Digital Single Market and the EU Competition Regime: An Explanation of Policy Change

ABSTRACT

Although the EU competition regime is well-established and highly effective, EU policy actors may still need to rethink their tried and tested approach to competition regulation. This is what happened in the context of the European Commission’s planned regulation of online platforms, embodied (in part) within the Digital Markets Act. This article reviews the interplay of the EU’s competition regime with its relatively new Digital Single Market strategy to ask how a traditional ex-post approach to competition regulation came to be supplemented by a (proposed) ex-ante regulatory approach. Informed by the literature on policy change, the article examines the policy context, the Commission’s experience gained in dealing with competition cases, and the input of lobbyists, advocacy groups and experts, to explain this shift in Commission policy.

KEYWORDS:

• Digital single market
• competition policy
• antitrust policy
• European Commission
• ex-post regulation
• European Union

Introduction

The proliferation of digital technologies is driving a ‘new global industrial revolution’ (European Commission , 2021a, 1). This digital transformation of business practices has been characterised by the emergence of large online (or digital) platforms that are able to dominate their own platform ecosystems. The activities of these online platforms are multi-sided. This allows them to act as gatekeepers for multiple business users and millions of consumers, while also offering services in their own right. End users may be tracked and profiled by the platforms. Business users are often dependent on them. At the same time, platforms may restrict competition by limiting access to core markets (European Commission 2020b, 1). In the case of the Single European Market (SEM), the conduct of these digital giants threatens free and fair competition, while the absence of effective EU-level regulation risks market fragmentation. The platforms, which include the ‘Big Five’, Google, Amazon, Facebook (now rebranded as Meta), Microsoft, and Apple, as well as smaller companies, generate considerable wealth, both for themselves and for the economies in which they operate. Yet, the profound political implications of their market power challenge policy-makers to develop appropriate responses.

This article examines the proposed European Union (EU) regulation of these powerful online platforms. Where regulation involves the oversight of anti-competitive practices, responsibility lies with the European competition (or antitrust) regime which allows the European Commission (DG Competition to be precise) to investigate and where necessary to sanction European-scale cartels, dominant practices, and anti-competitive mergers, as well as prohibiting unfair state aid granted by member states. Competition policy is governed by both primary and secondary EU legislation, with its regulatory approach largely characterised by ex-post enforcement. This ex-post approach sees the Commission and the Court of Justice, taking or ruling on decisions in individual cases after an infringement has occurred (Anderson and Mariniello 2021). Its purpose is to guarantee the effectiveness of the SEM by protecting and maintaining the EU’s level-playing field for business through the promotion of free and fair competition as a ‘necessary condition of [the Single Market’s] success’ (Cini and McGowan 2009, 32).

The enforcement of competition cases involving online platforms has become high-profile and controversial. In applying existing rules, senior Commission officials have emphasised the resilience of competition law and its ability to cope with new market circumstances. For much of the 2014–2019 period, European Commission leaders articulated a commitment to traditional competition instruments, which, it was argued, were able to meet the needs of the new digital economy and society (Laitenberger 2019a, 2019b). In 2018, Competition Commissioner Vestager stated that: ‘the fact is, the competition rules are designed to adapt. The principles they set out are valid for every market, for the ones we know today and the ones we’ll see in the future. And that’s why dealing with digital markets isn’t really about new rules. It’s about deepening our understanding of how those markets work’ (Vestager 2018). Yet, by mid-2020 the Commission had proposed new rules governing digital markets, opening the way to further market integration in this policy area.

The Commission’s position on ‘new rules’ changed gradually. In January 2019, Commissioner Vestager argued that competition policy did not have all the answers, but that she was unsure whether what was needed was to reinterpret existing rules or add new rules (European Commission 2019). In December 2020, official Commission documentation asserted that: ‘competition alone cannot address all the systemic problems that may arise in the platform economy. Based on the single market logic, additional rules may be needed to ensure contestability, fairness and innovation, and the possibility of market entry, as well as public interests that go beyond competition or economic considerations’ (European Commission , 2021a). This shift in approach was the basis for a legislative proposal, the Digital Markets Act (DMA). This draft legislation introduced a form of ex-ante regulation to supplement the Commission’s ex-post competition enforcement.

The puzzle that this article seeks to explain is what led to ‘renewed integration’ (see Raudla and Spendzharova, this volume) in this policy area?; why did the Commission’s position alter?; and why was the DMA proposed? More precisely, what were the drivers of policy change that led the European Commission to seek to supplement its traditional ex-post competition enforcement, directed at online platforms, with ex-ante regulation? To answer this question the article relies on an analysis of recent documentary sources largely drafted between 2018 and 2021, including legal texts and official documents of the EU, political speeches by EU Commissioners and their senior officials, reports and studies by stakeholders and experts, and commentary from reputable print and online media sources, primarily the Financial Times and Politico Europe. Following the hypotheses set out in the Introduction to this Special Issue (Raudla and Spendzharova, this volume), we are especially interested in what this research can tell us about the role of the Commission (H1) as a key protagonist within the multi-level governance architecture, and the interactions of public and private actors (H2) in the policy process.

Informed by policy change theories, we investigate four potential drivers of change which together help to explain the Commission’s decision to promote a new regulatory approach, thereby furthering integration in this policy area: (1) the political, industrial and digital contexts behind the proposed legislation; (2) the Commission’s own experience of using competition law to address the market power of online platforms; (3) the ideas and recommendations of experts; and (4) the influence and arguments of advocates and lobbyists. We find that policy change resulted from the opening of internal and external policy windows which then profited from both a consensus in support of ex-ante regulation among advisors and commentators, and member state opposition to other reform options. This consensus informed the approach taken by Commission decision-makers. By contrast the large online platforms – despite their reputation as lobbying power-houses – were unable to defend their interests and prevent the proposal of a new draft regulation.

The article is organised as follows: the first section identifies areas for investigation by drawing from policy change theories; the second section sets out the background to and political context of the legislative proposal. Section three reviews the EU experience of regulating online platforms using traditional competition instruments. Section four assesses both the outcome of the public consultation and other sources of expert advice. Section five reviews the extent and nature of stakeholder lobbying and advocacy. Section six provides a discussion of the findings. The article ends by restating the argument and drawing out some wider implications.

The Drivers of Policy Change

We start by defining policy narrowly, as ‘regulatory approach’, with policy change understood to be the introduction of a new regulatory approach and a renewed process of market integration. Our drivers of policy change are relatively broad factors which involve some form of impetus to change. And our policy subsystem, which draws a virtual boundary around the policy and the core institutional actors involved in it (differentiating ‘internal’ from ‘external’ drivers) is the field of digital competition policy. How, then, might public policy theories help to identify possible drivers of change? We draw on four theoretical approaches which helps inform our empirical investigation: an historical institutionalist approach; a theory of endogenous change; theories of expertise; and theories of interest group influence.

First, while historical institutionalism emphasises the importance of path-dependent institutional continuities, the theory of punctuated equilibrium highlights the significance for policy of breaks with those past trajectories (Baumgartner and Jones 1993). This theory normally refers to the way in which dramatic shocks to the system, such as those that arise from crises or from the introduction of technological innovation, can lead to policy change. The emphasis on technology within this approach speaks to our interest in the regulation of relatively new digital markets (Cartwright 2019).

Second, a more recent approach theorises ‘gradual institutional change’ (Mahoney and Thelen 2010, 4–7; see Eckert, this volume). This approach draws our attention to the internal factors (and actors) that might be as, if not more, important than external drivers of change. Mahoney and Thelen (Mahoney and Thelen 2010, 10) are particularly interested in the discretion that policy actors have over agreed public policy. They argue that ‘compliance [with rules] is inherently complicated by the fact that rules can never be precise enough to cover the complexities of all possible real-world situations. When new developments confound rules, existing policies may be changed to accommodate the new reality’(Mahoney and Thelen 2010, 11). This might entail rule creation or some other form of rule adaptation.

Third, theories of expertise come in various shapes and forms. Many of these theories seek to present policy change as a process that does not only concern conflict resolution, but that involves knowledge, beliefs, ideas, learning and debate (Radaelli 1999, 758). In his conceptualisation of an epistemic community, for example, Haas (1992) points to the role of experts and the importance of knowledge within public policy, acknowledging conditions under which knowledge or expertise might drive policy change. Sabatier and Jenkins-Smith (1993) focus on knowledge utilization and learning by communities sharing common beliefs and values; while other studies are interested in the role of policy entrepreneurs in brokering knowledge – including the European Commission (Zeilinger 2021).

Finally, theories of interest group influence seek to explain the effect of lobbying on policy outcomes, including on policy change. This literature often acknowledges the business bias found in interest group politics, which arises from the ability of corporations to mobilise effectively. One important approach argues that this bias results from the superior resources that are at the disposal of the business community (Dür and de Bièvre 2007). Yet, Falkner (2007) finds that even well-endowed businesses do not always win out, especially where there is conflict among business interests (Dür and Mateo 2016, 9).

Inspired by these theoretical insights, we first draw from punctuated equilibrium to address the wider external context–political, industrial and digital–which might have driven policy change; second, we privilege endogenous change by focusing on the European Commission’s experience of dealing with online platforms; third, we are informed by theories of expertise to reflect on the significance of knowledge and the inputs of experts as drivers of policy change; and fourth, we draw on theories of interest group influence to judge how lobbying might have informed the proposed policy.

Context and Background: Politics, Industry and the European Commission’s ‘Digital Turn’

As European governments began to regulate digital markets and services, the European Commission raised concerns as to the impact this might have on the SEM. The Commission response was the launch in 2014 of its initiative to create a Digital Single Market (DSM). Under Jean-Claude Juncker’s Commission presidency, the EU approved 28 separate legal acts over the next five years (European Commission , 2021c). Legislation which came into force in July 2020, the Platform to Business (P2B) Regulation, with its focus on fairness and transparency, formed a first step in protecting small businesses in their dealings with digital giants (European Commission , 2021b, , 2021d).

When Ursula von der Leyen took over as Commission President in late 2019, Margrethe Vestager was able to expand her portfolio from competition alone (in the Juncker Commission) to both competition and digital economy and society, signalling the significance and interplay of these two policy areas. The theme of ‘A Europe Fit for the Digital Age’ became one of the priorities of the new Commission (European Commission , 2021e) and Vestager was appointed Executive Vice-President to lead on this theme, with a supporting role for the Single Market Commissioner, Thierry Breton. In February 2020 the Commission launched its Digital Strategy, and a month later it announced a new industrial policy to support the twin transition to a green and digital economy, and to help promote ‘open strategic autonomy’ in the EU. The latter sought to build both an open economy and trade sovereignty, allowing the EU to protect itself against unfair and abusive practices by non-EU states and companies. US commentators saw this initiative as a revival of European interventionism and protectionism (Barshefsky 2020), which is particularly pertinent from a digital market perspective given the dominance of the US/Silicon Valley within the platform economy.

The Commission’s Digital Strategy sought to translate the broad goals of the DSM into specific targets (European Commission 2021f) to promote, amongst other things, a ‘fair and competitive digital economy’ (European Commission , 2021a). It was unclear initially what this would mean for digital market regulation as it related to the competition rules. The Commission had already started to explore the option of a New Competition Tool (NCT), which would allow it to investigate structural competition issues across markets (and not only in digital markets). While the original plan was for the Commission to be able to intervene and impose remedies in the case of certain market structures or market failures, without the need to prove an infringement of the competition rules, this market investigation instrument was met with a lukewarm response in the 2020 public consultation (European Commission , 2020c) including from certain large member states and their competition authorities (see, for example, CNMC 2020). The NCT was eventually folded into the proposed Digital Markets Act and watered down in the process. Alongside the NCT initiative ran the Digital Services package covering competition issues and content moderation (European Commission , 2021g). The DSA package was later presented as two separate legislative acts, the Digital Markets Act and the Digital Services Act.

The draft legislation was finalised on 15 December 2020. The proposal had had a difficult ride through the Commission’s Regulatory Scrutiny Board, where it had been rejected in early November. The revised version which was approved in mid-December set out prohibitions and obligations. It targeted large online platforms acting as gate-keepers, who were defined by their role, size and durability in the market (European Commission , 2020a, 9, , 2021g, , 2021h). Thus, the companies targeted by this new legislation would be large players acting as intermediaries between businesses and individual users who operated in several member states. The new Act would regulate: (1) the use of data gathered from businesses hosted by the platforms; (2) interoperability, such as where additional services only worked with the platforms’ systems; and (3) self-preferencing, that is, the treatment of the platforms’ own services more favourably than those of their competitors (European Commission 2020b). It comprised a list of ‘do’s and don’ts’ for online platforms and established a mechanism for ongoing market analysis to cope with the fast-changing situation in this policy area (European Commission , 2021h, , 2021i). It also foresaw large fines and periodic penalty payments for firms failing to comply with the rules. Although not stated explicitly in the text, the DMA proposal set up a regime of ex-ante regulation to create a Europe-wide level playing field for business and consumers involved in digital markets (European Commission , 2021a).

Learning from Experience: The Commission’s Ex-Post Competition Enforcement

The DMA was drafted after more than a decade of Commission competition enforcement in digital markets. Several of the largest cases to hit the headlines in that period were extremely controversial, especially those involving the American multi-national technology company, Google. Google is a company with diverse business interests, though most of its income comes from advertising. Typical of online platforms, Google’s activities involve gate-keeping, which means it can ‘have a major impact on, [and] control the access to digital markets. It can impose take-it-or-leave-it conditions on both […] business users and consumers’ (European Commission , 2021h). These activities have been the target of several European Commission investigations and decisions. The Commission has raised various concerns that include criticism of Google’s search algorithm, which had been used to promote its own products while downgrading the placement of those of its competitors (Google Shopping case; European Commission 2017). It was also found that Google had not applied its own system of penalties on the placement of products, to its own products. In the Google Adsense case, the tech giant was accused of treating its partners poorly by demanding exclusivity (i.e. no engagement with competitors) as the price of working with it (Vestager 2016, 2018; European Commission 2019). In all, in the decade after 2010, there were three large competition cases involving Google. These cases attracted fines of over €8 billion, the largest in the history of European competition policy (Espinoza 2021).

Critics have nevertheless questioned the effectiveness of the Commission’s competition enforcement approach, its remedies, and the process by which dominance in the relevant market has been analysed. Moreover, decision-making has often been very slow (Tirole 2019). The investigation in the Google Shopping case, which began in 2010, took five years before a Statement of Objections was issued, and even then, a final Decision was not forthcoming until June 2017, seven years after the opening of the case. Delays are perhaps understandable given both the technical complexity of these cases and the Commission’s limited experience in dealing with online platforms. Yet, by the time cases such as these are concluded, a great deal of harm may already have been done.

To make matters worse, Commission Decisions may have little effect on the conduct of online platforms, with the practices condemned (and prohibited) continuing with impunity, and appeals dragging on long after Decisions have been taken. Competitors have argued, for example, that Google continued to use its advertisements to favour its own service even after the case was formally resolved. Vestager acknowledged that conduct such as this, if left unchallenged, could undermine the credibility of the Commission (Vestager 2018).

The primary challenge for the Commission in addressing anti-competitive practices by online platforms is that it is only able to respond after the fact (that is, ex-post). This is not so problematic where markets are ‘well-ordered’ (Crémer, de Montjoye, and Schweitzer 2019, 125) and businesses organise their activities to avoid breaching the competition rules. Perhaps because digital markets are relatively immature and the Commission has not yet had built up enough of a track record, online platforms have tended to resist Commission control (Tirole 2019). Indeed, one of the functions of the new draft legislation would be to help nudge gatekeepers into cooperating with the Commission (De Streel et al. 2021, 26). Moreover, the Commission has acknowledged that problems associated with gatekeepers cannot always be resolved by Article 102 TFEU, given that a gatekeeper is not necessarily a dominant player (European Commission , 2020a).

This is not to claim that the Commission has been lacking in courage as it attempts to rein in these powerful companies. Whereas Competition Commissioners have come under substantial pressure to go gently on corporate actors that generate such wealth, they have not shied away from taking difficult decisions even when pressured to do otherwise (ALTER-EU 2020). Yet, the process of enforcing competition policy has been combative and adversarial, as well as having taken up inordinate amounts of DG Competition’s time (Espinoza 2020). While Commission efforts to tackle online platforms using traditional competition policy tools have not been entirely without their successes, it has become increasing clear to the Commission that it might be worth exploring the possibility of a new approach to competition regulation in this sector.

In this regard, the Commission was able to take inspiration from the UK Competition and Markets Authority’s market investigation tool. The latter informed the proposed NCT element of the draft legislation. Moreover, Vice-President Vestager was able to draw on the Commission’s own experience of ex ante regulation in support of its competition policy in other sectors. As she put it: this is ‘the same as we have been doing in banking, in telecoms, in energy – to realise that antitrust will have to work hand in hand with regulation. So that we have a complete set of tools’ (European Commission 2020b). Thus, the existence of these earlier examples offered the Commission a ready-made reform option.

Expert Advice: Ideas and Recommendations

With the window to reform open, the Commission was able to draw extensively on the advice of experts. While stakeholder arguments tended to rest on interests or on conceptions of the general good, expert advice relied more often on a technocratic logic. One source of information that cut across this distinction, however, was public consultation. Indeed the Commission ran two public consultations between June and September 2020, one on the NCT and the other on the Digital Services package. These consultations provided strong support for Commission regulation, necessary because of structural problems that could not be addressed by the competition rules. The Commission reported that although there were mixed opinions on the stand-alone NCT, all non-governmental organisations (NGOs) and public organisations such as trade unions, supported regulation, as did most business respondents (European Commission , 2020a, 7–8). On matters of substance most contributions accepted that the legislation should cover prohibitions and obligations for gatekeeper platforms and that remedies should be procedural, though there were differing opinions on the definition of a gatekeeper (European Commission , 2020a, 8).

The Commission actively sought out expert advice, as in January 2019 when DG Competition hosted a full-day conference: ‘Shaping competition policy in the era of digitisation’ (European Commission 2018, 2019). To signify its importance, the conference was introduced by Commissioner Vestager and concluded with a speech by DG Competition’s Director-General, Johannes Laitenberger (2019a). Many of the contributors pointed to the slow workings of competition policy while calling for better tools and the need to focus on systemic abuse across the wider digital environment, rather than simply picking on specific instances of anti-competitive practice. Participants highlighted the limits of traditional competition policy enforcement when faced with the market power and often recalcitrant conduct of online platforms. Out of around 100 written responses, ‘[t]he majority of the submissions laid out arguments for a pro-active competition enforcement and a competition-driven regulation’ (Laitenberger 2019b; see also European Commission , 2021j).

The Commission also sponsored its own expert report, ‘Competition policy for the digital era’, to explore how digital policy should evolve to promote pro-consumer innovation (Crémer, de Montjoye, and Schweitzer 2019, 2). Alongside a large number of technical recommendations, the report analysed the role of consumer welfare in the policy and suggested that competition policy should develop a stronger enforcement regime to better serve European citizens. This would help to identify and counter the anti-competitive strategies of online platforms, such as their zero-price policies, their market power and market dominance, as well as issues of transparency and data sharing (Crémer, de Montjoye, and Schweitzer 2019, 40–46, 48–50, 63–69). The report concluded by recognising that existing case law pointed to the need to ‘adjust the analytical tools, methodologies and theories of harm to better fit the new market reality’ (Crémer, de Montjoye, and Schweitzer 2019, 125). Their broad recommendation was that the Commission should devise new general legal rules to meet the new challenges of the digital economy; to forge a new kind of regulatory regime (Crémer, de Montjoye, and Schweitzer 2019, 126–127).

Alongside the expert report, the Commission was also able to draw, directly and indirectly, on in-house Commission bodies and groups for analysis and advice (European Commission , 2021g). The lead DGs on the dossier (namely DG Competition, DG Connect and DG Grow) commissioned their own support studies from external teams. For example, DG Connect supported a study on ‘Platforms with Significant Network Effects Acting as Gatekeeper’ from three consultancies, ICF, WiK and CEPS; and DG Competition commissioned three members of their Economic Advisory Group on Competition Policy to prepare an economic evaluation of the NCT. The Joint Research Centre (JRC), which is the Commission’s science and knowledge service, offered a range of analytical studies (European Commission , 2020a, 9) and a JRC high-level expert panel on platform issues was set up to offer advice to DG Connect. The 15-member expert group of the EU Observatory for the Online Platform Economy and the e-commerce expert group also provided relevant analyses (European Commission , 2020a, 9, , 2021g). While the content of these reports covered a wide range of issues, they were either explicitly or implicitly supportive of the introduction of some form of ex-ante regulation.

Expert advice also came from external sources (European Commission , 2021g). Of most significance were reports published by national governments and agencies, both EU and non-EU. The majority of non-EU national reports originated in the UK (such as the 2019 Furman Report) and the US, though Australian and Japanese contributions were also referenced by the Commission. Of the EU reports mentioned, and there were many of them, most were country-specific, but some were joint papers including a memorandum by the Belgian, Dutch and Luxembourg competition authorities and a paper from the German, French and Portuguese Economic Affairs Ministries on modernising competition policy (see European Commission , 2021g). The Commission also sought information on national positions through a targeted consultation within the framework of the multilateral European Competition Network (ECN) and bilaterally with National Competition Authorities (NCAs) across the European Economic Area (EEA) (European Commission , 2020a, 7, 9). In the DMA proposal, the Commission notes the support of the member states for the proposal (European Commission , 2020a, 8; see also Stolton 2020), though the Irish government (Government of Ireland 2020) and the Nordic states (Nordic Competition Authorities 2020) expressed some reservations.

While there was only one international organisation (OECD) report, dating from 2018, mentioned by the Commission as a source of expertise (OECD 2018), and only one consumer group report, from BEUC (2019), the Commission drew upon five reports from the think tank, Centre on Regulation in Europe (CERRE), a body dedicated to better regulation. Independent academic economists, lawyers and business experts were also consulted, including Massimo Motta, Luis Cabral, Richard Whish, Martin Peitz and Heike Schweitzer, to name but a few. Most individual experts contributed to the reports or initiatives mentioned above, though a small number were listed as having presented to the Commission in a workshop organised in January 2020. Most experts were European, which is significant if we consider that most European economists have been shown to favour tougher regulation of online platforms (Vaitilingam 2020).

Advocacy and Lobbying: Divide and Rule

It should come as no surprise that lobbyists have been active on the issue of digital regulation. Civil society groups mainly argued a strongly pro-regulation line, though some concerns were raised from within the legal community that regulation would substantially increase the Commission’s powers of intervention (Lübbig, Jensen, and Goyder 2020). NGO influence was modest, however. Only one civil society group, BEUC – the European Consumer Organisation – can be found in the list of top lobbyists on the DMA and DSA. Although some other NGOs, such as the European Digital Rights Network (EDRi), Reporters Sans Frontiers and the Electronic Frontier Foundation, were active, they appear to have gained limited access to EU policy-makers (CEO 2020).

By contrast, Big Tech have sought to defend their interests with vigour, placing the Commission under immense pressure. We know from unpacking the Commission’s 4,342 lobby meetings between 2017 and 2021 that large corporations, particularly those from Silicon Valley, had become more dominant in lobbying on digital issues (Kergueno 2018). Big Tech companies ranked highly in terms of their attendance at high-level Commission meetings (Pearson 2021). In a 12-month period from late 2019 and late 2020, 158 meetings ‘were logged as including discussions on the DMA or DSA’ (CEO 2020), and 13 out of 103 organisations, mostly companies and lobby groups, held three or more meetings with the Commission. Google was ahead of the rest when it came to lobby meetings, including those held with Commissioners and Directors-General, while Microsoft and Facebook were close behind, with Apple and Amazon ranking lower (CEO 2020). Other platforms also lobbied on digital market regulation, including Trivago, Allegro, Zalando and Airbnb (Hurley 2020a, 2020b). Digital content moderation and competition were the most common issues lobbied on by the Big Five tech companies in the Commission in 2020, ahead of COVID-19, climate change and artificial intelligence (Clarke 2021).

Online platforms spent a considerable amount of money addressing these issues. According to the European Transparency Register, the Big Five’s EU lobbying budgets amounted to approximately €20 million in 2019 (CEO 2020; see also Clarke 2021). Meanwhile, ‘Google’s in-house Brussels lobbying budget [had] increased by 360% since 2014ʹ (Pearson 2021). In all the Big Five had the biggest lobbying budgets in the EU (CEO 2020). With the platforms’ deep pockets in mind, civil society groups and digital rights campaigners also raised concerns about the extent and limited transparency of corporate lobbying on digital matters. In an open letter in 2020, the ALTER-EU Alliance, a group campaigning for more ethical lobbying, wrote to EU Commissioners Vestager and Breton about Google’s opaque lobbying practices and called for a stronger commitment to lobby transparency. However, as late as 2 December 2020, Vice-President Vestager and Commissioner Breton were discussing the proposed legislation with the Big Tech companies without consumer advocates and civil society organizations present. And the almost exclusive use of virtual channels after the onset of the COVID-19 pandemic from March 2020 ‘further reduced the transparency of an already opaque process closed off to the public’ (Scott and Kayali 2020).

While lobbying activity initially addressed both competition and content moderation often in the same meetings, by mid-2020 these lobbyists were almost exclusively focused on competition (Scott and Kayali 2020). Yet, beyond providing a richer information base for the Commission, there is no evidence that corporate lobbying affected the decision to introduce the ex-ante regulation of online platforms. After all, regulation was proposed. And even on the substance of regulation and most notably on the subjects about which the large online platforms were most exercised, such as the new rules applying exclusively to larger platforms (Google 2021), a one-size-fits-all approach undermining consumer protection, leading to less choice across Europe, and self-executing rules such as prohibition lists being too inflexible (Apple 2021; Facebook 2021), the Commission did not alter its position (CEO 2020).

Just as the Commission’s case-orientated involvement with the large online platforms was confrontational, so lobbying became at times rather aggressive in character. In October 2020 a leaked document exposed Google’s 60-day strategy to lobby the EU institutions and rally American allies. Google’s intention was to frame the proposed legislation as a threat to trans-Atlantic relations and to the European economy (Hurley 2020a) and to enlist US officials in their fight against the new European policy, to erode support and sow divisions within the Commission, whilst mobilising third parties such as think-tanks and academics (Satariano and Stevis-Gridneff 2020; CEO 2020). Google also intended to encourage smaller platforms such as Booking.com to join their campaign by arguing that these smaller European firms might also be hurt by the new legislation. While there was no official Commission response to Google’s leaked lobbying strategy, the publicity around this leak is likely to have weakened Google’s position (Scott and Kayali 2020).

In any case, not all online platforms were hostile to the Commission’s proposals. Indeed, the business users of the large platforms understood how they might benefit from the proposed legislation. As online platforms have different business models and antitrust priorities (Geradin 2018, 2–5), they quite often lobby against each other. This is especially the case for smaller platforms. However, even among the Big Five, Facebook was reported to have lobbied against ‘unfair trading practices’ by Apple’s app store (Scott and Kayali 2020), leading Facebook to state that the proposed legislation ‘was on the right track’ (Kelion 2020). Booking.com’s CEO also distanced himself from Google’s lobbying, arguing that Booking.com have smaller market shares and should not be subject to the policy in the same way (Scott and Kayali 2020). The tech environment was far from homogeneous on digital market issues.

Discussion

The narrative account in this article has focused on four factors that were likely to have influenced the Commission’s decision to regulate digital markets. How, then, did these factors matter? First, we found that the political, industrial and digital context set the scene for a change of policy. Concerns about the fragmentation of the Single Market, on the back of member state interest in regulating online platforms at a national level, was important in driving policy change. The fact that the member states seemed supportive of further digital regulation was relevant too; as was the more general emerging preference in the Commission for a more interventionist industrial policy, particularly where such a policy would be directed at anti-competitive conduct by non-EU actors within the European market. These developments were also taking place at a time when the EU was recognising the importance of digital markets to future European competitiveness, which had already begun to be addressed more widely in the context of the DSM. Second, the Commission’s experience of applying the competition rules in digital markets had not been an entirely happy one. Enforcement was complex, slow and not always effective, even where formal decisions had been taken. And while this alone was not a sufficient driver of change, it mattered that the Commission was itself having doubts about its own capacity to regulate online platforms using competition policy alone. Third, while experts differed in the recommendations they proposed (that is. the form regulation should take), a large majority supported the strengthening of the competition rules. To the extent that we can talk of the emergence of an epistemic consensus around the need for a new regulatory approach to digital markets. Nay-sayers were very much in the minority. This was not a case where the Commission selected experts to support its preferred agenda, as the consensus amongst expert commentators was as solid externally as internally. Finally, while lobbying sought to shape Commission decision-making in support of and in opposition to regulation, the focused nature of the legislation (targeting only Big Tech companies) created division amongst the online platforms, and the rather aggressive nature of corporate lobbying on the issue of regulation, which raised concerns of privileged access, did not have much of an effect. Despite claims of the influence of corporate lobbying within the Commission, the absence of information dependency on the part of the Commission (Coen 2009) and divided interests in the sector, militated against the power of the large online platforms.

Figure 1. Policy change: from competition enforcement to the Digital Markets Act.

Beyond the relevance of these four factors, we also note that the Commission’s impulse in responding to doubts about the effectiveness of competition enforcement in regulating digital markets had been to reform competition policy itself, that is, to keep control of the issue within DG Competition. This met with a lukewarm reception and did not find favour across the EU’s member states. Thus, this idea – embodied in the NCT – was ultimately subsumed within the DMA. In that sense, the drafting of a new regulation on digital markets was a second-best solution for the Commission. While regulation would strengthen the Commission’s hand in controlling the conduct of digital actors, in the world of internal Commission politics it might also be understood as a loss of control over digital market regulation for DG Competition. At the same time, as well as relying on the advice of experts, earlier examples of regulation of a similar kind in utilities markets proved an example of how regulation might work in practice (notwithstanding sectoral differences) and reassured the Commission that a competition enforcement plus regulation approach could work.

In sum, we find that contextual factors, the Commission’s own experience of competition enforcement in digital markets, and expert advice and commentary each play a part in accounting for the way in which the Commission’s position on digital markets regulation changed. Lobbying and advocacy are relevant in that they did not prevent the proposed regulation. A lack of enthusiasm for a wider competition reform, and the existence of a pre-existing model of regulation that worked alongside competition enforcement also helped to nuance the argument. But how did these factors work together to produce this outcome ? The point of departure is the Commission’s case experience over a period of more than a decade and its growing dissatisfaction with the limited impact of these cases. This opened an internal window of opportunity for policy reform – though the nature of that reform was yet to be determined. At the same time the external context was shifting, producing an external window of opportunity. National governments were planning their own responses to online platforms, threatening a fragmentation of the SEM in the process. At the same time the Commission was acknowledging the power that non-EU states and companies were already holding over the EU; and was keen to develop a home-grown digital economy and society to promote European competitiveness, based on a single market logic. This external window of opportunity together with the internal window opened the way for expert contributions on the form the EU and Commission’s response should take. As we saw, the Commission’s initial efforts were directed at competition reform. It was only after the Commission became aware of opposition the broader reform on the one hand and expert advice pointing to a different solution (EU regulation) on the other, that the institution shifted its position. This was not, however, the paradigm shift that some claimed it to be at the time. After all there already existed EU precedents in other sectors that prepared the ground for kind of regulatory approach (see Figure 1).

Conclusion

Why, after emphasising the merits of traditional competition enforcement as a way of controlling the market power of online platforms, did the European Commission choose to supplement its ex-post approach to digital competition with a new ex-ante regulatory regime? We have answered this question by identifying four drivers of policy change drawn from the policy theory literature, namely: the political, industrial and digital contexts, and the policies they inspired, which sought to address concerns regarding the fragmentation of the Single Market. This comprised a new industrial strategy for Europe in which the themes of digital economy and society became high-level Commission priorities; the Commission’s own troubled experience of using traditional competition tools to regulate online platforms; the ideas and knowledge presented by experts and stakeholders working within the specialised area of digital competition regulation, including those commissioned by the Commission itself; and the extent and influence of stakeholder lobbying. In addition, relevant to this explanation is the failure of a competing reform option and the existence of regulatory precedents that could act as a model for digital markets.

We found that policy change resulted from the opening of an internal and external window of opportunity provoked by the Commission’s case experience and the emergence of a new policy context. This window identified a set of problems, the solution to which was found in a consensus in support of ex-ante regulation amongst external and in-house expert advisors and commentators. This consensus informed the Commission’s thinking on the issue of digital market regulation and ultimately determined the approach taken by Commission decision-makers. The large online platforms – despite their reputation as lobbying power-houses – were unable to defend their interests and prevent the proposed regulation.

What then are the wider implications of this research? Returning to the two hypotheses that have informed this research (see Raudla and Spendzharova, this volume), what is first notable about this explanation of policy change is the central role of the European Commission, and the complexity of its role. In each of the four sections of this paper we see the Commission performing a different part – as agenda-setter on matters industrial, competition and digital; as a learning organisation, both learning-by-doing and learning from other actors and models; as a technocratic repository for expertise and information, as well as an active seeker of knowledge; and as a quasi-state actor subject to, but also able to resist, aggressive lobbying. In a sense, this case of policy change shows the Commission to be a hybrid entity embodying multiple functions (see, by way of comparison Stephenson, this volume and Cino-Pagliarello, this volume) and an organisation which uses the resources, capacity and discretion at its disposal to defend and push for further Single Market integration in which the Commission continues to play a leading role.

Second, we do not see any formalised co-regulation in this policy area as the DMA gives enforcement powers exclusively to the Commission. This means that the public-private impact is more nuanced than in other EU cases (see Raudla and Spendzharova, this volume). Yet, this research offers an interesting study of the relative roles of public and private actors in the EU decision-making process and the importance of information and expertise in the public-private relationship. While critics of the EU might have expected European institutions to be captured by large corporate interests, this case suggests that this is less likely where the Commission has access to extensive expertise beyond the lobbying environment and where that expertise points strongly in a particular direction. It helped too that the corporate lobbyists were divided. As such this article supports the view that we should not always assume the European Commission to be in thrall to large and powerful corporate lobbyists.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Education, Audiovisual and Cultural Executive Agency of the European Union, [612044-EPP-1-2019-1-NL-EPPJMO-NETWORK].