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

Learn More →

Big data and competition analysis under Australian competition law: comeback of the structuralist approach?

Big data and competition analysis under Australian competition law: comeback of the structuralist... Abstract Competition assessment in Australia has traditionally been based on an evaluation of the market structure relying on five factors, namely the degree of market concentration, the height of barriers to entry, the extent of product differentiation, the extent of vertical integration, and the nature of arrangements between firms. These factors, known as the ‘QCMA factors’, are characteristic of competition in the manufacturing industries of the ‘old economy’. Since the ascendancy of Chicago and Post-Chicago School thinking competition analysis in Australia has also taken into consideration non-structural factors. However, in light of the dominance of big tech companies in online markets, the so-called ‘Neo-Brandeisian School’ has advocated focusing on structural elements that are characteristic of online markets. This article examines to what extent the QCMA factors still a suitable structural framework for the assessment of competition in online markets. I. INTRODUCTION Throughout the past 40–50 years we have seen significant changes to market structures, most notably in the area of high-technology industries. ‘Old economy’ manufacturing industries are on the decline, paving the way for a rise in industries in the ‘platform economy’. The creation of digital platforms including search engines, social media and e-commerce platforms, has revolutionized almost everything we do, from the way we interact, to the way we shop, dine, and travel. Around the turn of the millennium some competition scholars and antitrust agencies recognized a distinction between the ‘old economy’ industries in which modern competition legislation was first adopted in industrialized countries and ‘new economy’ high-technology industries that started to make big waves around the late 1970s. For example, the Trade Practices Act 1974 (TPA) [now the Competition and Consumer Act 2010 (CCA)], Australia’s first effective competition legislation, entered into force before the first IBM computer came to market in 1981. Tech giants like IBM and Microsoft dominated during the 1980s and 1990s and eventually became the target of antitrust scrutiny. On 18 May 1998, when the US Department of Justice filed its complaint against Microsoft in the internet browser case, the two tech heavyweights of today, Google and Facebook, were not even founded yet. This essay proposes that the ‘platform economy’ is an even ‘newer’ economy. The defining characteristic of most markets of the digital economy1 largely relates to the role of data which makes these markets distinct from others, and this poses challenges to competition law enforcement. Competition assessment under the CCA has for a long time been based on an evaluation of the market structure. In 1976, Australia’s then Trade Practices Tribunal (now the Competition Tribunal) handed down its seminal decision in Re Queensland Co-Operative Milling Association Ltd (‘QCMA’).2 In that decision the Tribunal stressed that to determine whether a market is operating competitively requires the examination of five elements of market structure. The five factors, which are now referred to as the ‘QCMA factors’,3 can be summarized as: (i) the degree of market concentration; (ii) the height of barriers to entry or expansion; (iii) the extent of product differentiation; (iv) the character of vertical relationships and the extent of vertical integration; and (v) the nature of any formal, stable and fundamental arrangements between firms.4 These five factors are characteristics of competition in the traditional manufacturing industries of the ‘old economy’. A closer look at the digital economy reveals that many high technology-based industries and markets have changed substantially. Nonetheless, the Australian Competition and Consumer Commission (ACCC) in its 2018 ‘Guidelines on misuse of market power’ still refers to the QCMA factors as a relevant checklist for determining competitive constraints.5 The aim of this article is to determine whether the QCMA factors remain an adequate structural framework for determining market power in online markets. The article proceeds as follows: the next section outlines Australia’s legislation and the QCMA case to provide context to the current state of affairs relating to competition analysis in Australia. Section III distinguishes between the different types of the economy and explains why the ‘platform economy’ is distinct. Section IV looks at developments in relation to the assessment of competition in online markets and evaluates the relevance of each of the QCMA factors in the platform economy. Section V concludes. II. THE STRUCTURAL APPROACH FOR THE ASSESSMENT OF COMPETITION UNDER THE CCA The Trade Practices Tribunal’s 1976 decision in QCMA is a seminal one under Australian competition law. The case concerned the proposed acquisition by the applicants, QCMA and Defiance Holdings, of Barnes, a rival flour milling firm in Queensland. The Trade Practices Commission denied the merger, after which the applicants sought a review of this determination from the Trade Practices Tribunal.6 The significance of the case lies in the Tribunal’s discussion of the meaning and purpose of ‘competition’ as well as the meaning of ‘market power’—a term that was not defined under the TPA. The Tribunal found: [w]hether firms compete is very much a matter of the structure of the markets in which they operate. The elements of market structure which we would stress as needing to be scanned in any case are these: the number and size distribution of independent sellers, especially the degree of market concentration; the height of barriers to entry, that is the ease with which new firms may enter and secure a viable market; the extent to which the products of the industry are characterised by extreme product differentiation and sales promotion; the character of ‘vertical relationships’ with customers and with suppliers and the extent of vertical integration; and the nature of any formal, stable and fundamental arrangements between firms which restrict their ability to function as independent entities.7 Market power comes from a lack of effective competitive constraint.8 Scanning these five factors can help to determine whether a market is characterized by effective competition or whether one or more firms possess market power. The Tribunal’s explanation of ‘competition’ in QCMA highlights the structural aspects of competition.9 The QCMA factors form a structural framework for assessing competition on a market.10 This framework stems from the ‘structure-conduct-performance (S-C-P) paradigm’ that was proposed by industrial organization economists in the US in the 1950s and 1960s (also known as the ‘Structuralist School’).11 The drafting of the TPA and Australian competition policy at that time was heavily influenced by US antitrust legislation.12 The S-C-P paradigm implies that the structure of the market, particularly high concentration accompanied by high barriers to entry, dictate that firms in those markets will engage in certain types of conduct which would then lead to poor economic performance.13 Unfavourable market structures including monopolies and oligopolies can use their existing market power to block new entrants;14 and use their greater bargaining power against consumers and suppliers to raise prices and degrade quality while maintaining profits.15 While it was initially assumed that the causal flow running from structure to conduct to performance was unidirectional, it was later realized that there may also be feedback effects under which conduct may influence structure.16 This position was also accepted by the High Court.17 In 1994, on the 20th anniversary of the TPA, Brunt commented that the Australian approach to the assessment of competition has been to use a truncated S-C-P approach that highlights the causal importance of market structure but that also goes to evidence on market conduct and performance.18 The methodology is truncated only to the extent necessary to establish the existence of significant discretionary power and if so, whether it is sufficiently significant to qualify for a finding that is ‘substantial’ as it is required by the TPA.19 However, in its 2009 Re Application by Chime Communications20 judgment the Tribunal’s analysis deviated from the structuralist approach. It explained that this approach ‘has been overtaken by developments in economic theory and by empirical assessments of competition in modern markets which attest to the fact that [this causal flow from structure to conduct and then performance] is by no means the dominant mechanism to explain market behaviour’.21 The Tribunal’s broader approach took into account market dynamics and firms’ strategic behaviour by examining non-structural factors such as new sources of potential supply and the competitive nature of relevant market participants.22 While the structuralist approach basically suggests that the presence of dominant companies weakens the competitive process, economic theory and practical experience in the last four decades have shown that competitive dynamics can function well even if a market has some very large players.23 This shift reflects developments since the Chicago School and other schools of law and economic thought. The Chicago School found mainstream acceptance in the 1970s and early 1980s. Chicago School commentators rejected the structuralist view and instead sought to apply insights of neoclassical price theory to antitrust analysis.24 They presumed that market outcomes reflect the interplay of standalone market forces and the technical demands of production.25 The focus of their analysis was on consumer welfare as the sole outcome of competition.26 A key proposition made by Chicagoans was that the pursuit of economic efficiency (consisting of productive and allocative efficiency but not dynamic efficiency) should be the exclusive goal of competition law.27 The Post-Chicago School, which has grown out of the criticism of a range of aspects of the Chicago School, has gained significant influence since the mid-1980s.28 While Post-Chicago scholarship recognizes the importance of efficiency and free markets, contrary to the Chicago School it cautions that markets are not automatically effective and may be prone to market imperfections.29 Post-Chicagoans also rely less on neoclassical price theory and place greater emphasis on the strategic behaviour of firms. As such they also argue that actively attempt to influence market conditions though strategic conduct instead of reacting passively to structural market conditions.30 There is a very noticeable linguistic and conceptual difference in relation to these factors between the Tribunal in QCMA and the ACCC in the updated ‘Guidelines on misuse of market power’ of 2018. The Tribunal’s language explicitly says that these factors are elements of market structure and that need to be considered in any case. The ACCC, on the other hand, indicates that the list of factors in QCMA is discretionary and non-exhaustive.31 In fact, the last three QCMA factors have not come up in many cases since QCMA. Furthermore, the ACCC omits any reference to market structure and only speaks of competitive constraint. The ‘QCMA factors’ (elaborated below in Section IV) are now one of the many factors courts may turn to when establishing the competitive constraints faced by a company.32 III. THE OLD, THE NEW, AND THE NEWEST ECONOMY The QCMA factors were established almost 45 years ago. The economy at that time, not just in Australia, was fundamentally different to today’s fast-paced and technologically-based economy. This section sets out the different stages of the transformation of the economy since QCMA. The ‘Old’ economy In 1975, one year before the QCMA ruling, America’s largest Fortune 500 corporations were Exxon Mobil, General Motors and Ford Motor.33 Companies operating during this time period resided within the ‘old’ economy on which competition law analysis was founded. The old economy is generally considered to include industries that manufacture traditional goods such as steel and automobiles.34 According to Posner the old economy has a clear set of defining characteristics including multi-plant and multi-firm production, stable markets, heavy capital investment, modest rates of innovation, and slow and infrequent entry and exit.35 While it was possible for a potential competitor to enter a traditional industry market and displace an incumbent, those characteristics meant that the process for doing so was relatively slow, and required a level of innovation that was commonly absent or particularly slow in such industries.36 Based on the characteristics of traditional industries the focus on the five QCMA factors does not come as a surprise. In the absence of a statutory monopoly, those industries are characterized by competition in the market. In other words, the market leaves room for more than one company, and market shares indicate the economic strength of the actual competitors on the market. And those market shares may be relatively stable due to the existence of barriers to expansion and entry typically in the form of economies of scale and sunk costs.37 Melway38 is one Australian case in which the characteristics of the traditional economy were made clear. The High Court agreed with the trial judge that the Melway directory had substantial market power, holding in excess of 80–90 per cent of the retail market share in the relevant market.39 Barriers to entry were considered to be substantial due to the methods required to produce and compile the street directories. The court noted it was ‘neither rational nor likely’ for a new entrant to enter the market unless a new form of technology was created.40 The Melway example explains some of the key characteristics of a market in the traditional economy: substantial barriers to entry maintained by limited levels of innovation, and high costs of production which could lead to sunk costs in the event a new entrant was unsuccessful in competing effectively with the incumbent. The ‘New’ economy The term ‘new economy’ is commonly used to refer to high-tech industries that generate high rates of innovation such as computer software and hardware, internet-based businesses, electronic communication, biotechnology, and aerospace.41 Software was the paradigmatic new economy product (until the recent emergence of Big Data) and the relevant product market in Microsoft. New economy industries typically lack many of the abovementioned features of the old economy,42 and instead are often characterized by high rates of innovation, network effects, supply-side economies of scale, switching costs, and first-mover advantages.43 A more distinctive characteristic, as noted by Evans and Schmalensee, is that ‘[in] many of these industries, firms engage in dynamic competition for the market – usually through research-and-development (R&D) competition to develop the “killer” product, service or feature that will confer market leadership and thus diminish or eliminate actual or potential rivals’.44 In other words, the market has capacity for only one firm, and the market will tip towards that one firm whose product or service will become the standard. That one or more of those characteristics may be essential in the context of a new economy industry, however, does not mean that such characteristics never appear in other industries or that all of those characteristics always appear in new economy industries.45 It rather suggests that those characteristics are critical in industries in which innovation, intellectual property, and technological change are central features.46 The Microsoft case triggered a debate among competition law commentators about the application of competition law to the new economy, and particularly the relationship between competition law and intellectual property rights. The US Department of Justice among other things alleged that Microsoft violated section 1 of the Sherman Act by unlawfully tying its Internet Explorer to Windows and unlawfully maintained a monopoly position in the PC operating system market in violation of section 2 of the Sherman Act. Microsoft disputed these allegations and contended that traditional antitrust concepts are not applicable to the new economy.47 Microsoft argued that competition in a technologically dynamic market such as the PC operating system market is characterized by Schumpeterian competition in which firms compete through innovation for temporary market dominance until they are displaced by the next wave of product advancements. It was therefore not able to exercise market power as it was constrained by potential entrants.48 A less extreme position around the time of the Microsoft trial suggested that as antitrust law must preserve the incentives for firms to innovate, the application of the rules to new economy should be altered to fully accommodate the dynamic competition in those markets.49 On the other hand, some scholars argued that antitrust law was ‘sufficiently supple’ to adequately deal with new economy industries.50 Noting the lack of consensus about the need for amendment of the monopolization doctrine to account for competition in the new economy the D.C. Court of Appeals applied the conventional doctrine.51 The ‘Newest’ economy Since the Microsoft trials high-tech industries, and the IT industry in particular, have developed further rapidly. In 2011, Surblyte used the term ‘newest economy’ which relates to the latest shift in the computer industry, namely ‘cloud computing’.52 More broadly this economy can be said to comprise digital or online markets, and is most known for digital platforms, which is why it is often referred to as ‘platform economy’.53 The ACCC’s Digital Platforms Inquiry (DPI) has defined digital platforms as ‘applications that serve multiple groups of users at once, providing value to each group based on the presence of other users’.54 A characteristic feature of digital platforms is that they act as intermediaries between those user groups. Enabled by the mobile Internet and heavily relying on the use of Big Data and algorithms, digital platforms started to take off at the end of the noughties,55 and even more so since the launch of smartphones. At the same time this era is also epitomized by the rise of ‘GAFA’—an acronym that stands for Google, Apple, Facebook, and Amazon.56 The main products/services of Google (founded in 1998), Facebook (founded in 2004) and Amazon (founded in 1994) are non-hardware based and relate to things like search, social networking, and e-commerce respectively. While Microsoft still holds a strong position in some aspects of the newest economy (eg public cloud software), they were unable to attain a strong market position in other emerging platform markets (eg mobile operating systems). The ‘antitrust troublemakers’ of late are some of the aforementioned tech companies, first and foremost Google.57 Industries in the new and the platform economies are both driven by innovation and are characterized by competition for the market. Moreover, due to substantial R&D costs, those industries typically exhibit supply-side economies of scale.58 In contrast, the marginal costs of an additional user to the platform is relatively low. Additionally, digital products and services use inputs that often enable economies of scope in product development.59 Similar to the discussion around the time of the Microsoft trial, some commentators are now debating again whether in relation to digital platforms antitrust doctrine is in need of an overhaul,60 or even whether competition law is obsolete.61 Unlike in the aftermath of Microsoft, this time governments and antitrust agencies are more receptive to the idea that competition in digital markets might be different,62 and as eg in the case of Germany, a few countries have already amended their competition laws accordingly.63 One specific question is whether new factors should be taken into account when assessing the market structure of digital markets and the market power of players therein. Again, that one or more of the above characteristics are common to the platform economy does not suggest that such characteristics never appear in other industries, nor that all or most of them always apply to digital platforms.64 Furthermore, it should be noted that platform companies can be very different from each other in terms of the markets they operate in and the effects of their activities on competition.65 It is submitted in this article that the entirely different business model of most digital platforms, and in particular the role of data, makes their market structure decisively distinct from industries in the new economy, and therefore justifies a different competition analysis. Digital advertising platforms such as Google and Facebook operate by providing free services to consumers in exchange for consent to the collection of the users’ personal data.66 The companies then sell advertising businesses the opportunity to serve their ads to a targeted and highly personalized online audience.67 Although data collection is often a by-product of the usual functioning of these platforms, it can create a competitive advantage for incumbents simply from having a large pool of data to access.68 At the same time, platforms can use data as a sharable input to develop new products and services.69 The most characteristic features of the platform economy, and one closely connected to the collection of data, is the role of network effects. A network effect arises when value of a product or service increases with the number of users.70 The ‘classic’ type of network effects can also be seen in the old and the new economy. For instance, a telephone network’s value increases to one person from other people also joining the network. Another type of network effect is a two- (or multi-)sided network effect. It was Microsoft’s dominance in the software market that gave prominence to the economic theory of two-sided network effects.71 Microsoft was alleged of using its large user base to encourage software developers and computer hardware manufacturers to focus their efforts on Microsoft’s Windows Operating System. With an increasing number of software and hardware available for Windows, the operating system attracted more users, which in turn led to more software and computer hardware specific to Windows. This positive feedback loop can cause the market to tip and it becomes dominated by one firm. This ‘winner takes all’ effect helps to explain why in many new economy industries competition is for the market.72 According to Page and Lopatka ‘the new economic theory of network effects … provided a lens through which Microsoft’s victories over its rivals appeared anticompetitive. The development of the network effects theory parallels the development of the very software markets in which Microsoft competed’.73 Digital platform markets exhibit multi-sided network effects apart from classic network effects. In the DPI, the ACCC observes that there are different kinds of multi-sided network effects which ensure that an increase in the number of individual users increases the value of the platform for advertisers.74 Most importantly, an increase in the number of users increases the users exposed to an ad campaign, which may increase an advertiser’s return from that campaign.75 And also a platform with more users has access to more data, which can enhance the relevance of ads presented to users.76 Importantly, the collection of data on digital platforms can result in network effects that extend beyond the type of network effects seen in the previous economies. Stucke and Grunes identify three ‘data-driven network effects’ that stem from the scale, scope, and the spill-over effect of data.77 These data-driven network effects can be illustrated with the example of Google Search. The more people use this search engine, the larger the scale of Google’s collected data which the company can then use to expand trial-and-error experiments allowing Google’s search algorithm to enhance the relevance of its searches, thus attracting more users.78 Additionally, Google’s large scope of data (ie variety of data), which the company collects across its platform through its other service like Google Maps and Gmail, also allows the company to improve its search algorithm.79 Furthermore, Google’s collection of personal data has a spill-over effects into other sides of the market.80 Google is significant for advertisers who buy the opportunity to advertise to a targeted audience. Google can invest this revenue to improve the quality of its search engine which will attract more users, and therefore more advertisers, eventually resulting in a virtuous circle. Finally, as elaborated in the next section, data-driven network effects motivate digital companies to adopt conglomeration strategies.81 IN NEED FOR NEW ASSESSMENT CRITERIA IN ONLINE MARKETS? In light of the profound differences between the old and the platform economy, the purpose of this section is to determine which of the QCMA factors are still relevant in the assessment of competition in online markets. The comeback of the structuralist approach in online markets In Australia, and in other jurisdictions, there has been a debate in recent years about adopting new approaches to competition assessment in online markets.82 This debate has largely been sparked by rapid growth and expansion of tech giants like Google, Facebook and Amazon. In the final report of the DPI the ACCC states that ‘[t]he pace of technological change needs to be matched by the pace of policy review. As digital markets and the use of data continue to grow and change, governments need to continue to consider the appropriate level of oversight’.83 In antitrust scholarship the dominance of Big Tech has also triggered another debate about the ideology of competition law.84 A pertinent question in the Australian context is whether the QCMA framework is still suitable for digital platform markets. For a long time the structuralist approach has been regarded as out of date and was eventually replaced by the Chicago and the Post-Chicago School thinking which has become known as the ‘modern antitrust’ doctrine in the US and in other jurisdictions.85 However, in the past few years a movement referred to as the ‘Neo-Brandeisian School’86 emerged to criticize this modern antitrust doctrine and its failure to tackle the antitrust problems involving ‘Big Tech’.87 Khan argues that the outcome-oriented consumer welfare approach ‘is inadequate to promote real competition, a failure that is amplified in the case of dominant online platforms’.88 This approach ‘equates harm entirely with whether a firm chooses to exercise its market power through price-based levers, while disregarding whether a firm has developed this power, distorting the competitive process in some other way’.89 Letting firms accumulate market power does not only make it more cumbersome to adequately check that power when it is eventually exercised, but these firms may also exploit their market power through non-price based anticompetitive practices.90 A concept that was widely accepted among the Structuralist School but later discredited by the Chicago School, and now highlighted by the Neo-Brandeisians is the conglomerate model.91 A ‘conglomerate’ describes a company that consists of different and distinct parts that are under one roof and operate on neighbouring markets, offering non-substitutable products to the same group of customers. Industrial organization economists are particularly wary of conglomerate mergers as they are more likely to increase concentration in particular markets, dampen competition and raise barriers to entry.92 They are concerned that multi-market firms are able to shift supra-competitive profits from one market to another and may engage in predatory pricing, reciprocity or other exclusionary or exploitative practices.93 In addition, structuralists warn that conglomerates could wield non-economic power.94 Chicagoans, on the other hand, argue that conglomeration theories and predatory pricing in particular are implausible and therefore highly unlikely.95 According to Lim the ‘return of the conglomerate model has coincided with an intriguing change in the nature of competition in the high-tech industry, particularly where platform businesses that harness network effects are involved’.96 In addition to the past conglomeration theories, Bourreau and de Streel attribute the emergence of digital conglomerates mainly to two characteristics of the platform economy: economies of scope in product development and product ecosystems.97 From a supply-side perspective, the possibility of collecting large amounts of data on consumers, facilitated by the modular design of data, may incentivise digital companies to expand in other markets in order to acquire new data on unattached consumers or complementary data on attached consumers.98 From a demand-side perspective, a product ecosystem can develop when firms create ties between their different products (even possibly unrelated products) to enhance the complementarity between them.99 The expansion strategy could result in platforms becoming gatekeepers.100 It also gives platforms an incentive to pre-emptively acquire start-ups who would pose a threat if they continued to grow.101 In light of the developments involving digital platforms and the inadequacy of price-based measures of competition to capture market dynamics, particularly given the role and use of data, Khan argues antitrust enforcement in online markets should focus on the competitive process and market structure.102 She does not advocate a strict return to the S-C-P paradigm but rather claims that a competition analysis that ignores the role of structure is misguided since ‘the best guardian of competition is a competitive process, and whether a market is competitive is inextricably linked to – even if not solely determined by – how that market is structured’.103 The consideration of a more structuralist approach in practice as proposed by Khan involves an assessment of how a market is structured and whether a single firm had obtained sufficient power to distort competitive outcomes.104 This approach considers various factors that shed light on the neutrality of the competitive process and the openness of the market including entry barriers, conflicts of interest, and the emergence of gatekeepers or bottlenecks.105 Khan illustrates the significance of market structure with the example of Amazon. She claims that Amazon has established dominance on the e-commerce market owing to two closely linked elements of its business strategy, namely its willingness to forego profits and investing aggressively in order to achieve scale, and integration across multiple business lines.106 In establishing Amazon’s structural dominance Khan refers to conventional factors, in particular barriers to entry and vertical integration—which are also two of the most important QCMA factors. Apart from the need for significant investments, Khan highlights advantages stemming from data collection and network effects as barriers to entry.107 While the emphasis of establishing market power is on structural factors, strategic conduct may still be relevant in the Neo-Brandeisian approach given that there is no strict adherence to the S-C-P paradigm. In their pursuit of rapid growth digital platforms such as Amazon and Uber engage in predatory pricing to psychologically intimidate competitors and keep them out.108 How would the QCMA factors fare in the platform economy? Under Australian competition law there are no immediate plans to introduce new market power assessment criteria with regard to section 46 that are specifically tailored to the platform economy. The DPI is the first occasion in which the ACCC assesses the market power of digital platforms.109 In the final report of the inquiry the ACCC claims that in principle ‘the existing tools and goals of the competition law framework remain applicable for digital markets’.110 However, at the same time there is no explicit mention of the QCMA factors or the relevance of a structural factors. This section explores how useful the QCMA factors are when applied to competition analysis in online markets and sheds light on the ACCC’s approach in the DPI. The number and size distribution of independent sellers Under Australian competition law, like in many other competition regimes, the number and size distribution of independent sellers, ie the degree of market concentration and market shares, are generally regarded as a first indicator of market power.111 The higher a firm’s market share is, the more likely it will hold market power or at least market power at a substantial degree.112 The traditional methodology for determining market power involves identifying actual competitors in a market, ie the companies currently operating in that particular market, and to determine their respective market shares.113 Yet, unlike in some jurisdictions,114 there is no market share threshold for determining significant market power.115 In markets, including innovation markets, where a company has held a very large market share that is considerably larger than its rivals and maintained that position over many years, that company can be deemed to have substantial market power. In Queensland Wire, BHP produced nearly 97 per cent of the steel made in Australia and supplied about 85 per cent of Australia's requirements for steel and steel products.116 In the DPI, the ACCC points out the Google has been the dominant general search engine in Australia with a market share of 93–95 per cent since 2009.117 Nonetheless, market shares are merely a first indicator and never conclusive on their own. In Boral,118 Gleeson CJ, and Callinan J said in relation to market concentration and its relevance that a ‘large market share may, or may not, give power. The presence or absence of barriers to entry into a market will ordinarily be vital’.119 In particular, in the new and platform economy market shares are a less reliable indicator of market power because market share figures only measure actual but not potential competition. Yet, in those markets incumbents may be constrained by potential competitors who have not yet entered the market but who would do so if a profitable opportunity were to arise.120 Furthermore, as noted in the 1993 Hilmer Report, a large number of competitors in a market is not necessarily always optimal. Early economic work suggested that effective competition in a market requires many firms, each with a small market share. However, in some markets competition between a few large firms may yield more economic benefit than competition between a large number of small firms. Such oligopolistic competition may arise as a result of economies of scale and scope, not only in production but also in marketing, technology and, increasingly, in management.121 As stated above, competition in the digital economy is often characterized by competition for the market rather than competition in the market, and online markets therefore tend to be highly concentrated and are often dominated by one company. In addition, network effects result in a positive feedback loop which can accelerate the process of a firm capturing the market, ie these markets have a tendency towards tipping. Yet, despite a large market share a dominant platform is not immune to a relatively quick displacement by another company.122 Cognisant of the limited value of market shares in markets involving digital platforms, the ACCC examines to what extent dynamic competition place competitive constraints on Google and Facebook.123 For instance in the case of Facebook the ACCC notes that: in the market for social media services, Friendster was initially leapfrogged by MySpace, which, in turn, was rapidly replaced by Facebook. However, the barriers to entry and expansion may be substantially higher now than in the early phase of the social media market. In particular, the global size of Facebook now dwarfs the size of MySpace at its peak.124 There has not been any notable consideration of dynamic competition in previous cases, possibly because the traditional markets in question lacked the necessary potential for innovation. Market shares in online markets are more volatile than in traditional markets and the usual time period for deeming market shares to be stable is generally shorter. In light of the difficulty associated with network effects and market tipping, a new provision that addresses the largest platforms with superior growth rates in a market as mooted in relation to the 10th amendment of the German Competition Act as an alternative to market shares in online markets.125 A similar provision would also be desirable under Australian competition law, especially since there is no market share threshold. Hence, regardless of whether a platform company has substantial market power, if it experiences superior growth rates compared to its competitors it might be subject to closer antitrust scrutiny. Another reason to be cautious of market concentration and market shares is the inherent dependence of these two concepts on market definition—a notoriously difficult exercise that has been criticized in more recent years.126 In the digital economy, market definition is even more problematic. First, digital companies often offer free products and services which makes it hardly possibly to apply price-based test like the SSNIP test. Non-price based tests to define the relevant market are hardly workable, absent well-accepted, quantifiable metrics.127 Secondly, since online markets are often multi-sided it is not sufficient to only look at the relevant market on which the defendant is operating.128 It is necessary to pay close attention to network effects and how competition on related markets is impacted. Moreover, even if online markets could be defined accurately, the overall size of a company and conglomerate power aggregated on a number of related markets seem to be at least as relevant as a company’s market share on the actual market that is under investigation. Overall, even though market concentration and market shares continue to be relevant in the platform economy they are not as indicative of market power as they used to be in the old economy. The height of barriers to entry Barriers to entry have been defined as any advantage that allows an incumbent to earn above-normal profits without the threat of entry.129 More broadly, barriers to entry (and expansion) describe any material difficulties that an incumbent’s actual or potential competitors face when seeking to enter the market or expand within it.130 Barriers to entry, not market shares, are the key to any assessment of market power.131 A high market share may not be a reliable indicator of market power if barriers to entry are low since a new entrant might quickly take away market shares from the incumbent. Conversely, if barriers to entry are substantial, and new entry therefore less likely, a high market share will be a sign of market power.132 In QMCA, the Tribunal identified the height of barriers to entry to be the most significant element of market structure: Of all these elements of market structure, no doubt the most important is (2), the condition of entry. For it is the ease with which firms may enter which establishes the possibilities of market concentration over time; and it is the threat of the entry of a new firm or a new plant into a market which operates as the ultimate regulator of competitive conduct.133 The Hilmer Report discussed how potential competition can have a similar effect on the performance of a firm as actual competition. It noted that ‘a market which is highly open to potential rivals – known as a highly “contestable” market – may be of similar efficiency as a market with actual head-to-head competition’.134 Since innovation markets are characterized by competition for the market, the number and size of sellers are less relevant than in traditional markets, and barriers to entry play an even more vital role. The test for determining the existence of a significant barrier to entry is to assess whether there is a credible threat of entry that will constrain the incumbent to behave competitively.135 And this is typically done by considering whether entry is likely, timely, and sufficient.136 For entry barriers to be considered low, a potential competitor’s entry into the market must be likely as a matter of commercial reality, happen within a time horizon of three years, and not merely occur at the fringe of the market.137 In light of the pace of innovation in the platform economy a shorter time horizon might be more appropriate. Simply put, the essential test for whether or not a significant barrier to entry exists is whether the threat of entry of whatever kind will constrain incumbents to behave competitively.138 Barriers to entry come in various forms and can generally be categorized into structural and strategic ones.139 The former result from structural characteristics of the market and may include among other things regulatory requirements, access to essential facilities, intellectual property rights, economies of scale, and sunk costs.140 Identifying structural barriers to entry is quintessential to the S-C-P paradigm. Strategic barriers to entry, on the other hand, stem from the deterrent activities by an incumbent to purposely hinder the possibilities of entry, as opposed to ‘innocent’ entry barriers that unintentionally arise as a side effect of innocent profit maximization.141 In Re Qantas Airways, the Tribunal found that ‘it is entry as a behavioural phenomenon that is important in the analysis of competition rather than considering entry as a purely structural constraint’.142 Strategic barriers to entry exemplify the move from static economic models to dynamic market behaviour as advocated by Post-Chicago scholars.143 The novelty of the ACCC’s approach in the DPI is the consideration of strategic entry barriers. Different types of barriers to entry may be relevant in digital markets compared to the ones that more typically arise in traditional markets. It has been suggested that if competition authorities and courts consider only traditional barriers to entry, then they may falsely conclude that online markets are easy to enter.144 Despite winner-takes-all competition, from a demand perspective (ie attracting users) barriers to entry in many online markets are low since apps and services are offered for free or a very low price, thus in theory allowing users to multi-home easily.145 Yet, in reality high switching costs caused by positive network effects render multi-homing less likely.146 For example, pre-installed apps on users’ mobile phones can be a gateway barrier to entry. The status quo bias may lead users to use an app over a potentially more innovative rival app.147 This hindering of multi-homing raises switching costs and constitutes a strategic entry barrier.148 From a supply perspective many online markets lack entries barriers that are common to traditional markets such as land and raw materials. However, digital platforms face other forms of entry barriers, particularly with regard to the collection, storage, synthesis and analysis of data.149 Certain types of data may therefore amount to an essential facility.150 In that respect, another significant strategic barrier to entry and expansion that the ACCC focuses on in the DPI are Google and Facebook’s strategic acquisitions.151 The ACCC argues that some of those acquisitions have allowed Google and Facebook to entrench their respective incumbent positions. Those acquisitions have weakened the constraint from dynamic competition by eliminating potential competitors.152 The expansion into related markets has also provided Google and Facebook with advantages in terms of scope and network in order to collect more data. Network effects constitute by far the most significant barrier to entry in the platform economy. While network effect may also arise in the old and the new economies, as mentioned above, the three ‘data-driven network effects’ that stem from the scale, scope, and the spill-over effect of data are unique to the platform economy.153 Network effects (both same-side and cross-side) are the first barriers to entry that the ACCC analyses in relation to Google and Facebook. For instance in the case of Google, the ACCC finds that: all else being equal, a large amount of data improves the relevance algorithm in the search engine, increasing the quality of the search service. A greater quantity of user data, including data on user searches and user interactions with search results, allows the Google relevance algorithm to update in a timely fashion, improving its relevance ranking.154 Data-driven network effects may be reinforced by economies of scope and conglomeration effects, which the ACCC in the DPI discusses separately and not as a barrier to entry.155 A barrier to entry that is relevant to both traditional and online markets are economies of scale. In Queensland Wire, Mason CJ and Wilson J observed ‘[w]here the economies of scale in a market are such that the minimum size for an efficient firm is very large relative to the size of the market, it may be that potential competitors will be dissuaded from entering the market by the apprehension that only one firm would survive’.156 This is known as the supply-side of economies of scale. In old-economy industries, the high capital costs of manufacturing industries ensured that incumbents were protected from new players entering the protected market.157 In online markets, the demand-side of economies of scale (whereby the value of a product increases in line with the increasing number of users) may constitute an almost insurmountable barrier. In the DPI, the ACCC reiterates the finding in the report on digital platforms written for the European Commission stating that while economies of scale is a feature of a range of industries, ‘the digital world pushes it to the extreme and this can result in a significant competitive advantage for incumbents’.158 Since costs in the digital economy become marginal once the platform is set up, an incumbent is able to retain its market power relative to a new entrant. Its revenue will also continue to increase as new users and advertisers join the platform. This advertising revenue can be used to further invest into R&D. This could possibly lead to a positive-feedback loop. Two-sided markets require new entrants to achieve growth on both sides of the market in order to compete effectively. If there are high switching costs on top of demand-side economies of scale and network effects, potential entrants might be even more deterred as there is considerable uncertainty in innovation markets about recouping investment in R&D. Since competition is for the market, a potential entrant’s innovation must not just be a minor improvement. The innovation must rather be so substantial that people desert the incumbent and switch to the new entrant. In other words, the potential entrant must disrupt the market. Barriers to entry remain the most significant QCMA factor in the platform economy. The concept of barrier to entry is wide enough to cover a number of constraints—both structural and strategic. Moreover, many characteristics of the digital platforms—in particular network effects and economies of scale—may result in re-enforcing barriers to entry.159 Product differentiation Product differentiation denotes the ways in which firms seek to distinguish their products or services from those of their rivals in order to establish customer or brand loyalty.160 Where products are similar, they exercise constrain on each other. On the contrary, product differentiation gives a firm a degree of control on the price of its product and is thus an indicator of market power. In addition, the preference of buyers attached to an existing brand (ie brand loyalty) creates a barrier to entry for new firms.161 In QCMA, the Tribunal discussed that there was little product differentiation between the types of flour between the different firms.162 Since QCMA, new insights have altered our understanding of the implications of product differentiation on competition. Some of those insights are discussed in the Hilmer Report. First, the idea that firms provide identical products or services and compete mainly on price is the simplest notion of competition and the exception rather than the rule.163 In practice, work in business strategy suggests that competition arises when firms seek to offer consumers different mixes of benefits, some of which are already reflected in price and others of which are reflected in non-price elements such as service, quality, or timeliness of delivery.164 Secondly, competition is not always between identical products or services. The essence of competition is the striving to meet the same consumer need and this is reflected in the ways in which this is met by different market participants.165 Markets are also characterized by dynamic efficiency, ie the need for firms to make timely adjustments to their products in response to changes in consumer tastes and in productive opportunities.166 In the DPI, the ACCC still addresses product differentiation. The ACCC considers that the Google and Facebook’s branding constitutes a barrier to entry.167 In relation to Google the ACCC notes that ‘the strength of Google’s brand partly reflects the high quality of its search service. When first developed, the Google algorithm provided an innovative method for ranking the relevance of search results. Google invests a considerable sum each year on improving the quality of its service’.168 Moreover, with regard to display advertising, the ACCC argues that ‘the advertising inventory on social media platforms is a specific kind of display advertising – social media advertising – which is differentiated from other kinds of display advertising’.169 Dynamic efficiency is crucial in the platform economy and it arguably diminishes the role of product differentiation in some degree. Instead product imitation arguably plays a larger role than in the previous economies. Incumbent companies are generally wary of new entrants because they may challenge their market power. As explained above, in digital markets new entrants need to be highly innovative in order to displace an incumbent. Digital platforms compete on features, not price, and some of the major digital platforms offer combinations of services. The ACCC observes that the services provided by digital platforms are constantly changing due to technological advancement and shifts in consumer preferences.170 For instance, Facebook began as a consumer-facing social media service that allowed communication between networked users, but it now includes online marketplaces for goods and jobs, and Snapchat, which was initially a medium for creating and privately sharing photo-based content with other networked users, later expanded to include public content services.171 Major platforms are often easily able to introduce new features that mimic popular features of their rivals. For example, Facebook’s introduction of ‘stories’ to its social media platform was reminiscent of Snapchat’s core feature—the creation and sharing of multimedia messages referred to as ‘snaps’. For tech firms it is also relatively easy to remove a new feature if it is not well received by its users. The addition and removal of features on a digital platform is different to markets in the old economy where firms would more likely incur substantial sunk costs in the adaption of the production process as a consequence of when altering or updating a product. Moreover, it may be costly and time-consuming to reverse those changes. Similarly, tech start-ups may find it difficult to engage in product imitation as they might lack the necessary resources. Large platforms can use product differentiation, on the one hand, to build customer loyalty, and product imitation, on the other hand, to ward off smaller competitors. This strategy increases the likelihood of single-homing and users becoming locked in on a platform’s ecosystem. Single-homing and user lock-in are reinforced when a digital platform is able to use data to individualize products and services according to each individual user’s preference. Vertical relationships and vertical integration A company that at first sight appears to have SMP because of its considerable market share while shielded by high barriers to entry may in fact not be dominant due to its vertical relationships with customers and suppliers.172 For instance, a powerful buyer, who makes up a significant share of the upstream company’s sales, exercises competitive constraint over that company. If the upstream company were to raise the price of its product the buyer could threaten to switch to one of the latter’s competitors and the upstream company would probably not increase the price. Similarly, a company might be dependent on a powerful supplier that also exercises competitive constraint over it. An interesting vertical supply relationship in the platform economy can be seen between Google and Apple, who are competitors on some markets.173 Google was estimated to pay US$12 billion in 2019 for Apple to install Google Search as the default search engine on Apple’s Safari web browser, up from US$1 billion in 2014.174 In order to reduce dependence on upstream and/or downstream markets a company may opt to vertically integrate. Vertically integration arises when two or more successive stages of production and/or distribution are combined under the same control. A producer of goods or services can for instance carry out both the production and distribution functions itself, using its own employees or its own branches or through its wholly owned subsidiaries. Vertical integration may be vital for a platform company to gain more user data for better target advertising.175 For example in 2014 Google bought smart gadgets maker Nest Labs for US$3.2 billion.176 In 2018, Nest was merged into Google’s home devices unit. The brand Google Nest now offers products such as smart assistants that are integrated with Google Search. The integration of production and distribution processes as well as complementary products or services may ostensibly be considered as a pro-competitive efficiency. Yet, in some circumstances vertical integration may give rise to structural and strategic barriers to entry.177 Vertical integration constitutes a structural barrier to entry for instance where a firm has access to an essential facility through its parent firm or subsidiary, while other firms seeking to enter the market or expand within it lack that possibility and are dependent their competitor’s parent or subsidiary. In practice such a situation has often led to exclusionary abuses such as refusal to supply or margin squeeze.178 In Queensland Wire, Mason CJ and Wilson J observed that vertical integration makes it possible for the monopoly company to engage in discrimination, for instance in pricing between various users in the case of a distribution chain.179 When a firm has developed a reputation for engaging in exclusionary practices this can amount to a strategic barrier to entry. In addition to exclusionary abuses, vertical integration can lead to exploitative abuses. Even though production- and distribution chains are far less common in online markets, most of the largest firms are vertically integrated. After gaining enormous financial strength some of those firms achieved further vertical integration through major acquisitions.180 The ACCC has also recognized the potential anti-competitive effects of vertically integrated firms in the platform economy.181 For example, Google’s acquisition of online advertising agency DoubleClick for US$3.2 billion in 2008 significantly strengthened Google’s market power in search and search advertising.182 Hence, vertical relationships and vertical integration are still relevant in the digital economy. They allow digital companies to favour their own business interests above those of advertisers or other competing businesses,183 eg through the practice of sharing competitively relevant data between parent companies and their subsidiaries. Besides vertical integration, a regulatory authority should pay close attention to conglomeration which appears to be more common structural characteristic in online markets than in more traditional markets. Arrangements between firms The fifth and last QCMA factor requires examining the nature of any formal, stable and fundamental arrangements between firms which restrict their ability to function as independent entities. This goes beyond determining whether a company is competitively constrained by vertical relationships with their customers or suppliers and possibly relate to conglomerate effects. If arrangements exist that restrict the ability of a firm to operate as an independent entity, then they need to be taken into account in assessing market power.184 This QCMA factor is also reflected in section 46(3) CCA.185 This provision captures inter-corporate relationships and arrangements that give rise to substantial market power. To that end, it allows the aggregation of market power between related bodies corporate.186 While the mere fact of being part of a corporate group by itself does not indicate market power,187 it is possible for a subsidiary to exercise market power on the basis of its parent’s support. In TPC v Pioneer Concrete,188 the TPC argued that the defendant, as a subsidiary of Pioneer International Ltd and a member of the Pioneer International holding company, had substantial market power. The holding company gave Pioneer Concrete access to raw materials and the financial backing to sustain financial losses to maintain a substantial market share in all the pre-mixed concrete markets in Australia.189 In the DPI, the ACCC argues that some of Google and Facebook’s strategic acquisitions have enabled them to entrench their respective incumbent positions. Those acquisitions have not only weakened the constraint from dynamic competition by eliminating potential competitors,190 but the expansion into related markets has also provided Google and Facebook with advantages of scope in order to collect more data. Both section 46(3) and the last QCMA factor could therefore arguably be used to establish a firm’s conglomerate market power. CONCLUSION Despite the criticism over the structuralist approach since the rise of the Chicago School, this approach is still not obsolete. Neo-Brandeisians strongly argue that structural factors and concepts should be considered in the assessment of competition in online markets. In Australia, the structuralist approach has been predicated on the five structural factors in the QCMA case. These five factors, sometimes referred to as the QCMA factors, concern market concentration; barriers to entry; product differentiation; vertical integration, and the nature of arrangement between firms. The QCMA factors were borne out of the old economy which was characterized by competition in the market. Already competition in high-technology markets of the ‘new economy’ raised question marks over the usefulness of antitrust principles that were developed in old economy times. Those markets are characterized by competition for the market. Nonetheless, competition regimes were not amended to reflect this new type of competition and the underlying structural and dynamic factors. While the ‘newest economy’ shares many characteristics with its predecessor, the role of data-driven network effects justifies the application of modified assessment approach in online markets. The ACCC’s market power analysis deviates considerably from the conventional assessment approach. Already in the Guidelines on Misuse of Market Power (2018) the ACCC states that the list of factors in QCMA is discretionary and non-exhaustive.191 The Guidelines also omit any reference to market structure and only speak of competitive constraint. Moreover, in Re Application by Chime Communications the Competition Tribunal’s analysis deviated from the structuralist approach by adopting a broader approach that took into account market dynamics and firms’ strategic behaviour. While there is no explicit mention of the QCMA factors in the DPI, the ACCC still considers the structure of the relevant markets and discusses some of the QCMA factors albeit implicitly as part of other market aspects. However, in its assessment the ACCC supplements some of the structural factors with dynamic and strategic ones. This can mostly be seen in relation to market concentration and barriers to entry and expansion. Overall, even though market structure still matters, there is no complete comeback of the structuralist approach. Footnotes 1 This article uses the terms ‘digital economy’ and ‘platform economy’ interchangeably. 2 Re Queensland Co-Operative Milling Association Ltd (1976) 8 ALR 481 [hereafter ‘QCMA’]. 3 Maureen Brunt, Economic Essays on Australian and New Zealand Competition Law (Kluwer Law International 2003) 316. 4 QCMA, 512. 5 ACCC, Guidelines on Misuse of Market Power (31 August 2018), para 2.15 [hereafter ‘Guidelines on Misuse of Market Power’]. 6 QCMA, 481. 10 See Rhonda L Smith and David K Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5(1) Agenda 25, 25; Arlen Duke, Corones’ Competition Law in Australia (7th edn, Thomson Reuters 2019) 48. 11 George Raitt, ‘Competition and Efficiency Effects in Europe, North America and Australia’ (2016) 24 Competition & Consumer Law Journal 187, 189; Brunt (n 3) 315. 12 See Brunt, ibid 315; Russell V Miller, Miller’s Australian Competition Law and Policy (3rd edn, Thomson Reuters 2018) 28. 13 Edward Mason, ‘The Current Status of the Monopoly Problem in the United States’ (1949) 62 Harvard Law Review 1265; Joe S Bain, Industrial Organization (2nd edn, Wiley 1968); Kaysen and Turner (n 8). See also Brunt (n 3) 294. 14 See Maureen Brunt, ‘Economic Overview’, Mimeograph, Lecture No 11, Monash Trade Practices Lectures, 1975. 15 Lina M Khan, ‘Amazon’s Antitrust Paradox’ (2017) 126 Yale Law Journal 710, 718. 16 Frederic M Scherer, Industrial Market Structure and Economic Performance (3rd edn, Houghton Mifflin 1980) 4–5. 17 Queensland Wire, 200. See also Kemp (n 9) 56. 18 Brunt (n 3) 258 and 317–18. 19 ibid 317–18. Brunt argues: ‘Sometimes all that will be necessary is what the Americans call the “quick look”, that is, the highly truncated rule of reason approach by which one might seek to eliminate the existence of likelihood of any market power in any conceivably relevant market. Then, indeed, all that may be necessary is to establish an absence of significant barriers to entry, and market definition can be primitive. But where the question revolves around the measurement of market power or the true character of business conduct and its probable consequences, the evidence that is called for may well relate to elements of market structure, market conduct, market performance, and their interrelation.’ (ibid 219). 20 Re Application by Chime Communications Pty Ltd [No 3] [2009] ACompT 4 (24 August 2009). 21 ibid [11]. See also Smith and Round (n 10) 26; Alexandra Merrett, ‘Understanding Market Power’ in John Duns and others (eds), Comparative Competition Law (Edward Elgar 2015) 109. 22 ibid [11]–[12]. See also Smith and Round (n 10) 26; Caitlin Davies and Luke Wainscoat, ‘Not Quite a Cartel: Applying the New Concerted Practices Prohibition’ (2017) 25 Competition and Consumer Law Journal 173, 195; Lindsay Foster and Hanna Kaci, ‘Concerted Practices: A Contravention without a Definition’ (2018) 26 Competition and Consumer Law Journal 1, 13. 23 Gunnar Niels and others, Economics for Competition Lawyers (OUP 2011) 180. 24 See Andrew I Gavil and others, Antitrust Law in Perspective: Cases, Concepts and Problems in Competition Policy (3rd edn, West Publishing 2017) 71–72. 25 Khan (n 15) 719. 26 Robert H Bork, The Antitrust Paradox: A Policy at War with Itself (Basic Books 1978) 108. It has been widely argued that Bork defines consumer welfare not as consumer surplus but as total welfare. To date there is no consensus as to how consumer welfare is to be understood in US antitrust (see eg Barak Orbach, ‘Foreword: Antitrust’s Pursuit of Purpose’ (2013) 81 Fordham Law Review 215; Eleanor M Fox, ‘Against Goals’ (2013) 81 Fordham Law Review 2157). 27 Herbert Hovenkamp, Principles of Antitrust (West Academic Publishing 2017) 48. 28 Gavil and others (n 24) 75. 29 See Michael H Riordan and Steven C Salop, ‘Evaluating Vertical Mergers: A Post-Chicago Approach’ (1995) 63 Antitrust Law Journal 513. See also Herbert Hovenkamp, The Antitrust Enterprise: Principle and Execution (Harvard University Press 2005) 156. 30 M Sean Royall, ‘Symposium: Post-Chicago Economics - Editor’s Note’ (1995) 63 Antitrust Law Journal 445, 447. See also E Thomas Sullivan and others, Antitrust Law, Policy, and Procedure: Cases, Materials, Problems (7th edn, LexisNexis 2014) 56. Besides the Post-Chicago School other notable scholarships include ‘neo-Chicago’ and behavioural antitrust (see Hovenkamp (n 27) 52). 31 Guidelines on Misuse of Market Power, para 2.15. 32 s 50(3) CCA sets out a non-exhaustive list of factors that the courts must consider in any merger assessment. 33 Fortune 500, ‘Fortune500 Company Search: A database of 50 years of FORTUNE’s list of America’s Largest Corporations’, <https://archive.fortune.com/magazines/fortune/fortune500_archive/full/1960/> last accessed 1 December 2020. 34 Richard A Posner, ‘Antitrust in the New Economy’ (2001) 68 Antitrust Law Journal 925, 926. 35 ibid. 36 ibid. 37 See eg Arnotts Ltd v TPC (1990) 24 FCR 313. 38 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1. 39 ibid 11 [10]. 40 ibid. 41 Daniel Gifford and Robert Kudrle, ‘Antitrust Approaches to Dynamically Competitive Industries in the United States and the European Union’ (2011) 7 Journal of Competition Law and Economics 695, 695; Alison Jones and Brenda Sufrin, EU Competition Law: Text, Cases, and Materials (6th edn, OUP 2016) 48. 42 Posner (n 34) 926. 43 Antitrust Modernization Commission, Report and Recommendations (April 2007), 32–33; David S Evans and Richard Schmalensee, ‘Some Economic Aspects of Antitrust Analysis in Dynamically Competitive Industries’ in Adam B Jeffe and others (eds), Innovation Policy and the Economy, Volume 2 (MIT Press 2002) 8–13; Christian Ahlborn and others, ‘Competition Policy in the New Economy: Is European Competition Law Up to the Challenge?’ (2001) 5 European Competition Law Review 156, 158–161; J Gregory Sidak and David J Teece, ‘Dynamic Competition in Antitrust Law’ (2009) 4 Journal of Competition Law and Economics 581, 585. 44 Evans and Schmalensee, ibid 8–13; Damien Geradin and others, ‘DG Comp’s Discussion Paper on Article 82 Implications of the Proposed Framework and Antitrust Rules for Dynamically Competitive Industries’ (2006) 12, available at <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=894466> last accessed 1 December 2020. 45 Antitrust Modernization Commission (2002), 33. See also Jonathan M Jacobson, ‘Do We Need a “New Economy” Exception for Antitrust?’ (Fall 2001) Antitrust 89, 89. 46 Antitrust Modernization Commission (2002), 33. 47 United States v Microsoft Corp, 253 F3d 34, 49–50. 48 ibid, See Howard A Shelanski and J Gregory Sidak, ‘Antitrust Divestiture in Network Industries’ (2001) 68 University of Chicago Law Review 1, 11–12. 49 See Robert Pitofsky, ‘Challenges to the New Economy: Issues at the Intersection of Antitrust and Intellectual Property’ (2001) 68 Antitrust Law Journal 913, 916–17; J Gregory Sidak, ‘An Antitrust Rule for Software Integration’ (2001) 18 Yale Journal on Regulation 1, 27. See also Geradin and others (n 44) 20–24. 50 Posner (n 34) 92. 51 United States v Microsoft, 49–50; Gifford and Kudrle (n 41) 696. In the EU trial, the General Court also applied an unmodified approach to art 102 TFEU (Case T-201/04 Microsoft v Commission, ECLI:EU:T:2007:289). 52 See Gintare Surblyte, The Refusal to Disclose Trade Secrets as an Abuse of Market Dominance – Microsoft and Beyond (Stämpfli 2011) 143. 53 This article uses the terms ‘platform economy’ and ‘digital economy’ to distinguish them from the ‘new economy’. 54 ACCC, Digital Platform Inquiry, Final Report (June 2019), 41 [hereafter ‘DPI’]. 55 See Patrick Barwise and Leo Watkins, ‘The Evolution of Digital Dominance: How and Why We Got to GAFA’ in Martin Moore and Damian Tambini (eds), Digital Dominance: The Power of Google, Amazon, Facebook and Apple (OUP 2018) 21, 22. 56 ibid. See also Farhad Manjoo, ‘The Great Tech War of 2012: Apple, Facebook, Google, and Amazon battle for the Future of the Innovation Economy’ (Fast Company, 19 October 2011) <https://www.fastcompany.com/1784824/great-tech-war-2012> last accessed 1 December 2020. 57 See eg Google Search (Shopping) (Case COMP/AT.39740) [2018] OJ C9/11; Google Android (Case COMP/AT.40099) [2019] OJ C402/07; Google Search (AdSense) (Case COMP/AT.40411).. 58 See Carl Shapiro and Hal R Varian, Information Rules: A Strategic Guide to Network Effects (Harvard Business School Press 1999) 3. 59 Marc Bourreau and Alexandre de Streel, ‘Digital Conglomerates and EU Competition Policy’ (March 2019), 9, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3350512> last accessed 1 December 2020. 60 Rupprecht Podszun, ‘The More Technological Approach: Competition Law in the Digital Economy’ in Gintare Surblyte (ed), Competition on the Internet (Springer 2015) 101; Pablo Solani Díaz, ‘EU Competition Law Needs to Install a Plug-in’ (2017) 40 World Competition 393; Torsten Körber, ‘Analoges Kartellrecht für digitale Märkte?’ (2015) 65 Wirtschaft und Wettbewerb 120. 61 Thomas Jaeger, ‘Brauchen wir das Kartellrecht noch?’ Kartellrechtsanwendung im digitalen Umfeld am Beispiel Google (2015) 65 Wirtschaft und Wettbewerb 702. See also Körber, ibid 120. 62 See Autorité de la Concurrence and Bundeskartellamt, Competition Law and Data, Joint Report, 10 May 2016. 63 See 9th Amendment of the German Competition Act. 64 cf Antitrust Modernization Commission (2002), 33. See also Jonathan M Jacobson, ‘Do We Need a “New Economy” Exception for Antitrust?’ (Fall 2001) Antitrust 89, 89. 65 Kenneth A Bamberger and Orly Lobel, ‘Platform Market Power’ (2017) 32 Berkeley Technology Law Journal 1051, 1053. 66 Inge Graef, EU Competition Law, Data Protection and Online Platforms: Data as Essential Facility (Kluwer Law International 2016) 9–16; Antonio Capobianco and Anita Nyeso, ‘Challenges for Competition Law Enforcement and Policy in the Digital Economy’ (2018) 9 Journal of European Competition Law and Practice 19, 20; Lapo Filistrucchi and others, ‘Market Definition in Two-sided Markets: Theory and Practice’ (2013) 10 Journal of Competition Law and Economics 293, 333. 67 Graef, ibid 14; Rolf Weber, ‘Competition Law Issues in the Online World’ (Forum Paper, 20th St. Gallen International Competition Law Forum, 5 April 2013). 68 Jacques Crémer and others, Competition Policy for the Digital Era, European Commission Report No B-1049, 4 April 2019, 24; See also Daniel L Rubinfeld and Michal S Gal, ‘Access Barriers to Big Data’ (2017) 59 Arizona Law Review 339, 357. 69 Bourreau and de Streel (n 59) 10. 70 Michael L Katz and Carl Shapiro, ‘Network Externalities, Competition, and Compatibility’ (1985) 75 American Economic Review 424, 424–25. 71 US v Microsoft, 49; Case T-201/04 Microsoft v Commission, ECLI:EU:T:2007:289, para 106. 72 Schweitzer and others, ‘Modernisierung der Missbrauchsaufsicht für marktmächtige Unternehmen’, Projekt im Auftrag des Bundesministeriums für Wirtschaft und Energie (BMWi) Projekt Nr. 66/17 (29 August 2018) 12. 73 William H Page and John E Lopatka, The Microsoft Case: Antitrust, High Technology, and Consumer Welfare (University of Chicago Press 2007) 22. 74 DPI, 63. 75 ibid. 76 ibid 77 See Maurice E Stucke and Allen P Grunes, Big Data and Competition Policy (OUP 2016) 170–216. See also Jens Prüfer and Christoph Schottmuller, ‘Competing with Big Data’ (2017) TILEC Discussion Paper No 2017-006, <https://pure.uvt.nl/ws/portalfiles/portal/15514029/2017_007.pdf> last accessed 1 December 2020. 78 Stucke and Grunes, ibid 170. See also Geoffrey A Manne and Joshua D Wright, ‘Google and the Limits of Antitrust: The Case Against the Case Against Google’ (2010) 34 Harvard Journal of Law and Public Policy 171. 79 ibid 188. This dynamic is also referred to as demand-side economies of scale: See Pt 4B(2). 80 ibid. 81 Bourreau and de Streel (n 59) 11. 82 See eg DPI; Report of the UK Digital Competition Expert Panel, ‘Unlocking Digital Competition’ (March 2019); Crémer and others (n 68); Schweitzer and others (n 72). 83 DPI, 3. 84 See eg Ariel Ezrachi and Maurice E Stucke, ‘The Fight over Antitrust’s Soul’ (2019) 9 Journal of European Competition Law & Practice 1; Sandra Marco Collino, ‘The Antitrust F Word: Fairness Considerations in Competition Law’, The Chinese University of Hong Kong Faculty of Law Research Paper No 2018-09, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3245865> last accessed 1 December 2020. 85 See Raitt (n 11) 189; Smith and Round (n 10) 25; Herbert Hovenkamp, Federal Antitrust Policy: The Law of competition and its Practice (West Publishing 2005) 42–47. 86 Sometimes derogatively called ‘hipster antitrust’. For a critical view on the Neo-Brandeisian School see e.g. Joshua D Wright and Aurelien Portuese, ‘Antitrust Populism: Towards a Taxonomy’ (2020) 25 Stanford Journal of Law, Business & Finance 131; Joshua D Wright and others, ‘Requiem for a Paradox: The Dubious Rise and Inevitable Fall of Hipster Antitrust’ (2019) 51 Arizona State Law Journal 293; Seth B Sacher and John M Yun, ‘Twelve Fallacies of the “Neo-Anti000trust” Movement’ (2019) 26 George Mason Law Review 1491. 87 See eg Tim Wu, The Curse of Bigness: Antitrust in the New Gilded Age (Columbia Global Reports, 2018); Daniel A Crane, ‘How Much Brandeis Do the Neo-Brandeisians Want?’ (2019) 64 Antitrust Bulletin 531. 88 Khan (n 15) 744. 89 ibid 745. 90 ibid. 91 Yong Lim, ‘Tech Wars: Return of the Conglomerate – Throwback or Dawn of a New Series for Competition in the Digital Era?’ (2017), p 3, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3051560> last accessed 1 December 2020. 92 ABA Antitrust Section, Monograph No 7, Merger Standards under U.S. Antitrust Laws (1981), 162. 93 ibid. 100 ibid 19; Schweitzer and others (n 72) 8 and 85. 101 ibid 21; Lim (n 91) 11. 102 Khan (n 15) 737. 103 ibid 745. 104 ibid 746. 105 ibid. 106 ibid 746–47. 107 ibid 768 and 772. 108 ibid 772. 109 The Google/Fitbit merger is the first case on digital platforms in Australia but so far the ACCC has only published the statement of issues. 110 DPI, 138. 111 See eg Queensland Wire, 189–90; ACCC v Universal Music (2001) 115 FCR 442, 530 [381]; Case 85/76 Hoffmann-La Roche v Commission, ECLI:EU:C:1979:36, para 275; European Commission, Guidance on the Commission’s Enforcement Priorities in Applying art 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings [2009] OJ C45/2, para 22, (‘Guidance on the Commission’s Enforcement Priorities’). 112 ACCC v Universal Music, 536 [412]. 113 Sidak and Teece (n 43) 614. 114 See eg the AKZO presumption of dominance under art 102 TFEU where an undertaking has a market share of 50% or more. 115 Guidelines on Misuse of Market Power, para 2.16. See also Miller (n 12) 257. It could be argued that under Australian competition law market shares play a less central role than eg under US antitrust law or EU competition law. In the US monopolization is prohibited, while in the EU there is not only a market power threshold, but dominant companies also have a ‘special responsibility. Under Australian competition law, on the other hand, more than one company can have substantial market power as stipulated in s 46(7) CCA. 116 Queensland Wire, 183. 117 DPI, 65. 118 Boral Besser Masonry Ltd v ACCC (2003) 215 CLR 374. 119 ibid 423 [137]. 120 David Bailey and Laura E John (eds), Bellamy & Child – European Union Law of Competition (8th edn, OUP 2019), para 10.022. 121 National Competition Policy Review, 25 August 1993 (hereafter ‘Hilmer Report’), 3. 122 David S Evans and others, ‘An Analysis of the Government’s Economic Case in U.S. v. Microsoft’ in David S Evans (ed), Microsoft, Antitrust and the New Economy: Selected Essays (Kluwer Academic Publishers 2002) 29. 123 See DPI, 66 and 78. 124 DPI, 78. 125 See Oliver Budzinski and Annika Stöhr, ‘Competition policy reform in Europe and Germany – institutional change in the light of digitization’ (2019) European Competition Journal 15, 40. 126 See eg Louis Kaplow, ‘Why (Ever) Define Markets?’ (2010) 124 Harvard Law Review 437; Richard S Markovits, ‘Why One Should Never Define Markets or Use Market-Oriented Approaches to Analyze the Legality of Business Conduct Under U.S. Antitrust Law’ (2012) 57 Antitrust Bulletin 747; Rupprecht Podszun, ‘The Pitfalls of Market Definition: Towards an Open and Evolutionary Concept’ in Fabiana Di Porto and Rupprecht Podszun (eds), Abusive Practices in Competition Law (Edward Elgar 2018) 68. 127 Eg in cases where the quality of a product or service rather than price is decisive, competition authorities could probe a ‘Small but Significant, Non-transitory Decline in Quality’ (SSNDQ). 128 See Filistrucchi and others (n 66) 300–03. 129 Joe S Bain, Barriers to New Competition: Their Character and Consequences in Manufacturing Industries (Harvard University Press 1956) 3. 130 Bailey and John (n 120) para 10.031. 131 Franklin M Fisher, ‘Diagnosing Monopoly’ (1979) 19 Quarterly Review of Economics and Business 7, 18 and 23–27; Brunt (n 3) 7. 132 See Queensland Wire, 189–90. 133 QCMA, 512. See also Seven Network Ltd v News Ltd (2009) 182 FCR 160; Melway, 67. 134 Hilmer Report, 2. 135 ACCC, Merger Guidelines (November 2008), para 7.17; Brunt (n 3) 263. 136 See eg Guidance on the Commission’s Enforcement, para 16. 137 Duke (n 10) 114–123. See also Brunt (n 3) 263–64. 138 Brunt (n 3) 263. 139 Boral, [295]; Queensland Wire, 189–190; Re Chime Communications Pty Ltd (No 2) [2009] ACompT 2, 51-53; Ray Steinwall, Annotated Competition and Consumer Legislation (LexisNexis Butterworths 2018) 505; Miller (n 12) 97. 140 ibid; Arnotts, 338; Guidance on the Commission’s Enforcement Priorities, para 17. 141 Steven C Salop, ‘Strategic Entry Deterrence’ (1979) 69 American Economic Review 335. The Tribunal in Re Southern Cross Beverages recognized this type of strategic barriers to entry (Re Southern Cross Pty Ltd (1981) 50 FLR 176, 206). See also United Brands, paras 67–68. 142 Re Qantas Airways Ltd [2004] ACompT, [306]. 143 Duke (n 10) 128–29. 144 See Stucke and Grunes (n 77) 200. 145 See Facebook/WhatsApp, para 109; Schweitzer and others (n 72) 60. 146 Schweitzer and others (n 72) 28; Shapiro and Varian (n 58) 184–85. 147 Rubinfeld and Gal (n 68) 49. 148 Schweitzer and others (n 72) 60. 149 Rubinfeld and Gal (n 68) 49; Khan (n 15) 773 and 786. 150 See eg Graef (n 66). 151 Google and Facebook’s strategic acquisitions are not treated under the heading ‘barriers to entry and expansion’ but under a separate heading. Those acquisitions nonetheless constitute barriers to entry and expansion. 152 DPI, 74 and 80. 153 See above the section ‘The ‘Newest’ economy’. 154 DPI, 66. 155 DPI, 73–74 and 79–80. 156 Queensland Wire, 190. 157 ibid. 158 DPI, 73, citing Crémer and others (n 68) 2, 20. 159 Rubinfeld and Gal (n 68) 49. 160 Stephen G Corones, ‘Applying an Effects Test Under s 46 of the Competition and Consumer Act’ (Conference Paper, Bar Association QLD Event, 25 July 2016), p 4. 161 Arnotts, 338. 162 QCMA, 520. 163 Hilmer Report, 3. 164 ibid. 165 ibid. 166 ibid 4. In the context of merger analysis For the purposes of s 50(3)(g) CCA mentions that product dynamic characteristics include growth, innovation and product differentiation. The Tribunal in QCMA also discussed innovation (see QCMA, 187–88). 167 DPI, 72 and 79. 168 DPI, 72. 169 DPI, 98. 170 DPI, 42. 171 DPI, 41–42. 172 This is also reflected in see s 46(5)(b)(ii). 173 See DPI, 68–69. 174 Lisa M Segarra, ‘Google to Pay Apple $12 Billion to Remain Safari’s Default Search Engine in 2019: Report’ (Fortune, 30 September 2018) <https://fortune.com/2018/09/29/google-apple-safari-search-engine/> last accessed 1 December 2020. 175 See Schweitzer and others (n 72) 14. 176 From 2015 to 2018, Nest was part of Google’s parent company Alphabet and operated independently of Google. In 2018, Nest was merged into Google’s home devices unit. 177 See above. 178 See eg Queensland Wire; Melway; Universal Music; Taprobane Tours WA Pty Ltd v Singapore Airlines Ltd (1990) 33 FCR 158. 179 Queensland Wire, 190. 180 See OECD, ‘Data-Driven Innovation: Big Data for Growth and Well-Being’, 6 October 2015, <http://www.oecd.org/innovation/data-driven-innovation-9789264229358-en.htm> last accessed 1 December 2020. 181 DPI, 138–39. 182 DPI, 75. 183 ibid. 184 Duke (n 10) 143. 185 This factor has not been discussed in any misuse of market power case since QCMA, possibly because s 46(3) was added by the Trade Practices Amendment Act 1977 (Act 81 of 1977). 186 s 4A(5) CCA for the purposes of the Act deems a holding company and its subsidiaries to be related bodies corporate. 187 TPC v Ansett Transport Industries (Operations) Pty Ltd (1978) 32 FLR 305. 188 TPC v Pioneer Concrete (Qld) Pty Ltd (1994) 50 FCR 160. 189 ibid, para 14. 190 DPI, 74 and 80. See also Report of the Digital Competition Expert Panel, ‘Unlocking digital competition’ March 2019 (‘Furman Report’), para 1.49; Lear, ‘Ex-post Assessment of Merger Control Decisions in Digital Markets’ 9 May 2019. 191 Guidelines on Misuse of Market Power, para 2.15. 7 ibid 512. 8 See Carl Kaysen and Donald F Turner, Antitrust Policy: An Economic and Legal Analysis (Harvard University Press 1959) 75. 9 See also Queensland Wire Industries Pty Ltd v Broken Hill Pty Co Ltd (1989) 167 CLR 177; Katharine Kemp, Misuse of Market Power: Rationale and Reform (CUP 2018) 107. 94 See ibid 167. 95 See eg Robert H Bork, The Antitrust Paradox: A Policy at War with Itself (Basic Books 1978); Richard A Posner, ‘Conglomerate Mergers and Antitrust Policy: An Introduction’ (1969) 44 St. John’s Law Review 529. 96 Lim (n 91) 3. See also Nicolas Petit, ‘Technology Giants, the Moligopoly Hypothesis and Holistic Competition: A Primer’ (2016), <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2856502> last accessed 1 December 2020. 97 Bourreau and de Streel (n 59) 6. 98 ibid; Owing to the modularity of their product design, a digital output (eg a mapping service) can also be used as an input (eg for navigations systems). 99 Bourreau and de Streel (n 59) 11. For example, an Apple Watch can only be used together with an Apple iPhone and not with a smartphone from another manufacturer. This linkage between these two distinct but complementary products provides consumer with an additional benefit (a ‘consumption synergy’) when buying them together, even though the iPhone and the Apple Watch are not sold as a bundle by Apple. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Antitrust Enforcement Oxford University Press

Big data and competition analysis under Australian competition law: comeback of the structuralist approach?

Journal of Antitrust Enforcement , Volume Advance Article – Dec 21, 2020

Loading next page...
 
/lp/oxford-university-press/big-data-and-competition-analysis-under-australian-competition-law-AyR9aEHsaH
Publisher
Oxford University Press
Copyright
Copyright © 2021 Oxford University Press
ISSN
2050-0688
eISSN
2050-0696
DOI
10.1093/jaenfo/jnaa055
Publisher site
See Article on Publisher Site

Abstract

Abstract Competition assessment in Australia has traditionally been based on an evaluation of the market structure relying on five factors, namely the degree of market concentration, the height of barriers to entry, the extent of product differentiation, the extent of vertical integration, and the nature of arrangements between firms. These factors, known as the ‘QCMA factors’, are characteristic of competition in the manufacturing industries of the ‘old economy’. Since the ascendancy of Chicago and Post-Chicago School thinking competition analysis in Australia has also taken into consideration non-structural factors. However, in light of the dominance of big tech companies in online markets, the so-called ‘Neo-Brandeisian School’ has advocated focusing on structural elements that are characteristic of online markets. This article examines to what extent the QCMA factors still a suitable structural framework for the assessment of competition in online markets. I. INTRODUCTION Throughout the past 40–50 years we have seen significant changes to market structures, most notably in the area of high-technology industries. ‘Old economy’ manufacturing industries are on the decline, paving the way for a rise in industries in the ‘platform economy’. The creation of digital platforms including search engines, social media and e-commerce platforms, has revolutionized almost everything we do, from the way we interact, to the way we shop, dine, and travel. Around the turn of the millennium some competition scholars and antitrust agencies recognized a distinction between the ‘old economy’ industries in which modern competition legislation was first adopted in industrialized countries and ‘new economy’ high-technology industries that started to make big waves around the late 1970s. For example, the Trade Practices Act 1974 (TPA) [now the Competition and Consumer Act 2010 (CCA)], Australia’s first effective competition legislation, entered into force before the first IBM computer came to market in 1981. Tech giants like IBM and Microsoft dominated during the 1980s and 1990s and eventually became the target of antitrust scrutiny. On 18 May 1998, when the US Department of Justice filed its complaint against Microsoft in the internet browser case, the two tech heavyweights of today, Google and Facebook, were not even founded yet. This essay proposes that the ‘platform economy’ is an even ‘newer’ economy. The defining characteristic of most markets of the digital economy1 largely relates to the role of data which makes these markets distinct from others, and this poses challenges to competition law enforcement. Competition assessment under the CCA has for a long time been based on an evaluation of the market structure. In 1976, Australia’s then Trade Practices Tribunal (now the Competition Tribunal) handed down its seminal decision in Re Queensland Co-Operative Milling Association Ltd (‘QCMA’).2 In that decision the Tribunal stressed that to determine whether a market is operating competitively requires the examination of five elements of market structure. The five factors, which are now referred to as the ‘QCMA factors’,3 can be summarized as: (i) the degree of market concentration; (ii) the height of barriers to entry or expansion; (iii) the extent of product differentiation; (iv) the character of vertical relationships and the extent of vertical integration; and (v) the nature of any formal, stable and fundamental arrangements between firms.4 These five factors are characteristics of competition in the traditional manufacturing industries of the ‘old economy’. A closer look at the digital economy reveals that many high technology-based industries and markets have changed substantially. Nonetheless, the Australian Competition and Consumer Commission (ACCC) in its 2018 ‘Guidelines on misuse of market power’ still refers to the QCMA factors as a relevant checklist for determining competitive constraints.5 The aim of this article is to determine whether the QCMA factors remain an adequate structural framework for determining market power in online markets. The article proceeds as follows: the next section outlines Australia’s legislation and the QCMA case to provide context to the current state of affairs relating to competition analysis in Australia. Section III distinguishes between the different types of the economy and explains why the ‘platform economy’ is distinct. Section IV looks at developments in relation to the assessment of competition in online markets and evaluates the relevance of each of the QCMA factors in the platform economy. Section V concludes. II. THE STRUCTURAL APPROACH FOR THE ASSESSMENT OF COMPETITION UNDER THE CCA The Trade Practices Tribunal’s 1976 decision in QCMA is a seminal one under Australian competition law. The case concerned the proposed acquisition by the applicants, QCMA and Defiance Holdings, of Barnes, a rival flour milling firm in Queensland. The Trade Practices Commission denied the merger, after which the applicants sought a review of this determination from the Trade Practices Tribunal.6 The significance of the case lies in the Tribunal’s discussion of the meaning and purpose of ‘competition’ as well as the meaning of ‘market power’—a term that was not defined under the TPA. The Tribunal found: [w]hether firms compete is very much a matter of the structure of the markets in which they operate. The elements of market structure which we would stress as needing to be scanned in any case are these: the number and size distribution of independent sellers, especially the degree of market concentration; the height of barriers to entry, that is the ease with which new firms may enter and secure a viable market; the extent to which the products of the industry are characterised by extreme product differentiation and sales promotion; the character of ‘vertical relationships’ with customers and with suppliers and the extent of vertical integration; and the nature of any formal, stable and fundamental arrangements between firms which restrict their ability to function as independent entities.7 Market power comes from a lack of effective competitive constraint.8 Scanning these five factors can help to determine whether a market is characterized by effective competition or whether one or more firms possess market power. The Tribunal’s explanation of ‘competition’ in QCMA highlights the structural aspects of competition.9 The QCMA factors form a structural framework for assessing competition on a market.10 This framework stems from the ‘structure-conduct-performance (S-C-P) paradigm’ that was proposed by industrial organization economists in the US in the 1950s and 1960s (also known as the ‘Structuralist School’).11 The drafting of the TPA and Australian competition policy at that time was heavily influenced by US antitrust legislation.12 The S-C-P paradigm implies that the structure of the market, particularly high concentration accompanied by high barriers to entry, dictate that firms in those markets will engage in certain types of conduct which would then lead to poor economic performance.13 Unfavourable market structures including monopolies and oligopolies can use their existing market power to block new entrants;14 and use their greater bargaining power against consumers and suppliers to raise prices and degrade quality while maintaining profits.15 While it was initially assumed that the causal flow running from structure to conduct to performance was unidirectional, it was later realized that there may also be feedback effects under which conduct may influence structure.16 This position was also accepted by the High Court.17 In 1994, on the 20th anniversary of the TPA, Brunt commented that the Australian approach to the assessment of competition has been to use a truncated S-C-P approach that highlights the causal importance of market structure but that also goes to evidence on market conduct and performance.18 The methodology is truncated only to the extent necessary to establish the existence of significant discretionary power and if so, whether it is sufficiently significant to qualify for a finding that is ‘substantial’ as it is required by the TPA.19 However, in its 2009 Re Application by Chime Communications20 judgment the Tribunal’s analysis deviated from the structuralist approach. It explained that this approach ‘has been overtaken by developments in economic theory and by empirical assessments of competition in modern markets which attest to the fact that [this causal flow from structure to conduct and then performance] is by no means the dominant mechanism to explain market behaviour’.21 The Tribunal’s broader approach took into account market dynamics and firms’ strategic behaviour by examining non-structural factors such as new sources of potential supply and the competitive nature of relevant market participants.22 While the structuralist approach basically suggests that the presence of dominant companies weakens the competitive process, economic theory and practical experience in the last four decades have shown that competitive dynamics can function well even if a market has some very large players.23 This shift reflects developments since the Chicago School and other schools of law and economic thought. The Chicago School found mainstream acceptance in the 1970s and early 1980s. Chicago School commentators rejected the structuralist view and instead sought to apply insights of neoclassical price theory to antitrust analysis.24 They presumed that market outcomes reflect the interplay of standalone market forces and the technical demands of production.25 The focus of their analysis was on consumer welfare as the sole outcome of competition.26 A key proposition made by Chicagoans was that the pursuit of economic efficiency (consisting of productive and allocative efficiency but not dynamic efficiency) should be the exclusive goal of competition law.27 The Post-Chicago School, which has grown out of the criticism of a range of aspects of the Chicago School, has gained significant influence since the mid-1980s.28 While Post-Chicago scholarship recognizes the importance of efficiency and free markets, contrary to the Chicago School it cautions that markets are not automatically effective and may be prone to market imperfections.29 Post-Chicagoans also rely less on neoclassical price theory and place greater emphasis on the strategic behaviour of firms. As such they also argue that actively attempt to influence market conditions though strategic conduct instead of reacting passively to structural market conditions.30 There is a very noticeable linguistic and conceptual difference in relation to these factors between the Tribunal in QCMA and the ACCC in the updated ‘Guidelines on misuse of market power’ of 2018. The Tribunal’s language explicitly says that these factors are elements of market structure and that need to be considered in any case. The ACCC, on the other hand, indicates that the list of factors in QCMA is discretionary and non-exhaustive.31 In fact, the last three QCMA factors have not come up in many cases since QCMA. Furthermore, the ACCC omits any reference to market structure and only speaks of competitive constraint. The ‘QCMA factors’ (elaborated below in Section IV) are now one of the many factors courts may turn to when establishing the competitive constraints faced by a company.32 III. THE OLD, THE NEW, AND THE NEWEST ECONOMY The QCMA factors were established almost 45 years ago. The economy at that time, not just in Australia, was fundamentally different to today’s fast-paced and technologically-based economy. This section sets out the different stages of the transformation of the economy since QCMA. The ‘Old’ economy In 1975, one year before the QCMA ruling, America’s largest Fortune 500 corporations were Exxon Mobil, General Motors and Ford Motor.33 Companies operating during this time period resided within the ‘old’ economy on which competition law analysis was founded. The old economy is generally considered to include industries that manufacture traditional goods such as steel and automobiles.34 According to Posner the old economy has a clear set of defining characteristics including multi-plant and multi-firm production, stable markets, heavy capital investment, modest rates of innovation, and slow and infrequent entry and exit.35 While it was possible for a potential competitor to enter a traditional industry market and displace an incumbent, those characteristics meant that the process for doing so was relatively slow, and required a level of innovation that was commonly absent or particularly slow in such industries.36 Based on the characteristics of traditional industries the focus on the five QCMA factors does not come as a surprise. In the absence of a statutory monopoly, those industries are characterized by competition in the market. In other words, the market leaves room for more than one company, and market shares indicate the economic strength of the actual competitors on the market. And those market shares may be relatively stable due to the existence of barriers to expansion and entry typically in the form of economies of scale and sunk costs.37 Melway38 is one Australian case in which the characteristics of the traditional economy were made clear. The High Court agreed with the trial judge that the Melway directory had substantial market power, holding in excess of 80–90 per cent of the retail market share in the relevant market.39 Barriers to entry were considered to be substantial due to the methods required to produce and compile the street directories. The court noted it was ‘neither rational nor likely’ for a new entrant to enter the market unless a new form of technology was created.40 The Melway example explains some of the key characteristics of a market in the traditional economy: substantial barriers to entry maintained by limited levels of innovation, and high costs of production which could lead to sunk costs in the event a new entrant was unsuccessful in competing effectively with the incumbent. The ‘New’ economy The term ‘new economy’ is commonly used to refer to high-tech industries that generate high rates of innovation such as computer software and hardware, internet-based businesses, electronic communication, biotechnology, and aerospace.41 Software was the paradigmatic new economy product (until the recent emergence of Big Data) and the relevant product market in Microsoft. New economy industries typically lack many of the abovementioned features of the old economy,42 and instead are often characterized by high rates of innovation, network effects, supply-side economies of scale, switching costs, and first-mover advantages.43 A more distinctive characteristic, as noted by Evans and Schmalensee, is that ‘[in] many of these industries, firms engage in dynamic competition for the market – usually through research-and-development (R&D) competition to develop the “killer” product, service or feature that will confer market leadership and thus diminish or eliminate actual or potential rivals’.44 In other words, the market has capacity for only one firm, and the market will tip towards that one firm whose product or service will become the standard. That one or more of those characteristics may be essential in the context of a new economy industry, however, does not mean that such characteristics never appear in other industries or that all of those characteristics always appear in new economy industries.45 It rather suggests that those characteristics are critical in industries in which innovation, intellectual property, and technological change are central features.46 The Microsoft case triggered a debate among competition law commentators about the application of competition law to the new economy, and particularly the relationship between competition law and intellectual property rights. The US Department of Justice among other things alleged that Microsoft violated section 1 of the Sherman Act by unlawfully tying its Internet Explorer to Windows and unlawfully maintained a monopoly position in the PC operating system market in violation of section 2 of the Sherman Act. Microsoft disputed these allegations and contended that traditional antitrust concepts are not applicable to the new economy.47 Microsoft argued that competition in a technologically dynamic market such as the PC operating system market is characterized by Schumpeterian competition in which firms compete through innovation for temporary market dominance until they are displaced by the next wave of product advancements. It was therefore not able to exercise market power as it was constrained by potential entrants.48 A less extreme position around the time of the Microsoft trial suggested that as antitrust law must preserve the incentives for firms to innovate, the application of the rules to new economy should be altered to fully accommodate the dynamic competition in those markets.49 On the other hand, some scholars argued that antitrust law was ‘sufficiently supple’ to adequately deal with new economy industries.50 Noting the lack of consensus about the need for amendment of the monopolization doctrine to account for competition in the new economy the D.C. Court of Appeals applied the conventional doctrine.51 The ‘Newest’ economy Since the Microsoft trials high-tech industries, and the IT industry in particular, have developed further rapidly. In 2011, Surblyte used the term ‘newest economy’ which relates to the latest shift in the computer industry, namely ‘cloud computing’.52 More broadly this economy can be said to comprise digital or online markets, and is most known for digital platforms, which is why it is often referred to as ‘platform economy’.53 The ACCC’s Digital Platforms Inquiry (DPI) has defined digital platforms as ‘applications that serve multiple groups of users at once, providing value to each group based on the presence of other users’.54 A characteristic feature of digital platforms is that they act as intermediaries between those user groups. Enabled by the mobile Internet and heavily relying on the use of Big Data and algorithms, digital platforms started to take off at the end of the noughties,55 and even more so since the launch of smartphones. At the same time this era is also epitomized by the rise of ‘GAFA’—an acronym that stands for Google, Apple, Facebook, and Amazon.56 The main products/services of Google (founded in 1998), Facebook (founded in 2004) and Amazon (founded in 1994) are non-hardware based and relate to things like search, social networking, and e-commerce respectively. While Microsoft still holds a strong position in some aspects of the newest economy (eg public cloud software), they were unable to attain a strong market position in other emerging platform markets (eg mobile operating systems). The ‘antitrust troublemakers’ of late are some of the aforementioned tech companies, first and foremost Google.57 Industries in the new and the platform economies are both driven by innovation and are characterized by competition for the market. Moreover, due to substantial R&D costs, those industries typically exhibit supply-side economies of scale.58 In contrast, the marginal costs of an additional user to the platform is relatively low. Additionally, digital products and services use inputs that often enable economies of scope in product development.59 Similar to the discussion around the time of the Microsoft trial, some commentators are now debating again whether in relation to digital platforms antitrust doctrine is in need of an overhaul,60 or even whether competition law is obsolete.61 Unlike in the aftermath of Microsoft, this time governments and antitrust agencies are more receptive to the idea that competition in digital markets might be different,62 and as eg in the case of Germany, a few countries have already amended their competition laws accordingly.63 One specific question is whether new factors should be taken into account when assessing the market structure of digital markets and the market power of players therein. Again, that one or more of the above characteristics are common to the platform economy does not suggest that such characteristics never appear in other industries, nor that all or most of them always apply to digital platforms.64 Furthermore, it should be noted that platform companies can be very different from each other in terms of the markets they operate in and the effects of their activities on competition.65 It is submitted in this article that the entirely different business model of most digital platforms, and in particular the role of data, makes their market structure decisively distinct from industries in the new economy, and therefore justifies a different competition analysis. Digital advertising platforms such as Google and Facebook operate by providing free services to consumers in exchange for consent to the collection of the users’ personal data.66 The companies then sell advertising businesses the opportunity to serve their ads to a targeted and highly personalized online audience.67 Although data collection is often a by-product of the usual functioning of these platforms, it can create a competitive advantage for incumbents simply from having a large pool of data to access.68 At the same time, platforms can use data as a sharable input to develop new products and services.69 The most characteristic features of the platform economy, and one closely connected to the collection of data, is the role of network effects. A network effect arises when value of a product or service increases with the number of users.70 The ‘classic’ type of network effects can also be seen in the old and the new economy. For instance, a telephone network’s value increases to one person from other people also joining the network. Another type of network effect is a two- (or multi-)sided network effect. It was Microsoft’s dominance in the software market that gave prominence to the economic theory of two-sided network effects.71 Microsoft was alleged of using its large user base to encourage software developers and computer hardware manufacturers to focus their efforts on Microsoft’s Windows Operating System. With an increasing number of software and hardware available for Windows, the operating system attracted more users, which in turn led to more software and computer hardware specific to Windows. This positive feedback loop can cause the market to tip and it becomes dominated by one firm. This ‘winner takes all’ effect helps to explain why in many new economy industries competition is for the market.72 According to Page and Lopatka ‘the new economic theory of network effects … provided a lens through which Microsoft’s victories over its rivals appeared anticompetitive. The development of the network effects theory parallels the development of the very software markets in which Microsoft competed’.73 Digital platform markets exhibit multi-sided network effects apart from classic network effects. In the DPI, the ACCC observes that there are different kinds of multi-sided network effects which ensure that an increase in the number of individual users increases the value of the platform for advertisers.74 Most importantly, an increase in the number of users increases the users exposed to an ad campaign, which may increase an advertiser’s return from that campaign.75 And also a platform with more users has access to more data, which can enhance the relevance of ads presented to users.76 Importantly, the collection of data on digital platforms can result in network effects that extend beyond the type of network effects seen in the previous economies. Stucke and Grunes identify three ‘data-driven network effects’ that stem from the scale, scope, and the spill-over effect of data.77 These data-driven network effects can be illustrated with the example of Google Search. The more people use this search engine, the larger the scale of Google’s collected data which the company can then use to expand trial-and-error experiments allowing Google’s search algorithm to enhance the relevance of its searches, thus attracting more users.78 Additionally, Google’s large scope of data (ie variety of data), which the company collects across its platform through its other service like Google Maps and Gmail, also allows the company to improve its search algorithm.79 Furthermore, Google’s collection of personal data has a spill-over effects into other sides of the market.80 Google is significant for advertisers who buy the opportunity to advertise to a targeted audience. Google can invest this revenue to improve the quality of its search engine which will attract more users, and therefore more advertisers, eventually resulting in a virtuous circle. Finally, as elaborated in the next section, data-driven network effects motivate digital companies to adopt conglomeration strategies.81 IN NEED FOR NEW ASSESSMENT CRITERIA IN ONLINE MARKETS? In light of the profound differences between the old and the platform economy, the purpose of this section is to determine which of the QCMA factors are still relevant in the assessment of competition in online markets. The comeback of the structuralist approach in online markets In Australia, and in other jurisdictions, there has been a debate in recent years about adopting new approaches to competition assessment in online markets.82 This debate has largely been sparked by rapid growth and expansion of tech giants like Google, Facebook and Amazon. In the final report of the DPI the ACCC states that ‘[t]he pace of technological change needs to be matched by the pace of policy review. As digital markets and the use of data continue to grow and change, governments need to continue to consider the appropriate level of oversight’.83 In antitrust scholarship the dominance of Big Tech has also triggered another debate about the ideology of competition law.84 A pertinent question in the Australian context is whether the QCMA framework is still suitable for digital platform markets. For a long time the structuralist approach has been regarded as out of date and was eventually replaced by the Chicago and the Post-Chicago School thinking which has become known as the ‘modern antitrust’ doctrine in the US and in other jurisdictions.85 However, in the past few years a movement referred to as the ‘Neo-Brandeisian School’86 emerged to criticize this modern antitrust doctrine and its failure to tackle the antitrust problems involving ‘Big Tech’.87 Khan argues that the outcome-oriented consumer welfare approach ‘is inadequate to promote real competition, a failure that is amplified in the case of dominant online platforms’.88 This approach ‘equates harm entirely with whether a firm chooses to exercise its market power through price-based levers, while disregarding whether a firm has developed this power, distorting the competitive process in some other way’.89 Letting firms accumulate market power does not only make it more cumbersome to adequately check that power when it is eventually exercised, but these firms may also exploit their market power through non-price based anticompetitive practices.90 A concept that was widely accepted among the Structuralist School but later discredited by the Chicago School, and now highlighted by the Neo-Brandeisians is the conglomerate model.91 A ‘conglomerate’ describes a company that consists of different and distinct parts that are under one roof and operate on neighbouring markets, offering non-substitutable products to the same group of customers. Industrial organization economists are particularly wary of conglomerate mergers as they are more likely to increase concentration in particular markets, dampen competition and raise barriers to entry.92 They are concerned that multi-market firms are able to shift supra-competitive profits from one market to another and may engage in predatory pricing, reciprocity or other exclusionary or exploitative practices.93 In addition, structuralists warn that conglomerates could wield non-economic power.94 Chicagoans, on the other hand, argue that conglomeration theories and predatory pricing in particular are implausible and therefore highly unlikely.95 According to Lim the ‘return of the conglomerate model has coincided with an intriguing change in the nature of competition in the high-tech industry, particularly where platform businesses that harness network effects are involved’.96 In addition to the past conglomeration theories, Bourreau and de Streel attribute the emergence of digital conglomerates mainly to two characteristics of the platform economy: economies of scope in product development and product ecosystems.97 From a supply-side perspective, the possibility of collecting large amounts of data on consumers, facilitated by the modular design of data, may incentivise digital companies to expand in other markets in order to acquire new data on unattached consumers or complementary data on attached consumers.98 From a demand-side perspective, a product ecosystem can develop when firms create ties between their different products (even possibly unrelated products) to enhance the complementarity between them.99 The expansion strategy could result in platforms becoming gatekeepers.100 It also gives platforms an incentive to pre-emptively acquire start-ups who would pose a threat if they continued to grow.101 In light of the developments involving digital platforms and the inadequacy of price-based measures of competition to capture market dynamics, particularly given the role and use of data, Khan argues antitrust enforcement in online markets should focus on the competitive process and market structure.102 She does not advocate a strict return to the S-C-P paradigm but rather claims that a competition analysis that ignores the role of structure is misguided since ‘the best guardian of competition is a competitive process, and whether a market is competitive is inextricably linked to – even if not solely determined by – how that market is structured’.103 The consideration of a more structuralist approach in practice as proposed by Khan involves an assessment of how a market is structured and whether a single firm had obtained sufficient power to distort competitive outcomes.104 This approach considers various factors that shed light on the neutrality of the competitive process and the openness of the market including entry barriers, conflicts of interest, and the emergence of gatekeepers or bottlenecks.105 Khan illustrates the significance of market structure with the example of Amazon. She claims that Amazon has established dominance on the e-commerce market owing to two closely linked elements of its business strategy, namely its willingness to forego profits and investing aggressively in order to achieve scale, and integration across multiple business lines.106 In establishing Amazon’s structural dominance Khan refers to conventional factors, in particular barriers to entry and vertical integration—which are also two of the most important QCMA factors. Apart from the need for significant investments, Khan highlights advantages stemming from data collection and network effects as barriers to entry.107 While the emphasis of establishing market power is on structural factors, strategic conduct may still be relevant in the Neo-Brandeisian approach given that there is no strict adherence to the S-C-P paradigm. In their pursuit of rapid growth digital platforms such as Amazon and Uber engage in predatory pricing to psychologically intimidate competitors and keep them out.108 How would the QCMA factors fare in the platform economy? Under Australian competition law there are no immediate plans to introduce new market power assessment criteria with regard to section 46 that are specifically tailored to the platform economy. The DPI is the first occasion in which the ACCC assesses the market power of digital platforms.109 In the final report of the inquiry the ACCC claims that in principle ‘the existing tools and goals of the competition law framework remain applicable for digital markets’.110 However, at the same time there is no explicit mention of the QCMA factors or the relevance of a structural factors. This section explores how useful the QCMA factors are when applied to competition analysis in online markets and sheds light on the ACCC’s approach in the DPI. The number and size distribution of independent sellers Under Australian competition law, like in many other competition regimes, the number and size distribution of independent sellers, ie the degree of market concentration and market shares, are generally regarded as a first indicator of market power.111 The higher a firm’s market share is, the more likely it will hold market power or at least market power at a substantial degree.112 The traditional methodology for determining market power involves identifying actual competitors in a market, ie the companies currently operating in that particular market, and to determine their respective market shares.113 Yet, unlike in some jurisdictions,114 there is no market share threshold for determining significant market power.115 In markets, including innovation markets, where a company has held a very large market share that is considerably larger than its rivals and maintained that position over many years, that company can be deemed to have substantial market power. In Queensland Wire, BHP produced nearly 97 per cent of the steel made in Australia and supplied about 85 per cent of Australia's requirements for steel and steel products.116 In the DPI, the ACCC points out the Google has been the dominant general search engine in Australia with a market share of 93–95 per cent since 2009.117 Nonetheless, market shares are merely a first indicator and never conclusive on their own. In Boral,118 Gleeson CJ, and Callinan J said in relation to market concentration and its relevance that a ‘large market share may, or may not, give power. The presence or absence of barriers to entry into a market will ordinarily be vital’.119 In particular, in the new and platform economy market shares are a less reliable indicator of market power because market share figures only measure actual but not potential competition. Yet, in those markets incumbents may be constrained by potential competitors who have not yet entered the market but who would do so if a profitable opportunity were to arise.120 Furthermore, as noted in the 1993 Hilmer Report, a large number of competitors in a market is not necessarily always optimal. Early economic work suggested that effective competition in a market requires many firms, each with a small market share. However, in some markets competition between a few large firms may yield more economic benefit than competition between a large number of small firms. Such oligopolistic competition may arise as a result of economies of scale and scope, not only in production but also in marketing, technology and, increasingly, in management.121 As stated above, competition in the digital economy is often characterized by competition for the market rather than competition in the market, and online markets therefore tend to be highly concentrated and are often dominated by one company. In addition, network effects result in a positive feedback loop which can accelerate the process of a firm capturing the market, ie these markets have a tendency towards tipping. Yet, despite a large market share a dominant platform is not immune to a relatively quick displacement by another company.122 Cognisant of the limited value of market shares in markets involving digital platforms, the ACCC examines to what extent dynamic competition place competitive constraints on Google and Facebook.123 For instance in the case of Facebook the ACCC notes that: in the market for social media services, Friendster was initially leapfrogged by MySpace, which, in turn, was rapidly replaced by Facebook. However, the barriers to entry and expansion may be substantially higher now than in the early phase of the social media market. In particular, the global size of Facebook now dwarfs the size of MySpace at its peak.124 There has not been any notable consideration of dynamic competition in previous cases, possibly because the traditional markets in question lacked the necessary potential for innovation. Market shares in online markets are more volatile than in traditional markets and the usual time period for deeming market shares to be stable is generally shorter. In light of the difficulty associated with network effects and market tipping, a new provision that addresses the largest platforms with superior growth rates in a market as mooted in relation to the 10th amendment of the German Competition Act as an alternative to market shares in online markets.125 A similar provision would also be desirable under Australian competition law, especially since there is no market share threshold. Hence, regardless of whether a platform company has substantial market power, if it experiences superior growth rates compared to its competitors it might be subject to closer antitrust scrutiny. Another reason to be cautious of market concentration and market shares is the inherent dependence of these two concepts on market definition—a notoriously difficult exercise that has been criticized in more recent years.126 In the digital economy, market definition is even more problematic. First, digital companies often offer free products and services which makes it hardly possibly to apply price-based test like the SSNIP test. Non-price based tests to define the relevant market are hardly workable, absent well-accepted, quantifiable metrics.127 Secondly, since online markets are often multi-sided it is not sufficient to only look at the relevant market on which the defendant is operating.128 It is necessary to pay close attention to network effects and how competition on related markets is impacted. Moreover, even if online markets could be defined accurately, the overall size of a company and conglomerate power aggregated on a number of related markets seem to be at least as relevant as a company’s market share on the actual market that is under investigation. Overall, even though market concentration and market shares continue to be relevant in the platform economy they are not as indicative of market power as they used to be in the old economy. The height of barriers to entry Barriers to entry have been defined as any advantage that allows an incumbent to earn above-normal profits without the threat of entry.129 More broadly, barriers to entry (and expansion) describe any material difficulties that an incumbent’s actual or potential competitors face when seeking to enter the market or expand within it.130 Barriers to entry, not market shares, are the key to any assessment of market power.131 A high market share may not be a reliable indicator of market power if barriers to entry are low since a new entrant might quickly take away market shares from the incumbent. Conversely, if barriers to entry are substantial, and new entry therefore less likely, a high market share will be a sign of market power.132 In QMCA, the Tribunal identified the height of barriers to entry to be the most significant element of market structure: Of all these elements of market structure, no doubt the most important is (2), the condition of entry. For it is the ease with which firms may enter which establishes the possibilities of market concentration over time; and it is the threat of the entry of a new firm or a new plant into a market which operates as the ultimate regulator of competitive conduct.133 The Hilmer Report discussed how potential competition can have a similar effect on the performance of a firm as actual competition. It noted that ‘a market which is highly open to potential rivals – known as a highly “contestable” market – may be of similar efficiency as a market with actual head-to-head competition’.134 Since innovation markets are characterized by competition for the market, the number and size of sellers are less relevant than in traditional markets, and barriers to entry play an even more vital role. The test for determining the existence of a significant barrier to entry is to assess whether there is a credible threat of entry that will constrain the incumbent to behave competitively.135 And this is typically done by considering whether entry is likely, timely, and sufficient.136 For entry barriers to be considered low, a potential competitor’s entry into the market must be likely as a matter of commercial reality, happen within a time horizon of three years, and not merely occur at the fringe of the market.137 In light of the pace of innovation in the platform economy a shorter time horizon might be more appropriate. Simply put, the essential test for whether or not a significant barrier to entry exists is whether the threat of entry of whatever kind will constrain incumbents to behave competitively.138 Barriers to entry come in various forms and can generally be categorized into structural and strategic ones.139 The former result from structural characteristics of the market and may include among other things regulatory requirements, access to essential facilities, intellectual property rights, economies of scale, and sunk costs.140 Identifying structural barriers to entry is quintessential to the S-C-P paradigm. Strategic barriers to entry, on the other hand, stem from the deterrent activities by an incumbent to purposely hinder the possibilities of entry, as opposed to ‘innocent’ entry barriers that unintentionally arise as a side effect of innocent profit maximization.141 In Re Qantas Airways, the Tribunal found that ‘it is entry as a behavioural phenomenon that is important in the analysis of competition rather than considering entry as a purely structural constraint’.142 Strategic barriers to entry exemplify the move from static economic models to dynamic market behaviour as advocated by Post-Chicago scholars.143 The novelty of the ACCC’s approach in the DPI is the consideration of strategic entry barriers. Different types of barriers to entry may be relevant in digital markets compared to the ones that more typically arise in traditional markets. It has been suggested that if competition authorities and courts consider only traditional barriers to entry, then they may falsely conclude that online markets are easy to enter.144 Despite winner-takes-all competition, from a demand perspective (ie attracting users) barriers to entry in many online markets are low since apps and services are offered for free or a very low price, thus in theory allowing users to multi-home easily.145 Yet, in reality high switching costs caused by positive network effects render multi-homing less likely.146 For example, pre-installed apps on users’ mobile phones can be a gateway barrier to entry. The status quo bias may lead users to use an app over a potentially more innovative rival app.147 This hindering of multi-homing raises switching costs and constitutes a strategic entry barrier.148 From a supply perspective many online markets lack entries barriers that are common to traditional markets such as land and raw materials. However, digital platforms face other forms of entry barriers, particularly with regard to the collection, storage, synthesis and analysis of data.149 Certain types of data may therefore amount to an essential facility.150 In that respect, another significant strategic barrier to entry and expansion that the ACCC focuses on in the DPI are Google and Facebook’s strategic acquisitions.151 The ACCC argues that some of those acquisitions have allowed Google and Facebook to entrench their respective incumbent positions. Those acquisitions have weakened the constraint from dynamic competition by eliminating potential competitors.152 The expansion into related markets has also provided Google and Facebook with advantages in terms of scope and network in order to collect more data. Network effects constitute by far the most significant barrier to entry in the platform economy. While network effect may also arise in the old and the new economies, as mentioned above, the three ‘data-driven network effects’ that stem from the scale, scope, and the spill-over effect of data are unique to the platform economy.153 Network effects (both same-side and cross-side) are the first barriers to entry that the ACCC analyses in relation to Google and Facebook. For instance in the case of Google, the ACCC finds that: all else being equal, a large amount of data improves the relevance algorithm in the search engine, increasing the quality of the search service. A greater quantity of user data, including data on user searches and user interactions with search results, allows the Google relevance algorithm to update in a timely fashion, improving its relevance ranking.154 Data-driven network effects may be reinforced by economies of scope and conglomeration effects, which the ACCC in the DPI discusses separately and not as a barrier to entry.155 A barrier to entry that is relevant to both traditional and online markets are economies of scale. In Queensland Wire, Mason CJ and Wilson J observed ‘[w]here the economies of scale in a market are such that the minimum size for an efficient firm is very large relative to the size of the market, it may be that potential competitors will be dissuaded from entering the market by the apprehension that only one firm would survive’.156 This is known as the supply-side of economies of scale. In old-economy industries, the high capital costs of manufacturing industries ensured that incumbents were protected from new players entering the protected market.157 In online markets, the demand-side of economies of scale (whereby the value of a product increases in line with the increasing number of users) may constitute an almost insurmountable barrier. In the DPI, the ACCC reiterates the finding in the report on digital platforms written for the European Commission stating that while economies of scale is a feature of a range of industries, ‘the digital world pushes it to the extreme and this can result in a significant competitive advantage for incumbents’.158 Since costs in the digital economy become marginal once the platform is set up, an incumbent is able to retain its market power relative to a new entrant. Its revenue will also continue to increase as new users and advertisers join the platform. This advertising revenue can be used to further invest into R&D. This could possibly lead to a positive-feedback loop. Two-sided markets require new entrants to achieve growth on both sides of the market in order to compete effectively. If there are high switching costs on top of demand-side economies of scale and network effects, potential entrants might be even more deterred as there is considerable uncertainty in innovation markets about recouping investment in R&D. Since competition is for the market, a potential entrant’s innovation must not just be a minor improvement. The innovation must rather be so substantial that people desert the incumbent and switch to the new entrant. In other words, the potential entrant must disrupt the market. Barriers to entry remain the most significant QCMA factor in the platform economy. The concept of barrier to entry is wide enough to cover a number of constraints—both structural and strategic. Moreover, many characteristics of the digital platforms—in particular network effects and economies of scale—may result in re-enforcing barriers to entry.159 Product differentiation Product differentiation denotes the ways in which firms seek to distinguish their products or services from those of their rivals in order to establish customer or brand loyalty.160 Where products are similar, they exercise constrain on each other. On the contrary, product differentiation gives a firm a degree of control on the price of its product and is thus an indicator of market power. In addition, the preference of buyers attached to an existing brand (ie brand loyalty) creates a barrier to entry for new firms.161 In QCMA, the Tribunal discussed that there was little product differentiation between the types of flour between the different firms.162 Since QCMA, new insights have altered our understanding of the implications of product differentiation on competition. Some of those insights are discussed in the Hilmer Report. First, the idea that firms provide identical products or services and compete mainly on price is the simplest notion of competition and the exception rather than the rule.163 In practice, work in business strategy suggests that competition arises when firms seek to offer consumers different mixes of benefits, some of which are already reflected in price and others of which are reflected in non-price elements such as service, quality, or timeliness of delivery.164 Secondly, competition is not always between identical products or services. The essence of competition is the striving to meet the same consumer need and this is reflected in the ways in which this is met by different market participants.165 Markets are also characterized by dynamic efficiency, ie the need for firms to make timely adjustments to their products in response to changes in consumer tastes and in productive opportunities.166 In the DPI, the ACCC still addresses product differentiation. The ACCC considers that the Google and Facebook’s branding constitutes a barrier to entry.167 In relation to Google the ACCC notes that ‘the strength of Google’s brand partly reflects the high quality of its search service. When first developed, the Google algorithm provided an innovative method for ranking the relevance of search results. Google invests a considerable sum each year on improving the quality of its service’.168 Moreover, with regard to display advertising, the ACCC argues that ‘the advertising inventory on social media platforms is a specific kind of display advertising – social media advertising – which is differentiated from other kinds of display advertising’.169 Dynamic efficiency is crucial in the platform economy and it arguably diminishes the role of product differentiation in some degree. Instead product imitation arguably plays a larger role than in the previous economies. Incumbent companies are generally wary of new entrants because they may challenge their market power. As explained above, in digital markets new entrants need to be highly innovative in order to displace an incumbent. Digital platforms compete on features, not price, and some of the major digital platforms offer combinations of services. The ACCC observes that the services provided by digital platforms are constantly changing due to technological advancement and shifts in consumer preferences.170 For instance, Facebook began as a consumer-facing social media service that allowed communication between networked users, but it now includes online marketplaces for goods and jobs, and Snapchat, which was initially a medium for creating and privately sharing photo-based content with other networked users, later expanded to include public content services.171 Major platforms are often easily able to introduce new features that mimic popular features of their rivals. For example, Facebook’s introduction of ‘stories’ to its social media platform was reminiscent of Snapchat’s core feature—the creation and sharing of multimedia messages referred to as ‘snaps’. For tech firms it is also relatively easy to remove a new feature if it is not well received by its users. The addition and removal of features on a digital platform is different to markets in the old economy where firms would more likely incur substantial sunk costs in the adaption of the production process as a consequence of when altering or updating a product. Moreover, it may be costly and time-consuming to reverse those changes. Similarly, tech start-ups may find it difficult to engage in product imitation as they might lack the necessary resources. Large platforms can use product differentiation, on the one hand, to build customer loyalty, and product imitation, on the other hand, to ward off smaller competitors. This strategy increases the likelihood of single-homing and users becoming locked in on a platform’s ecosystem. Single-homing and user lock-in are reinforced when a digital platform is able to use data to individualize products and services according to each individual user’s preference. Vertical relationships and vertical integration A company that at first sight appears to have SMP because of its considerable market share while shielded by high barriers to entry may in fact not be dominant due to its vertical relationships with customers and suppliers.172 For instance, a powerful buyer, who makes up a significant share of the upstream company’s sales, exercises competitive constraint over that company. If the upstream company were to raise the price of its product the buyer could threaten to switch to one of the latter’s competitors and the upstream company would probably not increase the price. Similarly, a company might be dependent on a powerful supplier that also exercises competitive constraint over it. An interesting vertical supply relationship in the platform economy can be seen between Google and Apple, who are competitors on some markets.173 Google was estimated to pay US$12 billion in 2019 for Apple to install Google Search as the default search engine on Apple’s Safari web browser, up from US$1 billion in 2014.174 In order to reduce dependence on upstream and/or downstream markets a company may opt to vertically integrate. Vertically integration arises when two or more successive stages of production and/or distribution are combined under the same control. A producer of goods or services can for instance carry out both the production and distribution functions itself, using its own employees or its own branches or through its wholly owned subsidiaries. Vertical integration may be vital for a platform company to gain more user data for better target advertising.175 For example in 2014 Google bought smart gadgets maker Nest Labs for US$3.2 billion.176 In 2018, Nest was merged into Google’s home devices unit. The brand Google Nest now offers products such as smart assistants that are integrated with Google Search. The integration of production and distribution processes as well as complementary products or services may ostensibly be considered as a pro-competitive efficiency. Yet, in some circumstances vertical integration may give rise to structural and strategic barriers to entry.177 Vertical integration constitutes a structural barrier to entry for instance where a firm has access to an essential facility through its parent firm or subsidiary, while other firms seeking to enter the market or expand within it lack that possibility and are dependent their competitor’s parent or subsidiary. In practice such a situation has often led to exclusionary abuses such as refusal to supply or margin squeeze.178 In Queensland Wire, Mason CJ and Wilson J observed that vertical integration makes it possible for the monopoly company to engage in discrimination, for instance in pricing between various users in the case of a distribution chain.179 When a firm has developed a reputation for engaging in exclusionary practices this can amount to a strategic barrier to entry. In addition to exclusionary abuses, vertical integration can lead to exploitative abuses. Even though production- and distribution chains are far less common in online markets, most of the largest firms are vertically integrated. After gaining enormous financial strength some of those firms achieved further vertical integration through major acquisitions.180 The ACCC has also recognized the potential anti-competitive effects of vertically integrated firms in the platform economy.181 For example, Google’s acquisition of online advertising agency DoubleClick for US$3.2 billion in 2008 significantly strengthened Google’s market power in search and search advertising.182 Hence, vertical relationships and vertical integration are still relevant in the digital economy. They allow digital companies to favour their own business interests above those of advertisers or other competing businesses,183 eg through the practice of sharing competitively relevant data between parent companies and their subsidiaries. Besides vertical integration, a regulatory authority should pay close attention to conglomeration which appears to be more common structural characteristic in online markets than in more traditional markets. Arrangements between firms The fifth and last QCMA factor requires examining the nature of any formal, stable and fundamental arrangements between firms which restrict their ability to function as independent entities. This goes beyond determining whether a company is competitively constrained by vertical relationships with their customers or suppliers and possibly relate to conglomerate effects. If arrangements exist that restrict the ability of a firm to operate as an independent entity, then they need to be taken into account in assessing market power.184 This QCMA factor is also reflected in section 46(3) CCA.185 This provision captures inter-corporate relationships and arrangements that give rise to substantial market power. To that end, it allows the aggregation of market power between related bodies corporate.186 While the mere fact of being part of a corporate group by itself does not indicate market power,187 it is possible for a subsidiary to exercise market power on the basis of its parent’s support. In TPC v Pioneer Concrete,188 the TPC argued that the defendant, as a subsidiary of Pioneer International Ltd and a member of the Pioneer International holding company, had substantial market power. The holding company gave Pioneer Concrete access to raw materials and the financial backing to sustain financial losses to maintain a substantial market share in all the pre-mixed concrete markets in Australia.189 In the DPI, the ACCC argues that some of Google and Facebook’s strategic acquisitions have enabled them to entrench their respective incumbent positions. Those acquisitions have not only weakened the constraint from dynamic competition by eliminating potential competitors,190 but the expansion into related markets has also provided Google and Facebook with advantages of scope in order to collect more data. Both section 46(3) and the last QCMA factor could therefore arguably be used to establish a firm’s conglomerate market power. CONCLUSION Despite the criticism over the structuralist approach since the rise of the Chicago School, this approach is still not obsolete. Neo-Brandeisians strongly argue that structural factors and concepts should be considered in the assessment of competition in online markets. In Australia, the structuralist approach has been predicated on the five structural factors in the QCMA case. These five factors, sometimes referred to as the QCMA factors, concern market concentration; barriers to entry; product differentiation; vertical integration, and the nature of arrangement between firms. The QCMA factors were borne out of the old economy which was characterized by competition in the market. Already competition in high-technology markets of the ‘new economy’ raised question marks over the usefulness of antitrust principles that were developed in old economy times. Those markets are characterized by competition for the market. Nonetheless, competition regimes were not amended to reflect this new type of competition and the underlying structural and dynamic factors. While the ‘newest economy’ shares many characteristics with its predecessor, the role of data-driven network effects justifies the application of modified assessment approach in online markets. The ACCC’s market power analysis deviates considerably from the conventional assessment approach. Already in the Guidelines on Misuse of Market Power (2018) the ACCC states that the list of factors in QCMA is discretionary and non-exhaustive.191 The Guidelines also omit any reference to market structure and only speak of competitive constraint. Moreover, in Re Application by Chime Communications the Competition Tribunal’s analysis deviated from the structuralist approach by adopting a broader approach that took into account market dynamics and firms’ strategic behaviour. While there is no explicit mention of the QCMA factors in the DPI, the ACCC still considers the structure of the relevant markets and discusses some of the QCMA factors albeit implicitly as part of other market aspects. However, in its assessment the ACCC supplements some of the structural factors with dynamic and strategic ones. This can mostly be seen in relation to market concentration and barriers to entry and expansion. Overall, even though market structure still matters, there is no complete comeback of the structuralist approach. Footnotes 1 This article uses the terms ‘digital economy’ and ‘platform economy’ interchangeably. 2 Re Queensland Co-Operative Milling Association Ltd (1976) 8 ALR 481 [hereafter ‘QCMA’]. 3 Maureen Brunt, Economic Essays on Australian and New Zealand Competition Law (Kluwer Law International 2003) 316. 4 QCMA, 512. 5 ACCC, Guidelines on Misuse of Market Power (31 August 2018), para 2.15 [hereafter ‘Guidelines on Misuse of Market Power’]. 6 QCMA, 481. 10 See Rhonda L Smith and David K Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5(1) Agenda 25, 25; Arlen Duke, Corones’ Competition Law in Australia (7th edn, Thomson Reuters 2019) 48. 11 George Raitt, ‘Competition and Efficiency Effects in Europe, North America and Australia’ (2016) 24 Competition & Consumer Law Journal 187, 189; Brunt (n 3) 315. 12 See Brunt, ibid 315; Russell V Miller, Miller’s Australian Competition Law and Policy (3rd edn, Thomson Reuters 2018) 28. 13 Edward Mason, ‘The Current Status of the Monopoly Problem in the United States’ (1949) 62 Harvard Law Review 1265; Joe S Bain, Industrial Organization (2nd edn, Wiley 1968); Kaysen and Turner (n 8). See also Brunt (n 3) 294. 14 See Maureen Brunt, ‘Economic Overview’, Mimeograph, Lecture No 11, Monash Trade Practices Lectures, 1975. 15 Lina M Khan, ‘Amazon’s Antitrust Paradox’ (2017) 126 Yale Law Journal 710, 718. 16 Frederic M Scherer, Industrial Market Structure and Economic Performance (3rd edn, Houghton Mifflin 1980) 4–5. 17 Queensland Wire, 200. See also Kemp (n 9) 56. 18 Brunt (n 3) 258 and 317–18. 19 ibid 317–18. Brunt argues: ‘Sometimes all that will be necessary is what the Americans call the “quick look”, that is, the highly truncated rule of reason approach by which one might seek to eliminate the existence of likelihood of any market power in any conceivably relevant market. Then, indeed, all that may be necessary is to establish an absence of significant barriers to entry, and market definition can be primitive. But where the question revolves around the measurement of market power or the true character of business conduct and its probable consequences, the evidence that is called for may well relate to elements of market structure, market conduct, market performance, and their interrelation.’ (ibid 219). 20 Re Application by Chime Communications Pty Ltd [No 3] [2009] ACompT 4 (24 August 2009). 21 ibid [11]. See also Smith and Round (n 10) 26; Alexandra Merrett, ‘Understanding Market Power’ in John Duns and others (eds), Comparative Competition Law (Edward Elgar 2015) 109. 22 ibid [11]–[12]. See also Smith and Round (n 10) 26; Caitlin Davies and Luke Wainscoat, ‘Not Quite a Cartel: Applying the New Concerted Practices Prohibition’ (2017) 25 Competition and Consumer Law Journal 173, 195; Lindsay Foster and Hanna Kaci, ‘Concerted Practices: A Contravention without a Definition’ (2018) 26 Competition and Consumer Law Journal 1, 13. 23 Gunnar Niels and others, Economics for Competition Lawyers (OUP 2011) 180. 24 See Andrew I Gavil and others, Antitrust Law in Perspective: Cases, Concepts and Problems in Competition Policy (3rd edn, West Publishing 2017) 71–72. 25 Khan (n 15) 719. 26 Robert H Bork, The Antitrust Paradox: A Policy at War with Itself (Basic Books 1978) 108. It has been widely argued that Bork defines consumer welfare not as consumer surplus but as total welfare. To date there is no consensus as to how consumer welfare is to be understood in US antitrust (see eg Barak Orbach, ‘Foreword: Antitrust’s Pursuit of Purpose’ (2013) 81 Fordham Law Review 215; Eleanor M Fox, ‘Against Goals’ (2013) 81 Fordham Law Review 2157). 27 Herbert Hovenkamp, Principles of Antitrust (West Academic Publishing 2017) 48. 28 Gavil and others (n 24) 75. 29 See Michael H Riordan and Steven C Salop, ‘Evaluating Vertical Mergers: A Post-Chicago Approach’ (1995) 63 Antitrust Law Journal 513. See also Herbert Hovenkamp, The Antitrust Enterprise: Principle and Execution (Harvard University Press 2005) 156. 30 M Sean Royall, ‘Symposium: Post-Chicago Economics - Editor’s Note’ (1995) 63 Antitrust Law Journal 445, 447. See also E Thomas Sullivan and others, Antitrust Law, Policy, and Procedure: Cases, Materials, Problems (7th edn, LexisNexis 2014) 56. Besides the Post-Chicago School other notable scholarships include ‘neo-Chicago’ and behavioural antitrust (see Hovenkamp (n 27) 52). 31 Guidelines on Misuse of Market Power, para 2.15. 32 s 50(3) CCA sets out a non-exhaustive list of factors that the courts must consider in any merger assessment. 33 Fortune 500, ‘Fortune500 Company Search: A database of 50 years of FORTUNE’s list of America’s Largest Corporations’, <https://archive.fortune.com/magazines/fortune/fortune500_archive/full/1960/> last accessed 1 December 2020. 34 Richard A Posner, ‘Antitrust in the New Economy’ (2001) 68 Antitrust Law Journal 925, 926. 35 ibid. 36 ibid. 37 See eg Arnotts Ltd v TPC (1990) 24 FCR 313. 38 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1. 39 ibid 11 [10]. 40 ibid. 41 Daniel Gifford and Robert Kudrle, ‘Antitrust Approaches to Dynamically Competitive Industries in the United States and the European Union’ (2011) 7 Journal of Competition Law and Economics 695, 695; Alison Jones and Brenda Sufrin, EU Competition Law: Text, Cases, and Materials (6th edn, OUP 2016) 48. 42 Posner (n 34) 926. 43 Antitrust Modernization Commission, Report and Recommendations (April 2007), 32–33; David S Evans and Richard Schmalensee, ‘Some Economic Aspects of Antitrust Analysis in Dynamically Competitive Industries’ in Adam B Jeffe and others (eds), Innovation Policy and the Economy, Volume 2 (MIT Press 2002) 8–13; Christian Ahlborn and others, ‘Competition Policy in the New Economy: Is European Competition Law Up to the Challenge?’ (2001) 5 European Competition Law Review 156, 158–161; J Gregory Sidak and David J Teece, ‘Dynamic Competition in Antitrust Law’ (2009) 4 Journal of Competition Law and Economics 581, 585. 44 Evans and Schmalensee, ibid 8–13; Damien Geradin and others, ‘DG Comp’s Discussion Paper on Article 82 Implications of the Proposed Framework and Antitrust Rules for Dynamically Competitive Industries’ (2006) 12, available at <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=894466> last accessed 1 December 2020. 45 Antitrust Modernization Commission (2002), 33. See also Jonathan M Jacobson, ‘Do We Need a “New Economy” Exception for Antitrust?’ (Fall 2001) Antitrust 89, 89. 46 Antitrust Modernization Commission (2002), 33. 47 United States v Microsoft Corp, 253 F3d 34, 49–50. 48 ibid, See Howard A Shelanski and J Gregory Sidak, ‘Antitrust Divestiture in Network Industries’ (2001) 68 University of Chicago Law Review 1, 11–12. 49 See Robert Pitofsky, ‘Challenges to the New Economy: Issues at the Intersection of Antitrust and Intellectual Property’ (2001) 68 Antitrust Law Journal 913, 916–17; J Gregory Sidak, ‘An Antitrust Rule for Software Integration’ (2001) 18 Yale Journal on Regulation 1, 27. See also Geradin and others (n 44) 20–24. 50 Posner (n 34) 92. 51 United States v Microsoft, 49–50; Gifford and Kudrle (n 41) 696. In the EU trial, the General Court also applied an unmodified approach to art 102 TFEU (Case T-201/04 Microsoft v Commission, ECLI:EU:T:2007:289). 52 See Gintare Surblyte, The Refusal to Disclose Trade Secrets as an Abuse of Market Dominance – Microsoft and Beyond (Stämpfli 2011) 143. 53 This article uses the terms ‘platform economy’ and ‘digital economy’ to distinguish them from the ‘new economy’. 54 ACCC, Digital Platform Inquiry, Final Report (June 2019), 41 [hereafter ‘DPI’]. 55 See Patrick Barwise and Leo Watkins, ‘The Evolution of Digital Dominance: How and Why We Got to GAFA’ in Martin Moore and Damian Tambini (eds), Digital Dominance: The Power of Google, Amazon, Facebook and Apple (OUP 2018) 21, 22. 56 ibid. See also Farhad Manjoo, ‘The Great Tech War of 2012: Apple, Facebook, Google, and Amazon battle for the Future of the Innovation Economy’ (Fast Company, 19 October 2011) <https://www.fastcompany.com/1784824/great-tech-war-2012> last accessed 1 December 2020. 57 See eg Google Search (Shopping) (Case COMP/AT.39740) [2018] OJ C9/11; Google Android (Case COMP/AT.40099) [2019] OJ C402/07; Google Search (AdSense) (Case COMP/AT.40411).. 58 See Carl Shapiro and Hal R Varian, Information Rules: A Strategic Guide to Network Effects (Harvard Business School Press 1999) 3. 59 Marc Bourreau and Alexandre de Streel, ‘Digital Conglomerates and EU Competition Policy’ (March 2019), 9, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3350512> last accessed 1 December 2020. 60 Rupprecht Podszun, ‘The More Technological Approach: Competition Law in the Digital Economy’ in Gintare Surblyte (ed), Competition on the Internet (Springer 2015) 101; Pablo Solani Díaz, ‘EU Competition Law Needs to Install a Plug-in’ (2017) 40 World Competition 393; Torsten Körber, ‘Analoges Kartellrecht für digitale Märkte?’ (2015) 65 Wirtschaft und Wettbewerb 120. 61 Thomas Jaeger, ‘Brauchen wir das Kartellrecht noch?’ Kartellrechtsanwendung im digitalen Umfeld am Beispiel Google (2015) 65 Wirtschaft und Wettbewerb 702. See also Körber, ibid 120. 62 See Autorité de la Concurrence and Bundeskartellamt, Competition Law and Data, Joint Report, 10 May 2016. 63 See 9th Amendment of the German Competition Act. 64 cf Antitrust Modernization Commission (2002), 33. See also Jonathan M Jacobson, ‘Do We Need a “New Economy” Exception for Antitrust?’ (Fall 2001) Antitrust 89, 89. 65 Kenneth A Bamberger and Orly Lobel, ‘Platform Market Power’ (2017) 32 Berkeley Technology Law Journal 1051, 1053. 66 Inge Graef, EU Competition Law, Data Protection and Online Platforms: Data as Essential Facility (Kluwer Law International 2016) 9–16; Antonio Capobianco and Anita Nyeso, ‘Challenges for Competition Law Enforcement and Policy in the Digital Economy’ (2018) 9 Journal of European Competition Law and Practice 19, 20; Lapo Filistrucchi and others, ‘Market Definition in Two-sided Markets: Theory and Practice’ (2013) 10 Journal of Competition Law and Economics 293, 333. 67 Graef, ibid 14; Rolf Weber, ‘Competition Law Issues in the Online World’ (Forum Paper, 20th St. Gallen International Competition Law Forum, 5 April 2013). 68 Jacques Crémer and others, Competition Policy for the Digital Era, European Commission Report No B-1049, 4 April 2019, 24; See also Daniel L Rubinfeld and Michal S Gal, ‘Access Barriers to Big Data’ (2017) 59 Arizona Law Review 339, 357. 69 Bourreau and de Streel (n 59) 10. 70 Michael L Katz and Carl Shapiro, ‘Network Externalities, Competition, and Compatibility’ (1985) 75 American Economic Review 424, 424–25. 71 US v Microsoft, 49; Case T-201/04 Microsoft v Commission, ECLI:EU:T:2007:289, para 106. 72 Schweitzer and others, ‘Modernisierung der Missbrauchsaufsicht für marktmächtige Unternehmen’, Projekt im Auftrag des Bundesministeriums für Wirtschaft und Energie (BMWi) Projekt Nr. 66/17 (29 August 2018) 12. 73 William H Page and John E Lopatka, The Microsoft Case: Antitrust, High Technology, and Consumer Welfare (University of Chicago Press 2007) 22. 74 DPI, 63. 75 ibid. 76 ibid 77 See Maurice E Stucke and Allen P Grunes, Big Data and Competition Policy (OUP 2016) 170–216. See also Jens Prüfer and Christoph Schottmuller, ‘Competing with Big Data’ (2017) TILEC Discussion Paper No 2017-006, <https://pure.uvt.nl/ws/portalfiles/portal/15514029/2017_007.pdf> last accessed 1 December 2020. 78 Stucke and Grunes, ibid 170. See also Geoffrey A Manne and Joshua D Wright, ‘Google and the Limits of Antitrust: The Case Against the Case Against Google’ (2010) 34 Harvard Journal of Law and Public Policy 171. 79 ibid 188. This dynamic is also referred to as demand-side economies of scale: See Pt 4B(2). 80 ibid. 81 Bourreau and de Streel (n 59) 11. 82 See eg DPI; Report of the UK Digital Competition Expert Panel, ‘Unlocking Digital Competition’ (March 2019); Crémer and others (n 68); Schweitzer and others (n 72). 83 DPI, 3. 84 See eg Ariel Ezrachi and Maurice E Stucke, ‘The Fight over Antitrust’s Soul’ (2019) 9 Journal of European Competition Law & Practice 1; Sandra Marco Collino, ‘The Antitrust F Word: Fairness Considerations in Competition Law’, The Chinese University of Hong Kong Faculty of Law Research Paper No 2018-09, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3245865> last accessed 1 December 2020. 85 See Raitt (n 11) 189; Smith and Round (n 10) 25; Herbert Hovenkamp, Federal Antitrust Policy: The Law of competition and its Practice (West Publishing 2005) 42–47. 86 Sometimes derogatively called ‘hipster antitrust’. For a critical view on the Neo-Brandeisian School see e.g. Joshua D Wright and Aurelien Portuese, ‘Antitrust Populism: Towards a Taxonomy’ (2020) 25 Stanford Journal of Law, Business & Finance 131; Joshua D Wright and others, ‘Requiem for a Paradox: The Dubious Rise and Inevitable Fall of Hipster Antitrust’ (2019) 51 Arizona State Law Journal 293; Seth B Sacher and John M Yun, ‘Twelve Fallacies of the “Neo-Anti000trust” Movement’ (2019) 26 George Mason Law Review 1491. 87 See eg Tim Wu, The Curse of Bigness: Antitrust in the New Gilded Age (Columbia Global Reports, 2018); Daniel A Crane, ‘How Much Brandeis Do the Neo-Brandeisians Want?’ (2019) 64 Antitrust Bulletin 531. 88 Khan (n 15) 744. 89 ibid 745. 90 ibid. 91 Yong Lim, ‘Tech Wars: Return of the Conglomerate – Throwback or Dawn of a New Series for Competition in the Digital Era?’ (2017), p 3, <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3051560> last accessed 1 December 2020. 92 ABA Antitrust Section, Monograph No 7, Merger Standards under U.S. Antitrust Laws (1981), 162. 93 ibid. 100 ibid 19; Schweitzer and others (n 72) 8 and 85. 101 ibid 21; Lim (n 91) 11. 102 Khan (n 15) 737. 103 ibid 745. 104 ibid 746. 105 ibid. 106 ibid 746–47. 107 ibid 768 and 772. 108 ibid 772. 109 The Google/Fitbit merger is the first case on digital platforms in Australia but so far the ACCC has only published the statement of issues. 110 DPI, 138. 111 See eg Queensland Wire, 189–90; ACCC v Universal Music (2001) 115 FCR 442, 530 [381]; Case 85/76 Hoffmann-La Roche v Commission, ECLI:EU:C:1979:36, para 275; European Commission, Guidance on the Commission’s Enforcement Priorities in Applying art 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings [2009] OJ C45/2, para 22, (‘Guidance on the Commission’s Enforcement Priorities’). 112 ACCC v Universal Music, 536 [412]. 113 Sidak and Teece (n 43) 614. 114 See eg the AKZO presumption of dominance under art 102 TFEU where an undertaking has a market share of 50% or more. 115 Guidelines on Misuse of Market Power, para 2.16. See also Miller (n 12) 257. It could be argued that under Australian competition law market shares play a less central role than eg under US antitrust law or EU competition law. In the US monopolization is prohibited, while in the EU there is not only a market power threshold, but dominant companies also have a ‘special responsibility. Under Australian competition law, on the other hand, more than one company can have substantial market power as stipulated in s 46(7) CCA. 116 Queensland Wire, 183. 117 DPI, 65. 118 Boral Besser Masonry Ltd v ACCC (2003) 215 CLR 374. 119 ibid 423 [137]. 120 David Bailey and Laura E John (eds), Bellamy & Child – European Union Law of Competition (8th edn, OUP 2019), para 10.022. 121 National Competition Policy Review, 25 August 1993 (hereafter ‘Hilmer Report’), 3. 122 David S Evans and others, ‘An Analysis of the Government’s Economic Case in U.S. v. Microsoft’ in David S Evans (ed), Microsoft, Antitrust and the New Economy: Selected Essays (Kluwer Academic Publishers 2002) 29. 123 See DPI, 66 and 78. 124 DPI, 78. 125 See Oliver Budzinski and Annika Stöhr, ‘Competition policy reform in Europe and Germany – institutional change in the light of digitization’ (2019) European Competition Journal 15, 40. 126 See eg Louis Kaplow, ‘Why (Ever) Define Markets?’ (2010) 124 Harvard Law Review 437; Richard S Markovits, ‘Why One Should Never Define Markets or Use Market-Oriented Approaches to Analyze the Legality of Business Conduct Under U.S. Antitrust Law’ (2012) 57 Antitrust Bulletin 747; Rupprecht Podszun, ‘The Pitfalls of Market Definition: Towards an Open and Evolutionary Concept’ in Fabiana Di Porto and Rupprecht Podszun (eds), Abusive Practices in Competition Law (Edward Elgar 2018) 68. 127 Eg in cases where the quality of a product or service rather than price is decisive, competition authorities could probe a ‘Small but Significant, Non-transitory Decline in Quality’ (SSNDQ). 128 See Filistrucchi and others (n 66) 300–03. 129 Joe S Bain, Barriers to New Competition: Their Character and Consequences in Manufacturing Industries (Harvard University Press 1956) 3. 130 Bailey and John (n 120) para 10.031. 131 Franklin M Fisher, ‘Diagnosing Monopoly’ (1979) 19 Quarterly Review of Economics and Business 7, 18 and 23–27; Brunt (n 3) 7. 132 See Queensland Wire, 189–90. 133 QCMA, 512. See also Seven Network Ltd v News Ltd (2009) 182 FCR 160; Melway, 67. 134 Hilmer Report, 2. 135 ACCC, Merger Guidelines (November 2008), para 7.17; Brunt (n 3) 263. 136 See eg Guidance on the Commission’s Enforcement, para 16. 137 Duke (n 10) 114–123. See also Brunt (n 3) 263–64. 138 Brunt (n 3) 263. 139 Boral, [295]; Queensland Wire, 189–190; Re Chime Communications Pty Ltd (No 2) [2009] ACompT 2, 51-53; Ray Steinwall, Annotated Competition and Consumer Legislation (LexisNexis Butterworths 2018) 505; Miller (n 12) 97. 140 ibid; Arnotts, 338; Guidance on the Commission’s Enforcement Priorities, para 17. 141 Steven C Salop, ‘Strategic Entry Deterrence’ (1979) 69 American Economic Review 335. The Tribunal in Re Southern Cross Beverages recognized this type of strategic barriers to entry (Re Southern Cross Pty Ltd (1981) 50 FLR 176, 206). See also United Brands, paras 67–68. 142 Re Qantas Airways Ltd [2004] ACompT, [306]. 143 Duke (n 10) 128–29. 144 See Stucke and Grunes (n 77) 200. 145 See Facebook/WhatsApp, para 109; Schweitzer and others (n 72) 60. 146 Schweitzer and others (n 72) 28; Shapiro and Varian (n 58) 184–85. 147 Rubinfeld and Gal (n 68) 49. 148 Schweitzer and others (n 72) 60. 149 Rubinfeld and Gal (n 68) 49; Khan (n 15) 773 and 786. 150 See eg Graef (n 66). 151 Google and Facebook’s strategic acquisitions are not treated under the heading ‘barriers to entry and expansion’ but under a separate heading. Those acquisitions nonetheless constitute barriers to entry and expansion. 152 DPI, 74 and 80. 153 See above the section ‘The ‘Newest’ economy’. 154 DPI, 66. 155 DPI, 73–74 and 79–80. 156 Queensland Wire, 190. 157 ibid. 158 DPI, 73, citing Crémer and others (n 68) 2, 20. 159 Rubinfeld and Gal (n 68) 49. 160 Stephen G Corones, ‘Applying an Effects Test Under s 46 of the Competition and Consumer Act’ (Conference Paper, Bar Association QLD Event, 25 July 2016), p 4. 161 Arnotts, 338. 162 QCMA, 520. 163 Hilmer Report, 3. 164 ibid. 165 ibid. 166 ibid 4. In the context of merger analysis For the purposes of s 50(3)(g) CCA mentions that product dynamic characteristics include growth, innovation and product differentiation. The Tribunal in QCMA also discussed innovation (see QCMA, 187–88). 167 DPI, 72 and 79. 168 DPI, 72. 169 DPI, 98. 170 DPI, 42. 171 DPI, 41–42. 172 This is also reflected in see s 46(5)(b)(ii). 173 See DPI, 68–69. 174 Lisa M Segarra, ‘Google to Pay Apple $12 Billion to Remain Safari’s Default Search Engine in 2019: Report’ (Fortune, 30 September 2018) <https://fortune.com/2018/09/29/google-apple-safari-search-engine/> last accessed 1 December 2020. 175 See Schweitzer and others (n 72) 14. 176 From 2015 to 2018, Nest was part of Google’s parent company Alphabet and operated independently of Google. In 2018, Nest was merged into Google’s home devices unit. 177 See above. 178 See eg Queensland Wire; Melway; Universal Music; Taprobane Tours WA Pty Ltd v Singapore Airlines Ltd (1990) 33 FCR 158. 179 Queensland Wire, 190. 180 See OECD, ‘Data-Driven Innovation: Big Data for Growth and Well-Being’, 6 October 2015, <http://www.oecd.org/innovation/data-driven-innovation-9789264229358-en.htm> last accessed 1 December 2020. 181 DPI, 138–39. 182 DPI, 75. 183 ibid. 184 Duke (n 10) 143. 185 This factor has not been discussed in any misuse of market power case since QCMA, possibly because s 46(3) was added by the Trade Practices Amendment Act 1977 (Act 81 of 1977). 186 s 4A(5) CCA for the purposes of the Act deems a holding company and its subsidiaries to be related bodies corporate. 187 TPC v Ansett Transport Industries (Operations) Pty Ltd (1978) 32 FLR 305. 188 TPC v Pioneer Concrete (Qld) Pty Ltd (1994) 50 FCR 160. 189 ibid, para 14. 190 DPI, 74 and 80. See also Report of the Digital Competition Expert Panel, ‘Unlocking digital competition’ March 2019 (‘Furman Report’), para 1.49; Lear, ‘Ex-post Assessment of Merger Control Decisions in Digital Markets’ 9 May 2019. 191 Guidelines on Misuse of Market Power, para 2.15. 7 ibid 512. 8 See Carl Kaysen and Donald F Turner, Antitrust Policy: An Economic and Legal Analysis (Harvard University Press 1959) 75. 9 See also Queensland Wire Industries Pty Ltd v Broken Hill Pty Co Ltd (1989) 167 CLR 177; Katharine Kemp, Misuse of Market Power: Rationale and Reform (CUP 2018) 107. 94 See ibid 167. 95 See eg Robert H Bork, The Antitrust Paradox: A Policy at War with Itself (Basic Books 1978); Richard A Posner, ‘Conglomerate Mergers and Antitrust Policy: An Introduction’ (1969) 44 St. John’s Law Review 529. 96 Lim (n 91) 3. See also Nicolas Petit, ‘Technology Giants, the Moligopoly Hypothesis and Holistic Competition: A Primer’ (2016), <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2856502> last accessed 1 December 2020. 97 Bourreau and de Streel (n 59) 6. 98 ibid; Owing to the modularity of their product design, a digital output (eg a mapping service) can also be used as an input (eg for navigations systems). 99 Bourreau and de Streel (n 59) 11. For example, an Apple Watch can only be used together with an Apple iPhone and not with a smartphone from another manufacturer. This linkage between these two distinct but complementary products provides consumer with an additional benefit (a ‘consumption synergy’) when buying them together, even though the iPhone and the Apple Watch are not sold as a bundle by Apple. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

Journal of Antitrust EnforcementOxford University Press

Published: Dec 21, 2020

There are no references for this article.