

    Consultation activities in the context of the European Competition Network (National Competition Authorities of the European Economic Area):

    ·Summary of the contributions of the responses: https://ec.europa.eu/competition/consultations/2020_new_comp_tool/summary_contributions_NCAs_responses.pdf  

    Expert advice reports commissioned in the context of the New Competition Tool:

    ·Prof. Massimo Motta and Prof. Dr. Martin Peitz (2020), Intervention triggers and underlying theories of harm, https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420575enn.pdf  

    ·Prof. Pierre Larouche and Prof. Alexandre de Streel (2020), Interplay between the New Competition Tool and Sector-Specific Regulation in the EU, https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0120577enn.pdf  

    ·Prof. Dr. Heike Schweitzer (2020), The New Competition Tool: Its institutional set up and procedural design, https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420574enn.pdf  

    ·Prof. Richard Whish (2020), Legal comparative study of existing competition tools aimed at addressing structural competition problems, with a particular focus on the UK’s market investigation tool: https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420573enn.pdf  

    Report by the Economic Advisory Group on Competition Policy (EAGCP) (Gregory Crawford, Patrick Rey and Monika Schnitzer) on an economic evaluation of the NCT: 

    · https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0320680enn.pdf  

    Annex 5.2: Summary of the EU Observatory work supporting the initiative

    The Observatory for the Online Platform Economy supported by its expert group and the support study has produced a number of analytical papers and reports that confirm the international consensus on the need for new rules for digital platforms in order to complement the competition law enforcement.

    1.The reports by the expert group for the Observatory on the Online Platform Economy

    Firstly, the expert group for the Observatory on the Online Platform Economy  has produced 3 preliminary reports published for feedback on 9 July:

    §Measurement of the Online Platform Economy

    §Differentiated treatment

    §Data in the Online Platform Economy

    §Market Power and Transparency Issues in Open Display Advertising – a case study (to be published in December)

    §Market Power (to be published in December)

    This feedback will form part of the Final Report to be published by the end of 2020. It will also include two further reports: on the transparency in the online advertising and market power.

    1.1.Report on Measurement and Economic Indicators

    The Report on Measurement and Economic Indicators identified the indicators that could be used to monitor the online platform economy for the purposes of policy making and further regulation, (e.g. in order to identify platforms in scope of the regulation). The report breaks down the problem of observation into three broad areas that cut across policy domains.

    The first is economic significance of platforms in the context of the broader economy. The report identifies three measurement indicators: volume of trade mediated by platforms; platform size and importance; and data on data. It offers suggestions as regards new, more conceptual approaches to measuring platform size and ‘data on data’.

    The second area of observation is the platforms’ power over their users. The report identifies three indicators for measurement: business dependence on platforms; platform’s share of consumer attention; and acquisitions as a competitive strategy.

    Regarding acquisitions as a competitive strategy, including so-called ‘killer acquisitions’ designed to pre-empt future competition, the report suggests automated market intelligence data feeds and recommends to consider new obligations on major platforms to report M&A activities to the European Commission, for ex-post research and monitoring purposes.

    The third area of observation covered in the Measurement report relates to the alleged effects of platforms’ power: how to measure platform volatility (e.g. continuous changes in terms and conditions or algorithms); platform transparency; and other potentially problematic and thus policy-relevant practices. The report stresses that platform transparency would benefit from further conceptual research to better understand the trade-offs between a public’s need for transparency of powerful actors vs. the legitimate private business interests of a platform company.

    As for other potentially problematic practices, the report recommends that the data generated by the internal complaint-handling procedures, as mandated by the P2B Regulation, should be analysed with a view to identifying and assessing any need for further public policy intervention.

    1.2.Report on differentiated treatment

    The Report on differentiated treatment focuses on differentiated treatment as a potential source of ‘unfairness’ in the relationship between platforms and their business users in the online platform economy. It distinguishes between practices of self-favouring, whereby a platform gives preferential treatment to its own vertically integrated activities over those of rivals, and more general practices of differentiated treatment where one or more business users are treated more favourably than others.

    The report provides guidance on how to assess the impact of differentiated treatment by online platforms from a technical, economic and legal perspective. It also identifies areas requiring further scrutiny because of the particularly problematic nature of certain practices implemented by platforms. Given that instances of differentiated treatment are not necessarily limited to cases where a platform holds a ‘dominant position’ within the meaning of EU competition law, the report looks beyond the application and interpretation of competition law.

    The report stresses that for assessing what practices can be considered ‘unfair’, more transparency and oversight are needed into the practices in which platforms engage. In this respect, the Platform-to-Business Regulation 97 provides a good starting point to facilitate the more concrete identification of forms of differentiated treatment that can be considered unfair and might, as such, need to be regulated.

    The report concludes that it is desirable to keep monitoring the sector closely and conduct focused studies to scrutinise the impact of problematic practices.

    1.3.Report on Data in the Online platform ecosystem

    The Report on Data in the Online platform ecosystem provides a structured overview of how data is generated, collected and used in the online platform economy. It maps out the diversity and heterogeneity of data-related practices and expands on what different types of data require a careful examination in order to better understand their importance for both the platforms and their users as well as the issues and challenges arising in their interactions. The report concludes with a range of issues, which deserve, in the view of the authors, further policy attention and analysis in the light of the limited evidence available and/or the importance and impact they entail.

    2.Support Study for the Observatory

    The consortium supporting the work of the Observatory composed of PPMI (lead), Open Evidence, IW and Rand Europe 98  have produced the following analytical papers (AP):

    AP1: Differentiated treatment (IW)

    AP2: Platform data access and secondary data sources (PPMI)

    AP3: Transparency in the business-to business commercial relations in the online advertising market (Open Evidence)

    AP4: Significant Market Status (RAND) 

    AP5: Business user and third-party access to digital platform data (PPMI)

    AP6: Structure of the online platform economy post COVID-19 outbreak (Open Evidence)

    AP7: The main obstacles and opportunities for multihoming (PPMI)

    AP8: Developments concerning B2B platforms and emerging issues (RAND)

    1.4.Analytical paper #1: Business user access to platform data and alternative data sources

    The paper argues that online platforms create value by using data to facilitate interactions (for example, commercial transactions) between users. This means that data is at the core of the platforms’ business model and they use it to provide and improve their services.

    Data in the possession of platforms allows them to understand the preferences of customers and their reactions to market signals, including changes in prices and product characteristics. This puts online platforms in a unique position as they are able to observe the functioning of the market in real time 99 . Data is thus a key source of market power. In other words, platforms’ decisions on what data to share, with whom, and under which conditions have far-reaching consequences to all the participants in the market.

    Business users: data needs and access to data

    In the paper the contractor identified three general dimensions of data relevant to platform business users:

    ·the type of data by object (customers, businesses, user behaviour, markets, transactions, etc.);

    ·whether the data is about an individual business which receives it, or other businesses on the platform (competition)/ whole marketplace.

    ·by the level of data processing and its value, from raw datasets to insights guiding business decision-making.

    The paper shows that the kinds of data provided or not provided by the platforms (Amazon, eBay, Google Play and Booking.com) are rather similar. Access to data as well as advanced analytics are granted to the extent that it could generate more income for the platform as well as the business users. In such a case the key question is whether the business users can take full advantage of the data provided to them. Further, a significant share of businesses signal that they are experiencing data access problems. This was very visible in interviews where businesses, especially the bigger or stronger ones, felt strongly about the data access. Their key concern was getting access to data so that they could use it to innovate and keep up with the competition. A recurring issue was also the power of the vertically integrated platforms and especially the extent to which such platforms may use data to develop their own competing products.

    The paper explored firstly, the findings concerning taking advantage of the available data and, secondly, the evidence concerning data that is not shared with the business users.

    The research showed that a significant share of business users express dissatisfaction with regard to the level of data access provided to them by online platforms. The business user survey showed that access to data possessed by online platforms is of concern to around a third of surveyed business users who reported that they cannot access at least some data that is essential to their business. Generally, although the platforms collect and analyse loads of data, only a fraction of this is provided to other players. The platforms do not share the raw big data on day-to-day activities, as well as detailed data on customers and competitors.

    The paper identified three groups of concerns that business users express with regard to data sharing.

    The first is related to lack of access to personal data, such as customers’ e-mail address. Some business users, especially in the hospitality and e-commerce sectors consider such data of key importance to them so that they could establish a more direct client relationship. Other personal data collected by platforms, but usually not provided to business users include, for example: telephone, address, credit-card data. As confirmed by our desk research and interviews with the platforms themselves, this data is not provided for a number of reasons. Firstly, this is not considered compatible with the platform’s business model as business users may use direct communication to bypass platforms in the future. Secondly, platforms consider that a consistent client relationship and data protection is part of the client experience. They are wary that direct access to the clients by business users may result in a surge of unwanted marketing messages (this argument was not supported by the interviewed business users). Finally, the personal data protection regulation (including GDPR) puts obligations on platforms in terms of data sharing and management, including the obligation of getting explicit consent from consumers to collect and share their data. According to the business user survey, legislative or regulatory restrictions is indeed acknowledged as the key reason for not getting access to data.

    Secondly, business users need data that help them to stay competitive, innovate and develop their products and services. Partly this is related to data on, for example, search keywords, search volumes, consumer behaviour in reaction to different price signals. Platforms do share such data to a certain extent (e.g. three most important key words), however some interviewed business users felt that this does not give them sufficient level of detail. Partly, this is also related to data about competitors and their products and services. In this case however, both the interviewed platforms as well as business users expressed understanding that the level of detail is naturally limited as businesses would not want their individual business performance information to be made available to others.

    The third concern is that platforms are taking advantage of data to promote their own products that are very similar to those offered by their business users. This is primarily pertinent to vertically integrated platforms with significant market power. So, in the business user survey, 58% of respondents reported that the platform itself offers the same (or very similar) goods or services to those that their businesses offer on the platform. Among these respondents, 55% argued that online platforms are favouring their own goods or services vis-à-vis the same (or very similar) goods or services offered by their businesses. Unique and comprehensive datasets on all the firms and their consumers operating in the marketplace can give a huge business advantage to the platform operators. The key ways of favouring include ranking, placement of advertisement, pricing and other – all of these are enabled by the data collected by platforms. Some interviewed business users argued that platforms (specifically – Amazon) are using data to monitor which goods have the best margins in the market and then move into offering such goods themselves.

    The paper also analysed the role of data companies. The businesses that need more data than they can get from platforms pursue two broad strategies: (1) collect and analyse data themselves, from sources available to them; (2) rely on third party providers (data brokers). Four-fifths of the business user survey respondents (81%) indicated that they collect some data themselves; the most prevalent data type is identification details of own customers (57% of respondents collect this data), followed by business performance data (55%) and analysis of market trends/ developments (55%). Further, a third of the surveyed companies (33%) reported that they use third-party sources (data brokers). Most of the interviewees - especially in the e-commerce sector - reported that they use the services of third-party data and analytics providers.

    Companies specialising in data and insights fill important data gaps, especially with regard to data on competition and actionable business insights. According to PwC estimations, data companies earned USD 21 billion in 2018 100 . Some data brokers specialise to cover specific sectors, such as applications (App Annie, App figures, Apptopia, Mobile Action, Sensor Tower), e-commerce (Algopix, Jungle Scout, Sellics, Teikametrics, Terapeak) or hospitality (AirDNA, Beyond Pricing, Uplisting, Wheelhouse, Skift). Other data brokers, such as Similar Web and Zirra provide data on multiple sectors.

    Data brokers use highly advanced technical methods to extract data, or they buy data from online and offline sources. A lot of data is scraped from the platforms. Another key source is crowdsourcing business user account data. Some third-party data providers ask online sellers to share their marketplace information, and then link the data of thousands of users to draw market insights. For example, Jungle Scout collects data from a large number of sellers (over 225,000) who have opted in to share their sales information 101 . When merged with the data gathered by scraping the platform’s front-end (e.g., Best Seller rank on Amazon), this can yield quite precise estimations and extrapolations. Similarly, if AirDNA users wish to receive performance analytics, they will be asked to upload their Airbnb host IDs. After doing this, they can see their performance trends, comparative and financial analysis on all vacation rental listings.

    The key value proposition of the data brokers lies in their ability to bring together a combination of sources as well as superior technical and analytical capacities, innovative tools and approaches. Data brokers allow their users to learn about their competitors, get a detailed market overview, obtain actionable insights. According to the analysis presented in this paper, this is the kind of information that is most in demand by the business users and/or platforms do not provide to a sufficient extent. Further, business users themselves do not need to invest into any analytics or IT, but rather buy products tailored to their needs.

    For instance, as explained by several interviewed Amazon sellers, Jungle Scout and other providers, such as Helium10, AMZScout and Unicorn Smasher, supply them with comprehensive market insights and competitor overviews. Obviously, these data brokers provide estimations based on what data they could gather rather than exact information. Nevertheless, the estimations are said to be ‘spookily accurate’ 102 . Similarly, data providers for app developers, such as AppAnnie and SensorTower, offer comprehensive app market data, including performance of specific apps and markets. Interviewed app developers mentioned that they use the sources together with the app store data extensively. In the accommodation/ hospitality sector, companies such as AirDNA provide insights based on data that the OTAs do not share. For example, in late 2015 Airbnb stopped providing the overall real-time reservation data. AirDNA, in turn, uses an algorithm based on 16 indicators picked up in historical data to determine the reservation status for each listing. They argue that their algorithm has an error margin of only 5%.

    The data companies’ market is very dynamic and fast-paced. This paper identified a number of issues, illustrating the key challenges and limitations of data brokers. Firstly, the data companies remain highly dependent on data sharing policies of platforms. For example, Amazon until recently provided exact and broad match search volume and product relevance data via one of its APIs. It was feeding several third-party software providers such as Viral Launch and Helium10 until late 2018, when the platform removed these metrics from the API. Another platform, Allegro made significant investment to develop new data products (Allegro Statistics) that are now provided to its sellers; this is endangering the business model of third-party analytics providers.

    Secondly, the data needs of platform business users are often very specific and concern platforms that they use. Such data cannot be easily scraped or estimated by the third-party data providers 103 . It includes information on real-time of activities on the platform (e.g. X currently has product Y added to the shopping cart), which would allow to effectively address the customer; transaction-related data about the customers, sales activities and listings of specific business user.

    Thirdly, the huge amounts of data that data brokers collect, store, possibly re-personalise and disseminate and are of interest from the regulatory perspective, first and foremost due to privacy concerns. Most individuals or companies are unaware of what information data brokers collect on them or even that they collect information at all 104 . Due to this asymmetry, the data broker industry has been often characterised as opaque, non-transparent, arbitrary, biased, unfair and unaccountable 105 . Interviewees from the data brokers argued that they are taking actions to make sure they are compliant with data protection and privacy laws, such as the GDPR. However, other sources show that such compliance has not always been properly ensured. For example, a few months after the GDPR came into force, Privacy International filed a complaint against seven data brokers: Acxiom, Oracle, Criteo, Quantcast, Tapad, Equifax, and Experian 106 . The main argument was their failure to comply with data protection principles (such as acquiring consent, providing detailed and transparent information for the data subject access requests) and exploitation of data in unknown ways.

    As a final point, the analysis pointed out that some business users are exploring innovative approaches that would allow them to joint forces and be less dependent on big platform companies. One example includes cooperative marketplaces, such as Fairmondo.de, which belongs to its business users and employees. Through a cooperative structure, the users can share the platform as a resource for mutual benefit and decide on the rules for data sharing and access.

    1.5.Analytical paper #2: Differentiated treatment of business users by online platforms

    Differentiated treatment of business users is one way in which online platforms can distort competition. It refers to the application of dissimilar conditions to (or preferencing of) similar business users, goods or services. Differentiated treatment can affect competition in two ways. First, if a platform’s differentiated treatment disadvantages certain business users, it influences competition between business users. Second, competition can also be influenced by so-called ‘self-preferencing’ on the part of vertically integrated platforms. Such businesses not only operate the platform but are also business users of the platform – for instance, they sell their own products via the marketplace. Vertical integration is desirable for online platforms because it enables them to develop new revenue streams and exploit opportunities that arise from analysing data generated by the platform. The inherent danger of vertical integration lies in the opportunity it provides the platform to abuse its favourable position. Since the platform directly controls the ecosystem in which it competes alongside independent business users, it could employ the rules to its own advantage.

    This analytical paper on differentiated treatment demonstrates that differentiated treatment by online platforms – defined as applying dissimilar conditions to similar business users – can occur for different reasons. On the one hand, the technical or regulatory framework can make such platform behaviour necessary. On the other hand, online platforms can use differentiated treatment to increase their revenues. This mainly includes platform behaviour that aims at increasing the benefits for the consumers, e.g. by offering individualised services or ensuring a high quality of the facilitated transactions. However, differentiated treatment can also aim at increasing revenue for the platform without benefits for the consumers. In such cases, differentiated treatment obstructs competition between the business users of online platforms and – in case of vertically integrated platforms – between business users and the platform itself.

    According to the data collected for this paper, vertically integrated platforms seem to possess a stronger incentive to apply such behaviour than non-integrated platforms. However, based on the available evidence, differentiated treatment of business users is not widespread in the EU.

    Reasons for the differentiated treatment of business users by online platform can generally be grouped in two categories:

    ·Regulatory or technical necessities: the legal framework within which the platform operates, or the specific technical requirements of different business users (such as specific hardware or software) may give rise to differentiated treatment. In such cases, differentiated treatment may not constitute intentionally discriminatory behaviour on the part of the platform, but may instead be a response to these specific circumstances.

    ·Increasing revenue: a platform may engage in differentiated treatment in an attempt to increase its revenue via a rise in market share or sales, or by expanding into other markets, improving its gatekeeping position, lowering its own costs, increasing the fees paid by business users, as well as offering loyalty rewards or ‘mainstreaming’, i.e. adjusting content to match the preferences of the majority of users. These motivations can explain many types of differentiating behaviour, including: blocking listings or accounts; manipulating rankings or prices; restricting access to data or installing technical barriers to business users; and differentiated terms and conditions or customer support.

    Differentiated treatment of app developers

    Applications for mobile devices (‘apps’) are developed for a specific operating system and must be distributed to the users of mobile devices. The distribution of apps is to a large extent carried out via ‘app stores’. App stores and operating systems can both be characterised as digital platforms. Apple produces both the hardware and software for its devices, and hence has a great influence on the distribution of apps for its devices. In fact, the Apple App Store is the only (and hence dominant) app store for iOS. Every app that a consumer wishes to install must first be certified by Apple. Android, in contrast, has been developed through the cooperation of large manufacturers of mobile devices, among others. Accordingly, there exist multiple app stores for Android, e.g. independent app stores and app stores implemented by device manufacturers. Google Play Store, however, remains the dominant app store. The development of dominant platforms within the app store market is due to the reinforcing of positive indirect network effects. The more consumers use an app store, the more attractive it becomes for developers to distribute their apps through this store, and vice versa. High market shares, and the fact that the platforms offer their own apps, can make differentiated treatment a serious problem for individual app developers, as well as distorting competition and harming innovation.

    Since Apple and Google offer their own apps in their app stores, both platforms are vertically integrated. Self-preferencing, as well as other forms of differentiated treatment, are therefore possible.

    To gain qualitative insights into differentiated treatment for the analytical paper, 23 interviews were conducted. App developers and publishers accounted for 15 of these interviews 107 ; their respective associations accounted for six. The remaining two interviews were conducted with Google and Apple, as the largest providers of app stores. Small app developers in particular acknowledged the opportunities platforms offered them to distributing their apps to consumers. However, 16 interviewees mentioned problems with differentiated treatment by platforms. Among these 16 interviewees, 13 were app store businesses users. 108 Furthermore, 12 interviewees claimed the platform favoured its own products or services. Ten of the interviewees that reported cases of platform self-preferencing were business users and two represented developer’s associations. 109 Generally, app developers in the interviews feared being blocked by the platform and, hence, losing customers. They also feared that the platform could enter and dominate their market. Other forms of differentiated treatment mentioned by interviewees included impeding business users that offer substitutes to the platform’s own products or services; denying access to data; or the mandatory use of platform services. Technical barriers, better customer support for large business users, and terms and conditions that favour the platform were also reported as issues. The interviewees generally claimed that larger businesses enjoyed greater opportunities to reach out to the platform in order to have their problems solved.

    Differentiated treatment of e-commerce business users

    Generally, two types of business models used by online marketplaces can be distinguished. Platforms can be either vertically integrated or non-vertically integrated. The former includes a retail arm in addition to the platform. In contrast, non-vertically integrated online marketplaces are pure platform businesses. While market shares are difficult to determine, vertically integrated Amazon is the most important online marketplace in several European countries, as well as the United States.

    In the online survey, nearly two-thirds of e-commerce respondents stated that they were completely or very dependent on online platforms. However, a clear majority of all respondents 68% strongly agreed or agreed that the online platform which was most important for their business treated its business users in a fair and unbiased manner. This is in line with the results for the entire sample (see above). The statement “My business can easily access the data collected by the platform that is important for my business”, which focusses on data access as a specific type of differentiated treatment, yields a similar result. This result does not point to a widespread occurrence of differentiated treatment. Furthermore, around two-thirds of respondents strongly agreed or agreed that many other business users on the platform offered products similar to their own. Hence, competition among business users appears high.

    Of those respondents who indicated that they had experienced differentiated treatment by a platform, the placement of advertising was the type most frequently cited (specified by around 62% of this group). The second most common type was the ranking of listings. The pricing of the platform’s services came in third. The interviews conducted with e-commerce business users confirmed the relevance of differentiated treatment in the form of manipulated ranking results, as well as a lack of access to data.

    Vertical integration of e-commerce platforms can have an influence on differentiated treatment. Nearly 53% of e-commerce respondents in our sample whose main platform offered the same or similar products reported self-preferencing by the platform. While there are limitations to this result given the survey sample, it provides a strong indication of it in markets with vertically integrated online marketplaces. The e-commerce business users interviewed did not provide unified views on differentiated treatment of vertically integrated platforms, however. While some said they had observed self-preferencing, others stated that they had not.

    According to the survey, conflicts sometimes occur between e-commerce platforms and their business users: 57% of the surveyed e-commerce business users had experienced a disagreement with the platform they most frequently used at least once. These conflicts range from disputes over technical problems or a lack of transparency in the platform’s data policy, to sudden price changes or discrimination through pricing. Many of the business users affected, namely 47%, had complained to the online platform in order to resolve the problem. Overall, the survey showed that 87% respondents had the conflicts, experienced with e-commerce marketplaces, completely resolved. Challenges mentioned by the interviewees regarding the redress process generally centred on the standardised way in which platforms dealt with complaints or requests. Several interviewees mentioned that their complaints or requests were answered by automated systems instead of humans, the replies often not capturing the essence of the complaint or request completely.

    1.6.Analytical paper #3: Transparency in the business-to business commercial relations in the online advertising market

    The paper focused on the perceived lack of transparency and accountability in business-to-business (B2B) commercial relations in online advertising. Transparency issues have been observed especially for ad exchanges and ad placements in programmatic advertising, as well as concerns about the gatekeeping role of large online platforms towards business users in the market.

    The analytical paper analysed the level and means of transparency in the online advertising value chain, through collection of evidence and facts about various business models, advertising practices and stakeholders.

    It identified three inter-related challenges affecting business to business (B2B) commercial relations in online advertising:

    ·Significant imbalances of market power in the ad ecosystem, resulting from the dominance of a few platforms that occupy strategic positions across the ad value chain and have the ability to act as gatekeepers with business users.

    ·The transparency issues in B2B relations, some of which are linked to the market power of platforms while others result from the complexity of programmatic advertising.

    ·The issues of ad fraud, exacerbated by the ad ecosystem opacity.

    The paper argued that the distribution of digital ad revenue shows that the online market is increasingly dominated by a few large online platforms (Google, Facebook) that occupy strategic positions across the ad value chain and can take advantage of their vertical integration.

    It further points out that Google, Facebook, and to a lesser extent Amazon benefit from a vast ad inventory on their own websites and operated services, which they can monetise to generate most of their ad revenues. They have extensive proprietary user data from their consumer facing services, which they can use to improve targeting but to which they restrict access. Platforms such as Google and Facebook can also benefit from network effects and economies of scale from their vertical integration in the ad supply chain. As argued by the paper, due to these advantages, platforms have the ability to engage in potentially anti-competitive practices such as self-preferencing, leveraging of their market power to other markets, and they can act as gatekeepers with the ability to charge higher fees and set their own terms for access to businesses.

    Secondly, the contractor analysed the transparency of the online advertising environment. The paper concluded that this environment is characterised by opacity, partly linked to the practices of a few platforms, and to the complexity of programmatic advertising. On the one hand within walled gardens, online platforms can use their economic power to impose their terms and limit the disclosure of information on the costs, profits and effectiveness of placement of ads. This undermines the decision making of advertisers and publishers regarding spending and their ability to refine targeting. Privacy legislation has been considered as an additional driver to reduce data disclosure to advertisers and publishers. The authors also argue that the removal of third-party cookies will also affect advertisers’ ability to do audience targeting and may incentivise them to shift more to walled gardens where first-party cookies are still available, further decreasing publishers’ revenues. On the other hand on the open web, the sharing of information depends on the positions and strategies of players along the supply chain, which results in fragmented information but also in user data leakage in RTB.

    In addition, there is a lack of transparency over the functioning and matching process of auctions, due to the use of algorithms and potential influence of vertically integrated platforms. Stakeholders also reported an opacity on the fees charged across the supply chain due to the number of intermediaries. The lack of transparency on money flows leads advertisers and publishers to question the efficiency of the online ad supply chain. The opacity of the ad tech value chain, including the reliance on algorithms and the vast array of service firms, also makes open programmatic advertising rife with fraud, at the expense of advertisers.

    Proposed solutions

    The contractor also suggests possible solutions to address these different issues at policy, industry and individual level. Potential regulatory responses to address transparency issues include focused monitoring and enforcement of existing legislation by specific regulatory units, international cooperation, the development of codes of conduct with the main online platforms and regulatory reform based on evidence-based recommendations from the different inquiries and market studies commissioned by regulatory authorities, that can include requirements for information disclosure and interoperability and structural remedies.

    In addition, several industry initiatives offer solutions for more trustworthy, transparent and verifiable ad trading. These include standards and practices for ad quality and measurement, charters or guides, innovative solutions to increase transparency on fees and bidding data, and programmes on user privacy and consent. They note though that effectiveness of self-regulatory initiatives depends on their adoption and implementation across the industry.

    Finally, the paper points to the academic literature that provides a range of methods and models to help advertisers and publishers mitigate programmatic advertising opacity by enabling them to take more informed decisions and optimise their strategy and revenue.

    Overall conclusion is that no single regulatory, industry or individual measure in isolation may sufficiently address the various issues identified but that better implementation of the existing initiatives and a combination of the proposed measures could be more effective in tackling these issues.

    1.7.Analytical paper #4: Online platforms with significant/strategic market status

    This analytical paper examined the evidence in relation to better understanding the various issues, and strengths and weaknesses of emerging approaches to identify online platforms with significant/strategic market status.

    The potential for these platforms to act as barriers to a competitive market, has resulted in an increasing need for new policy approaches to assess whether online platforms have significant or strategic market status. A key part of this discussion has focussed on whether traditional approaches, based around assessing market shares, are adequate. Increasingly, it has been thought that current policy approaches should be extended or adapted to consider the dynamic, varied, and constantly changing nature of the online platform economy ecosystem.

    The findings from the research run by the contractor suggests that emerging approaches to assessing online platforms with significant/strategic market status could be generally categorised as follows:

    ·Emerging approaches which draw on the traditional market share-based tests for application to online platforms; and

    ·Emerging approaches which appear to be devised specifically for online platforms.

    Emerging approaches based on the traditional market share-based tests include: revenue share; user share; barriers to entry; mark-up index; and network effects.

    Emerging approaches devised specifically for online platforms include: gatekeeper power; leveraging power; information/data exploitation power; prevalence of positive feedback loops; prevalence of indirect network effects; and the extent to which single- and multi-homing exists in the market.

    The paper argues that the key challenges to the use of emerging approaches based on the traditional market share-based tests include factors such as: the fact that user share can be identified in several ways; barriers to entry may be hard to measure; and a zero-price market poses challenges for assessing market power of online platforms.

    Amongst the emerging approaches devised specifically for online platforms, the contractor examined two approaches in further detail: gatekeeper power and leveraging power. Available evidence suggests that gatekeeper power - the level of power a platform can exert on its users through acting as a ‘gatekeeper’ – is a dynamic phenomenon. The main challenge with identifying gatekeeper power is likely to be in effectively establishing where the ‘gates’ are in relation to online platforms. Additionally, it may not be possible to assess gatekeeper power without considering it with other emerging approaches. The evidence also highlights leveraging power – the ability of platforms to establish an advantageous position in a separate or ancillary market – as potentially important. However, experts suggest that leveraging is a common business practice and as a result leveraging power may not be a decisive indicator of market power on its own.

    The main strengths of emerging approaches identified in the literature and suggested by the experts are that they seem to offer a more flexible instrument to market analysis and provide more dynamic indicators of market power suitable to the online platform ecosystem. The main challenges related to emerging approaches include a lack of reliable datasets to use some of the approaches and that due to their insufficient use in practice, the viability of these approaches is not yet clear. A comparison of traditional and emerging approaches suggests that traditional approaches appear to be more reliant on static indicators and stringent market definitions with a focus on single-sided market transactions. In contrast, emerging approaches may be more effective at recognising transactions on all sides of the market and thus better suited to the online platform ecosystem.

    At present, the emerging approaches appear to be focussed on economic, regulatory, and competition aspects of the online platform economy. Experts suggest that the emerging approaches also need to consider broader social and political impacts of the online platform ecosystem when identifying whether an online platform has strategic/significant market status. When the systemic interdependencies within the online platforms are considered, a single emerging approach is unlikely to be effective in practice. Using the emerging approaches in conjunction with each other is likely to be more effective due to the complex, multi-sided interactions of the online platforms.

    According to the paper, in order to identify whether an online platform has significant/strategic market status, policy makers would need to consider how the emerging approaches can be integrated into existing policy frameworks to adopt an open and flexible approach.

    1.8.Analytical paper #5: Business user and third-party access to online platform data

    This analytical paper investigated the state of the art of data sharing by digital platforms with third parties. The analysis covered three sectors of the platform economy: e-commerce, online tourism services and app stores. It was based on a detailed research of secondary sources, 61 interview and 15 platform-specific case studies that included Amazon, AliExpress, eBay, Google Play, Apple App Store, Booking.com and others. Specifically, the paper strived to answer the following questions:

    ·What data, collected and held by platforms, is important for their business users and other businesses active in their respective sectors?

    The analysis concludes that all data types collected by platforms are or could be important for business users for re-use. This includes data about transactions concerning own products and services, own clients/customers, and own business performance. Next, information concerning the broader market trends is also of key importance. It includes listings of other businesses, their customers, performance of different businesses in a specific market. Further, customer characteristics and customer profiles are of interest to all businesses, for example, behavioural data, such as browsing habits, search terms, purchasing decisions. The businesses using OTAs and e-commerce platforms underlined the importance of getting access to customer identification details e.g. for direct marketing. Finally, many companies, especially the smaller ones, expressed their preference for data analytics and insights as they do not have sufficient infrastructure and skills to take advantage of raw data.

    Some businesses also use platform data as an input to develop or improve data-based products or services (upstream process). In particular, the datasets of online platforms are of interest to two types of companies: app developers and data brokers or marketplace/app store optimisation companies. All types of data are pertinent to them, however they have a preference for granular and raw data that could be combined with other data sources and could be used to train algorithms, develop insights and provide value to their customers. More specifically, datasets and real-time data feeding into software and mobile applications can cover various areas and technologies, such as images for image recognition, audio files for speech recognition, weather or traffic data, health data, geolocation data and so on.

    ·What kinds of data do platforms provide and what data they refuse to share?

    Analysis carried out for this study shows that platforms provide data to their business users, which is sufficient to process transactions and manage their business. The businesses receive detailed data about their own listings, prices, sales, transactions and business performance. Platforms also provide some data about direct customers. Further, most major platforms share some data about the broader market, including overall market trends, best-selling products, customer profiles, although the type and granularity of such information differs from platform to platform. Overall, the major platforms compete for their business users and thus various metrics and dashboards are part of their value proposition. These metrics and dashboards are designed to help the business users to know their customers, monitor their own business performance, and understand the broader market trends.

    However, some data usually is not provided by the platforms, despite demand from their business users. Firstly, this concerns customer nominal data and contact details (especially pertinent in e-commerce and for OTAs). Secondly, the granularity of data concerning the customer profiles is also often considered insufficient by businesses.

    Businesses also demand more data about competing products and businesses on the platform. They also expressed a need for data about customer behaviour, such as search keywords, search volumes, buying patterns, responses to pricing signals. The platforms usually provide such data in a highly aggregated form and draw on it to develop analytics and insights that are offered or sold to business users. Nevertheless, many business users argue that such information is not sufficiently granular. Businesses that operate on the vertically integrated platforms (among online marketplaces, first and foremost, Amazon) also assume that the platform uses data from its marketplace to gain an unfair advantage over its own business users.

    The analysis also revealed power imbalances among platforms that are reflected in data sharing arrangements. Google and Facebook have the central position in online marketing and advertising, to the extent that they are unavoidable trading partners, including other platforms from the analysed sectors. This puts them in a position to determine the terms and conditions of data access and data reuse. Whereas Google and Facebook receive data from platforms concerning their listings, customers and business users, they do not share detailed data gained through the advertising activities. Further, some platforms also signalled that data sharing arrangements put them at risk of being pushed out of the market by Google and Facebook that are developing their own business verticals in travel and e-commerce.

    Finally, data brokers and online optimisation tool providers play an important role in data markets by offering data which is not accessible directly from the platforms. They usually pool platform data from multiple sources, including publicly available data, crowdsourced business user account data, data provided by platforms through APIs and data scraped from platform websites. The platforms that were analysed in this study argue that they do not have direct contractual relationship with the data brokers/online optimisation tool providers and thus are not responsible for quality or accuracy of the data. Nevertheless, the platforms see value in this market because it is useful for their business users; however, they may take action if, for example, they see that traffic from online optimisation tools providers start interfering with platforms’ services. Platform-specific case studies also revealed several examples when decisions by online platforms (e.g. changing APIs, development of their own analytical services) undermined the business model of specific data brokers/ online optimisation tools providers.

    Generally, all platforms claim that the only intended recipients for their data for re-use are their direct business users. Web-scraping is the main way to get access to platform data for all the other organisations interested in it. This is enabled by the fact that to generate transactions platforms must make a lot of information available for the customers on their websites.

    ·What are the incentives and constraints for platforms to share data?

    The analysis shows that when taking decisions to share or not to share data, online platforms must reconcile several competing and potentially conflicting imperatives. On the one hand, the success of the business users is important because it generates revenues for the platform. In this sense, online platforms have a strong incentive to provide access to data that could help businesses to understand their customers and to improve their product. On the other hand, online platforms must maintain trust of their clients (business users and customers of the business users), which means that they should avoid sharing data that these clients are unwilling to share, for example, personal information, sensitive business information.

    Online platforms have also designed their terms and conditions to comply with the applicable regulatory frameworks, including P2B regulation, personal data protection, competition law, regulation forbidding trade in illegal and counterfeit products, and others. Generally, interviews with platforms revealed that they feel that they operate in an environment of legal uncertainty, which makes them reluctant to open more data. For example, they face different data protection regimes globally, as well as diverging interpretations of GDPR in EU member states. Further, whereas data sharing is usually considered as a measure to ameliorate power imbalances in the online platform economy, sharing seller-specific revenue information among sellers can be interpreted as providing a competitive advice under the national anti-trust law.

    Several groups of players operate within the data ecosystem surrounding each online platform. These include other platforms, large and small businesses, customers of the business users, data brokers or companies providing online optimisation tools, regulatory and other public authorities. Sometimes these groups have diverging interests and competing demands concerning data access. As mentioned earlier, the platforms see personal data protection as part of their value proposition, however this claim is not always accepted by some businesses who argue that platforms use data protection as an excuse for not sharing important data. If platforms decide to open more raw data to business users, this could benefit large businesses at the expense of the smaller ones, because the big companies have the necessary infrastructure and know-how to take advantage of such information.

    If a specific dataset is at the core of a platform’s business model, it is unlikely to be shared. Due to this reason platforms will be reluctant to share datasets that could be used to undermine their role as leading intermediaries in two-sided markets. Vertically integrated platforms are not likely to share detailed market-level data, which could help the emergence of new competitors in their market. Yet these platforms also make internal decisions on what information from their marketplace/app store can or cannot be shared with the retail/app development division. Such decisions are of crucial importance to many businesses that compete with goods and services sold by the platform itself. Next, when taking decisions on data sharing, platforms consider the global competition. For example, several platform interviewees pointed out that they detect abusive bots originating from China, crawling their pages or trying to use their APIs. Platforms see Chinese marketplaces as serious competitors that are not competing on a level playing field as they are in the position to disregard many regulations that European companies must comply with.

    Finally, the lack of technical interoperability between different platforms is also a constraint impeding data sharing and data portability. Introducing interoperability is costly, because it requires the development of common standards and revision of back-end code. From the perspective of platforms, investing into interoperability does not necessarily provide a clear commercial gain. Interoperability also has its downsides because it may make the system slower and limit the development of new or innovative products.

    ·What are the possible solutions to address platform refusals to share data important to other users?

    The paper concludes that there is a clear public interest to encourage more data sharing, to the extent it could promote competition, offer more choices to businesses and their customers, foster innovation and help alleviate the market power of big online platforms. At the same time, the principles of personal data protection, business secrets’ and intellectual property protection should also be taken into consideration.

    Various solutions have been put forward by various stakeholders that could potentially facilitate data sharing. They include both public-sector led initiatives, as well as market-based ones, focusing specifically on the incentives and constraints for data sharing stemming from the analysis. Public sector led solutions include mandated access; mandated interoperability and data portability; prohibition of certain business practices (for example, mandatory ‘walls’ prohibiting vertically integrated platforms from sharing data between their marketplaces and product development / retail departments); and reversal of the burden of proof (i.e. platforms may be required to demonstrate that their data practices are beneficial for their users). Market-based or self-regulatory solutions considered include offering access to data based on FRAND (Fair, Reasonable, And Non-Discriminatory terms) principles; data pools or data trusts; as well as company-led incentives for interoperability and data portability.

    Annex 5.3: International consensus on the need to act

    1.Summary

    At the international level, a number of countries have already started to discuss how to best address certain harmful behavior by gatekeepers. The problems they point to and conclusions they draw are to a big extent similar to the ones that are to be addressed in this initiative.

    In the UK, the Furman Report reflects on the need to regulate platform companies “in position to exercise market power or a gateway or bottleneck in the digital market, where they control others’ market access” (defined as companies with ‘strategic market status’). 110 It points to persistent dominance of these platforms, exerting significant market power over their users and not being required to deliver the same level of positive outcomes as they would if facing normal competitive market conditions. In terms of solutions it suggests the use of ex-ante tools that should help to prevent negative outcomes before they occur. They should be based on three key pro-competition functions that can deliver benefits beyond core competition: (i) binding Digital Platforms Code of Conduct promoting fair, pro-competitive conduct by platform companies with strategic market status; (ii) personal data mobility and (iii) data openness. The monitoring and enforcement of the rules would be assigned to the new regulator - pro-competition digital markets unit. Its new powers should allow it to impose remedies and to monitor, investigate and penalise non-compliance. This call is further reinforced in the Competition and Markets Authority report 111 calling on the UK Government to establish a new pro-competition regulatory regime with strong and clear ex ante rules for those firms deemed to have ‘Strategic Market Status’ (SMS), overseen by a Digital Markets Unit.

    In a similar vain, in the US, Stigler Centre Report points to insufficient entry (and therefore insufficient competition) in digital platforms caused by companies with ‘bottleneck power’ - meaning companies that have incentive and ability to develop and preserve a single-homing environment. It suggests setting up a new digital regulator -that the Digital Authority that would have the sole authority to define bottleneck power and update the definition regularly. The Digital Authority would enforce two sets of rules: (i) broadly applicable to all platforms, such as data portability, open standards to promote competition, interoperability and (ii) rules applicable only to companies with bottleneck power, such as non-discrimination and foreclosure or bundling. 

    In Australia, its competition authority (the ACCC) in its Digital Platforms Inquiry Report 112 set out its views on the market power of the two leading digital platforms –Google and Facebook- considering that both platforms have substantial market power thanks to their advertising businesses, that are extended well beyond their core owned and operated platforms. In terms of solutions, the ACCC considers that opening up the data, or the routes to data, held by the major digital platforms may reduce the barriers to competition in existing markets and assist competitive innovation in future markets. Increasing portability of data held by digital platforms may deliver significant benefits to current and potential future markets, including through innovation and the development of new service. The ACCC recommends to put in place frameworks that enable adverse consequences to be addressed and that reduce the likelihood of new issues arising. The report also proposes the creation of a branch within the ACCC to focus on digital platforms.

    In China, its market regulator, published on November 2020 draft rules aimed at preventing monopolistic behavior by internet platforms, so as to increase scrutiny on the country's e-commerce marketplaces and payment services. 113 The draft rules would look to prevent e-commerce practices such as ‘choose one between two’, under which a marketplace restricts brands from selling on multiple platforms. The draft rules would also cover differentiate treatment based on big data, payment ability, consumption preferences, and usage habits.

    As demonstrated above, a number of non-EU countries point to the same problems taking place in the digital markets and come up with similar solutions as the ones advocated by this initiative. The ex-ante rules targeting platforms with market /bottleneck power and ensuring fair and contestable digital markets, are perceived as the way to address these problems. Most of them also envisage setting up a specialised regulator responsible for monitoring and enforcement of the new rules in the digital markets. However, even if this initiatives in third countries will be further pursued and will lead to some form of (national) regulation of gatekeeper platforms in these countries, these regulations will most likely be tailored to the most salient needs and problems in the respective jurisdictions passing the regulation, and can therefore not be expected to effectively address the gatekeeper related problems as they manifest themselves in the EEA.

    2.The Furman Report (UK)

    The Furman report 114  reflects on the need to regulate platform companies with ‘strategic market status’, defined as those “in position to exercise market power or a gateway or bottleneck in the digital market, where they control others’ market access”.

    Problems:

    The report points out to the following problems that should be addressed by new ex ante rules:

    -A handful of powerful platform companies dominate a number of digital markets and this dominance is persistent. The position of the largest firms is getting stronger, and this strength and their positions are not imminently under threat. This means that they can exert significant market power over their users and are not required to deliver the same level of positive outcomes as they would if facing normal competitive market conditions.

    -Lack of contestability: Due to the barriers to entry that exist in established digital platform markets they cannot generally be considered freely contestable. The significant amounts of data held by incumbent firms considered the single biggest barrier to entry in the digital economy.

    -Gatekeeper position fostering dependency: The result is that one, or in some cases two firms in certain digital markets have a high degree of control and influence over the relationship between buyers and sellers, or over access by advertisers to potential buyers. As these markets are frequently important routes to market, or gateways for other firms, such platforms are then able to act as a gatekeeper between businesses and their prospective customers.

    Impact on consumers

    According to the report, in terms of impact on consumers, these market dynamics will lead to business users of platforms accepting worse terms than they would face if multiple platforms were competing with one another in each market. The consequences of these terms will ultimately feed through to consumers in the prices they pay, the quality they receive, and the range of innovative new products and services they are able to choose from.

    Impact on innovation

    The Report pointed to the stifling effect of the above practices on invitation. In particular it noted that to killer acquisitions by big platform companies “at best, absorb innovation to protect themselves from potential competition and, at worst, use acquisitions to kill off or distort innovation, creating a ‘killzone’ around their positions.”

    Who should be in scope

    Platform companies with ‘strategic market status’, defined as those in position to exercise market power or a gateway or bottleneck in the digital market, where they control others’ market access.

    Designation of platform companies with ‘strategic market status

    According to the report, it would be up to the regulator (the Digital Markets Unit) to determine which markets have companies able to hold a strategic market status, where a high and enduring market share or other factors lead to market power. To do so the regulator needs to develop a clear test for the characteristics of a company’s market position above which regulatory powers are appropriate.

    Every 3 to 5 years the regulator would conduct a statutory review of both markets and the companies with strategic market status.

    Aspects of market power particularly relevant to platforms and their potential to act as a bottleneck should also be considered for incorporation: economic dependence, relative market power and access to markets.

     Solutions/Remedies

    The report argues that the use of ex-ante monitoring and enforcement of a detailed set of pro-competition rules should help to prevent negative outcomes before they occur. Pro-competition policy tools will tackle the factors that lead to winner-takes-most outcomes and to that position becoming entrenched. Pro-competitive rules and frameworks should be based on three key pro-competition functions that can deliver benefits beyond core competition:

    1.a binding pro-competitive code of conduct promoting fair, pro-competitive conduct by platform companies with strategic market status

    Digital Platform Code of Conduct should be based around a set of core principles that would be required for of digital platforms deemed to have strategic market status. For the business side of platforms with a strategic market status, the principles should ensure that business users are:

    • provided with access to designated platforms on a fair, consistent and transparent basis

    • provided with prominence, rankings and reviews on designated platforms on a fair, consistent, and transparent basis

    • not unfairly restricted from, or penalised for, utilising alternative platforms or routes to market

    2.personal data mobility (giving consumers greater control of their personal data, e.g. their profile, purchase history or content) and systems with open standards and

    3.data openness

    These pro-competition tools will be implemented by a digital markets unit, with powers to regulate and enforce these functions.

    Implementation and Enforcement

    The pro-competition digital markets unit is to be responsible for monitoring and enforcing of the pro-competitive rules and frameworks. Its new powers should allow it to impose remedies and to monitor, investigate and penalise non-compliance.

    To avoid burdens on smaller companies, its enforcement powers should be focused on companies with ‘strategic market status’.

    The unit’s approach should combine participation and consultation (with a wide range of stakeholders) with the scope for regulatory enforcement, necessary to overcome incentives against compliance and make its solutions operate effectively and quickly. It should only intervene where doing so is effective and proportionate to achieve competitive aims. 

    The Code should be set up to achieve fast resolutions (in multiples of weeks or months). This approach would be supported by strong powers to formally request information from designated platforms within tight deadlines set by law when it suspects a breach of codes. It would also need power to enforce legally binding decisions and penalties for contraventions of the code where a participative approach is not effective.

    The Digital Markets Unit should also have the powers to implement (ii) personal data mobility and systems with open standards as well as pursue data openness as a tool to increase competition.

    3.CMA study (UK)

    The Competition and Markets Authority (CMA) Report 115 on online platforms and digital advertising focused in particular on whether rival providers of search and social media services can no longer compete effectively with Google and Facebook because of their size, and a range of concerns in the digital advertising market, including in particular a lack of transparency and conflict of interest (self-preferencing).

    Problems:

    The report pointed to the following problems:

    -Conflict of interest

    The report points out that the extent of vertical integration by Google and Facebook that has taken place in the open display market raises numerous concerns as it may give result in the conflicts of interest and allow companies with market power at one stage of the value chain to use it to undermine competition at other stages. There are concerns whether Google can use its market power in inventory and data to advantage its DSP services and use its market power as an ad server to favour its SSP.

    The extensive amount of data available to Google and Facebook provide these platforms with a competitive advantage and assist with entry into related markets. After entering the market, the role of Google or Facebook as a host or gateway then enables these platforms to advantage their own related businesses. Google and Facebook have the ability and incentive to favour a business with which they have an existing relationship (and through which additional revenue may be generated), such as websites that are members of their display or audience network or use their ad tech services. For example, when operating on behalf of the publisher, Google may have an incentive to favour bids coming through its own advertiser-side intermediaries, rather than those that are best for the publisher. When operating on the buy-side, it might have an incentive to channel advertiser’s spend to its publisher clients, rather than to the publishers that are best for the advertiser. Given the substantial market power of each of Google and Facebook, their presence in a significant number of related markets and the opacity of their key algorithms, there is significant potential for self-preferencing by Google and Facebook to substantially lessen competition.

    -Lack of transparency and asymmetric information

    The findings of the report identify a series of issues relating to lack of transparency and the data advantages of the large platforms which could limit competition in digital advertising:

    üthe large platforms’ processes for auctioning inventory are not transparent and there is limited ability to independently verify the effectiveness of advertising because of lack of access to data; and

    üthe data advantages of the large platforms in targeting advertising mean they can monetise their content much more effectively than other platforms/publishers, increasing their market power.

    The lack of transparency exists mainly in the open display market where publishers and advertisers rely on intermediaries to manage the process of real-time bidding and ad serving. The CMA report points out that they cannot observe the actions of the intermediaries directly and do not see how the fees are charged along the supply chain. Hence, it undermines their ability to make optimal choices concerning buying and selling their inventory.

    CMA believes that extensive data that is collected in the sector could address some of these concerns, but this data is held by a few parties, which leads to concerns on the asymmetric information. The report recalls the views of advertisers and publishers that Google and Facebook enjoy significant competitive advantages in both measuring effectiveness and targeting because of their extensive access to user data. Google offers in-depth targeting options, driven by its unique and vast sources of data while Facebook has the advantage of providing the ability to target specific audiences based on demographic characteristics, interests and location. However, the two platforms do not allow independent verification of their inventory.

    Given the lack of transparency over fees and bids through the intermediation chain, there might be a legitimate concerns about any operator having positions on both the buy and sell side of the market, whether or not that operator is in fact acting in its clients’ best interests.

    Solutions:

    In terms of potential interventions it supports ex- ante regulatory regime to regulate the activities of online platforms funded by digital advertising and recommends a number of solutions. It also reflects on the need to launch market investigation on the open display advertising market, with focus on the conflict of interest Google faces at several parts of its vertically integrated chain of intermediaries.

    The final report recommends that the UK Government establishes a new pro-competition regulatory regime with strong and clear ex ante rules for those firms deemed to have ‘Strategic Market Status’ (SMS), overseen by a Digital Markets Unit. CMA is now leading a Digital Markets Taskforce to consider the design and implementation of the procompetitive framework for digital markets.

    The CMA’s Digital Markets Taskforce is currently considering the test which might be used to identify which firms may have SMS and therefore would be subject to additional rules. A variety of factors could indicate that a firm has a strategic position including:

    - evidence of the ability of the firm to leverage one market position into a variety of other markets

    - the firm’s size and scale; or

    - its position as an access point to customers for businesses across a diverse range of markets.

    It is when a firm has obtained such a position that the effects of its market power are likely to be particularly significant and existing tools are unlikely to be adequate in addressing this market power.

    The new regime proposed in the market study would be comprised of two sets of tools:

    ·The first, an enforceable code of conduct to mitigate the effects of the market power of SMS firms by governing their behaviour.

    ·The second, a range of ‘pro-competitive interventions’ to tackle the sources of market power and promote competition.

    The types of remedies that the market study outlines include data-related remedies, consumer choice and default remedies, and separation remedies.

    4.The Stigler Center Report (US)

    I.Problem definition

    According to the report the general harm identified is insufficient entry (and therefore insufficient competition) in digital platforms.

    Increased concentration levels, market power, network effects, and control over data and analytics have in many digital markets tipped the market in favour of the incumbents. Many digital markets feature large barriers to entry. Once the incumbent is established, entry into digital platform businesses is very difficult. The winner often has a large cost advantage from its scale of operations and a large benefit advantage from the scale of its data.

    The role of data in digital sectors is particularly critical. The new entrant starved of data relative to a tech giant, is at a significant competitive disadvantage.

    Problems arising in the digital markets:

    ·Harms to investment and innovation

    By excluding competitors, dominant firms do not need to innovate as hard as they otherwise would be required to keep their customers. Likewise, when platforms do not face competition, they will be able to reduce quality, for example, by decreasing privacy protections, without losing customers or revenue.

    ·Harms to entry, including disintermediation

    There is growing evidence that conglomerate digital platforms are in an advantaged position to stop or block entry by more focused rivals when compared to traditional businesses. A platform that has total control of demand can steer customers to content and complements it owns rather than to those provided by independent firms that might challenge its market power.

    Platforms have bluntly moved to prevent disintermediation and have engaged in foreclosure to block potential rivals. For example, Facebook acted to suppress the growth video-capture-and-sharing app Vine when Vine attempted to link its users to their Facebook friends.

    II.Who should be in scope

    Companies with ‘bottleneck power’ - meaning companies that have incentive and ability to develop and preserve a single-homing environment.

    The Digital Authority should have the sole authority to define bottleneck power and should update the definition regularly or on an ‘as needed’ basis.

    Stigler report refers here to Furman report to explain the meaning of bottleneck power:

    [O]ne, or in some cases two firms in certain digital markets have a high degree of control and influence over the relationship between buyers and sellers, or over access by advertisers to potential buyers. As these markets are frequently important routes to market, or gateways for other firms, such bottlenecks are then able to act as a gatekeeper between businesses and their prospective customers.

    The finding of bottleneck power will employ consideration of the forces that tend to impede entry and lead to foreclosure. The Furman Report similarly explains that this single-homing foreclosure tends to happen when users experience high switching costs, such as loss of valued personal data or reputational indicators at the point of switching; contract terms that deter switching; technical barriers to switching, such as complex switching processes or a lack of interoperability between the old service and the new or second service; tying services, which can be by contract or technical; and the inertia of defaults.

    III.Solutions/remedies

    The reports proposes the following solutions:

    ·Improved antitrust enforcement:

    1)Reform of antitrust law to adequately deliver competition to consumers

    2)The establishment of a specialist competition court to hear all private and public antitrust cases

    ·Regulatory measures:

    3)A specialist regulator – the Digital Authority and

    4)new broadly applicable rules such as: 

    a.data portability

    b.open standards to promote competition (in particular in micro-payments and digital identities)

    c.interoperability

    5)new rules applicable to companies with bottleneck power:

    a.mergers - DA could be given merger review authority over all transactions involving companies with bottleneck power

    b.non-discrimination and foreclosure

    as discrimination is an important tool in a foreclosure strategy by a digital bottleneck market power

    Platform strategies to prevent multi-homing are an important category for DA to include in its analysis of foreclosure. The DA could promulgate regulations prohibiting the foreclosure of a competing content provider on a platform that is vertically integrated.

    c.bundling

    A digital platform with bottleneck power may have a contract with complementors (e.g., retailers on an ecommerce platform) that bundles together access to their transaction data along with logistics services. This could have harmful anticompetitive effects. The business may also compete against those sellers on its e-commerce site, using the retailer’s data to learn about which products are selling well and expropriate the ideas and strategies of the seller.

    The DA could establish regulations that prohibit anticompetitive bundling by firms with bottleneck power. Such a firm would be required to demonstrate that its bundle was on balance procompetitive if foreclosure was alleged. The DA could require unbundling and an offer to business customers of a choice of contracts in the case of anticompetitive bundling. The DA would need to enforce such contracts.

    DA- Enforced Remedies for Antitrust violations:

    When a company has been found liable for violating the antitrust laws, the regulator, in conjunction with the antitrust authority, could apply the following remedies in order to restore competition:

    -data sharing,

    -full protocol interoperability,

    -non-discrimination requirements, and

    -the unbundling of content from a platform.

    5.The Australian Competition and Consumer Commission (ACCC): Digital Platforms Inquiry

    The ACCC’s Inquiry focussed on the three categories of digital platforms: online search engines, social media platforms and other digital content aggregation platforms. A large part of the Report focuses on Google and Facebook, reflecting their influence, size and significance as well as the fact that Google and Facebook are the two largest digital platforms in Australia. The Report focuses on the impact of the digital platforms on competition in the advertising and media markets and on advertisers, media content creators and consumers.

    Chapter 2 of the Report sets out the ACCC’s views on the market power of the two leading digital platforms, Google and Facebook, with a focus on the markets most relevant to the Inquiry.

    Problem:

    The report considers that both Google and Facebook have substantial market power thanks to their advertising businesses, that are extended well beyond their core owned and operated platforms.

    Google 116 has substantial market power in the supply of general search services in Australia (95% of market) and in performing search advertising revenues in Australia (96%). It enjoys advantages of scope in accumulating data from consumers using its wide range of services (Google Search, Google Maps, YouTube, Gmail) and the Android OS, so it is able to track consumers on the more than two million websites that use Google advertising services. According to the report, Google also benefits from its position as the default search engine on both the Chrome browser (owned by Google), and the Safari browser (owned by Apple), which together account for more than 80 per cent of the Australian market for browsers. The substantial amount paid by Google to Apple for default status on Safari (estimated at approximately US$12 billion in 2019) reflects the value of this default status. Google Chrome is pre-installed on nearly all Android devices.

    Due to the market dynamic – strategic acquisitions – Google has obtained further advantages of scope and reduced potential competition and his position on Australian market is very unlikely to change in the middle time. Report also recognises Google’s importance to news media businesses, which is an unavoidable trading partner, and presumes significant loss of revenue if Google users could no longer click on links to their website in search result. The ACCC therefore considers that Google has also significant bargaining power in its dealings with these media businesses.

    The report also concludes that Facebook has substantial market power in the supply of social media services and display advertising services. This is caused by a fact that Facebook has three time larger audience that Snapchat has (the closest competitor to Facebook) and similarly as in Google case creates a significant barrier to entry and expansion of its (possible) competitors. It benefits from the fact that its consumers are using another platforms owned by Facebook, mostly Instagram, Messenger and WhatsApp; other numerous strategic acquisitions are likely to even increase Facebook’s advantage of scope and market power. Regarding display advertising market, Facebook and Instagram’s combined share of the market is estimated to be 51% while the other suppliers don’t hold more than 5%. Also similarly to Google, ACCC considers Facebook to has substantial bargaining power over news media businesses; Facebook’s strength is in being a vital distribution channel for a number of media businesses targeting particular demographic groups.

    Implications of substantial market power:

    The Report concludes that a firm with substantial market power could damage the competitive process by preventing or deterring rivals, including potential rivals, from competing on their merits. That is, a firm with substantial market power could maintain or advance its position by restricting or undermining its rivals’ ability to compete, rather than by offering a more attractive product.

    ACCC also recognises that there is a lack of transparency in the online advertising markets. In particular, it is unclear how Google and Facebook rank and display advertisements and the extent to which each platform self preferences their own platforms or businesses in which they have interests.

    Who should be in scope of the rules:

    ·online platform with substantial market power in the [observed] market;

    ·market dynamic lowered by acquisitions of potential competitors due to which

    ·potential of new entry in the market is low.

    Rules/procedures to be applied:

    ·ACCC recommends the merger framework in Australia to be updated to make it clearer so that acquisition of potential competitors and economies of scope created via control of data sets are taken into consideration in assessing whether an acquisition has the effect or likely the effect of substantially lessening competition.

    ·Currently the notification of M&A to the ACCC is voluntary in Australia, but ACCC considers it appropriate that the large digital companies would each agree to a protocol to notify the ACCC of proposed acquisitions that may impact competition in Australia.

    ·As regards addressing default bias, ACCC considers that offering Australian consumers the choice that Google is forced to implement in Europe after the EC decision 117 would have the effect of improving competition in the search services market and recommends that Google also implement these changes also in Australia.

    ·As regards the role of data in market power, the ACCC considers that opening up the data, or the routes to data, held by the major digital platforms may reduce the barriers to competition in existing markets and assist competitive innovation in future markets. This could be achieved by requiring leading digital platforms to share the data with potential rivals.

    ·One potential mechanism is the application of the Consumer Data Right, another is to require the platforms to provide interoperability with other services.

    ·Incentives for portability, privacy concerns and identification of the extent of data to be shared have to carefully considered.

    ·Particularly increasing portability of data held by digital platforms may deliver significant benefits to current and potential future markets, including through innovation and the development of new service. If data portability or interoperability would be identified to be beneficial in addressing the issues of market power and competitive entry or switching, the ACCC could recommend this to the Government.

    ·The creation of a branch within the ACCC to focus on digital platforms

    ·Proactive investigation, monitoring and enforcement of issues in markets in which digital platforms operate

    ·Inquiry into the supply of ad tech services and advertising agencies

    6.Chinese draft competition rules for online platforms

    China's market regulator on published on 10 November 20220 draft rules aimed at preventing monopolistic behavior by internet platforms, so as to increase scrutiny on the country's e-commerce marketplaces and payment services. The draft rules would look to prevent e-commerce practices such as ‘choose one between two’, under which a marketplace restricts brands from selling on multiple platforms. The draft rules would also consider whether a transaction treats different customers in different ways based on big data, payment ability, consumption preferences, and usage habits.

    China's State Administration for Market Regulation (SAMR), which issued the draft, said 118 it wanted to prevent platforms from dominating the market or from adopting methods aimed at blocking fair competition. The definitions it provided for internet platforms mean the new rules could apply to e-commerce sites, such as Alibaba Group's Taobao and Tmall marketplaces or JD.com, as well as payment services like Ant Group's Alipay or Tencent Holding's WeChat Pay.

    The draft comes after China's Financial Stability and Development Committee, a cabinet-level body headed by Vice Premier Liu He, flagged in October 2020 the need to improve mechanisms to ensure fair competition and called for the strengthening of anti-monopoly law enforcement.



    Annex 5.4: Overview of laws and proposed legislation in Member States related to the initiative

    This annex summarises existing and forthcoming regulation by the Member States addressing economic power of digital platforms. It then compares those frameworks with the aim to evidence the already existing and the forthcoming fragmentation as specified under Article 114 TFEU.

    1.Notion of fragmentation

    Article 114 (1) TFEU forms the basis to act at EU level where the approximation of provisions in Member States have as their object the establishment and the functioning of the internal market. The internal market objective is met where the EU act aims at abolishing obstacles to the freedoms of the treaty and/or to remedy the disadvantages resulting from disparities and different conditions of competition. 119 This also covers the prevention of expected obstacles/prevent distortions to competition that may arise from expected action at MS level. Where reliance on Article 114 (1) TFEU is based on preventing forthcoming fragmentation it must be demonstrated that it is likely that the measures proposed at the level of MS will materialise. 120 The threshold for fragmentation to be relevant under Article 114 (2) TFEU regarding the first alternative under Article 114 (1) - i.e., on obstacles to freedoms of the treaty - is met by the sole fact that there are diverging rues in place or likely to be put in place. There is no minimum quantitative level to be demonstrated as to the importance of those differences. This is because the differences in law are indicative for demonstrating obstacles to the freedoms. 121 Regarding the second alternative under Article 114 (1) TFEU - i.e. on distortion of competition - the threshold to be met in order to justify intervention is that the distortion must be appreciable. The distortion is appreciable where the different national rules lead to different production costs or, where they affect the freedom of the treaties or systemic competition. 122 Where the conditions of Art. 114 are fulfilled and where other provision of the TFEU could also possibly cover the objectives of harmonisation, there is no need to take those other legal basis into consideration. 123  

    2.Existing fragmentation resulting from divergences in the laws of Member States addressing economic power of digital platforms

    Currently, MS already apply divergent frameworks to address the problems arising from the dependency of businesses on enterprises with relative market power and the resulting cases of unfairness. Those rules in most Member Stated of a horizontal nature, i.e., applicable also outside of digital platforms. For instance, in Belgium, the prohibition of abuse in dependency relationships was introduced by law of 4 April 2019 defining dependency by reference to absence of alternatives for the business and the possibility to impose conditions which could not be obtained under market conditions. 124  Bulgaria introduced regulation against abuse of economic dependence providing that undertakings with ‘superior bargaining position’ (‘SBP’), are prohibited to act in a way which contradicts good faith business practices and harms or threatens the legitimate interests of the weaker contractual party and the consumers. 125  In Cyprus, the Competition Act addresses relationships of economic dependency by qualifying the imposition of unfair trading conditions, the application of discretionary treatment, or of sudden and inexcusable interruption of long-term trade relationships as unfair. 126  In France, currently the Commercial Code addresses unfairness in imbalanced B2B relationships. 127  

    In Germany, currently the Competition Act rules out certain abuses of relative market power to impede smaller competitors in an unfair manner. 128 For instance, for enterprises with superior bargaining power in relation to an SME is prohibited to price below costs and placed under an internal non-discrimination obligation, i.e., it cannot offer services to itself at better conditions than to the SMEs, for example delivery. 129 Furthermore, Section 58a of the German Payment Services Supervisory Act 130 (introduces a right for payment service providers and e-money issuers to directly access technical infrastructure providing payment services, such as the near-field communication (NFC) interfaces. This provision has been described in doctrine as the ‘Lex Apple Pay’ and aims at regulating a gatekeeper position in the field of payment services. 131 In Hungary the Competition Act prohibits abuse of superior bargaining position, the abuse consisting in fixing purchase or sales prices unfairly in business relations, including where general contract terms and conditions are applied; stipulating unjustified advantages by any other means; or forcing the acceptance of detrimental terms and conditions on the other party. In addition, the rules prevent undertaking with superior bargaining position from influencing the other party's business decisions for the purpose of gaining unjustified advantages; creating a market environment that is unreasonably disadvantageous for the competitors; or influencing their business decisions for the purpose of gaining unjustified benefits. 132  In Italy, an asymmetric B2B law results from the extension of the protection under the unfair commercial practices law to cover also the protection of micro enterprises. 133  

    Those dependency and relative market power rules are divergent as to the threshold for intervention. For instance, the superior market power is often defined by reference to superior bargaining power, but not in all cases. Furthermore, the dependency rules also diverge as to the protected enterprises; those are not in all cases SMEs but also in some cases microenterprises (Italy). Finally, those rules also differ as to the specific prohibited abuses.

    As has already been set out in the Impact Assessment to the proposal for the P2B Regulation, 134 EU Competition law grants to a certain degree a basis for overcoming fragmentation, at least regarding the so called wide MFN clauses – i.e., the prohibition to sell at lower prices on other distribution channels. Those wide parity clauses were removed in a large part of Member States. However, the usage of so called narrow MFN clauses - i.e. the offering of better conditions via the sales channel of the hotel if compared to the conditions it offers on the platform - is still common and currently subject to monitoring by the competition authorities. However, in certain Member States also narrow MFN clauses were banned via legislative action. This is the case in France 135 , Austria 136 , Italy 137  and Belgium 138 . The laws of those MS prohibit all most favoured nation (MFN) clauses thereby allowing hotels to grant any discount or pricing advantages to their customers via other sales channels and via their own channel. In addition, France also prescribes that the room prices shall be specified in a ‘mandate contract’. The national legislations in place bans narrow parity clauses beyond the level of harmonisation achieved under competition law. Consequently, costs of providing the service differ among Member States, either on the side of the platform or the side of the hotels.

    In conclusion, the current rules in place already create a certain degree of distortion of competition between Member States insofar as the rules on tackling unfairness in dependency relationships diverge as to the preconditions to intervene and as to the depth of intervention. Regarding MFNs a dual type of fragmentation exists: on the one hand an obvious ‘first level fragmentation’ results from the fact that some MS have legislative bans in place and some MS do not. On the other hand there is also fragmentation observable due to differences in the MFN-legislations in place.

    3.Forthcoming fragmentation likely to emerge due to initiatives at MS’s level aiming at redressing unbalanced situations of businesses in relation to digital gatekeeper

    The divergences in regulation of economic power are likely to deepen due to the current initiatives at MS level to address specifically imbalanced relationships between digital platform and their business users. In a series of Member States legislative projects are under discussion and/or have been proposed within the legislative process.

    In Germany, new rules, are likely to be imposed on undertakings with paramount significance for competition across markets. The proposed rules cover prohibitions/obligations in relation to discrimination, leverage, usage of data, portability, interoperability and information on quality and performance. The governmental draft bill for the 10th amendment to the Competition Act (GWB-Digitalisierungsgesetz) of 9 September 2020 139 contains profound changes to the Competition Act and introduce a set of rules specifically applicable to undertakings active to a significant extent on multi-sided markets or with networks. In order to extend the existing notion of abuse and to partly prevent competition problems on digital markets the ministerial draft bill contains two proposals: Firstly the proposed § 19a GWB introduces new forms of abuses for undertakings with paramount significance; and secondly the proposed § 20(3a) GWB makes it abusive for a company with superior market power to prevent the creation of network effects to the benefits of competitor.

    Pursuant to §19a of the draft bill, the Competition Authority (Bundeskartellamt, BKartA, herein after: “NCA”) would acquire the powers to issue a decision stating such status of a company. The criteria for paramount significance across markets are dominance in one or several markets, financial strength and access to other resources, vertical integration and activities in related markets, access to data relevant for competition, its importance for other companies in order to access sales and supply markets and its impact on their business activity. In the explanatory part to the governmental bill it is indicated “that the determination of a paramount significance for competition across markets can only be made for a few companies and the rule will therefore have a narrowly limited circle of addressees.” 140  

    As to the imposition of obligations, the draft bill §19a (2) GWB provides that the NCA can impose specific prohibitions on digital platforms found to have paramount significance unless the behaviour is shown to be objectively justified, while the burden of proof relies with the platform. The NCA may, for instance, impose (1) a non-discrimination obligation, (2) a prohibition of exclusionary conduct in adjacent competitive markets, (3) a prohibition to use data collected in the dominated or other markets for the purpose of creation of market entry barriers or other exclusionary conduct and the imposition of conditions allowing for such a use, (4) a prohibition to impede interoperability of portability, (5) a prohibition to insufficiently inform users about quality and success of their services or obstruct their the possibilities of assessment of their performance by other means. Generally speaking, the new § 19a aims at preventing digital platforms to use their market position and the economic power in certain markets strategically to restrict competition in other markets. This is intended to address problems that may arise when certain companies establish anti-competitive structures, for example in new markets, without these companies necessarily being already dominant in all these markets. 141  

    The second set of key provisions, § 20 (1) (2) (3a) GWB, are part of the framework abuses of relative market power to impede smaller competitors in an unfair manner under § 20 GWB. While currently only small and medium sized companies may benefit from the prohibitions in § 20, the draft proposes to remove the SME-condition, thereby extending the protection to all companies independent of their size. This is based on the findings that also large companies may now encounter situations of imbalanced bargaining power vis-à-vis gatekeeper platforms. 142 A further novelty would be that the source of superior relative market power could also result from intermediation power. Finally, where competitors of companies with relative or superior market power are prevented by the gatekeeper from achieving economies of scale themselves, these practices are to be pursued as unfair impediments. This new prohibition under § 20 (3a) aims to prevent tipping. 143  

    In Germany, those proposed amendments of the Competition Act are likely to be adopted. It is currently discussed in Parliament and the Federal Council (Bundesrat). 144 .

    In France, a report has been submitted to the Parliament proposing to set the criteria to define platforms with structuring power (‘platformes structurantes’), with a view of the establishment of a list covering those platforms and imposing on those platforms ex ante rules on transparency on algorithm for the purpose of audit, interoperability and portability, access to data with an essential facility feature, device neutrality for access to apps and a prohibition of self-preferencing.13 

    Although in France, a legislative proposal has not yet been tabled it seems likely that this will be the case in the near future. The political will to proceed in this direction is evidenced by the facts that

    -the French Ministry for the Economy and Finance has been calling 145  for asymmetric regulation at EU level allowing for targeted and proportionate rules and obligations to complement competition law.

    oAccording to the Ministry asymmetric regulation of structuring platforms should be enforced on a case-by-case basis, when competitive problems related to a platform appear to be structural and lasting, therefore requiring continuous intervention. Possible remedies could include obligations on data mobility and data portability to help reducing switching costs from one platform to another. The ultimate goal should be access to data potentially constituting barriers to entry (example: obligation to develop technical standards that facilitate interoperability of services and migration options for users).

    oDesignation of the most structuring platforms, to whom the new regulatory framework should apply, should be based on a set of economic characteristics and conditions that justify regulation. Mechanism to identify companies and define obligations need to be sufficiently agile to react to the rapid development of tech companies and their practices.

    oAs regards oversight and enforcement, a dedicated entity at the European level, to be coordinated with the Commission’s existing series, could be created to implement this regulatory framework and establish supervision of structuring platforms.

    -The French competition authority 146 argues there is a need for a solution to address the behaviour of structuring platform in markets where they are not dominant. They call for a new legal regime for ‘quasi-dominant’ operators to impose on them enforceable obligations in terms of interoperability, non-discrimination and access to data. Relevant competition authority could thus, on a case-by-case basis, either accept commitments and make them mandatory, or order the company to modify its behaviour in response to the identified competition concern.

    -As regards ‘killer acquisitions’, the competition authority report proposes the introduction of mandatory information requirements for every merger carried out by a structuring platform. The Autorité further proposes to assess whether substantive merger control rules should be adapted to digital challenges, especially in terms of potential competition, conglomerate effects, the relevant time scale of the analysis, and the impact of data and the creation of large user communities.

    In Italy, the Competition Authority (AGCM), the Data Protection Authority (DPA) and the National Regulatory Authority AGCOM have issued a report on policy recommendations 147 :

    -the data-driven approach in the analysis of the platform economy, and the analysis of data gathering, management and profiling from a multi-purpose angle encompassing consumer protection, privacy and competition objectives.

    -IT authorities stresses the risk of competitive barriers in existing and adjacent but also possibly completely new markets due to network effects and economies of scale/scope in data gathering, as well as in particular zero-pricing policies.

    -They also stress the importance, but also limits of privacy rules to achieve an optimal competitive amount of data protection granted by platforms (due to high information asymmetries between consumers/individuals and platforms, costs in switching and porting).

    -They indicate privacy and consumer protection breaches, including in particular lack of transparency on purpose of data gathering, as well as conglomerate effects due to extent of data sources and analysis particularly relevant also for the analysis of antitrust breaches.

    In the Netherlands, the Dutch government 148  is calling for ex ante intervention in addition to competition enforcement in order to prevent anti-competitive behaviour by dominant companies acting as gatekeeper to the relevant online ecosystem (to prevent that ex post enforcement comes too late to keep markets competitive and contestable). By adding an extra tool to Regulation 1/2003, both at EU and national levels respectively, the new instrument will preserve the single market and national enforcement (to reflect heterogeneity of platforms/markets). Platforms in scope are platforms with gatekeeper role/bottleneck power, not necessarily dominant under competition rules but presenting risk of permanent dominance in the future due to ecosystem control (identifying factors: network effects, data collection, scale and scope effects, platform-of-platforms/ecosystems). As regards remedies, they propose SMP-type remedies to keep them targeted, such as platform access, data portability/sharing, non-discriminatory ranking; by adding an extra tool to Regulation 1/2003; both at EU and national levels to respectively preserve the single market and national enforcement (to reflect heterogeneity of platforms/markets). They call for the notification thresholds to be amended to take account of the deal value.

    Regarding studies carried out in the Netherlands supportive for action to be taken the following should are to be mentioned:

    -The Dutch competition authority market study on app stores 149 points in particular to bottleneck power over app providers and unilateral conduct of Google and Apple that can be used to expand their platform-ecosystems. As specific problems, they point to differentiated treatment, self-preferencing and lack of transparency.

    -The report commissioned by the Dutch Ministry of Economic Affairs and Climate Policy on digital gatekeepers of October 2019. 150  

    The Belgian, Dutch and Luxembourg competition authorities issued a position paper on the challenges faced by competition authorities in the digital world. Besides proposals to modernise the EU Merger Control Regulation and to (re-) introduce case-by-case guidance letters upon request, the three NCAs advocate for the introduction of an ex-ante instrument similar to the Dutch government proposal.

    -As regards addressees, they argue that the concept and interpretation of ‘dominance’ under Art. 102 TFEU should be closely followed for reasons of legal certainty and predictability.

    -COM Guidelines should be updated, clarifying e.g. the role of data, consumer behaviour and network effects.

    -As regards the nature of remedies, the new tool could be modelled along (1) UK CMA power to impose remedies following market studies and/or (2) MS’ telecom authorities to impose remedies on companies with significant market power.

    -Only behavioural remedies should be used, e.g. platform access, data portability, data-sharing and on-discriminatory ranking.

    -As regards procedural aspects, they argue for ‘voluntary’ commitments similar to Art. 9 of Regulation 1/2003, but without intention by the Commission to adopt a decision and no accusation of any wrongdoing. Rebuttable presumption that remedies are proportionate.

    -With respect to competent authorities, Commission is best-placed to impose remedies on EU-wide dominant companies. MS should enforce at national level in cases where company is dominant only in one MS.

    In Romania on 20 June 2020 a draft law on relative bargaining power has been published for public consultation. The dependency criteria are defined by reference to the existence of an imbalance of power due to elements such as the considerably larger dimension or market position, the importance of the commercial relationship for the dependent enterprise and the difficulty. 151

    To summarise, the current legislative projects differ as to the threshold for intervention and as to the concept of scoping the services to be covered. While some project stay within the competition logic of market power within relevant markets and adjacent markets, some other proposals go for a larger intervention logic (Germany, France, relying on cross-market significance). More importantly, the proposed set of obligations differ with respect to the proposed prohibitions and obligations. For instance, regarding the proposed ex ante regulation on data, the French proposal is to provide access to data while the German proposal is only to prohibit cross platform usage. Another example for likely forthcoming discrepancies of obligations is illustrated by the fact that the French proposal contains further reaching obligations regarding device neutrality, while the German proposal does not contain such an obligation.

    Against those divergences in the legislative projects it is foreseeable that the existing divergences between MS described above are most likely to deepen even if not all proposed concepts are going to be maintained within the legislative processes. First of all it is very likely that the national rules will be scoped differently as to the types of power of digital platforms captured and that therefore the list of platforms covered will divergent. Finally, the legislative projects under way in MS will most likely result in the imposition of diverging ex ante obligations.

    Finally it should be born in mind that although in some Member States (BE, NL, LUX) there is the political will and the supporting studies to address the issues covered by the present initiative those Member States prefer to support harmonisation at EU level rather than to proceed at national level. However, absent Community action Member States are likely to start a legislative process with the resulting likelihood of further fragmentation. Different national legislation within the EU may lead to increased fragmentation and compliance costs for large market players and the business users that rely on them. At the same time, start-ups and smaller businesses are also negatively impacted by this situation, as it impedes them from scaling-up and from cross-border expansion, in order to grow into challengers of established players in the digital sector.

    Therefore, action at Community level is covered by Article 114 (1) TFEU also with the aim to prevent future fragmentation.

    Annex 5.5: Cost of No-Europe

    Gatekeepers may be legally established in one Member State and provide their services to almost the entire EU population. Given the intrinsic cross-border nature of platforms, measures at national level cannot be effective in addressing issues in the digital space. On the contrary, the proliferation of national laws would result in a range of different rules, which puts at risk the scale-up and competition capacity of smaller and start-up online platforms, thus further cementing gatekeepers’ entrenched position. Lacking any EU-wide regulation, national solutions are likely to lead to conflicting outcomes where they are implemented by platforms operating at a pan-European scale. A multiplication of national rules and a lack of coordination only benefits the largest companies that are able to deal with 27 different legal systems. At the same time, larger platforms would also be negatively impacted by a fragmented legal landscape since it undermines legal certainty and regulatory predictability. For businesses using online platforms it would be even harder to apply different set of rules within the EU so fragmentation would discourage them to trade across the EU.

    Not addressing issues raised by gatekeepers would thus lead to stronger legal fragmentation undermining the potential of the Digital Single Market. As further explained under Section 6 on impacts, the online platform economy contributes heavily to EU cross-border trade and the EU economy as a whole. The top 50 online platforms represent 60% of the traffic share 152 in Europe reaching revenues for about EUR 276 billion in 2018 and employing almost 600 000 people. In addition, the platform economy is expected to grow 153 and represents an opportunity for EU platforms and businesses using their services. It is therefore necessary to address obstacles to a properly functioning online platform economy in order to ensure its positive contribution to the Digital Single Market. This is well illustrated by the following figures: cross-border e-commerce in Europe was worth EUR 143 billion in 2019. 59% of this market, i.e. EUR 84 billion, is generated by online marketplaces. Consequently, in an extreme scenario, where barriers between Member States are established that inhibit all cross-border sales by marketplaces, 59% of total turnover in 2019 would have been lost. Given that this figure is projected to increase to 65% in 2025, the lost cross-border sales would increase over time. Marketplaces with European capital represent 11% of the market.

    The size of online cross-border trade in Europe reached EUR 108.75 billion of turnover in 2019, representing 14.4% annual growth compared to 2018. However, if there is no EU intervention there is a risk of fragmentation in the Digital Single Market, which might reverse the positive trends in cross-border online trade.

    Assuming a 10% decrease per year in online cross-border trade, the opportunity cost of the digital market fragmentation would be EUR 1.76 trillion after 10 years.

     

    Annex 5.6: List of antitrust decisions and investigations in core platforms services

    1.Overview

    This annex highlights examples of conduct by some of the largest platforms in the core platform services that are being, or have been, investigated in recent years. These conducts range from anti-competitive use of third party data, through to practices which limit the ability of application and service providers to advertise or offer subscriptions outside a given platform for a lower price (anti-steering and most favoured nation clauses), through to practices which artificially raise the profile of the platforms’ own services and applications at the expensive of competitors.

    1.1.Misuse of third party data for competitive gain

    Procedure, authority, date

    Case reference

    Description

    Amazon

    European Commission

    Abuse of dominance or restrictive agreement

    Statement of objections on 10 November 2020

    AT.40462 Amazon Marketplace

    The Commission has informed Amazon of its preliminary view that very large quantities of non-public seller data are available to employees of Amazon's retail business and flow directly into the automated systems of that business, which aggregate these data and use them to calibrate Amazon's retail offers and strategic business decisions to the detriment of the other marketplace sellers.

    The Commission opened a second antitrust investigation into Amazon's business practices that might artificially favour its own retail offers and offers of marketplace sellers that use Amazon's logistics and delivery services (the so-called “fulfilment by Amazon or FBA sellers”).

    1.2.Other data practices

    Procedure, authority, date

    Case reference

    Description

    Facebook

    German NCA

    Abuse of dominance

    Decision of 6 February 2019

    Case B6-22/16

    The German NCA found that Facebook abused its dominance by applying terms and conditions, which made the use of its social network conditional upon Facebook’s possibility to collect and combine user data from multiple sources, including sources beyond the Facebook platform itself. In its decision, the competition authority prohibited Facebook and its other group companies from using those terms and conditions and from processing data accordingly insofar as private users in Germany are affected.

    The Bundeskartellamt’s decision was suspended upon appeal by the Duesseldorf Regional Court but on appeal, the German Federal Supreme Court supported the Bundeskartellamt’s case and found that each increase in the quantity and quality of the data and data analysis provided by Facebook, which is already very large in terms of the number of users, also reduces the chances of both actual and potential competitors to compete with this offer, with the risk that (potential) competitors will lose the competition for advertising contracts necessary to run the network. 154  

    Italian NCA

    Abuse of dominance

    11 May 2017

    CV154

    The Italian NCA found that WhatsApp de facto forced the users of its service to accept in full the new Terms of Use, and specifically the provision to share their personal data with Facebook, by inducing them to believe that without granting such consent they would not have been able to use the service anymore.

    1.3.Preferential display, restrictions in interoperability

    Authority, procedure, date

    Case reference

    Description

    Amazon

    German and Austrian NCAs

    Abuse of dominance – settled

    17 July 2019 (closed subject to commitments)

    Case B2-88/18 Amazon (Germany)

    Case Amazon (Austria)

    The German and Austrian competition authorities closed their investigations into several of Amazon’s terms and conditions for third-party sellers – ranging from liability rules to the duration of the notice period for the termination of a seller – following Amazon’s decision to amend them worldwide.

    In Germany, Amazon’s behaviour could have qualified as one of the following abusive conduct under the GWB: (i) An exploitative abuse of dominance in the form of abnormal business terms and discrimination, (ii) An abuse in the form of unjustified advantages, because there seemed to be no objective justification to the advantages required by Amazon, and (iii) An exclusionary abuse in the form of unfair business terms that favour its own downstream business, Amazon Retail, to the detriment of third-party sellers.

    Italian NCA

    Abuse of dominance

    10 April 2019 (opening of investigation)

    A528 Possible abuso di posizione dominante in marketplace e-commerce e servizi di logistica

    The NCA investigates whether Amazon abuses its dominance by discriminating on its e-commerce platform in favour of third-party merchants that use Amazon’s logistics services.

    The NCA’s theory of harm is that this practice may create barriers to expansion or entry for Amazon’s competitors: (i) in the e-commerce logistics market (leveraging), and (ii) possibly in the e-commerce platform market as well.

    Luxembourg NCA

    Abuse of dominance

    21 June 2017 (non-infringement)

    Case 2017-C-02 Amazon

    Amazon allegedly used its dominant position in the upstream market for platform services to foreclose the unnamed complainant from the downstream online retail market, where it competed with Amazon.

    The NCA concluded that Amazon’s platform service is not an essential facility for the complainant’s business and that Amazon does not have the incentive to pursue the alleged foreclosure strategy.

    Apple

    European Commission

    Restrictive agreement and/or abuse of dominance

    16 June 2020 (opening of investigation)

    AT.40452 Apple - Mobile payments

    The Commission is investigating the possible impact of the following on competition in mobile payments solutions: (i) Apple’s terms, conditions and other measures related to the integration of Apple Pay for mobile purchases (iPhones and iPads) on merchant apps and websites; (ii) Apple Pay is the only mobile payment solution that can access the NFC ‘tap and go’ technology that is embedded in iPhones for payments in stores; and (iii) Apple allegedly restricts access to Apple Pay for specific products of rivals on iPhones and iPads.

    Dutch NCA

    Abuse of dominance

    11 April 2019 (opening of investigation)

    Apple alleged app store abuse

    The NCA opened an investigation into whether Apple abused its dominance in the mobile app store market, “for example, by giving preferential treatment to its own apps”.

    The NCA opened the investigation after a market study into mobile app stores of Apple and Google, which identified self-favouring as one of the conduct that might warrant further investigation.

    Google

    European Commission

    Abuse of dominance

    Decision of 27 June 2017

    Case AT.39.740 Google Search (Shopping)

    Pending General Court judgment in Case T-612/17 Google and Alphabet

    Google's appeal is pending before the EU General Court against the Commission’s 2017 decision fining the company EUR 2.42bn.

    According to the Commission, Google leveraged its dominance in general internet search services to the separate comparison shopping service (CSS) market by favouring Google Shopping on its general web search results page. The Commission concluded that this had the potential to foreclose competing CSSs, which could lead to: (i) higher fees for merchants, (ii) higher prices and search costs for consumers, (iii) less innovation.

    To support this finding, the Commission among other things relied on statistics showing that: (i) Google Shopping significantly increased its traffic (45-fold in the UK, 35-fold in Germany, 29-fold in the Netherlands, 19-fold in France, 17-fold in Spain and 14-fold in Italy) and (ii) traffic to rival CSSs dropped significantly (by 92% for a rival in Germany and by 85% for a rival in the UK) without an alternative explanation than Google’s behaviour.

    Italian NCA

    Abuse of dominance

    8 May 2019 (opening of investigation)

    A529 Google/

    compatibilità app Enel X Italia con sistema Android Auto 

    The NCA investigates whether Google abuses its dominance in the market for licensable smart mobile operating systems by refusing to integrate energy company Enel’s X Charge app in its Android Auto app.

    According to the NCA, this conduct may hinder competition on the merits and limit consumer choice by: (i) excluding Enel’s smartphone app for users of electric vehicles (EVs) which provides a location service that competes with Google Maps; and consequently, and (ii) “considerably compromising” several parameters of competition, including innovation, quality and diversity of services and diversity of business models.

    Italian NCA

    Abuse of dominance

    20 October 2020 (opening of investigation)

    A542 Google online display advertising

    The NCA is investigating an alleged exclusionary abuse of dominance by Google in the form of “internal-external discrimination”, which consists in: (i) its display advertising intermediation services relying on user data that Google collects from other, unrelated services or applications in which it is dominant; and (ii) not making this data available for competing providers of display ad intermediation services.

    1.4.Anti-steering and MFN clauses

    Authority, procedure, date

    Case reference

    Description

    Apple

    European Commission

    Restrictive agreement and/or abuse of dominance

    16 June 2020 (opening of investigations)

    AT. 40437 Apple – App Store practices (music streaming)

    AT.40652 Apple – App Store practices (e-books/audiobooks)

    AT.40716 Apple – App Store practices (all other apps that compete with Apple)

    As Apple charges third-party app developers a 30% commission on all subscription fees through its mandatory in-app purchase (IAP) system, “Apple's competitors have either decided” to disable the in-app subscription possibility or pass this fee on to consumers. 

    The IAP obligation also appears to give Apple full control over the relationship with the subscribers of its competitors. Therefore, it is “dis-intermediating its competitors from important customer data while Apple may obtain valuable data about the activities and offers of its competitors”. In addition, Apple’s rules prevent third-party app developers from informing users of alternative options to purchase paid content.

    European Commission

    Restrictive agreement and/or abuse of dominance

    4 May 2017 (closed subject to commitments)

    AT.40153 E-book MFNs and related matters (Amazon)

    Commission decision that renders legally binding the commitments offered by Amazon. The Commisison had concerns about clauses included in Amazon's e-books distribution agreements that required publishers to offer Amazon similar (or better) terms and conditions as those offered to its competitors and/or to inform Amazon about more favourable or alternative terms given to Amazon's competitors. The clauses covered not only price but many aspects that a competitor can use to differentiate itself from Amazon, such as an alternative business (distribution) model, an innovative e-book or a promotion.

    The Commission considered that such clauses could make it more difficult for other e-book platforms to compete with Amazon by reducing publishers' and competitors' ability and incentives to develop new and innovative e-books and alternative distribution services.

    1.5.Tying

    Authority, procedure, date

    Case reference

    Description

    Apple

    European Commission

    Restrictive agreement and/or abuse of dominance

    Decision of 18 July 2018

    AT.40099 Google Android

    Google's appeal is pending before the EU General Court against the Commission’s 2017 decision fining the company EUR 4.34bn.

    According to the Commission, Google has ensured that its Google Search app is pre-installed on practically all Android devices sold in the EEA by tying it pre-installation with the pre-installation of the Google Play Store. The Commission found that pre-installation can create a status quo bias. Users who find search and browser apps pre-installed on their devices are likely to stick to these apps.

    Google's practice has reduced the incentives of manufacturers to pre-install competing search apps, as well as the incentives of users to download such apps. This reduced the ability of rivals to compete effectively with Google.

    (1)

         Massimo Motta is a professor at the Pompeu Fabra University in Barcelona and served as Chief Competition Economist of the European Commission from 2013 to 2016. Martin Peitz is a professor of economics at the University of Mannheim.

    (2)

         Alexandre De Streel is professor of European law at the Universities of Namur and Louvain, Professor Larouche is professor in law and innovation at the Faculty of Law at the Université de Montréal.

    (3)

         Heike Schweitzer is a professor in the Humboldt University of Berlin and was one of the special advisers authoring the Competition policy for the digital era report.

    (4)

         Richard Whish is emeritus professor of Law at King's College London and one of the leading competition law scholars.

    (5)

         Gregory S. Crawford is a professor of Applied Microeconomics at the University of Zurich, Patrick Rey is Professor of Economics at the Toulouse School of Economics, Monika Schnitzer is a member of the German Council of Economic Experts and a professor of comparative economics at the Ludwig-Maximilian-University Munich.

    (6)

      https://ec.europa.eu/digital-single-market/en/expert-group-eu-observatory-online-platform-economy  

    (7)

      https://ec.europa.eu/digital-single-market/en/news/commission-expert-group-publishes-progress-reports-online-platform-economy  

    (8)

      https://publications.jrc.ec.europa.eu/repository/handle/JRC122896  

    (9)

          https://publications.jrc.ec.europa.eu/repository/handle/JRC122897  

    (10)

          https://publications.jrc.ec.europa.eu/repository/handle/JRC122898

    (11)

          https://publications.jrc.ec.europa.eu/repository/handle/JRC122899  

    (12)

          https://ec.europa.eu/jrc/sites/jrcsh/files/jrc109185.pdf .

    (13)

          https://ec.europa.eu/jrc/sites/jrcsh/files/jrc109186.pdf .

    (14)

          https://ec.europa.eu/jrc/sites/jrcsh/files/jrc106299.pdf .  

    (15)

          https://ec.europa.eu/jrc/sites/jrcsh/files/JRC101501.pdf .  

    (16)

          https://cerre.eu/events/contestability-digital-markets-role-essential-data .  

    (17)

          https://cerre.eu/events/designing-eu-intervention-standard-digital-gatekeepers .  

    (18)

          https://cerre.eu/sites/cerre/files/2019_cerre_market_definition_market_power_platform_economy.pdf .

    (19)

          https://www.cerre.eu/sites/cerre/files/cerre_intermediationbiasremedies_report.pdf .  

    (20)

          https://cerre.eu/publications/big-data-and-competition-policy .  

    (21)

          https://cerre.net/publications/internet-platforms-non-discrimination/ .  

    (22)

          https://www.acm.nl/sites/default/files/documents/market-study-into-mobile-app-stores.pdf .

    (23)

          https://www.government.nl/documents/reports/2019/10/07/digital-gatekeepers .  

    (24)

          https://www.government.nl/documents/letters/2019/05/23/future-proofing-of-competition-policy-in-regard-to-online-platforms .  

    (25)

          https://www.tresor.economie.gouv.fr/Articles/7690058a-00e4-44a7-8aed-9a2ee5a04d51/files/c888861f-5516-4e4e-b3ce-a96af66b3c34 .  

    (26)

          https://www.autoritedelaconcurrence.fr/fr/communiques-de-presse/lautorite-publie-sa-contribution-au-debat-sur-la-politique-de-concurrence .  

    (27)

          https://monopolkommission.de/images/HG23/HGXXIII_Gesamt.pdf#page36 . 

    (28)

      https://www.bmwi.de/Redaktion/EN/Downloads/a/a-new-competition-framework.pdf?__blob=publicationFile&v=2 .

    (29)

      https://www.bmwi.de/Redaktion/DE/Downloads/Studien/modernisierung-der-missbrauchsaufsicht-fuer-marktmaechtige-unternehmen-zusammenfassung-englisch.pdf?__blob=publicationFile&v=3 .  

    (30)

    https://www.bwb.gv.at/news/detail/news/rtr_praesentiert_methodenpapier_in_enger_kooperation_mit_bwb_zu_monitoring_digitaler_plattformen/ .  

    (31)

      https://www.agcom.it/documents/10179/17633816/Allegato+10-2-2020+1581347457837/c4139504-3777-4674-ad6d-ac6b9d501608?version=1.0 .

    (32)

          https://ec.europa.eu/competition/information/digitisation_2018/contributions/comision_nacional_de_los _mercados_y_la_competencia.pdf .

    (33)

          US House of Representatives Majority Staff Report, Investigation of Competition in Digital Markets, October 2020 .

    (34)

          https://www.gov.uk/government/publications/unlocking-digital-competition-report-of-the-digital-competition-expert-panel .

    (35)

          https://www.ofcom.org.uk/__data/assets/pdf_file/0025/174634/online-market-failures-and-harms.pdf .  

    (36)

          https://www.gov.uk/cma-cases/online-platforms-and-digital-advertising-market-study#final-report .  

    (37)

          https://research.chicagobooth.edu/-/media/research/stigler/pdfs/market-structure-report.pdf?la=en&hash=E08C7C9AA7367F2D612DE24F814074BA43CAED8C .  

    (38)

          https://www.accc.gov.au/system/files/Digital%20platforms%20inquiry%20-%20final%20report.pdf .  

    (39)

          https://www.accc.gov.au/publications/serial-publications/digital-platform-services-inquiry-2020 -2025/digital-platform-services-inquiry-september-2020-interim-report .

    (40)

          https://www.jftc.go.jp/en/pressreleases/yearly-2019/October/191031Report.pdf .  

    (41)

          https://www.kantei.go.jp/jp/singi/digitalmarket/pdf_e/documents_200616-1.pdf .

    (42)

    https://www.beuc.eu/publications/beuc-x-2019-054_competition_policy_in_digital_markets.pdf .

    (43)

          http://www.oecd.org/daf/competition/Rethinking-antitrust-tools-for-multi-sided-platforms-2018.pdf .  

    (44)

          https://publications.jrc.ec.europa.eu/repository/handle/JRC122910

    (45)

          https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3469624 .  

    (46)

          https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3544694 .

    (47)

          https://platformobservatory.eu/about-observatory/group-of-experts/ .  

    (48)

          Inception Impact Assessment for the Digital Services Act package .

    (49)

          Inception Impact Assessment of the New Competition Tool .

    (50)

          Inception Impact Assessment of the New Competition Tool , at page 3; and Inception Impact Assessment for the Digital Services Act package , at page 4.

    (51)

    See for a more elaborate synopsis report the OPC results on the Ex Ante Tool Annex 2.1 below.

    (52)

          https://ec.europa.eu/competition/consultations/2020_new_comp_tool/index_en.html .  

    (53)

         See https://ec.europa.eu/competition/consultations/2020_new_comp_tool/index_en.html .

    (54)

         See https://ec.europa.eu/competition/consultations/2020_new_comp_tool/index_en.html .

    (55)

         This synopsis report of the open public consultation is based on the analysis of the replies performed by College of Europe contracted by the Commission to support in the qualitative and quantitative analysis.

    (56)

    Sample size: 2863.

    (57)

      https://ec.europa.eu/info/sites/info/files/better-regulation-toolbox.pdf .

    (58)

     i.e. we weighted the term frequency for term i and document j by the natural log of the number of documents over the number of other documents containing word i.

    (59)

    See section 5 below.

    (60)

         See Annex B, Table B.1 for the overview of responses among different stakeholders’ categories. E.g. about 73% of the 11 telecom operators, 67% of the 52 civil society organisations and 44% of the 9 scale up and startup platforms responding to the question, fully disagree that consumers have sufficient choices to the offering of the online platforms. Looking at large platforms, 33% of the 33 respondents somewhat disagree that consumers have sufficient choices.

    (61)

         Positions of publishers on media plurality are further assessed in the next section.

    (62)

         I.e. large user base, wide geographical coverage, large share of total market revenue, impact on a certain sector, exploitation of strong network effects, leverage of assets to enter new areas of activity, raising of barriers to entry, accumulation of valuable and diverse data and information, lack of alternative services, lock-in of users.

    (63)

         Figure 3. “Accumulation of valuable & diverse data and information”, sample size 1304; “Large user base”, sample size 1300; “Creation of barriers to entry for competitors”, sample size 1293.

    (64)

         See Annex B, Table B.2 for the overview of responses among different stakeholders’ categories. Among the respondents to this question, 25 out of 32 large platforms and 5 out of 9 scaleup/startup platforms replied that they agree or fully agree with the need to consider dedicated regulatory rules for gatekeeper platforms, that is 30 out of 41 platforms (73%).

    (65)

         Sample Size: 1346. The only exception are business associations representing the interest of platforms among which 38% agree with the need to prohibit certain practices and 50% replied that they do not know.

    (66)

         Sample Size: 1274.

    (67)

         Sample Size: 1216.

    (68)

    See Annex B, Tables B.3 and B.4 for the overview of responses among different stakeholders’ categories.

    (69)

         See Section 1 on ‘unfair practices by gatekeeper platforms’.

    (70)

    See Annex B, Table B.6 for the overview of responses among different stakeholders’ categories.

    (71)

    Sample size: 1295.

    (72)

    The only respondent categorized as ‘caching services’ also expressed the preference for oversight at EU level only.

    (73)

         Sample Size: 1715.

    (74)

         Sample Size: 1476.

    (75)

         Sample Size: 1346.

    (76)

         Sample Size: 1274.

    (77)

         Sample Size: 1216.

    (78)

         Sample Size: 1295.

    (79)

         OECD, 2014, Fact-sheet on how competition policy affects macro-economic outcomes.

    (80)

         See G. Federico, F. Scott Morton & C. Shapiro (2019), Antitrust and Innovation: Welcoming and Protecting Disruption, NBER Working Paper No. 26005 and P. Aghion, N. Bloom, R. Blundell, R. Griffith & P. Howitt (2005), Competition and Innovation: an Inverted-U Relationship, The Quarterly Journal of Economics, volume 120(2), pages 701-728. On empirical work, see P. Aghion, S. Bechtold, L. Cassar & H. Herz (2014), The causal effects of competition on innovation: Experimental evidence, Journal of Law, Economics, and Organization, volume 34 (2), pages 162-195 and C. Shapiro (2012), Competition and innovation. Did Arrow hit the bull’s eye?, chapter 7 of Josh Lerner and Scott Stern (eds.), The Rate and Direction of Inventive Activity Revisited, pages 361-404.

    (81)

         As evidenced by Gutmann & Voigt (2014), there is a significant relationship between the introduction of competition law and annual growth arising mainly from more investment. J. Gutmann & S. Voigt (2014), Lending a Hand to the Invisible Hand? Assessing the Effects of Newly Enacted Competition Laws.

    (82)

         See for instance S. Ahn (2002), Competition, Innovation and Productivity Growth: A Review of Theory and Evidence, OECD Economics Working Paper No. 317. See also for example, a study by the European Commission (2015) on The Economic Impact of enforcement of competition policies in the functioning of EU energy markets , which found that the Commission's decision finding an abuse of dominance by E.ON lead to a reduction in prices for both wholesalers and retailers to the benefit of consumers. See also the Note by the UNCTAD Secretariat (2014), The benefits of competition policy for consumers .

    (83)

         Online platforms benefit from asymmetry of information (they dispose of large data sets compared to consumers). Platforms’ analytical capacity gives them the possibility to use advanced algorithms and machine learning techniques to facilitate targeting, discriminatory practices and behavioural manipulation. BEUC considers that such practices can have an impact on demand and distribution of wealth – “the most vulnerable consumers might end up paying higher prices than under a competitive price scenario (when personalisation is combined with commercial practices seeking to increase the individual consumer’s willingness to pay). They may also be used to target biases and reinforce existing or desired viewpoints with the aim of keeping users engaged with the firm’s platform so as to generate advertising revenues.” BEUC (2019), The Role of Competition Policy in Protecting Consumers’ Well-being in the Digital Era .

    (84)

         In fact, higher advertising prices represent increased costs to the companies producing goods and services which are purchased by consumers. These costs are expected to be passed through to consumers in terms of higher prices for goods and services, even if the downstream market is highly competitive.

    (85)

         See Section 6 of CMA report on Online platforms and digital advertising .

    (86)

         See Section 6 of CMA report on Online platforms and digital advertising .

    (87)

         Cross-border e-commerce in Europe was worth EUR 143 billion in 2019, with 59% of this market being generated by online marketplaces. This is projected to increase to 65% in 2025 ( Ecommerce News Europe (2020 )).

    (88)

    The data used in the input-output modelling come from three sources: (a) The 2014 world input-output table (WIOT) publicly available from the World Input-Output Database (WIOD, www.wiod.org ), (b) Employment (number of persons engaged) and compensation of employees obtained from the Socio-Economic Accounts (SEAs) of WIOD, and (c) private R&D investments in information and communication (and its subitems represented by NACE Rev.2’s Section J’s divisions and/or groups), which were downloaded from Eurostat (rd_e_fundgerd).www.wiod.org), (b) Employment (number of persons engaged) and compensation of employees obtained from the Socio-Economic Accounts (SEAs) of WIOD, and (c) private R&D investments in information and communication (and its subitems represented by NACE Rev.2’s Section J’s divisions and/or groups), which were downloaded from Eurostat (rd_e_fundgerd).www.wiod.org), (b) Employment (number of persons engaged) and compensation of employees obtained from the Socio-Economic Accounts (SEAs) of WIOD, and (c) private R&D investments in information and communication (and its subitems represented by NACE Rev.2’s Section J’s divisions and/or groups), which were downloaded from Eurostat (rd_e_fundgerd).www.wiod.org), (b) Employment (number of persons engaged) and compensation of employees obtained from the Socio-Economic Accounts (SEAs) of WIOD, and (c) private R&D investments in information and communication (and its subitems represented by NACE Rev.2’s Section J’s divisions and/or groups), which were downloaded from Eurostat (rd_e_fundgerd).

    (89)

         The most recent available input-output matrix is for 2014, yet the matrix does not change significantly across time.

    (90)

    Leontief, W.W. (1986), Input-Output Economics, Second edition. Oxford: Oxford University Press.

    (91)

    Miller, R.E. and Blair, P.D., 2009. Input-output analysis: foundations and extensions. Cambridge university press. Available at: http://digamo.free.fr/io2009.pdf .

    (92)

         Business expenditure on R&D (BERD) by NACE Rev. 2 activity and source of funds [rd_e_berdfundr2], metadata accessible at: https://ec.europa.eu/eurostat/cache/metadata/en/rd_esms.htm l.  

    (93)

         The R&D expenditure data cover part of ICT services (but not ICT manufacturing), along with other subitems of Information and communication sector. These ICT services include Software publishing (NACE Group 58.2), Telecommunications (NACE Division 61), Computer programming, consultancy and related activities (NACE Division 62), and Data processing, hosting and related activities; web portals (NACE Group 63.1).

    (94)

         Alternatively, one can see this as the difference between (comparative) static and dynamic approaches.

    (95)

         Alternative definitions of the market size give similar results. See Duch-Brown and Martens (2016) for further details.

    (96)

      https://ec.europa.eu/competition/consultations/2020_new_comp_tool/index_en.html .  

    (97)

    Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services (‘ Platform-to-Business Regulation ’) [2019] OJ L 186/57.

    (98)

    Support study to the Observatory for the Online Platform Economy, SMART 2018/0034.

    (99)

         Crémer, J., de Montjoye, Y.-A., & Schweitzer, H. (2019). Competition policy for the digital era. European Commission .

    (100)

         Gröne, F., Péladeau, P., & Samad, R. A. (2019). Tomorrow’s data heroes. Strategy+business. Retrieved from https://www.strategy-business.com/article/Tomorrows-Data-Heroes?gko=5f270 .  

    (101)

         Rohler, M. (2019). How are Estimated Sales and Revenue Calculated? What are AccuSales? Jungle Scout Help Center. Retrieved from https://support.junglescout.com/hc/en-us/articles/360008616814-How-are-Estimated-Sales-and-Revenue-Calculated-What-are-AccuSales-#:~:targetText=We%20gather%20it%20from%20a,their%20sales%20info%20with%20us.&targetText=This%20means%20that%20as%20we,analyze%20and%20test%20it%20daily.  

    (102)

         Gerber, S. (2018). Want to sell on Amazon? 15 strategies for success. The Next Web (TNV). Retrieved from https://thenextweb.com/contributors/2018/01/19/want-sell-amazon-15-strategies-success/ .  

    (103)

         Graef, I. (2016). EU Competition Law, Data Protection and Online Platforms: Data as Essential Facility. Kluwer Law International BV.

    (104)

         Christl, W. (2017). How companies use personal data against people. Working paper by Cracked Labs, Institute for Critical Digital Culture.

    (105)

         Christl, W. (2017). How companies use personal data against people. Working paper by Cracked Labs, Institute for Critical Digital Culture.

    (106)

         Privacy International. (2018). Why we’ve filed complaints against companies that most people have never heard of – and what needs to happen next. Retrieved from https://privacyinternational.org/advocacy/2434/why-weve-filed-complaints-against-companies-most-people-have-never-heard-and-what .

    (107)

         Five of these 15 business users represented businesses with less than 10 employees. Two interviewees represented a business with between 10 and 49 employees and three from a business with between 50 and 249 employees. Large businesses with more than 249 employees accounted for five interviews.

    (108)

         The two interviewed business users that did not experience differentiated treatment spoke for a company with less than 10 employees and one with between 50 and 249 employees, respectively.

    (109)

         Among the business users that experienced self-preferencing were five businesses with more than 249 employees, two with between 50 and 249 employees, two with between 10 and 49 employees and one with between 1 and 9 employees.

    (110)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019.

    (111)

         CMA report on Online platforms and digital advertising .

    (112)

         ACCC report, Digital Platforms Inquiry, Final Report , June 2019.

    (113)

          http://www.samr.gov.cn/hd/zjdc/202011/t20201109_323234.html .  

    (114)

         Furman report,  Unlocking digital competition, Report of the Digital Competition Expert Panel , March 2019. 

    (115)

         CMA report on Online platforms and digital advertising .  

    (116)

    The ACCC has not undertaken a detail assessment of non-dominant markets where Google offers services (markets for advertising technology services or programmatic display ads, but it recognises that EC has found Google to be dominant in both mobile operating system and app store markets.

    (117)

         European Commission, Google Android decision, 18 July 2018; European Commission, Antitrust: Commission fines Google €4.34 billion for illegal practices regarding Android mobile devices to strengthen dominance of Google’s search engine, 18 July 2018, accessed 4 June 2019.

    (118)

    Reuters, 10 November 2020.

    (119)

         ECJ, Judgment of the Court of 13 May 1997, C-233/94, para 19; ECJ, Judgment of the Court of 5 October 2000, C-376/98, Federal Republic of Germany v European Parliament and Council of the European Union, paras 95, 96.

    (120)

         ECJ, Judgment of the Court (Grand Chamber) of 10 February 2009, C301/06, Ireland v European Parliament and Council of the European Union, paras 62-72.

    (121)

         ECJ, Judgment of 13 July 1995, C-350/92, paras 33-40.

    (122)

         ECJ, Judgment of 5 October 2000, C-376/98, para 106-109;

    (123)

         ECJ, Judgment of the Court of 9 October 2001, C-377/98, Kingdom of the Netherlands v European Parliament and Council of the European Union, para 28; ECJ, Judgment of the Court of 5 October 2000, C-376/98, Federal Republic of Germany v European Parliament and Council of the European Union, para 88; ECJ, Judgment of the Court of 10 December 2002, C-491/01, British American Tobacco, paras 62, 75.

    (124)

         La loi du 4 avril 2019 modifiant le Code de droit économique en ce qui concerne les abus de dépendance économique, les clauses abusives et les pratiques du marché déloyales entre entreprise. Article I.6.4 Code de Droit Economique (CDE, Code of Economic Law) defines dependency ; Article IV.2/1 CDE describes the types of prohibited abuses. Following Royal Decree amending the Code of Economic Law regard g abuses of economic dependence published on 12 August 2020, this prohibition on abuse of economic dependence has entered into force on 22 August 2020.

    (125)

         Article 37A Competition Act introduced by the amendment to the Protection of Competition Act of 9 July 2015.

    (126)

    Competition Act 2008 and 2014 Part 2, Chapter 6, para 2: Competition Act II 6 (2).

    (127)

    The key provision to regulate significant imbalance was introduced in 2008. It provides in Article 442-6 -2 Code de Commerce, that any producer, trader, manufacturer or person recorded in the trade register who commits the following offences shall be held liable and obliged to make good the damage caused … 2° Subjecting or seeking to subject a trading partner to obligations that create a significant imbalance in the rights and obligations of the parties; see also the description of the ‘petit droit de la concurrence’ on dealing with unfainess in dependency relationships and the cases dealt with by the DGCRF under Article 442-1 and Article 442-6 -2 Code de Commerce, in Rapport d’information par la Commission des Affaires Economiques sur les platformes numériques, présenté par MME Valeria Fauré-Muntian and M. Daniel Fasquelle, a l‘Assembléé Nationale, 24 June 2020, Rapport No 3127, p. 42-44.

    (128)

    Article 20 Competition Act.

    (129)

    Article 20 (3) of the Competition Act provides that “Undertakings with superior market power in relation to small and medium-sized competitors may not abuse their market position to impede such competitors directly or indirectly in an unfair manner. An unfair impediment within the meaning of sentence 1 exists in particular if an undertaking 1 offers goods or commercial services not just occasionally below cost price, or 2. demands from small or medium-sized undertakings with which it competes on the downstream market in the distribution of goods or commercial services a price for the delivery of such goods and services which is higher than the price it itself offers on such market, unless there is, in each case, an objective justification."

    (130)

    PSSA Gesetz über die Beaufsichtigung von Zahlungsdiensten – ZAG, https://www.gesetze-im-internet.de/zag_2018/__58a.html.

    (131)

         Franck/Linartardos, EPOS collaborative Research Center, discussion paper no 173, https://www.crctr224.de/en/research-output/discussion-papers/archive/2020/germanys-lex-apple-pay-payment-service-regulation-overtakes-competition-enforcement-jens-uwe-franck-dimitrios-linardatos.

    (132)

    Hungarian Competition Acm Section 21 paragraphs b), c) and i).

    (133)

    Decree Law no. 214 /2011 (‘Salva Italia’ Decree) of 22 December 2011 and Decree law no. 1/2012 (‘Cresci Italia’ Decree) 24 January 2012.

    (134)

    Impact assessment - Proposal for a Regulation on promoting fairness and transparency for business users of online intermediation services, COM(2018) 238 final –Part I/2, Chapter 1/2 2.1.1.6 Most-favoured nation (MFN) clauses.

    (135)

    Article L311-5-1 of the Tourisme Law (Code du Tourisme) as modified by the Law Macron provides that an hotel operator maintains his freedom to agree with the client any rebate of tariff advantage of any kind while any clause stipulating otherwise must be considered as unwritten and is void. The « LOI n° 2015-990 du 6 août 2015 pour la croissance, l'activité et l'égalité des chances économiques » (so called ‘Loi Macron’)

    (136)

    The ban of narrow MFN clauses was introduced by a modification to the Unfair Commercial Practices Law by adding those clauses to the blacklisted practes-Z32, modification to the UWG, östBGBl I 2016/99, for a complete picture of situation in Austria on MFN see Chapter 5.1 of the publication on MFN’s in in Österreichische Zeitschrift für Kartellrecht 13, 127–140)

    (137)

    In Italy, all parity clauses are banned by the Competition Act,, Legge annuale per il mercato e la concorrenza, adopted on 2 August 2017 prohibit any MFN clauses in agreements between OTAs and hotel operators (i.e. wide and narrow MFN clauses, and regardless of the size of the OTA).

    (138)

    In Belgium, since August 2018 wide and narrow MFN clauses are banned by the law on freedom of hotels to set prices in their relationship with reservation platforms- Loi du 30 Juillet 2018 relative à la liberté tarifaire des exploitant d’hebergements touristiques dans les contrats conclus avec les operateurs des platformes de reservation en ligne, Moniteur Belge (M.B.) 10 Aout 2018.

    (139)

    Gesetzesentwurf der Bundesregierung, Entwurf eines Gesetzes zur Änderung des Gesetzes gegen Wettbewerbsbeschränkungen für ein fokussiertes, proaktives und digitales Wettbewerbsrecht 4.0 und anderer wettbewerbsrechtlicher Bestimmungen ( GWB-Digitalisierungsgesetz ).

    (140)

    Governmental draft bill for the 10th amendment to the Competition Act of 9 September 2020 Explanatory part, p. 84.

    (141)

    Governmental draft bill for the 10th amendment to the Competition Act of 9 September 2020 Explanatory part, p. 83.

    (142)

     Governmental draft bill for the 10th amendment to the Competition Act of 9 September 2020, Explanatory part, p. 89

    (143)

     Governmental draft bill for the 10th amendment to the Competition Act of 9 September 2020 Explanatory part, p. 94.

    (144)

    The governmental draft of 9 September 2020 is currently discussed in the German Parliament (Deutscher Bundestag Drucksache 19/23492 19. Wahlperiode 19.10.2020, https://dip21.bundestag.de/dip21/btd/19/234/1923492.pdf ) and in the German Council of the regions Bundesrat, Drucksache 568/20 of 6 November 2020, https://www.bundesrat.de/SharedDocs/TO/995/erl/32.pdf?__blob=publicationFile&v=1 ).

    (145)

    Autorité de la concurrence: Contribution of the Autorité de la concurrence to the debate on competition policy and digital challenges (2020).

    (146)

      https://www.agcm.it/dotcmsdoc/allegati-news/Big_Data_Lineeguida_Raccomandazioni_di_policy.pdf .

    (147)

      Dutch position on modernising competition policy in relation to digital platforms .

    (148)

      https://www.acm.nl/sites/default/files/documents/market-study-into-mobile-app-stores.pdf .

    (149)

      https://www.government.nl/documents/reports/2019/10/07/digital-gatekeepers .

    (150)

    See description of the public consultation: Romania: Draft law sanctioning the abuse of superior bargaining position published for public consultation at https://www.lexology.com/library/detail.aspx?g=5be9d7e5-8e41-4d38-b36c-17f7370e245f .

    (151)

    Traffic share is one of the most important proxies of the sector.

    (152)

       ICF support study for the IA shows that the size of EU28 online cross-border trade in Europe for 2019 represents a 14.4% increase in comparison to 2018. Also, according to Cross-Border Commerce Europe 2019 study online marketplaces will represent 65% of cross-border online sales in Europe by 2025.

    (153)

         Decision by the Bundesgerichtshof of 23 June 2020, KVR 69/19, available at http://juris.bundesgerichtshof.de/cgi-bin/rechtsprechung/document.py?Gericht=bgh&Art=en&nr=109506 .

    Top