We are delighted to present this first edition of Digital Competition in the Lexology Getting The Deal Through series, and we hope that it will be a useful resource for businesses and their advisers in successfully navigating the myriad complex issues that can arise when applying competition law in digital markets. We would like to thank all those who have contributed their valuable insights.
The term ‘digital markets’ is generally used as a shorthand for all business activities that make use of digital technology, either to support existing business ventures or to grow new ones. Increasingly, this can encompass almost any industry sector, as a result of the rapid growth in, and importance of, digitalisation in our everyday lives, and the exponential growth of e-commerce over the past decade. The growth of the digital economy and related technological innovation has been a major driver of increased competition across a wide range of markets, and has in many cases resulted in significant benefits for consumers in terms of lower prices and better quality goods and services, as well as greater innovation. However, many digital markets also have features that make it more challenging to apply competition law to ensure that they remain competitive as they develop further, particularly those involving large multisided digital platforms.
For example, digital markets are often dynamic and fast-moving, and may also be ‘zero price’, at least on one side, for instance, where a digital platform user receives services in exchange for data. This can make delineation of the relevant market definition – traditionally a fundamental part of both antitrust and merger analysis – more challenging. The multisided nature of many digital markets, and the impact of network effects and ‘big data’, can also complicate the analysis, alongside other issues, such as increased use of algorithms, and the growth of ‘digital ecosystems’ through vertical integration.
Notwithstanding these difficulties, competition authorities worldwide have identified enforcement of competition law in digital markets as a priority, given the importance to consumers of ensuring digital markets remain competitive. All leading competition authorities are proactively pursuing both antitrust and merger investigations involving digital markets, and this trend shows no sign of abating anytime soon.
We highlight below some of the current ‘hot topics’ and key issues that are explored further from the perspective of individual competition authorities in the jurisdiction chapters that follow. However, we first wish to address briefly a fundamental question that all leading competition authorities are currently grappling with: are existing competition law frameworks fit for purpose in the context of digital markets, or are new tools needed (whether by way of new legislation, soft law or other options)?
Are existing competition law frameworks ‘fit for purpose’ in the context of digital markets?
To help determine the extent to which existing competition law frameworks may need to be adapted to the increasingly digital era, a flurry of reports have been published in the past few years that have sought to identify the key challenges in applying competition law in digital markets and to improve competition authorities’ understanding of the complex issues that can arise, as well as exploring potential ways in which to address them. The findings of these reports, including the UK’s Furman Report (March 2019), the EU’s Special Advisors’ Report (April 2019) and the report of the German Commission Competition Law 4.0 (September 2019), are all explored in more detail in the relevant individual jurisdiction chapters of this guide.
In addition, many competition authorities have set up dedicated units focusing on digital markets. For example, the UK’s Competition and Markets Authority (CMA) formally launched a new Digital Markets Taskforce on 1 July 2020, and it also has a specialist Data, Technology and Analytics unit. The CMA’s recent market study into online platforms and digital advertising (completed in July 2020) also backed the flagship proposal of the earlier Furman Report that a new dedicated body, the Digital Markets Unit, should be created to exercise oversight with new powers, including the ability to enforce a code of conduct governing the behaviour of digital platforms and significant intervention powers to promote competition. In the United States, the Federal Trade Commission has a specialist unit, the Technology and Enforcement Division, which monitors and investigates potential anticompetitive conduct and transactions in digital markets. Similarly, within the Australian Competition and Consumer Commission (ACCC), there is a Digital Platforms Branch, responsible for the ongoing scrutiny of digital platform markets, including conducting relevant inquiries as well as competition and consumer law enforcement cases. China appears to represent a notable exception to this trend, with no dedicated digital markets unit within the State Administration for Market Regulation at the time of writing, although the focus of the Chinese authorities on digital markets is increasing.
The global consensus to date appears to be that existing competition law frameworks generally provide a sound and sufficiently flexible basis for protecting competition in the digital era, albeit with some complications in areas such as market definition and assessment of market power (mainly due to the fast-moving and dynamic nature of many digital markets), and a need to develop new theories of harm or types of abuse within existing frameworks (such as self-preferencing by dominant digital platforms).
However, some jurisdictions have recently been more vocal than others in proposing a degree of ‘rethinking’ of the traditional tools of analysis and enforcement, and in certain cases, specific reforms. For example, the European Commission is currently considering amending the Vertical Agreements Block Exemption to reflect the impact of online sales and online platforms on distribution models, as well as revising its 1997 Market Definition Notice to address the intricacies of digital markets. In December 2020, it also announced far-reaching proposals for regulation of the digital sector in the form of a Digital Services Act and a Digital Markets Act. The first of these contains proposed new rules, including on illegal content and transparency, applicable to all online platforms, with additional rules for very large platforms. The second proposes ex-ante regulation of digital ‘gatekeepers’ providing ‘core platform services’. In Germany, the proposed 10th revision of the Act Against Restraints of Competition (ARC) – which is advertised as an ARC Digitalization Act – will equip the Federal Cartel Office with new tools in the area of unilateral conduct, with particular relevance for companies active in the digital sector. Addressing concerns related to the market power of digital platforms is also a key focus in Australia, where the recommendations of the recent Digital Platform Inquiry Report include a proposal to require large digital platform businesses (including, but not necessarily limited to, Google and Facebook) to notify the ACCC of any proposed acquisition potentially impacting competition in Australia. A similar recommendation was also made by the Furman Report in the UK.
Historically, most competition authorities have been understandably nervous about the possibility of the over-enforcement of competition law having a negative effect on nascent digital markets. This has resulted in limited intervention, particularly in relation to mergers. As highlighted in the UK Furman Report, between 2009 and 2019, the five largest firms active in digital markets made over 400 acquisitions globally, none of which was blocked, and very few of which had conditions attached to approval, or were even investigated by competition authorities, in the UK or elsewhere. However, there are signs that the tide is turning, and that the greater concern is now that under-enforcement may lead to worse outcomes overall (even taking into account the risk of ‘false positives’, eg, blocking a merger that should have been allowed). For example, the UK CMA has expressly acknowledged that earlier UK decisions, such as the unconditional clearance of Facebook’s acquisition of Instagram, did not fully consider evidence as to how the relevant markets were likely to evolve in the future, and a review of previous merger decisions in digital markets commissioned by the CMA concluded that this may have represented a ‘missed opportunity’ for the emergence of a credible challenger. The CMA has indicated that it intends to take a more interventionist approach going forward, and a similar policy shift is also noticeable from other authorities. For example, the ACCC has demonstrated a clear recent focus on strategic acquisitions in digital markets, with its proactive review of Facebook’s proposed acquisition of Giphy (also now under investigation in the UK) and its ongoing review of Google’s proposed acquisition of Fitbit (having rejected undertakings offered by Google in December 2020).
Key issues arising in relation to anticompetitive agreements
Anticompetitive agreements between competitors tend to take fundamentally similar forms in digital markets to non-digital markets, and in many cases, it will be relatively straightforward to apply the existing competition law framework. However, there are nonetheless some interesting challenges, including for example ‘hub-and-spoke’ information exchange concerns (as illustrated by the investigation by the UK Office of Fair Trading, predecessor of the CMA, into indirect information exchanges between private motor vehicle insurers using a specialist market analysis software tool) and information collection issues (as illustrated by the ongoing European Commission investigation into Amazon’s use of data collated from independent third-party sellers that use its Marketplace platform).
One other particularly interesting issue that arises in this context is the potential role of algorithms in facilitating collusion between competitors (as illustrated by the 2016 Posters case in the UK, and the 2018 Consumer Electronics case in the EU, both discussed in more detail in the relevant jurisdiction chapters). To date, algorithms appear to have been used simply as a means of implementing and monitoring a fairly ‘classic’ price-fixing agreement, but a number of competition authorities have also started to consider the potential implications of ‘tacit’ collusion (ie, algorithms ‘learning’ to fix prices by themselves) for future competition law enforcement. In this regard, both the EU and Australian authorities have made it clear that they intend to take a strict approach: EU Competition Commissioner Margrethe Vestager stated in a 2018 interview that, when building algorithms, companies should ‘remember to send them to law school before they let them loose’, echoing the sentiments of ACCC Chair Rod Sims, who has emphasised that ‘you cannot avoid liability by saying “my robot did it”.’ In a similar vein, the president of Germany’s Federal Cartel Office (FCO) Andreas Mundt stated in March 2020 that ‘companies cannot hide behind an algorithm.’ That said, it remains a ‘live’ question whether there can be said to be an ‘agreement’ for the purposes of applying existing competition law rules where two algorithms coordinate pricing without any human input.
In relation to vertical agreements, competition authorities have unsurprisingly focused heavily on restrictions on online sales, including absolute bans on online sales; restrictions on the use of online platforms or marketplaces; geoblocking or imposing territorial restrictions (a particular focus for the European Commission in light of the EU’s internal market objectives); and restrictions imposed by users on online platform operators, such as selective distribution systems that threaten to delist platforms that do not comply. One further form of restriction on online sales that has attracted considerable attention is most-favoured-nation clauses (MFNs), in particular in the context of restrictions imposed by online platform operators on users, such as price parity clauses, imposed by various hotel booking platforms, which have been the subject of investigation by numerous competition authorities around the world, including the UK, France, Germany and Australia. Enforcement action against the use of MFNs by digital platform operators has generally been more limited in the US compared to Europe and Australia, but the Department of Justice’s case against Apple concerning e-books illustrates that these sorts of restrictions are also certainly ‘on the radar’ in the US (albeit that, as noted in the detailed US chapter, the US courts have generally been less willing to uphold antitrust claims against platform MFNs that are not based on evidence of a horizontal conspiracy).
Issues arising in relation to abuse of dominance or market power
Concerns surrounding the conduct of large digital players, in particular digital platforms, have been at the forefront of the debate as to the suitability of existing competition law frameworks for the digital era. While digital platforms may provide significant benefits to consumers, the ‘winner takes all’ nature of competition in many platform-based markets tends to result in a small number of platforms with a significant degree of market power, with high barriers to entry due to data-driven network effects, economies of scale and scope, and control of user data. The potential for anticompetitive behaviour in such circumstances is clear. However, competition authorities seeking to enforce existing rules prohibiting abuse of market power in digital markets face a number of challenges.
The first of these relates to market definition. This is clearly critical in determining the degree of market power enjoyed by a particular undertaking, which is the starting point for applying competition law rules prohibiting abuse of such power. However, as noted above, many of the features common to digital markets – including their fast-moving and dynamic nature, and the prevalence of multisided and zero-price markets – can often make delineation of the relevant market difficult, at least when using traditional tools of analysis such as the Small but Significant Non-transitory Increase in Prices test. These difficulties were expressly recognised by Margrethe Vestager, the European Commissioner for Competition, when announcing a review of the European Commission’s Market Definition Notice, which was formally launched in April 2020.
Even if the relevant degree of market power can be established, a further key hurdle remains: identifying abuse of that power. In this regard, competition authorities have demonstrated a willingness to be creative within the confines of the existing rules, including identifying new theories of harm based on specific features of digital markets. A good example of this is the German FCO’s 2019 Facebook decision, in which it held that user terms and conditions which (allegedly) are in breach of data protection principles could constitute an abuse of a dominant position under competition law. This decision is currently on appeal, but it is notable that in June 2020 the German Federal Supreme Court overturned a temporary injunction suspending the FCO’s decision, which had been granted by the Higher Regional Court of Düsseldorf (albeit on the basis of a different – less data protection-related – line of reasoning). A separate abuse of dominance investigation against Facebook launched by the FCO in December 2020 further illustrates its appetite for considering novel elements even when relying on a more traditional form of abuse: in the context of alleged tying of Oculus virtual reality products and Facebook’s social network, it appears that the FCO is arguing that Facebook would use a product where it is not dominant (virtual reality) to strengthen its position in a market where it is already dominant (rather than a company dominant in the tying market trying to leverage its dominant position into the tied market). Other potential new theories of harm identified in digital markets include self-preferencing (as illustrated by the European Commission’s Google (Shopping) infringement decision), limiting multihoming, and hampering data portability.
It remains a matter of ongoing debate whether enforcement of ex post competition laws prohibiting abuse of market power (even if adapted for digital markets) are the best way of addressing many of the concerns that arise in relation to digital platforms. Alternative forms of ex ante regulation are currently under consideration in a number of jurisdictions. For example, following a market study in online platforms and digital advertising, the UK CMA has recommended the adoption of new UK legislation to provide additional regulatory controls for online platforms funded by digital advertising. It has also backed the flagship proposal of the earlier Furman Report for the creation of a new dedicated body, the Digital Markets Unit, which would exercise oversight and have numerous powers, including the ability to enforce a code of conduct governing the behaviour of digital platforms and significant intervention powers to promote competition. As already noted above, the European Commission has recently announced a Digital Markets Act that would introduce ex ante regulation of digital ‘gatekeepers’ providing ‘core platform services’ (now under further consideration by the European Council and European Parliament for debate, amendment and adoption into law). The German legislator is also considering a tool addressing conduct by companies that are ‘active to a significant extent’ on multisided markets, which has been identified by the FCO as being of paramount significance for competition. And, in Australia, the government has recently developed its first stage roadmap of regulatory reforms following the ACCC’s recommendations in its Digital Platforms Inquiry Report, including a staged process to reform media regulation towards a platform-neutral regulatory framework.
However, it seems that – for now at least – the more draconian step of seeking to formally break up large global technology companies such as Amazon and Facebook is off the cards. This possibility attracted attention in 2019 as part of the presidential nominee election campaign of US Democratic Senator Elizabeth Warren, but has failed to gain traction elsewhere. EU Commissioner Vestager memorably stated during a visit to Washington in March 2019 that network effects would mean that even if such companies were split up, replacements would be likely to grow again very quickly, such that, like a mythological Hydra, ‘when you chop off the head, there are four more.’
Issues arising in relation to mergers
Merger control regimes tend to be focused on review of transactions which meet certain turnover or asset thresholds. As a result, many mergers in digital markets – where the target company may often be a start-up with low (or even no) turnover and limited assets – are not caught by existing thresholds.
This has led some jurisdictions to consider whether to amend the applicable thresholds to include a ‘size of transaction’ test, following in the footsteps of Germany, which amended its regime in 2017 to include an additional notification requirement wherever the consideration paid exceeds €400 million, irrespective of the turnover of the target business (in addition to existing purely turnover-based notification requirements). However, many jurisdictions, including the EU and the UK, have concluded that existing merger control thresholds do not need to be amended (or at least, that it is too early to do so at the moment). In the EU, it has been suggested that the EUMR referral system (whereby a transaction that does not meet the notification thresholds of the EU Merger Regulation (EUMR) but meets the thresholds of multiple national regimes and may be referred to the European Commission for consideration under the EUMR) may provide an adequate alternative to amending EUMR thresholds, based on cases such as Apple’s acquisition of Shazam and Facebook’s acquisition of WhatsApp. In September 2020, Commissioner Vestager indicated that the EUMR referral mechanism will be further amended by the middle of 2021 to enable even transactions that do not meet national regime notification thresholds to be referred to the European Commission, where competition concerns are identified.
A further concern is that many such acquisitions of small start-up businesses in digital markets are motivated by a desire to remove a potential competitor from the market, often referred to as ‘killer’ acquisitions. These acquisitions often occur at a relatively early stage in the development of the start-up, before it has had a chance to establish itself in the relevant market, such that it can be difficult to demonstrate that the exit of the start-up will be likely to have a significant impact on competition. The EU Special Advisers Report notably recommended that in cases where the acquisition is part of a possible strategy against partial user defection from a digital ecosystem, the burden of proof should be reversed, such that the merging parties bear the burden of showing that adverse effects on competition are offset by pro-competitive efficiencies (while emphasising that this would not be intended to create a presumption against the legality of such mergers).
Against this backdrop, competition authorities have begun to focus more on the impact of a transaction on potential competition in their substantive assessment (as illustrated by the UK CMA’s recent investigation of Amazon’s acquisition of a 16 per cent stake in Deliveroo), and may consider a more dynamic counterfactual. It is anticipated that this may lead to a more interventionist approach going forward, as competition authorities become more concerned about the potential negative consequences of under-enforcement than the risks of over-enforcement (which have historically been of greater concern, as noted above). It is interesting to note in this regard that the recommendations of the UK Furman Report included legislative amendments to allow the CMA to apply a ‘balance of harms’ test, which would take into account the scale as well as the likelihood of harm in merger cases involving loss of potential competition and harm to innovation. A similar approach has been advocated in Australia, where ACCC Chairman Rod Sims has expressed the view that if the prospect that the target will become an effective competitor is small, but the potential increase in competition and consumer welfare is large, greater weight should be put on the potential for competition. However, legislative changes have not yet been made to the applicable substantive test in either jurisdiction.
Finally, as already noted above, a number of jurisdictions, including Australia and the UK, are also considering mandating notification of all acquisitions of proposed acquisitions by certain large digital companies (described in the UK’s Furman Report as those designated with ‘strategic market status’). However, there is considerable debate as to whether this would be an efficient way to address competition concerns in digital markets, and it remains to be seen whether such proposals will be implemented.
Where do we go from here?
It is clear that applying competition law in digital markets will remain a challenge – and a priority – for competition authorities worldwide in the coming months (and indeed years). While considerable progress has been made in terms of improving competition authorities’ understanding of the complex issues that can arise, and starting to identify potential ways of addressing them, many of the proposals discussed in this guide are still at an early stage. It remains to be seen how they will be developed and implemented (if indeed they are implemented), and what the impact will be on companies active in digital markets. Like digital markets themselves, this is a fast-moving area, and we can expect to see significant developments over the next 12 months, both in terms of enforcement by competition authorities and national courts, and new legislation and guidance.
It does, however, seem likely – and indeed desirable – that a global approach will need to be agreed between leading competition authorities on many of the issues highlighted in this guide. Digital markets are international by nature, as recently emphasised by the UK CMA in its response to the European Commission’s consultation on its proposed Digital Services Act package and the ‘new competition tool’ (an idea for an investigation regime similar to the CMA’s market investigations regime that was ultimately not included in the European Commission’s final proposals except in a very slimmed-down version that found its way into the provisions relating to ex ante regulation of gatekeeper digital platforms included in the proposed Digital Markets Act). As such, enforcers should seek as far as possible to adopt a coherent approach across jurisdictions so as to provide clarity for businesses, investors and consumers.