|
|
Home
| Databases
| WorldLII
| Search
| Feedback
University of Otago Law Theses and Dissertations |
Last Updated: 21 September 2023
Jacob Bird
An Analogue System of Competition Law in a Digital World: How Should New Zealand Respond to the ACCC’s Digital Platforms Inquiry?
A dissertation submitted in partial fulfilment of the requirements of the degree of Bachelor of Laws (with Honours) at the University of Otago, Dunedin, New Zealand – Te Whare Wananga o Otago
2 October 2020
Acknowledgements
A very big thank you to my supervisor, Professor Rex Adhar, for his invaluable guidance over the course of a very messy year.
Thank you to my parents for far too many things.
Thank you to my siblings for helping me get this done in one way or another.
And finally, thank you to all the outstanding people I have met down here at Otago.
Table of Contents
I Introduction
Competition law is back in fashion. Not since its birth in the late 19th century has antitrust had such political salience, and not since then has it faced such overwhelming challenges. The digital economy is here, and with it, the digital platform. Platforms offer a new way of addressing the fundamental issue of industrial organisation, of how to coordinate supply and demand between millions of individuals in the absence of complete information.1 In doing so they have created an immense amount of welfare for consumers, allowing them to connect with one another and access information in ways inconceivable just a few decades ago.
Few could have predicted the extent of their ubiquity; today’s biggest platform businesses resemble powerful nation-states. With more than 1.5 billion users, Facebook oversees a ‘population’ larger than China’s.2 Google handles 64 percent of the online searches in the
U.S. and 90 percent of those in Europe.3 Platforms of this scale control economic systems bigger than all but the biggest global economies. The size and reach of the largest platforms come with some familiar consequences, such as economic clout and resources to invest in lobbying. However, they are also responsible for many consequences that are unfamiliar. Unique consequences arise for privacy, the news media and our political institutions, which are quickly becoming apparent.
Governments around the world are beginning to wake up. The concentration observed in digital markets is seen by some as representative of weak anti-trust policy, unsuited to the new challenges presented by the digital platforms economy.4 Lord Tyrie, the former chairman of the Competition and Markets Authority, laments that the UK is fitted with an “analogue system of competition and consumer law in a digital age.”5 Today’s concerns about corporate power are as fervent as ever, and represent an opportunity to strengthen competition policy throughout the world.
The Australian Competition and Consumer Commission (the ACCC) have taken up the mantle with their recent Digital Platforms Inquiry.6 In June 2019 they delivered an extensive report into the impact of online search engines, social media and other digital platforms on competition, with a special focus on the media and advertising services markets.7 In practice, the inquired focused on Google and Facebook as the two largest digital platforms in Australia. The report delivered wide-ranging recommendations to the Government on issues of data protection, competition and consumer law.8 This paper
1 Martin Moore and Damian Tambini Digital Dominance: The Power of Google, Amazon, Facebook, and Apple (Oxford University Press, New
York, 2018) at 52.
2 Geoffrey Parker, Marshall W. Van Alstyne and Sangeet Choudary Platform Revolution: How Networked Markets Are Transforming the Economy and How to Make Them Work for You (W.W.Norton, United States, 2016) at 159.
3 At 159.
4 Diana Coyle, “Practical Competition Policy Implications of Digital Platforms" (2005) 82 ALJ 3 at 840.
5 Paul Rogers “Consumer watchdog chairman Lord Tyrie quits over ‘limits’ of powers” The Times (London, June 18 2020)
6 Digital Platforms Inquiry: Final Report (Australian Competition and Consumer Commission, June 2019)
7 At 4.
8 At 5.
critically analyses only those recommendations made by the ACCC to strengthen competition law in Australia.
Before addressing these recommendations, chapter one looks at the digital platforms market in detail. It first explores the characteristics that drive high levels of concentration and market power, including the presence of strong network effects, extreme economies of scale, zero consumer-side prices and the importance of data and economies of scope. These features are not entirely conducive to competition in the market, leading to some ‘populist’ cries that competition is not possible in the platforms market, and that instead incumbents should be broken up. I reject that argument, and conclude that competition policy in a modified form should be at the heart of our regulation strategy for digital platforms.
Chapter one notes that competition in this context is more likely to be dynamic, or from competitors not currently operating in the incumbent’s market. This form of ‘potential competition’ ought to act as the primary competitive constraint on large incumbents, however a weak merger regime has prevented this from being so. Instead, firms are able to purchase potential competitors before they develop into substantial competitive threats. Chapter two analyses the recommendations made by the ACCC to strengthen merger laws and address this concern. It finds that although on the right track, more stringent reform will likely be necessary.
Finally, chapter three outlines the biggest set of policies missing from the antitrust toolbox. These are a range of pro-active, ex ante regulatory tools, that attempt actively to foster competition in the market. The chapter welcomes the new era of competition policy crossed with regulation, overseen in Australia by the newly established Digital Platforms branch. I will assess the desirability of various forms of these regulations, including codes of conduct, open standards and data portability.
The time has come for New Zealand’s Commerce Commission to consider its own position. Regulation of global digital firms requires a co-ordinated, global effort in order to be effective. New Zealand should join the many other jurisdictions tabling changes to their competition laws, to ensure that the global response is synchronised and productive.
II Chapter One
A Section ONE: Characteristics of the Digital Platforms Market
Network effects are described as the “most central element of digital platforms markets” by some commentators.9 They are perhaps the force driving concentration in the market, and can be divided into two forms; direct and indirect.10 Direct network effects arise where the benefits to a user on one side of the platform increase the value of the service to users on that side of the platform.11 For a social media service, “identity-based network effects” are often the key competitive advantage for an incumbent.12 This effect describes an increase in benefit for a user as the total number of other users they know, or wish to connect with, increases. The ACCC found identity-based network effects formed a significant barrier to entering the social media market, as Facebook’s large number of users meant it had a higher chance of attracting users than a smaller social media platform.13 In the case of Google, direct network effects result from data accumulation.
The ACCC found that the relevance algorithm of Google’s search engine improves as more data accumulates, increasing its ability to produce reliable results as more consumers use the search engine.14
In the case of two-sided platforms, indirect network effects also operate. Indirect network effects occur when the benefits to users on one side of a platform increase with the number of users on the other side of the platform.15 More specifically, value to advertisers on one side of a platform increase with the number of consumers on its other side.16
It is common for cross-subsidisation to accompany these effects, whereby revenue earned on one side of a platform subsidises the cost of providing a service to users on the other side of the platform.17 The subsidy goes to the side whose demand is more sensitive to price.18 In the social media and search markets, this effect has resulted in zero pricing for consumers, as firms seek to maximise value to advertisers by increasing their user base to
9 Elena Argentesi and Paolo Buccirossi and Emilio Calvano and Tomaso Duso and Alessia Marrazzo and, Salvatore Nava “Merger Policy in
Digital Markets: An Ex-Post Assessment” (December 2019). Social Science Research Network <https://www.ssrn.com> at 4.
10 at 4.
11 Jason Furman Unlocking digital competition: Report from the Digital Competition Expert Panel (UK government publication, HM Treasury, March 2019) at 35.
12 Germany’s Bundeskartellarmt deemed “identity-based network effects” to be the “decisive criteria” for the choice of a social network. “Background information on the Facebook proceeding” (19 December 2017) Bundeskartellamt <https://www.bundeskartellamt.de> at 3. 13 ACCC, above n 6, at 79.
14 at 66.
15 Furman, above n 11, at 35.
16 at 35.
17 ACCC, above n 6, at 63.
18 Diane Coyle. "Making the most of platforms: a policy research agenda." (16 October 2016) Social Science Research Network
<https://www.ssrn.com> at 6.
increase the reach of ads. Further, a platform with more users has more access to data, which can improve the relevance of ads presented to users.19
Network effects can cause a market to “tip” to monopoly.20 They heavily favour firms already established in the market, and as such act as a formidable barrier to entering the market.21 Strong network effects create an “incumbency advantage”, allowing an incumbent to generate higher profits than a new entrant to the market, despite that new entrant potentially offering the same or a better service.22 The size of this incumbency advantage is reinforced in part by the difficulty for users to coordinate migration to a new service.23 As demonstrated by Crèmer et al, even when all consumers stand to gain from a joint migration, plainly no consumer wants to migrate on their own and forgo the positive network effects from remaining with the original service.24 Migration will therefore not occur, and consumers are ‘locked-in’ to the inferior service.25
In Chapter three it is argued that competition agencies can and ought to reduce the size of this incumbency advantage by facilitating data portability and platform interoperability.26
A recent report on digital platforms commissioned by the European Commission’s Commissioner for Competition noted that while economies of scale are a common phenomenon in many industries, the digital platforms market “pushes it to the extreme”, resulting in a “significant competitive advantage for incumbents”.27 Platforms face a very high upfront investment cost, coupled with low or near-zero marginal costs of additional users.28 A platform’s average costs therefore reduce significantly as it grows its user base, incentivising growth at scale. To the extent that they remain “sunk” or unrecoverable, these high fixed costs form a significant barrier to entry.29 The ACCC noted both Google and Facebook “benefit from significant returns to scale.”30
Compounding this effect is the global nature of digital markets. While economies of scale have traditionally been constrained by barriers of physical production such as location or transport costs, digital markets are not.31 Economies of scale therefore support
19 ACCC, above n 6, at 64.
20 Howard Shelanski ““Information, Innovation, and Competition Policy For the Internet” (2013) 161 U Pa L Rev1663 at 1682.
21 Jacques Crémer, Yves-Alexandre de Montjoye and Heike Schweitzer Competition Policy For the Digital Era (Publications Office of the European Union, 2019) at 23.
22 At 23.
23 At 22.
24 Gary Biglaiser, Emilio Calvano and Jacques Crèmer “Incumbency Advantage and its Value” (2019) 28 J Econ Manage Strat 41 at 42.
25 Shelanski, above n 20, at 1682.
26 Crèmer, de Montjoye and Schweitzer, above n 21, at 23.
27 above, n 21, at 23.
28 ACCC, above n 6, p 32.
29 At 73.
30 At 11.
31 ACCC, above n 6, at 73.
concentration on a global, rather than local scale, facilitating the emergence of global firms of a size never before seen in history.32
Large economies of scale and network effects help to explain the zero-pricing strategy of market incumbents. The combination of near zero marginal costs and high fixed costs encourages a pricing strategy that attracts the maximum number of users to minimise average costs per user.33 The aforementioned cross-subsidisation effect also encourages a zero-pricing strategy, whereby firms attempt to maximise value for advertisers on one side of the platform by pricing in a manner that attracts the highest number of users on the other.34 The so-called ‘attraction of the free’ effect explains that zero-pricing is the most effective pricing strategy to attract consumers.35 Consumers are attracted by a zero- price as they irrationally perceive the benefits associated with free products to be higher than they actually are.36
Such a strategy acts as a major barrier to entry, as in the absence of a price to be undercut by a disruptive new entrant, prospective new entrants are forced instead to compete on quality.37 This is difficult for an entrant to any market which in relative terms lacks resources, however it is even more difficult in the digital sphere where quality often depends substantially on scale, due to the prominence of data, discussed in the next section.38
A critical asset to all digital platforms is data. Data informs businesses of consumer habits, preferences and activity, helping them to improve their services in a variety of ways.39 Data can be of substantial value, with one major study finding that companies that use data-driven innovation experience between 5% and 10% faster productivity growth than companies that do not.40 While businesses have traditionally sought to gather information about their customers, it is the scale and extent of today’s data accumulation that creates such a significant competitive advantage for large incumbents.41
32 Furman, above n 11, at 32. Facebook reported 2.70 billion active monthly users as of June 30, 2020. Facebook “Facebook Reports Second Quarter 2020 Results” (press release, July 30 2020)
33 Crèmer, de Montjoye and Schweitzer, above n 21, at 20.
34 ACCC p 63
35 for a discussion of this effect see Kristina Shampanier, Nina Mazar and Dan Ariely “Zero as a Special Price: The True Value of Free Products” (2007) 26 Mark Sci 742
36 at 742.
37 Damien Geradin “What Should EU Competition Policy do to Address the Concerns Raised by the Digital Platforms’ Market Power?” (12 December 2018) Social Science Research Network <https://www.ssrn.com> at 3.
38 At 3.
39 Furman, above n 11, at 23.
40 Ania Thiemann and Pedro Gonzaga Big Data: Bringing Competition Policy to the Digital Era (Directorate for Financial and Enterprise Affairs Competition Committee, DAF/COMP(2016)14, October 2016) at 8.
As mentioned, positive network effects result from data accumulation in the cases of Google and Facebook. The more users of each platform, the more data that can be collected to improve the quality of service offered to both consumers and advertisers.42 Collecting personal data helps the firms learn more about consumers, allowing them to tailor their service to better suit consumer needs. This creates value for the consumer, and also for advertisers, to the extent that it increases the attention the consumer devotes to the service and its ads.43 Data can also create value for advertisers by allowing them to better ‘target ’the right consumer with the right ad.44 The ACCC highlighted in particular that the size and impressive quality of Google and Facebook’s data sets formed a key competitive advantage for both firms.45
While greatly advantageous to their holders, it bears noting that data sets are non- rivalrous and easily duplicated.46 Significant opportunity therefore exists for competition agencies to explore ways in which data might be shared amongst rivals. Chapter three addresses such possibilities.
Platforms may increase the quality of their offerings and reduce costs by operating simultaneously across adjacent markets.47 Known as ‘advantages of scope’, this gives rise to conglomeration effects that help entrench the market power of an incumbent. Broadly, advantages of scope describe the positive effect on the quality and efficiency of a given service as an incumbent expands its offering of other services.48
The ACCC noted that Google and Facebook both enjoy the benefits of advantages of scope.49 As both firms offer a range of services, they are able to accumulate data relating to many different aspects of consumers ’lives.50 Competitors do not have access to such a rich range of data and are therefore at a relative disadvantage. For example, Google collects data from Search, YouTube, Google Shopping, and Gmail (among others), revealing valuable preferences relating to consumer viewing habits, shopping preferences and work-related habits among many other things.51 This couples with Google’s collection of data from third party websites whenever a customer uses an Android operating system or the Chrome Browser.52
42 ACCC, above n 6, at 66.
43 Furman, above n 11, at 22.
44 At 22.
45 ACCC, above n 6, at 87, 89.
46 Furman, above n 11, at 23.
47 At 23.
48 Crèmer, de Montjoye and Schweitzer, above n 21, at 33.
49 ACCC, above n 6, at 8, 9.
50 At 9.
51 At 73.
Advantages of scope are often pro-competitive, for they allow a large incumbent to provide a service more efficiently.53 However, they can also prevent competition on the merits.54 Where a conglomerate is able to gain privileged access to user data from a large range of sources, the use of this data might allow them to offer a far superior service. This leaves a competitor lacking in such data, but with a far superior underlying technology or algorithm being unable to compete on the merits.55
Furthermore, benefits arise to consumers who use just one firm’s entire range of products as significant issues of interoperability between platforms currently exist. This can prevent a new superior service from gaining traction, as a consumer is wary of departing from the established ‘ecosystem’.56 Defining a market for an ‘ecosystem of services’ can alleviate some of these anti-competitive effects, as is canvassed in section three.
53 Crèmer, de Montjoye and Schweitzer, above n 21, at 34.
54 At 34.
55 At 34.
56 At 34.
B Section TWO: Implications of These Market Characteristics
A combination of the factors described above explains the market we see today; one with very high barriers to entry, high levels of concentration and enduring market power for a small number of firms. Google and Facebook in particular have become economic titans, consuming 49.2% of all Australians’ time spent online.57 Facebook captures 51% of all online display advertising revenue, while Google 96% of all search advertising revenue in Australia.58 Both have been dominant players for well over a decade, and to quote the ACCC, their growth shows “no signs of slowing”, with around half the value of both companies’ share prices attributable to future growth.59
Strong network effects and extreme economies of scale in particular drive the market to concentration.60 Economic theory predicts that because of these features, an efficient market will consist of only a small number of large firms.61 These effects are so strong in the digital platforms market that sometimes it will completely ‘tip’ in the favour of a single winner, as has occurred in Google’s favour in the market for search engines.62
Competition ‘in’ the market is therefore likely to be limited. Instead, competition is primarily ‘for’ the market.63 Although incumbents face few worthy competitors for the time being, there is a continuous threat that a radical new entrant will altogether displace them. This is particularly so in markets characterised by high levels of innovation, such as the digital platforms market. The Schumpeterian school of thought, named after the influential economist Joseph Schumpeter, explains that in such markets, competition is more likely to come from entirely new innovative products, rather than from competition on existing ones.64 On this account, competition is dynamic and comes in waves of “creative destruction”, with each superior innovation allowing a new entrant to entirely capture the market previously dominated by an incumbent.65
57 ACCC, above n 6, at 65.
58 At 65.
59 At 6.
60 Crèmer, de Montjoye and Schweitzer, above n 21, at 36.
61 At 36.
62 Cédric Argenton and Jens Prüfer “Search Engine Competition With Network Externalities” (2012) 8 JCL & E 73 at 76.
63 Crèmer, de Montjoye and Schweitzer, above n 21, at 23.
64 Joseph Schumpeter Capitalism, Socialism and Democracy (George Allen & Unwin Ltd, UK, 1943) at 83. For a discussion of this effect in the context of modern digital markets, see Michael Katz & Howard Shelanski, “Schumpeterian Competition and Antitrust Policy in High-Tech Markets” (2005) 14 COMPETITION 18 at 5.
65 At 83.
Baumol, Panzar, and Willig’s controversial Theory of Contestable Markets states that this threat from future, or ‘potential’ competitors, can deliver the benefits of a highly competitive market, despite few competitors existing in the market itself.66 As firms are wary of being displaced from their dominant position by the next wave of technology, they continue to innovate and strive to provide high-quality services. They have limited possibilities to exploit their market power, as they attempt to fend off competitors who attempt to take the whole market from them.67 The large digital platforms suggest this effect is extremely intense, and as such their monopoly position in the market ought not to concern antitrust regulators.68 Instead, antitrust agencies ought to tolerate monopoly positions for periods of time, instead focussing on creating the conditions for dynamic innovation to flourish.69
Apologists for the platforms may argue that their own rise to prominence in this manner evidences this form of dynamic competition at work. Facebook displaced Myspace in the US in 2009 with a far superior social networking site.70 Google offered a far superior relevance algorithm for its search engine which saw it displace incumbents such as Yahoo! and Altavista, both of whom who have since drifted into relative obscurity.71
The ACCC recognised that although the threat of dynamic competition is in theory a competitive constraint on Google and Facebook, in practice both firms are largely insulated from this threat.72 Factors contributing to this were the extremely high barriers to entering the market, the firms’ advantages of scope and their acquisition strategies.73 The ACCC found that both firms tended to acquire potential competitors before they grew to a sufficient extent to act as competitive constraints.74 Merger laws have failed to prevent these acquisitions from going ahead, despite their clear anti-competitive potential. Chapter two assesses the suggestions made by the ACCC to reverse this trend, and finds that even stricter measures may be required than those they suggest.
However, the theory of contestable markets has been largely discredited since its inception.75 Further, Crèmer et al. have questioned whether its rationale could apply at all to this context. 76 It relies on relatively free and frequent entry to the market, and as such does not readily apply to a market marred by high entry barriers and strong network effects. 77 It therefore seems unlikely that the threat of potential competition alone is sufficient to constrain a dominant incumbent in the digital platforms market, in the same way that a competitive market would. The ACCC noted their concern that digital
66 Baumol, William J. "Contestable Markets: an Uprising in the Theory of Industry Structure" (1982) 72 American Economic Review 1
67 Crèmer, de Montjoye and Schweitzer, above n 21, at 36.
68 For example, see this argument in this Google funded research, David Evans “Multisided Platforms, Dynamic Competition, and the Assessment of Market Power for Internet-Based Firms” (10 March 2016) Social Science Research Network <https://www.ssrn.com> at 23. 69Katz and Shelanski, above n 26, at 1693.
70 ACCC, above n 6, at 78.
71 At 68.
72 At 58.
73 At 58.
74 At 10.
75 Jose Alberro and Rainer Schwabe The Theory of Contestable Markets and its Legacy in Antitrust (ABA Economics Committee Newsletter, Volume 16, number 1, Cornerstone Research, United States of America, 2016) at 23.
76 Crèmer, de Montjoye and Schweitzer, above n 21, at 36.
77 At 36.
platforms currently have the ability to misuse their substantial market power and impede the efficient operation of markets.78
To really unlock the opportunities of the digital market, stimulating more competition in the market must be a goal for regulators. It is widely accepted that this will require increased intervention of some form.79 The debate is now over what this exact form will be.80 Some have suggested that stimulating competition in the market is not possible, as platforms exhibit natural monopoly characteristics. Instead, a mixture of divestiture and natural monopoly regulation is proposed as the only viable means of regulation. As I will outline in the next section, that suggestion is unlikely to succeed.
That approach also unnecessarily concedes defeat for competition in the market. Greater competition between platforms is both necessary, and possible, provided two key policy changes occur. First, many of the traditional competition law tools of analysis require modification to remain applicable to the market. The end of the chapter suggests some practical changes that can be made. Secondly, as the ACCC recognised, ex post competition policy alone is inherently unsuited to addressing harms in the digital sector. For one it is slow, with proceedings sometimes stretching more than a decade.81 It also relies on waiting until anti-competitive harms have occurred, meaning sometimes irreparable damage is done to the precious few competitors in the market. It is for these, and other reasons, that the ACCC has opted to pursue a more pro-active, ‘pro- competitive’ ex ante approach to competition law in the digital platforms market. Chapter three will set out a more detailed justification for the change in approach, to be administered by the newly established Digital Platforms branch.
78 ACCC, above n 6, at 137.
79 This is the premise of the ACCC’s report, the EU report and the Furman Report.
80 Mike Walker "Competition policy and digital platforms: six uncontroversial propositions." (2020) 16 Eur Competition J 1 at 8.
81 Reuters Staff “TIMELINE-Google's decade-long antitrust battle in Europe” (21 March 2019) Reuters <https://www.reuters.com>
C Section Three: An alternative to Competition Policy: Divestiture & Natural Monopoly Regulation
Some large digital platforms exhibit natural monopoly characteristics. Google Search for example has very high fixed costs, and low or zero marginal costs, creating economies of scale similar to network industries.82 Data is also a key driver of concentration, argued by some to lead to a natural monopoly market structure.83 It has been considered by some authors whether, as an alternative to full scale regulation, the natural monopoly segment of a firm could be divested and regulated as an essential facility.84 The segment would remain regulated and constrained to provide a fair and non-discriminatory access to competitors in other markets than can sustain competition.85 This was the rationale behind the 1984 AT&T divestiture, where local services had natural monopoly status and were left in the hands of the “baby bells”, but long distance segments were divested and turned over to competition.86
This approach has some appeal, but faces immediate difficulties in the digital platforms context. First, Nobel Prize winning economist Jean Tirole points to the difficulty of identifying a stable essential facility in such a dynamic market.87 Even if an essential facility could initially be identified, its nature is likely to change as time progresses, making it difficult to isolate.88 As a divestiture takes some years and considerable cost to implement, if the location of the essential facility kept migrating it might make the divestiture worthless.89 Previous successful divestitures occurred in relatively stagnant markets such as the electricity or rail markets, however digital markets are fast moving, meaning any essential facility is continually morphing.90 Further, Hovenkamp suggests that identifying an essential facility may not even be possible for many digital platforms, as many are not truly natural monopolies.91 It is still far from certain whether even a likely candidate such as Google Search is rightly considered a natural monopoly or not.92
A second concern with divestiture is that it may destroy the benefits of network effects enjoyed by consumers.93 Breaking a social network into multiple parts is likely to lower consumer welfare, for consumers may find it more difficult to connect with users
82 Francesco Ducci Natural Monopolies in Digital Platform Markets (Cambridge University Press, Cambridge, 2020) at 5.
83 At 5.
84 Jean Tirole Competition and the Industrial Challenge for the Digital Age
(background paper for the IFS Deaton Review on “Inequalities in the Twenty-First Century”, April 2020) at 6.
85 At 6.
86 United States v AT&T Co 552 F Supp 131, 142 n 42 (D DC 1982) For a discussion see Herbet Hovenkamp “Antitrust and Platform Monopoly” (2020) 130 Yale.L.J. 20 at 59.
87 Tirole, above n 84, at 6.
88 At 6.
89 At 6.
90 At 6.
91 Hovenkamp, above n 86, at 68.
92 At 60. Please note, the full debate is outside the scope of this research.
93 Tirole, above n 84, at 6.
sprinkled across multiple networks. As Melamed and Petit point out, welfare losses are also likely from breaking-up a search engine: 94
“we invite readers to travel back to the world of Internet search before 2000, when users searched web pages through multiple platforms and then had to compare search results. The upshot was lost time for users and costly campaigns for advertisers...”
Moreover, this difficulty makes it unlikely the divestiture will be successful. There is a high chance that the market will drift back to concentration, as the benefit of network effects incentivise consumers to re-populate just one of the broken-up sites, thereby creating a new monopoly platform in course.95
Finally, divestiture is practically very difficult. Extricating two firms is typically described as a messy process akin to attempting to ‘unscramble an omelette or egg’.96 Hovenkamp notes that although it might not be difficult to break a multi-divisional firm such as Alphabet Inc. along its divisional lines, (e.g. into Google search, YouTube, Google Nest for its home products, Waymo for its autonomous driving technology and so on), breaking apart noncompeting units does not increase competition.97 Instead, to have any effect a regulator would have to break apart each unit, for example Google Search itself.98 This is very practically difficult, particularly where the unit is heavily interconnected to other segments of the company and across the world.99 It also risks harming consumer welfare and being ineffective in the long run due to the presence of network effects, as described above.
The ACCC briefly addressed and rejected submissions that called for structural remedies.100 Along with noting the points made above, they mentioned their concern that such an approach might reduce incentives for investment and efforts to improve productivity.101 Certainly for now, it appears the case for divestiture has not yet been made out. Instead, the ACCC’s attempt to stimulate competition in the market appears justifiable. Even some academics who are of the opinion that digital platforms are natural monopolies believe that ex ante regulation may be a first-best solution.102 However in time, if ex ante regulation proves to be ineffective, this conclusion could conceivably change.
94A. Douglas Melamed and Nicholas Petit “The Misguided Assault on the Consumer Welfare Standard in the Age of Platform Markets” (2019)
54 Rev Ind Organ 741 at 765
95 Carl Shapiro “Antitrust in a Time of Populism” (2018) 61 IJIO 714 at 744.
96 Chris Noonan, “Horizontal Mergers in New Zealand: Drifting Towards Structuralism” (2019) 26 CCLJ 263, 264: “A wait-and-see approach is not practical. Once consummated, a merger is usually impossible to unscramble.”
97 Hovenkamp, above n 87, at 60. Please note Alphabet Incorporated is Google’s parent company.
98 At 60.
99 At 60. Another interesting issue not touched on here is that absent a supranational regulator, it is difficult to predict who would ensure a divestiture policy was properly carried out for Google or Facebook’s operations across globe.
100 ACCC, above n 6, at 116.
101 At 116.
102 For example, see Ducci, above n 82, at 8.
D Section Four: Modifications to the Traditional Antitrust Framework
Chapter three will discuss the importance of building competition into digital services. Central to this is a regime that can effectively implement new regulation ex ante, to pro- actively foster new competition where it is currently lacking.
Optimising the competition framework for digital platforms is a critical complement to this policy change, in order to ensure that competition law is effectively enforced ex post. The traditional competition law framework has been stretched by the unique characteristics of the digital platforms market. The market tends to concentrate strongly, lacks consumer prices, is highly innovative and evolves very quickly, making traditional ex post assessments difficult.
Despite this, it is widely accepted that the traditional approach to competition policy ought to largely remain, albeit in a modified form. This section sets out only a limited number of the changes that are required, including changes to the consumer welfare standard, our conceptions of innovation and to market definition.
The purpose of the Commerce Act 1986 is to promote competition in markets for the
long-term benefit of consumers within New Zealand.103 Conceptually underpinning this is the consumer welfare standard. This is a standard adopted around the world, that broadly dictates that competition decisions should be based on their benefit to consumers.104 The standard has a slippery history in antitrust, and has been at the centre of an ideological war between the Chicagoan and Harvard Schools.
This war has spilled over into the digital economy, where the standard’s efficacy is often questioned.105 A commonly launched criticism is that it is overly focused on short term metrics of price and output to measure effects on consumer welfare.106 For some, this makes it wholly unsuited to the digital platforms market, where the lack of consumer prices means competitive pressure is instead exerted through service quality and innovative.107
The accusation is largely unfounded. Though traditionally the consumer welfare standard may have been “price centric” as some claim, the law has already demonstrated it is more than capable of shifting its focus to quality and innovation effects where they are more
103 Commerce Act 1986, s 1A.
104 Richard Bork The Antitrust Paradox: A Policy at War with Itself (New York, Basic Books, 1978) at 90 “Antitrust is about the effects of business behaviour on consumers.” Please note that Bork’s conception of the standard is at odds with its accepted understanding.
105 See for example Lina Khan “Amazon’s Antitrust Paradox” (2016) 126 Yale.L.J. 710
106 At 744. For further discussion, multiple issues with the CWS are mentioned in Furman Report at p 87.
107 Khan, above n 105, at 737.
relevant.108 New Zealand’s Merger and Acquisition Guidelines 2019 explicitly considers all “dimensions of competition, including quality, range, and level of innovation”.109 In the US, the landmark Microsoft case entirely concerned threats to innovation from conduct that might create barriers to entry.110
Although other dimensions of competition are included in the standard, these are commonly measured through price and output for practical reasons. Anti-trust agencies are constrained in their ability to measure less tangible competitive effects, particularly where they are imperfectly observable (e.g. deterring entry), measurable (e.g. product quality) or unpredictable (e.g. innovation effects).111 Data regarding price and output effects are naturally easier to obtain and are thus often proxies for the ultimate investigation into broader welfare effects.112
A related criticism of the consumer welfare standard is that it demands an unrealistic standard of proof in antitrust proceedings. Professor Tim Wu argues that the standard requires competition agencies to show measurable anticompetitive outcomes for consumers, rather than whether conduct causes damage to the competitive process.113 Particularly as traditional metrics of anticompetitive conduct (e.g. price and output) fade into insignificance in digital markets, proof that certain actions result in anticompetitive outcomes becomes more difficult, and has arguably led to a widespread underenforcement of competition law.114 For example, in the digital platforms market the effects of anticompetitive behaviour on innovation and service quality are likely to be the most important, in the absence of consumer prices. Both effects are difficult to measure in concrete terms, however some academics argue that the consumer welfare standard requires competition agencies to do so. Wu has argued that it is very possible that the Microsoft case would be thrown out today, because the Government had failed to demonstrate concrete consumer harm.115
To solve the problem, Wu proposes an alternative “protection of the competitive process” standard, which he describes as a process-oriented or procedural standard, as opposed to a goal-focused standard.116 On this approach, an action is considered anticompetitive if it is a “sufficient deviation” from the competitive process, regardless of its outcome.117 The problem of producing concrete evidence about harms to innovation is therefore overcome, for the inquiry focuses on harms to the competitive process instead. Maurice Stucke argues that the D.C. Circuit in the Microsoft case accepted that harm to the competitive process was sufficient to meet the consumer welfare standard, and that it was only because of this that harms to innovation were properly considered by the Court.118
108 Tim Wu “The “Protection of the Competitive Process” Standard” (Research Paper no 14-612, Columbia Law School, 2018) Columbia
Public Law Research Paper No. 14-612 (2018).
109 Mergers and Acquisitions Guidelines 2019 [2.18]
110 United States v. Microsoft Corporation, [2001] USCADC 115; 253 F.3d 34 (D.C. Cir. 2001).
111 Melamed and Petit, above n 94, at 753.
112 ICLE Final Report on Competition and Consumer Protection in the 21st Century: The Weakness of Interventionist Claims (ICLE, June 2019) at 34.
113 Wu, above n 108, at 2.
114 At 1.
115 Tim Wu, Julius Silver Professor of Law Columbia University “Hearing #5” (Federal Trade Commission Hearings, Washington, D.C. November 2018) at 190.
116 Wu, above n 108, at 2.
117 Wu, above n 108, at 2.
118 Wu, above n 115, at 190.
Professor Carl Shapiro agrees that the growing difficulty for regulators to prove that conduct is anticompetitive is a concerning evolution.119 However, in his view, attempting to extricate competition law analysis from its ultimate effects on consumer welfare is problematic. To use his example, standard setting amongst competitors is not per se illegal, because it ultimately leads to many consumer benefits, such as lower prices, or other benefits of compatibility between products. This is despite standard setting having clear collusive elements that are contrary to the ideals of a competitive process. In this way, Shapiro demonstrates that a sole focus on the competitive process can lead to uncertain outcomes.120 Instead, he argues that an anticompetitive action ought to be defined as one that disrupts the competitive process, and that has negative consequences for the affected trading parties.121 He describes this as a “protecting competition standard”, effectively a more accurately branded consumer welfare standard.122
Shapiro convincingly argues that most of the difficulty with the consumer welfare standard lies not in its content, but in where the burden of proof lies.123 If the current difficulty with standard is that it requires too onerous proof on the part of agencies, rather than “throwing out the standard” for something vaguer, we should instead shift the burden of proof onto the commercial party in certain circumstances.124 The particular challenges of proving anticompetitive harms in the digital economy justifies this shift in some areas. Chapter two examines the possibility in the context of merger law.
We have encountered a couple of times the immense role innovation plays in the digital platforms market. It is well established that the dynamic, Schumpeterian benefits of competition are more important to economic growth and social welfare than static competition among existing products, and therefore ought be afforded supremacy under the law.125 However the emphasis in practice to date has been on static efficiency.126 There is no settled approach to weighing static efficiencies against dynamic efficiencies in the economic literature.127 In 2010, Manne and Wright found “the theoretical literature relating to competition and innovation remains insufficient to instil any great confidence in our ability to determine what antitrust policies will encourage innovation and result in
119 Carl Shapiro, Professor of Economics at the University of California at Berkeley (“Hearing #5” (Federal Trade Commission Hearings,
Washington, D.C. November 2018) at 273.
120 At 273.
121 At 275. ‘Trading parties’ is preferred by Shapiro to the term ‘consumer’, to recognise that parties affected by anticompetitive behaviour can be upstream or downstream from the anticompetitive actor. However in his view, the case law already recognises that those potentially affected by anticompetitive behaviour includes more than just ‘consumers’.
122 At 273.
123 At 282.
124 At 277.
125 see, for example Jonathan Baker “Beyond Schumpeter vs. Arrow: How Antitrust Fosters Innovation” (2007) 74 ANTITRUST LJ 575, 601 as cited in Howard Shelanski ““Information, Innovation, and Competition Policy For the Internet” (2013) 161 U Pa L Rev 1663
126 Coyle, above n 18, at 5.
127 Coyle, above n 18, at 6.
net consumer welfare gains”.128 In the intervening years, little appears to have changed.129 A key development to come in competition law will be reliable, clear theories of harm in this regard.
One area of immediate importance is innovation spending. Regulators ought to distinguish between incumbent spending resulting from the pressure of dynamic competition, and spending attempting to reduce such pressure.130 Several studies have noted the high R&D spending of Facebook and Google, taking this as a sign of their commitment to innovation.131 However, simply noting a high R&D figure is insufficient for a conclusion that innovation is alive and well in the market. Coyle argues that we must also “explore whether this is spending on novel research to serve users” or is in fact “predatory innovation”, that reduces interoperability with competitors.132 Schrepel advocates for predatory innovation to become formalised as a concept, in the same vein as predatory pricing.133 However defined, developing clearer theories of innovation harms will be crucial to antitrust in the years to come.
The difficult process of market definition is even more so in the current context. To start, market boundaries are not as clear as in the ‘old economy’. Market definition and the tools for identifying a market were built for standard goods and services, not the more ‘fluid ’services that exist today (such as a social network).134 Some of these tools are totally ineffective in the digital platforms market, for example the SSNIP test.135 Moreover, the dynamic digital market often leads to quickly changing relationships of substitutability and partial overlaps between different services.136 Many experts argue, for example, that digital platforms compete for the attention of users across market boundaries that would ordinarily be defined based on the features of products and services.137
The response to such difficulties has been to place less emphasis on market definition generally.138 Scholars are instead urging regulators to focus “first, and more directly, on the competitive effects of conduct.”139 This argument is a familiar one of those opposed to market definition, and appears to have gained even more traction in this context. At
128 Geoffrey Manne and Joshua Wright “Innovation and the Limits of Antitrust” (2010) JCL & E 153 at 166.
129 Coyle, above n 18, at 5.
130 At 5.
131 Nicolas Petit “Technology Giants , the “Moligopoly” Hypothesis and Holistic Competition: A Primer” (20 October 2016) Social Science Research Network < https://www.ssrn.com> at 31.
132 Coyle, above n 18, at 13.
133 Thibault Schrepel “Predatory Innovation: the Definite need for Legal Recognition” (2018) 21 SMU Sci & TECH. L. REV. 19
134 Crèmer, de Montjoye and Schweitzer, above n 21, at 47.
135 At 44. This test defines the smallest market in which a hypothetical monopolist could impose a “small but significant non-transitory increase in price”, usually a price increase of 5% for 12 months. It does not readily apply to two sided markets, as it is difficult to know how to apply the price increase across both sides of a multi-sided market. The task becomes a near impossibility in the cases of Google and Facebook where the consumer faces side charges a zero price.
136 At 45.
137 David Evans “Attention Rivalry Among Online Platforms” (2013) 9(2) JCL & E 313 at 346.
138 Crèmer, de Montjoye and Schweitzer, above n 21, at 46.
139 Shelanski, above n 20, at 1673.
risk of over-simplification, the argument goes that defining a market diverts attention away from the key issue, that being defining likely competitive effects, that are more directly gauged through means other than defining a market.140 Given the increased difficulties of market definition in the digital platforms context, relying less on the exercise appears sensible.
Classic market definition must also be broadened to account for the rise of the ‘ecosystem’. Digital firms compete to draw consumers into more or less comprehensive ecosystems, comprising of a range of products and services.141 Firms compete for “access points” to consumers through different products, with the goal of locking consumers into their broader range of offerings.142 They encourage this in various ways, including by ensuring their own products interoperate better which each other than with the products of another firm, or by steering demand towards their own products through biased rankings, or the use of default settings.143 For example, Google’s search engine provides links to its other services Youtube and Gmail, making it relatively easier for consumers to remain solely in the ‘Google world’ than to use services from multiple different companies.144 A “classical” definition of markets alone fails to fully appreciate the harms of such behaviours. In response, Cremer et al. suggest that a market for ecosystems should be defined and operate alongside various markets for specific products or services.145 This is particularly relevant when assessing the harms of conglomerate mergers, as is discussed in chapter two.
E Conclusion
The traditional competition policy framework is more than capable of responding to the demands of the digital platforms market. Provided several modifications are made, ex post competition assessments are likely to continue to be an important part of antitrust policy. Many changes are required, and three are focused on in this paper. First, the consumer welfare standard must look past the short-term metrics of price and output, to more important effects on innovation and service quality. Second, theories of harm must be developed that better protect the dynamic forces of innovation. Finally, market definition may need to be relied on less, and be considered in conjunction with the concept of an ecosystem for services.
140 see, for example Louis Kaplow “Why (Ever) Define Markets” (2010) 124 Harv L Rev 437
141 Crèmer, de Montjoye and Schweitzer, above n 21, at 47.
142 At 48.
143 At 47.
144 John Newman “Antitrust in Digital Markets” (2019) 72 VAND L REV 1497 at 1508.
145 Crèmer, de Montjoye and Schweitzer, above n 21, at 48..
III Chapter TWO
A Introduction
We have now seen that because of certain features of the digital platforms market, it tends towards concentration. We have seen that competition ‘in’ the market is likely to be limited, with competition instead regularly occurring ‘for’ the market. The primary source of competition in the market thus comes from potential competitors, or firms not yet acting as competitive constraints, but with the potential to do so in the future. A stringent merger regime is essential to ensuring potential competition delivers meaningful pro-competitive benefits. Where merger controls are too relaxed, incumbents are able to insulate themselves from future competitive threats by acquiring young up-and-coming firms. Their subsequent avoidance of the competitive process harms consumers, who no longer profit from increased quality, innovation and lower prices (to name but a few benefits) that would have otherwise come from the competitive battle. The ACCC concluded that the acquisition strategies of both Google and Facebook insulate them from potential competitors in this exact regard.146
The current law does not adequately protect potential competition. Merger controls are far too weak, and too seldom enforced. The Furman Report found that over the past decade, Amazon, Apple, Facebook, Google, and Microsoft made over 400 acquisitions globally.147 Of these, none were blocked in the UK by the Competition and Markets Authority (the CMA), and none investigated since 2013.148 The EU Commission similarly failed to block any mergers in the sector over the same period.149 This means that to date, there have been no false positive errors involving the major digital platforms (i.e. no mergers were blocked that should not have been). This has come almost certainly at the expense of some false negative errors, as argued by the Furman Report.150 In other words, merger restrictions have been chronically under-enforced. In response, the ACCC recommended several modifications to toughen up the law. This chapter outlines these, and finds that although useful, in many respects they are insufficient. To cure the law of its current defects requires far bolder measures, some of which I outline at the end of the chapter. First, I will look at the current law on mergers in NZ and the underlying theory of harm to potential competition, and why these make it difficult for mergers to be challenged in the digital platforms context.
146 ACCC, above n 6, at 58.
147 Furman, above n 11, at 91.
148 At 91.
149 At 91.
150 At 91.
B The Current Law
In New Zealand, s 47 of the Commerce Act 1986 prohibits mergers or acquisitions that “have, or would be likely to have, the effect of substantially lessening competition in a market”.151 The Commerce Commission’s Merger and Acquisition Guidelines 2019 (The Guidelines) confirm that when making this assessment, the Commission will look to unilateral effects that might emerge “when an existing firm merges with a potential or emerging competitor.”152 Further, it ensures s 47 covers not only mergers between competitors in the same market (“horizontal mergers”), but also those effected by non- competitors in different markets (“non-horizontal mergers”) in certain circumstances. The Guidelines envisage possible anti-competitive harm from a conglomerate merger, albeit in the context of foreclosure from the tying or bundling of products from different markets.153 In theory therefore, the ‘substantial lessening of competition’ assessment of s 47 accounts for the doctrine of potential competition. Below I will set out the doctrine, before explaining why it - as currently conceived - fails to block harmful mergers in the digital platforms market.
C The Doctrine of Potential Competition
The doctrine of potential competition generally addresses non-horizontal mergers.154 It has been divided into two sub-theories of harm. First, the actual potential competition theory assesses the future effects a merger may have on competition.155 This theory states that a merger might be anti-competitive where it removes the possibility that two firms would compete within the same market in the future.156 Secondly, the perceived potential competition theory assesses the present pro-competitive effects a non-horizontal merger may threaten.157 This theory states that a transaction may remove or reduce present- procompetitive influences that the acquired firm has on the incumbent, which stems from the incumbent firm’s perceptions of the acquired firm’s ability to enter its market.158
The actual potential competition theory did not gain significant traction, while the perceived potential competition theory was prevalent in the US 1960s and 70s.159 The doctrine has since fallen out of use.160 This is in large part due to the extraordinarily high
151 Commerce Act 1986, s 47.
152 Mergers and Acquisitions Guidelines 2019, [3.63].
153 At [5.11-5.15].
154 Henry Klimowicz “Reinvigorating the Perceived Potential Competition Theory: An Analysis of the Potential Competition Doctrine and FTC
v. Steris Corp” (Law School Student Scholarship, Seton Hall University, 2019) at 4. However, the doctrine of potential competition can sometimes be used to assess horizontal mergers.
155 Darren Bush and Salvatore Massa “Rethinking the Potential Competition Doctrine” (2004) 2004 WIS L REV 1035 at 1046.
156 At 1046.
157 At 1046.
158 Klimowicz, above n 154, at 6.
159 Mark Glick and Catherine Ruetschlin “Big Tech Acquisitions and the Potential Competition Doctrine: The Case of Facebook” (8 November 2019) Social Science Research Network <https://www.ssrn.com> at 14, 26.
160 The last Supreme Court ruling on a potential competition case was in United States v Marine Bancorporation Inc [1974] USSC 155; 418 U.S. 602 (1974)
requirements of proof required by the US Supreme Court in a 1974 decision.161 There the Court required “unequivocal proof” that a young firm was likely to enter the incumbent’s market.162 Even if this standard was met, the Court then required the likely future entry to significantly reduce the dominance of the incumbent firm, and also that the entrant was uniquely situated to enter, rather than one of many potential entrants.163 This required complex predictions about the future conduct and performance of a potential entrant that could seldom be made with sufficient certainty to meet the standard’s requirements. The doctrine was therefore rendered a nearly useless tool for regulators, one that has been gathering dust since 1974.164
D A Resulting Underenforcement of Mergers
It is unsurprising then, that agencies have found it difficult to challenge Big Tech mergers in the US. Commentators frequently point to the Facebook/Instagram and Facebook/Whatsapp mergers as precisely the kind of harmful mergers that a functioning doctrine of potential competition would prohibit.165 However, despite being assessed in the US and EU, neither were analysed under the potential competition doctrine. Instead, the mergers tended to be analysed within the confines of the relevant market, for example the market for consumer communications services in the case of Facebook/Whatsapp.166 This is understandable, given the unrealistic standard of proof required by the potential competition doctrine. Recent academic work has analysed the Facebook/Instagram merger and found that even a potential competition challenge by the Department of Justice would have failed under its own guidelines.167 Some commentators have called for a reinvigoration of the theory given the importance of potential competition to markets today, which is welcomed.168 Whether or not NZ would be impaired by the doctrine in the same manner is unclear, as it is not mentioned in the Commission’s current Merger and Acquisition Guidelines.169 However, it is reasonable to assume that NZ would face similar difficulties when attempting to assess mergers between potential competitors.
Evidence for this proposition comes from the global failure to adequately protect potential competition. That none of the last 400 Big Tech mergers were blocked demonstrates a cross-jurisdictional failure of merger law in this regard, as many academics have argued.170 As the Furman Report found, “merger assessment in digital
161 Marine Bancorporation, above n 160.
162 Glick and Ruetschlin, above n 159, at 26.
163 At 26.
164 As in footnote 160, the last case in the US Supreme Court was Marine Bancorporation.
165 see Glick and Ruetschlin, above n 159.
166 At 32, 47. The only mention of ‘potential competition’ during either merger assessment was by the UK’s Office of Fair Trading in the Facebook/Instagram merger. They considered Instagram a potential competitor in the digital advertising market, but found that Instagram was an insignificant competitive constraint on Facebook compared to other rivals Google, Yahoo and Microsoft.
167 At 41.
168 see, for example Glick and Ruetschlin, above n 159
169 Mergers and Acquisitions Guidelines 2019
170 see e.g. Walker, above n 80.
markets needs a reset.” More “frequent and firm action is required” to challenge harmful mergers.171 This failure can in part be attributed to the decline of the potential competition doctrine.
It can also be partly attributed to the types of merger that occur in the digital sphere. Research commissioned by the CMA from the Laboratory of Economics, Antitrust and Regulation (LEAR) assessed acquisitions by Google, Amazon and Facebook, of which there were 299 between 2008-2018.172 As mentioned above, none were blocked.
However, two important observations were made. First, most of the acquisitions were of young firms. Nearly 60% of the acquired firms were four years old or less.173 Relevant for present purposes, the median age of the firm when acquired was 4 years for Google acquisitions and 2.5 years for Facebook acquisitions.174 The relative youth of such firms makes any assessment of future competitive importance very difficult, given the difficulty of predicting how they might have developed if not acquired. To take the case of Instagram, that firm was only 18 months old, had only 13 employees and generated no revenue when it was purchased for US$1b by Facebook.175 It would be very difficult to predict that the firm would reportedly generate US$20 billion in advertising revenue just seven years later.176
Secondly, LEAR’s research noted that “in most cases” the acquisitions were of firms producing complements to the products of the acquiring firm.177 Traditionally, mergers of firms producing complements were believed to lead to welfare enhancing efficiencies and were not subject to scrutiny.178 However, Crèmer et al. argue that such mergers can be anticompetitive in certain circumstances, particularly where an incumbent’s acquisition of complementary technology increases the quality of their service and therefore their dominant market position.179 A conglomerate merger can also remove potential competition, by removing the possibility that a firm in an adjacent market may attack the incumbent’s core market.180 Crèmer et al. suggest updating our theories of harm accordingly, a suggestion that is examined in more detail at the end of the chapter.
Both types of merger are not usually anticompetitive, and can therefore not be blocked by the ‘substantial lessening of competition’ test used in New Zealand.181 A merger between a large incumbent and a small start-up is unlikely to substantially lessen competition, because it is unlikely the start-up will grow to successfully compete with the incumbent. A merger with a firm in an adjacent market is similarly unlikely to substantially lessen
171 Furman, above n 11, at 91.
172 Elena Argentesi, Paolo Buccirossi, Emilio Calvano, Tomaso Duso, Alessia Marrazzo and Salvatore Nava Merger Policy in Digital Markets: An Ex-Post Assessment (LEAR, DIW Berlin Discussion Paper No. 1836, December 2019) at 15. Please note LEAR is the Laboratory of Economics, Antitrust and Regulation.
173 At 19.
174 At 19.
175 Glick and Ruetschlin, above n 159, at 29.
176 Sarah Frier and Nic Grant “Instagram Brings In More Than a Quarter of Facebook Sales” (February 5 2020) Bloomberg News <
<https://www.bloomberg.com>
177 Argentesi, Buccirossi, Calvano, Duso, Marrazzo and Nava, above n 172, at 1.
178 see, for example Uğur Akgün Cristina Caffarra Federico Etro Robert Stillman “On the welfare impact of mergers of complements: Raising rivals’ costs versus elimination of double marginalization” (2020) 195 Econ Lett
179 Crèmer, de Montjoye and Schweitzer, above n 21, at 122.
180 At 122.
181 New Zealand’s merger laws require the Commerce Commission to block any transaction likely to substantially lessen competition. see s 47(1) Commerce Act 1986.
competition, because it is unlikely the firm will switch into the incumbent’s market. However, in a small number of cases, both of these types of merger will turn out to be anticompetitive, because the acquired firm would have actually developed into a substantial competitor of the incumbent. That is, under a substantial lessening of competition test, some false negative errors are inevitable for mergers in the digital platforms market.
These false negative errors are likely to be particularly costly in the digital platforms market.182 The economist Carl Shapiro has rightly noted that the greater the market power of an incumbent, the greater the cost to society of false negative errors.183 Given the astronomical market power of the large platforms, the payoff from preventing false negative errors in the market is likely to be particularly large.
This justifies a different approach to our tolerance of error in digital mergers.184 Shapiro argues that this tolerance for error should be increased, stating that “sound competition policy would tolerate some false positives - blocking mergers involving targets, only to find that they do not grow to challenge the incumbent - in order to avoid some false negatives.”185 Furthermore, an implication of taking more account of dynamic competition is that the law will need to be more forward looking and accept more uncertainty if it is to remain effective.186 Support for a general reset in our tolerance of error can be found in both the Furman report and the report by Crèmer et al. for the European Commission’s Commissioner for Competition.187 As will be examined at the end of the chapter, this may require a lowering of the standard of proof. First, the ACCC’s own recommendations are set out below.
E The ACCC’s Recommendations
The ACCC recommended two changes to merger law in Australia. Australia shares New Zealand’s substantial lessening of competition test for assessing mergers, codified in s 50 of the Competition and Consumer Act 2010.188 Section 50(3) provides the decision maker with a non-exhaustive list of matters to consider when making this assessment.189 The ACCC’s first recommendation is that two factors be added to the s 50(3) list:190
(j) the likelihood that the acquisition would result in the removal from the market of a potential competitor;
182 Crèmer, de Montjoye and Schweitzer, above n 21, at 51.
183 Shapiro, above n 95, at 741.
184 Walker, above n 80, at 7.
185 Shapiro, above n 95, at 741.
186 Walker, above n 80, at 7.
187 see Crèmer, de Montjoye and Schweitzer, above n 21, at 51 and Furman, above n 11, at 100.
188 Competition and Consumer Act 2010 (Cth), s 50.
189 At s 50(3).
190 ACCC, above n 6, at 105.
(k) the nature and significance of assets, including data and technology, being acquired directly or through the body corporate
Recommendation two of the report suggested a notification protocol that requires large digital platforms to provide advance notice to the ACCC of any proposed acquisitions potentially impacting competition in Australia. Though the exact details are forthcoming, the finalised protocol would specify:191
a) the types of acquisitions requiring notification (including any applicable minimum transaction value), and
b) the minimum advance notification period prior to completion of the proposed transaction to enable the ACCC to assess the proposed acquisition.
If such a commitment is not forthcoming from the large digital platforms, the ACCC will make further recommendations to the Government that address this issue.
F Have they done enough?
I have discussed the benefit of protecting potential competition as the primary source of competition in the digital platforms market. To see it addressed in direct legislative change is encouraging. The ACCC stated that it already considers effects on potential competition in a market when assessing mergers.192 As such, this legislative change sought to clarify the matters to which they looked and to "emphasise the importance of something that is already considered in merger analysis.”193 Some submitters were concerned that the change might see an overemphasis on potential competition, resulting in more innovative-chilling false positives.194 The ACCC assured these submitters that this was neither likely, nor its intention, suggesting their assessments were unlikely to change substantially.195
Overall, the recommendation certainly appears a useful reformulation of the law, however whether it alone offers the meaningful protection of potential competition desperately required is unlikely, given as the ACCC mentioned, it is not intended to fundamentally change its analysis. It is perhaps notable that Google supported the legislative change, which indicates that it did not consider the change of great substantive effect.196
191 ACCC, above n 6, at 30.
192 At 107.
193 At 107.
194 see, for example REA Group “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 2 and Microsoft “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 5.
195 ACCC, above n 6, at 107.
196 Google “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 17.
The addition of data and technology to the s 50(3) list is welcome. Chapter one demonstrated that data is a particularly valuable asset for incumbent firms, and a key contributor to the strength of the barriers to entering the market. The addition of data to the s 50(3) list recognises that data can be uniquely valuable to firms, and may warrant special consideration in certain mergers. Many submitters were in favour of the change, and arguments to the contrary were weak.197 These tended to argue that data are a class of asset that vary in their competitive significance and do not warrant special consideration, an argument without academic traction.198
Recommendation two signals a return of sorts to the compulsory premerger notification system promulgated in New Zealand under the former 1975 Commerce Act.199 Professor Rex Ahdar argues that for such a system to operate effectively, it ought to only require the notification of mergers with anti-competitive potential.200 If the bar is set too low, competition agencies risk being inundated with proposals of a benign nature. Too low a bar is also a double-edged sword, for it also places a burden on the commercial parties to which it would apply for no apparent gain.201 The ACCC appear alert to this second risk, recognising “the burden that this recommendation may place on digital platforms” and the concomitant “unintended effects on innovation and investment” in digital markets.202
To alleviate both risks, as noted above the ACCC recommends establishing certain “minimum commitments” relating to the types of transaction that require notification.203 The Furman Report in the UK suggested companies with “strategic market status” ought to notify all mergers to the CMA.204 Strategic market status would be designated by the CMA to only large platforms that function as "competitive gatekeepers” to the market for numerous other firms, such as Google and Facebook.205 This proposal seems sensible. It is unlikely that it places too high a burden on firms such as Google and Facebook who are abundantly resourced. Provided the ACCC can process the required volume of transactions, a similar setting would be appropriate for Australia and New Zealand.
G Conclusion
The recommendations thus move the law in the right direction. However, they are underwhelming in the face of the fundamental changes required to merger law in this context. To better protect potential competition, merger law must be more forward
197 For submissions in favour see: The Guardian Australia “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February
2019) at 6-7; Oracle Corporation “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (March 2019) at.2; Australian Privacy Foundation “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 3; Outdoor Media Association “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 6; Media, Entertainment and Arts Alliance “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 1.
198 see for example, Microsoft “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019) at 6.
199 Rex Adhar The Evolution of Competition Law in New Zealand (Oxford University Press, 2020) at 197.
200 At 197.
201 At 197.
202 ACCC, above n 6, at 110.
203 At 110.
204 Furman, above n 11, at 139.
205 At 59.
looking, and accept that this analysis will necessarily be more uncertain than a static assessment of competition. It does not appear that the changes by the ACCC will either increase the tolerance for error in merger review, or help the ACCC with their difficult burden of proving that a transaction is anticompetitive. Mergers involving potential competitors will likely continue to be approved, despite the increasing evidence that some have reduced consumer welfare. The following chapter suggests that additional measures may be necessary, including lowering the standard of proof, shifting the burden of proof and improving our theories of harm.
H Some Bolder Suggestions
At present, in most jurisdictions the law only considers how likely a merger is to reduce competition. New Zealand’s s 47 of the Commerce Act requires a “substantial lessening of competition” in order for a merger to be blocked.206 Although this is often a reasonable approach, it does not adequately account for the magnitude of any harm or benefit to be considered alongside the likelihood it will occur.207
For digital mergers, this can be a crucial gap. We have seen that acquisitions are frequently of young firms responsible for new technology or an innovation that might see that firm eventually grow into a viable competitor.208 Under current legal settings, the Commerce Commission can only block a large incumbent from acquiring a smaller firm if that smaller firm is more likely than not to be able to succeed as a competitor to the incumbent. Given this is almost impossible to prove at an early stage of the small firm’s life, merger enforcement is next to impossible.
The Furman report recommended that the assessment should be able to test “whether a merger is expected to be on balance beneficial or harmful, taking into account the scale of impacts as well as their likelihood.”209 Merger assessment would shift to a more economic assessment of the expected outcome of intervention, rather than the single measure of likelihood currently used. To use the Furman Report’s example, authorities would have the option of acting where there was a 20% chance of serious harms to consumers, set against an 80% chance of relatively small benefits occurring.210
To apply this new framework, an assessment of Facebook’s merger with Whatsapp would require a competition agency to consider the harms of removing a potential future rival with Facebook’s network, along with the likelihood of Whatsapp actually entering the
206 Commerce Act 1986, s 47.
207 Furman, above n 11, at 13.
208 Argentesi, Buccirossi, Calvano, Duso, Marrazzo and Nava, above n 172, at 15.
209 Furman, above n 11, at 13.
210 At 120.
social media market.211 Harms might include things such as forgone improvements to the quality of Facebook’s service, less chance that Facebook would pursue new innovations and higher costs of digital advertising. Importantly, because of Facebook’s immense size, the large scale of these potential harm would be considered. This would then be assessed against the likelihood that Whatsapp would one day become a direct competitor with Facebook. Even if this was only moderately likely, because the potential benefits of it occurring are so high, the merger might be blocked.
This approach is attractive in the digital platforms context. It is inherently uncertain whether a start-up will grow to become a successful competitor, or whether a firm in an adjacent market will change lanes. Yet this uncertainty ought not to disable competition agencies altogether as it currently does. Of course, it would not be desirable to tip the scales totally in the opposite direction, and cause a flurry of false positive errors.
Interestingly, it was this fear that caused the CMA to oppose the suggestion. In their view, it would represent a “fundamental shift in merger policy” that should “not be underestimated.”212 However, it is arguable that a “fundamental shift” is exactly what is required. As mentioned, competition policy ought to tolerate some false positive errors, particularly in a context where false negatives errors are of such high cost.
Further, the CMA cautioned against “practical challenges” in applying the test in a “transparent and robust way” to avoid unintended consequences.213 The test admittedly adds another layer to the counterfactual assessment by requiring a consideration of benefits and harms as well as their likelihood. But the promise of a more accurate outcome justifies this added difficulty. A comprehensive cost-benefit analysis is already used in authorisation proceedings in New Zealand and common law guidance developed in that area is capable of informing a cost-benefit analysis in this area too.214
Separate to the Inquiry, the ACCC is considering whether to advocate for legislative change in favour of a rebuttable presumption against certain mergers in Australia.215 The starting point for the court would be that an acquisition will substantially lessen competition, absent clear and convincing evidence put by the merger parties.216 This would reverse current settings, where the ACCC has the burden of proving that an acquisition will contravene s 50.217 Similarly, it would reverse the current burden of proof
211 Note that the EU Commission primarily analysed the merger within the confines of the relevant market for consumer communications services. See Glick and Ruetschlin, above n 159, at 47.
212 Letter from Andrea Coscelli (Chief Eecutive of the Conduct Markets Authority) to Alex Chisholm (Permanent Secretary, Business, Energy and Industrial Strategy) regarding the CMA’s view of the recommendations of the Digital Competition Expert Panel (21 March 2019)
213 Letter from Andrea Coscelli, above n 212.
214 Mergers and Acquisitions Guidelines 2019 [2.44]
215 ACCC, above n 6, at 109.
216 At 109.
217 Competition and Consumer Act 2010 (Cth), s 50.
in New Zealand, where the Commerce Commission is currently obliged to prove that a merger is likely to substantially lessen competition under s 47.218
The ACCC briefly explained their rationale for this proposal. It recognised that proving that a merger is anti-competitive is particularly difficult in the digital market, where “market dynamics are particularly fast moving” and predicting the future direction of the market is likely to be more challenging.219 In addition, the ACCC noted that they will, by nature, have less knowledge of the transaction than the parties to it, and face a difficult time gathering the relevant information.220 Crèmer et al. noted that in the Facebook/Whatsapp merger, the Commission found no documentary evidence that Whatsapp was planning to become a fully-fledged social network in the future, placing particular reliance on that fact.221 They go on to note, that by its nature “such proof...will generally be difficult to obtain in other cases.”222 It makes sense then, for the merging parties (often already in possession of the required detail) to bear the burden of bringing it forth to the court.
This suggestion builds on recent criticism by the Australian Competition Tribunal and the Australian courts over the sufficiency of the ACCC’s evidence.223 The ACCC noted they had taken steps in response, such as “a greater focus on gathering documentary and other evidence to assist in its decision making”, but that they remained concerned at the hurdles it faced in opposing anti-competitive mergers in court.224
Responses to the proposal have been mixed. A recent study by Motto and Peitz found in favour of the concept of shifting the legal presumption, particularly in the context of transactions involving one party in an entrenched dominant position.225 They expected that in the majority of cases, the merging firms would be able to discharge the burden, ensuring it would not act too onerously.226 The Stigler Report for the US was strongly in favour, citing the gains to be made from speeding up merger review by placing the burden on the merging parties “who have the incentive, data, and resources to quickly deliver the right information to the authority.”227 Cremer et al. are in favour of shifting the burden of proof to a lesser degree, to only require the merging parties to prove the efficiencies of their deals.228
Most notably in opposition is the Furman Report. Furman et al. expressly considered a legal presumption against all acquisitions by large digital companies as an alternative to their “balance of harms” approach discussed in the preceding section.229 Ultimately, they found it was “not a proportionate response” to the challenges posed by mergers in the
218 Commerce Act 1986, s 47.
219 ACCC, above n 6, at 109.
220 At 109.
221 Crèmer, de Montjoye and Schweitzer, above n 21, at 119.
222 At 119.
223 ACCC, above n 6, at 108.
224 At 108.
225 Massimo Motta and Martin Peitz “Big Tech Mergers” (2020) Inf Econ Policy (in press)
226 At 14.
227 Stigler Committee on Digital Platforms Final Report (George J. Stigler Center for the Study of the Economy and the State, September 2019) at 111.
228 Crèmer, de Montjoye and Schweitzer, above n 21, at 11.
229 Furman, above n 11, at 101.
digital economy, “because the majority of such mergers are likely to be either benign or beneficial for consumers”.230 They also cited the importance of being acquired as an exit- strategy, implying their concern that the presumption would close this possibility and therefore be too far reaching.231
I share the view of the Furman Report that a legal presumption that all mergers in the digital platforms market substantially lessen competition is a very fundamental shift in the law, and may not be a proportionate response. It would unnecessarily burden merging parties in the majority of transactions that are not of concern to competition. However, this is not a reason to abandon the proposal altogether. Instead, the ACCC would be best served to vary the form of the presumption, to ensure it applied to a narrower range of transactions. As an example, the ACCC stated that in the US, mergers that result in a significant increase in market concentration are presumed likely to substantially lessen competition.232 Setting an appropriate concentration ratio above which a merger is subject to a rebuttable presumption might be an optimal response.
The Merger and Acquisition Guidelines 2019 recognise that though usually benign, a conglomerate merger can sometimes substantially lessen competition.233 However, the harm that the Guidelines recognise is limited to harm arising from a merged firm’s ability to foreclose competitors.234 Crèmer et al. propose a modification to this existing theory where a dominant, multi-product platform that benefits from strong network effects attempts to acquire a successful start-up in a related market.235 They propose that the transaction may have “strategic relevance” in shielding the acquiring firm’s ecosystem from competitive threats from fringe markets.236
Viewing the market in broader terms as an ‘ecosystem’ (as defined in chapter one), might justify a finding of a substantial lessening of competition in the circumstances outlined above. This is because the transaction can strengthen the acquiring firm’s ecosystem, and subsequently its market dominance in a number of ways. First, the transaction builds on the firm’s existing economies of scope, enabling it to collect a greater variety of data and to offer a higher quality core service to consumers.237 Secondly, the acquirer attempts to expand the scope of network effects that protect its core service by appropriating the network effects of the target firm.238 These stronger network effects both increase the benefit consumers gain from the ecosystem, making them less likely to leave, and raise
230 Furman, above n 11, at 101.
231 At 101.
232 ACCC, above n 6, at 109.
233 Mergers and Acquisitions Guidelines 2019, [5.12].
234 At 5.12.
235 Crèmer, de Montjoye and Schweitzer, above n 21, at 121.
236 At 124.
237 At 121.
238 At 122.
barriers to entering the market, making other firms less likely to enter. Crèmer et al. argue that such a transaction will substantially impede effective competition (as is the European test), where the target firm could have succeeded as a stand-alone business, or would realistically have been bought up by another competitor.239
New Zealand would be well served to update the Guidelines to reflect this sort of harm. It quells some of the uncertainty that arises from analysing this sort of transaction through the lense of harm to potential competition, for in this case the competitive threat is already present.240 However, some uncertainty persists in formulating the appropriate counterfactual. It requires analysing whether the target will survive and grow into a self- standing competitor if it were not acquired by the incumbent, or if other companies are realistically interested in buying the target. This naturally is a difficult, forward looking inquiry for which there is no easy recipe. Nevertheless, the theory appears to be a promising plug for the hole that persists in our theory of conglomerate harms.
I Conclusion
New Zealand would be best served by considering the bolder options suggested above. As matters stand, there is real concern that the reforms put forth by the ACCC will not deliver the fundamental changes required to merger law. It is promising that they intend to consider introducing a legal presumption against certain mergers, and that should be encouraged.
The Furman Report made an important point regarding the need for international co- operation. As it rightly suggested, “co-ordination across national borders would be the ideal way to address companies that operate at a global scale.”241 Avoiding a fragmented regulatory landscape across jurisdictions will be essential to ensure reforms deliver their promised benefits, and to ease the burden of adjustment on firms of adjusting to different laws. Given this, it is disappointing that fractures are already appearing between different countries’ responses. Australia is considering a legal presumption against certain mergers, the UK a lowering of the standard of proof and the EU a modification of existing theories of harm. It is now difficult to predict the jurisdiction with the tougher regime, and it creates a needless headache for firms attempting global mergers.
It also leaves New Zealand in a tough position. For a large multi-national firm New Zealand is but a blip on their radar. We would be poorly served by a merger regime that operated far more strictly than those overseas, for we would arguably be at risk of such a firm pulling out of the country altogether. Also interesting to consider is that New Zealand is far more likely to produce the kind of small start-up looking to be acquired than the next Facebook. From a selfish perspective, this might call for a more relaxed
239 Crèmer, de Montjoye and Schweitzer, above n 21, at 121.
240 At 122.
241 Furman, above n 11, at 119.
regulatory environment to allow our firms to enjoy the benefits of being acquired by a dominant incumbent. Nevertheless, as a member of the global community New Zealand ought to at least keep pace with the changes occurring throughout the world. In my view, this necessitates our adopting the recommendations put forth by the ACCC and at least one of the bolder suggestions relating to the standard and burden of proof. Ideally, all suggestions above would be considered.
IV Chapter THREE
A Introduction
Chapter one argued that competition policy ought to remain as the dominant control on large digital platforms, albeit in an updated form. However, irrespective of those updates, it is increasingly clear that ex post enforcement alone may not be enough to address all concerns arising in the digital platforms market. A new form of regulation is emerging, one of competition policy crossed with regulation, or “light touch regulation” as it is sometimes termed by Tirole.242 This is a more proactive approach to antitrust, one that attempts to pre-empt competitive harms and evolve at pace with the dynamic market. It promises to make competition law more agile and effective in the face of a complex and rapidly changing environment. Economist Jean Tirole argues the shift in favour of such regulation should be considerable, arguing the “balance between anticipating evolutions and reacting ex post should tilt more toward the former.”243
The ACCC recommended that this new form of ex ante regulation was administered by a specialist digital platforms branch.244 They recommended that the branch was given a range of enforcement powers and the ability to use various pro-competitive tools in addition to the existing ex post antitrust system.245 This chapter first outlines the shortcomings of the ex post competition regime, to justify the shift towards ex ante measures, before setting out the ACCC’s recommendation to establish a digital platforms branch in detail. It is worth noting that on the 12th of December 2019 the Australian Government committed to setting up the branch, and have since done so.246 Given this, and the fact that the case for ex ante regulation appears strongly made out, this chapter focuses on its exact form. It covers various issues relating to the desirable scope of the branch, its powers and the question of its independence. I end the chapter by examining some important pro-competitive tools the branch ought to keep in its arsenal, before asking whether New Zealand would be wise to create its own digital platforms branch.
242 Tirole, above n 84, at 3. The new approach is not to be confused with the “disastrous” ‘light-handed regulation that was trialled in New
Zealand in the 1980s and 90s, though it does share some characteristics (see Adhar, above n 199, at 293). Light-handed regulation was a decision to deregulate several natural monopoly industries, and instead subject those industries to competition legislation and regulations controlling information. (see Louise Hill “Litigation Versus Regulation: How Successful Is Light-Handed Regulation?” [1999] AukULawRw 7; (1999) 8(4) Auckland U L Rev 1119 at 1132.) The new approach of light-touch regulation is more similar to industry specific regulation, where a new regulator administers regulations, the regulations themselves designed to stimulate competition. Regulation under the former approach was designed solely to provide the Commerce Commission or private parties with the requisite information to investigate and, where relevant, prove anti- competitive behaviour.242 Furthermore, compared to the industries subjected to light-handed regulation, the structure of the digital platforms market is not clearly naturally monopolistic, meaning attempting stimulating competition in the market is, at this stage, still a feasible approach. 243 At 7.
244 ACCC, above n 6, at 31.
245 At 31.
246 Regulating in the digital age: Government Response and Implementation Roadmap for the Digital Platforms Inquiry (The Australian Government The Treasury, December 2019) at 3.
B Competition Policy’s Shortcomings
Even if competition policy is optimised for the digital economy, by their nature, ex post assessments are constrained in how effective they are in the digital platforms market. The reasons are various, and some are outlined below.
First, ex post assessments are slow. The average duration of antitrust cases completed by the CMA between 2014-2017 was 39 months.247 The complexity and opacity of the digital platforms market can make this even longer. Eight years elapsed between the 2009 complaint against Google Shopping and the European Commission’s eventual 2017 decision.248 This means the effects of anticompetitive conduct linger for extended periods of time, causing greater damage to rival firms and ultimately, consumer welfare. In a digital market where the issues are sometimes particularly complex, and court proceedings relatively longer than usual, companies exposed to anti-competitive behaviour may go out of business before a remedy is delivered. As Furman et al. argue, the use of a team of experts such as the digital platforms branch, as well as their ongoing collection of data relevant to investigations, appear promising means to shorten the duration of proceedings.249
Secondly, ex post competition policy is mostly backward looking.250 This exposes incumbents to legal uncertainty, for they must operate without clear guidelines as to what they can and cannot do.251 Furman et al. note that this is particularly pertinent in the digital platforms context, where the issues are often “so nuanced and technical that it can be hard to draw clear forward-looking guidance from them for wider behaviour.”252 It also means anti-competitive behaviour tends to be treated, rather than prevented.
Although clear theories of harm might deter future anticompetitive conduct, an ex post assessment can do little to help a competitor already savaged by the conduct.253 They can also do little to improve competition in digital markets that have already ‘tipped’ in favour of one incumbent. These concerns could be remedied by a more prospective approach, based on a code of competitive conduct that clarified acceptable conduct between platforms and their users. At the end of the chapter I examine the required settings for an effective code of conduct.
Thirdly, the ACCC noted that existing competition law is insufficient to deal with market failures that arise due to a lack of transparency or due to externalities in other markets.254 While section 46 of the Competition and Consumer Act 2010 allows investigation of abuses of market power, the ACCC stated they are “unable to investigate issues stemming from a lack of transparency or from other market failures.”255 Further, the
247 Competition and Markets Authority Annual Report and Accounts 2016/17 (Competition and Markets Authority, Annual report, July 2017) at
30. This compares to the average figure for all cases over that period of 25 months.
248 Case AT.39740 European Commission v Google Search (Shopping) ECLI:EU:C:2017:4444
249 Furman, above n 11, at 104.
250 Tirole, above n 84, at 7. Tirole rightly notes that an exception to this is merger policy, which “is a reason to give it a bigger role in preventing further concentration.”
251 At 7.
252 Furman, above n 11, at 58.
253 Tirole, above n 84, at 6.
254 ACCC, above n 6, at 139.
255 At 139.
opacity of the platforms market means that many breaches likely go undetected.256 This is especially so given that the ACCC currently rely on conduct to be brought to their attention by market participants. The ACCC is concerned these participants might be increasingly unable to detect discrimination where platforms operate as opaque “black boxes”.257 Pro-active investigation of the market by a specialist agency might remedy these concerns.
Finally, remedies are difficult to get right. An effective remedy is one that is a clear deterrent against future anticompetitive conduct. As Furman et al. argue, this requires that fines are extremely high in the case of large incumbent platforms, which would be difficult to impose on a consistent basis and could make fines disproportionately harsher than a firm’s anticompetitive conduct in some instances.258 Even an effective remedy does not restore competition to markets in which competition has been harmed. An antitrust authority that wants to restore lost competition must induce entry and/or impose structural remedies.259 Such remedies are not favoured in New Zealand due to their complexity and required ongoing monitoring. It is not difficult to see why pre-empting competitive harms altogether is preferable.
C The ACCC’s Recommendation
Recommendation four of the ACCC’s report provides the proposal to establish a specialist digital platforms branch. This branch has the purposes of:260
The proposal accords with international efforts. In the UK, the Furman Report recommended the establishment of a similar Digital Markets Unit. This unit was formally established on the 1st of July 2020.261 In the US, the Federal Trade Commission’s Bureau of Competition established a Technology Task Force early last year.262 Also of note is that this approach to competition has a number of precedents. The new specialist branch is highly similar in design to several existing specialist regulatory bodies including the Financial Markets Authority.
256 At 139.
257 ACCC, above n 6, at 139.
258 Furman, above n 11, at 38.
259 Stigler Report, above n 227, at 99.
260 ACCC, above n 6, at 142.
261 Jesse Norman BUDGET 2020 (Her Majesty’s Treasury, Copy of the Budget Report, March 2020) at 99.
262 Federal Trade Commission “FTC’s Bureau of Competition Launches Task Force to Monitor Technology Markets” (press release, 26 February 2019)
D Specifics of the Digital Platforms Branch
An important preliminary detail to be determined is exactly whom the branch should regulation. Furman et al. convincingly argued that the scope of the branch should be kept narrow, to minimise the burden on smaller companies and reduce unnecessary action.263 That report favoured focusing solely on firms with “strategic market status”, or those that control a gateway or bottleneck in a digital market, meaning they control others’ market access.264 The ACCC were somewhat less precise, preferring their branch to focus on all digital platforms, as in their view “issues relating to market failure are not dependent on market power.” In practice, they would focus their investigations on the larger digital platforms, being Facebook and Google at present.265
Some academics have advocated for a clearer legal test to determine the characteristics of a company’s market position above which regulatory powers are appropriate. Alexiadis and de Streel recently articulated one possibility.266 Their simple, three-criteria test not only clarifies to whom specialist regulation ought to apply, but also of what sort. The first criterion is a non-contestable, concentrated market structure.267 This requires the existence of a digital platform in a market that is characterised by network effects, high entry barriers and that is prone to tipping, identical to the markets of Google and Facebook. The second is that the platform must act as a digital “gatekeeper” which renders it an unavoidable trading partner.268 This requirement encapsulates the concept of a firm with “strategic market status” from the Furman Report, as earlier discussed. Where the first two criteria of the test are met, certain types of behaviour are likely to justify ex post intervention in the market.269 The third criterion requires that ex post enforcement is unlikely to be effective. Factors suggesting an ex post intervention’s ineffectiveness include the impossibility of establishing an abuse of dominance claim, or a lack of a suitable, timely remedy for a given harm, making the changes to the marketplace irreversible.270 Only when this third criterion is also met is ex ante regulation necessary.
The ACCC’s attempt to monitor all digital platforms might be manageable for now, however as time progresses and their number grows, a more streamlined intervention test such as that above could be necessary.
263 Furman, above n 11, at 10.
264 Furman, above n 11, at 55. The Digital Markets Unit would designate “strategic market status”, periodically assessing their designation.
265 ACCC, above n 6, at 141.
266 see Peter Alexiadis and Alexandre de Streel “Designing an EU Intervention Standard for Digital Platforms” (Working paper, Florence School of Regulation, 2020). Please note their test is produced here in a simplified form for clarity.
267 At 35.
268 At 36.
269 At 39. Behaviours suggested as being likely to be anti-competitive by the authors were bundling and envelopment strategies, refusals to provide access or interoperability to key inputs and innovation capabilities, discrimination/self-preferencing, and violations of key regulatory principles.
270 At 37.
The salient feature of the branch is its proactive monitoring and investigation power. The ACCC indicated that the opacity of the market makes the detection of anticompetitive behaviour difficult.271 In response, the branch is endowed with the ability to gather information and investigate breaches even where no complaint has been made by market participants.272 Relatedly, the branch has the ability to compel data from digital platforms.273 This gives the branch similar powers to entities in the telecommunications industry for example, and allows the ACCC to build an evidence base to assist with future investigations, rather than relying on collecting evidence after an investigation begins.274 Recent cases such as Google Shopping demonstrate the importance of data in court proceedings, with that case relying substantially on data collected by a third party to show the effect of Google’s changes to its algorithm.275 Third party data cannot always be relied upon with confidence by antitrust agencies, making it welcome to see the ACCC attempt to build their own data trove.
The branch appears set to become a key policy informant. It is furnished with the ability to publicly report to Government on issues of concern that might not breach the Competition and Consumer Act 2010.276 Further, the branch is empowered by Ministerial direction to hold extended public inquiries of at least five years in duration.277 On the 10th of February 2020 the Australian Government directed the branch to conduct its first such inquiry; a comprehensive inquiry into markets for the supply of digital platform services. This final report of this inquiry is not due until the middle of 2025.278
Further details regarding the powers of the branch are limited, though there is an assurance that the branch may take “action to enforce competition and consumer laws”.279 Presumably, this means the branch will have the same general enforcement powers as the ACCC. I welcome this and would suggest similar settings for an equivalent New Zealand regulator, to ensure the reforms that are pursued have legal force.
271 ACCC, above n 6, at 140.
272 At 141.
273 At 141.
274 for example, the Australian Prudential Regulation Authority has powers to compel data collection.
275 Google Search (Shopping), above n 248.
276 ACCC, above n 6, at 141.
277 At 142.
278 Josh Frydenberg Competition and Consumer (Price Inquiry— Digital Platforms) (Australian Government, Ministerial Direction, February 2020)
279 ACCC, above n 6, at 141.
The branch sits as a division of the ACCC, rather than as an independent regulator. The ACCC proposed this to capitalise on their existing skills and expertise in relation to the digital platforms market. Further, in their view it allows the branch to benefit from the existing relationships of the ACCC with other Australian Government entities.280
This proposal seems practical. One of the major oppositions to this new regulator in submissions to the ACCC was the risk of regulatory capture, or that rivals might attempt to lobby the regulator, to obtain through regulation what they could not achieve through the market.281 The Stigler Report rightly notes that the risk of regulatory capture is reduced in the case of a cross-industry regulator as compared to one that stands alone.282 The case is thus strong for a New Zealand equivalent to sit within our Commerce Commission in a similar fashion.
E Pro-competitive Regulatory Tools
The code of conduct is likely to be the predominant site of platform regulation in the near future. Economist Jean Tirole is vocal in his support for a new age of “participative antitrust”, in which industry and antitrust authorities develop regulations in conjunction.283 Participation between industry and regulators promises to make new regulation better informed, however it must be noted that the interests of both parties are by nature not aligned, or else regulation would not be required in the first place. Codes of conduct will therefore have to be mandatory and legally binding to be effective. The EU uses codes of conduct to assist in the operation of their Platform-to-Business regulations, to regulate the relationship between large platforms and dependent business users.284 The ACCC are also in the processing of drafting a mandatory code of conduct, this to manage the relationship between large digital platforms and the news media.285 Australia and New Zealand are in need of a code of conduct to govern anti-competitive conduct, and ought to look to the European Union for precedents.
The justification for the code of conduct is strong. As ex ante, black letter law it promises to establish clearer obligations of all parties in the market, thereby acting in the interests
280 ACCC, above n 6, at 141.
281 ICLE “Submission on the final report of the Australian Competition and Consumer Commission’s Digital Platforms Inquiry” (September 2019) at 11.
282 Stigler Report, above n 220, at 18.
283 Alison Schrager “A Nobel-winning economist’s guide to taming tech monopolies” Quartz (New York City, June 27 2018)
284 Regulation 0112/2018 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L 186/57
285 Treasury Laws Amendment (News Media and Digital Platforms Mandatory Bargaining Code) Bill 2020 (Cth) (Exposure Draft)
of commercial predictability, certainty and transparency.286 Codes are ideally created in partnership with industry, lessening the information asymmetry between tech firms and antitrust agencies that can make solving issues difficult.287 Crucially, they promise quick and efficient redress.288 Clearer obligations established in a code promise to make it easier to determine when they are breached, allowing issues to be more quickly resolved. Even more importantly, codes of conduct promise to avoid disputes altogether, because parties are more certain of their obligations to each other, and more careful not to breach these.289 The result, argue Furman et al., are “better and more competitive outcomes for consumers.”290
It all sounds rather promising. Whether these benefits will fully be realised remains unclear, as the approach is still very much in its infancy. Resistance from firms already appears to be a barrier to success, and highlights the limited extent to which this new approach can truly be ‘participatory’. As I write this, Facebook announces their plan to stop users from sharing news content in Australia, in retaliation to the ACCC’s draft code of conduct requirement that Facebook remunerate news outlets for posting their content.291 Co-ordinating codes cross-jurisdictionally will reduce the effectiveness of these corporate tantrums, but for now they remain problematic.
The European Competition Commission has developed useful principles to guide the effective formulation and implementation of codes of conduct.292 They advise codes to be prepared openly, by as many relevant actors as possible. They ought to be principle based, starting with principles of a sufficiently high level to apply across different contexts. Under each principle, a successful code would then set out more detail to clarify the forms of conduct that are fair and reasonable, and those that are not.293 I avoid prejudging precisely what those details would be in the digital platforms context, bearing in mind they ought to be established in collaboration with industry.
In terms of the implementation of the code, the EU Commission advise its monitoring in an open and autonomous way. They argue that it ought to be improved in an iterative manner and that non-compliance should be subjected to sanctions.294 The Furman Report emphasised the importance of quick resolutions, ideally achieved before an anticompetitive action has either debilitated a firm, or harmed consumers.295 To this end, it is very useful the digital platforms branch is endowed with powers to compel information from firms, and to impose legally binding sanctions, measures that should hasten the resolution of issues.
The time has come for New Zealand and Australia to implement codes of conduct to govern anti-competitive behaviours of digital platforms. Codes can and should be
286 Furman, above n 11, at 60.
287 Alexiadis and de Streel, above n 266, at 42.
288 Council of the European Union “EU introduces transparency obligations for online platforms” (press release, 14 June 2019).
289 Furman, above n 11, at 63.
290 At 60.
291 Radio New Zealand “Facebook threatens to halt news sharing in Australia” (1 September 2020) RNZ < https://www.rnz.co.nz >
292 European Commission Principles for Better Self- and Co-Regulation (28 November 2018) < https://ec.europa.eu/ >
293 At 62.
294 European Commission, above n 292.
295 Furman, above n 11, at 63.
complementary to each other, meaning different relationships with the platforms can feasibly be governed by different codes. The EU has already produced an excellent template to govern the platform-to-business relationship.296 The ACCC are in the process of finalising a code to govern the relationship between platforms and the news media.297 The Commerce Commission should strongly consider these templates for use in New Zealand. It makes sense to carefully co-operate with Australia in this regard. The management teams of Facebook and Google both manage Australia and New Zealand as a single area, meaning a single code for both areas would be more effective. New Zealand is a relatively insignificant market for both Facebook and Google, meaning firms would sooner pull out of the country than comply with onerous regulation. Drawing up codes of conduct in conjunction with Australia is essential therefore. Coordinating codes internationally is even more desirable.
Significant competitive opportunities could be created through systems built on open standards. A system built on open standards is one built using commonly agreed technical specifications that are freely available.298 Open standards are building blocks for interoperability, allowing products from different firms to interoperate, or connect with each other. They are widely regarded as the foundation of the internet as we know it, with web content transmitted via an open standard known as the Hypertext Transfer Protocol (HTTP), a standard that makes it possible for anyone to share a webpage for free.299 Email from different providers is also made possible via an open standard (Simple Mail Transfer Protocol, or SMTP).300
Open and interoperable standards have already proven themselves as valuable enablers of competition. In 2016, the UK’s Open Banking System developed an open API system that allowed a common technical standard for fintech innovators to access information about bank services, prices, service quality and customer usage.301 There is evidence the system reduced barriers to entering the market, with a relative surge of 200 organisations seeking accreditation to provide bank services in the subsequent years.302 Open standards are thought to positively affect competition by allowing innovators to more easily build new services to compete directly with existing ones.303 Further, innovators may build ancillary services to connect with existing systems, allowing ‘fringe’ competition in a market dominated by an incumbent.304
296 Regulation 0112/2018 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L 186/57
297 Treasury Laws Amendment (News Media and Digital Platforms Mandatory Bargaining Code) Bill 2020 (Cth) (Exposure Draft)
298 Furman, above n 11, at 71.
299 At 71.
300 At 72.
301 Competition and Markets Authority Retail Banking market investigation: overview of the final report (August 2016) at 6. Note that API is the acronym for Application Programming Interface, which is a software intermediary that allows two applications to talk to each other.
302 Bill Roberts “Celebrating the first anniversary of Open Banking” (11 January 2019) GOV.UK <https://competitionandmarkets.blog.gov.uk>
303 Furman, above n 11, at 72.
Both the Furman Report and Crèmer et al.’s report for the European Commission of Competition advocate for the introduction of open standards of some form to the digital platforms market. However, the extent to which they are desirable in this context is unsettled. Both reports recognise that departing from a closed system requires some sacrifice.305 As Furman et al. argue, a closed system can better protect intellectual property and offer a more secure business investment, can help ensure technical consistency and crucially, can be more easily updated and developed.306 Crèmer et al. argue that requiring strong standardisation across competition platforms may dampen their ability to innovate and differentiate their types of service.307 This would then eliminate one of the most important grounds for competition between platforms.
To that end, Crèmer et al. suggest an attractive middle ground. They differentiate between what they term “full protocol interoperability” and the lesser “protocol interoperability.”308 Full protocol interoperability requires two services to fully interoperate, demanding deep standardisation that, in the authors view, would likely dampen a platform’s ability to innovate and differentiate their service.309 The connection of telecommunications networks would be the most prominent example of this. In our context, it might require Facebook’s users to be able to message and interact with users of another social media platform. Instead, a lesser form of interoperability, or ‘protocol interoperability’ is preferred. This ensures that two services can interconnect with each other, and has already been imposed as a duty in antitrust in the Microsoft case.310 Examples in the digital platforms context include a platform interconnecting with the operating system of a computer, or with complementary services. Although full protocol interoperability more effectively decreases the extent to which consumers are locked in by network effects, the harms to incentives to innovate and the ability to differentiate a service, as outlined above, make it a less attractive prospect in my view.
The ACCC would be wise in future to consider how open standards might best be used in the digital platforms market. They ought to proceed with some caution, for any standard defined too strictly or too early could hinder innovation.311 A New Zealand regulator should work together with the ACCC to establish a common standard for the Australia/New Zealand region to maximise the chances of compliance. Co-ordination with competition agencies further afield is desirable, provided it is possible.
305 A closed system is the alternative to an open system. A closed system can be thought of as Apple’s OSX operating system, which only runs
on specific hardware produced by the manufacturer. In comparison, Windows’ open operating system allows interoperability with a range of hardware.
306 Furman, above n 11, at 72. Many in the tech world lament that the open standards elements of the internet are “stuck in time”. For example, see moxie0 “Reflections: The ecosystem is moving” (10 May 2016) Signal <https://signal.org> “We got to the first production version of IP, and have been trying for the past 20 years to switch to a second production version of IP with limited success. We got to HTTP version 1.1 in 1997, and have been stuck there until now. Likewise, SMTP, IRC, DNS, XMPP, are all similarly frozen in time circa the late 1990s.”
307 Crèmer, de Montjoye and Schweitzer, above n 21, at 85.
308 At 58.
309 At 85.
310 Case COMP/C-3/37.792 European Commission v Microsoft [2004] OJ L 32/23
311 Crèmer, de Montjoye and Schweitzer, above n 21, at 85.
Consumers ‘pay’ for free online services through their attention and their provision of personal data.312 When aggregated, this data is immensely valuable to firms, allowing them improve the quality of their service in a way akin to ‘learning-by-doing’.313 Network effects increase the value of data, meaning the more users a firm has, the greater the ‘incumbency advantage’ secured by their data trove.314 Yet datasets are non- rivalrous, meaning that opening them up to competing users does not deplete their use by the original user.315 Significant opportunities therefore exist to enable consumers to port their data between competing providers.
The ACCC briefly addressed the desirability of data portability. They considered that the application of the Consumer Data Right (CDR) to enable data portability could improve competition by facilitating competitive entry and consumer switching.316 Despite this finding, the ACCC chose not to recommend increased data portability obligations, as they felt there were more pressing threats to competition to address.317 Although this may be so, designing a data portability regime in the future does promise to benefit competition.
In New Zealand, the Ministry of Business, Innovation and Employment is currently deciding whether to develop a CDR in New Zealand.318 Although still in the consultation process, the Ministry’s preferred option is to introduce the right on a “sectoral- designation approach”, allowing it to be tailored to different sectors.319 This option is welcomed, insofar as it allows a right to be defined that is specific to the digital platforms market, and designed in the interests of improving competition.
In terms of an appropriate design, some inspiration can be sought from the European Union’s General Data Protection Regulation (GDPR).320 Article 20 of the GDPR codifies a limited right to data portability, allowing users to receive a copy of their personal data in a “‘structured, commonly used and machine-readable format” and manually migrate it to another service.321 However, both Crèmer et al. and the Furman Report favour an even more stringent regime in the digital platforms context.322 For data portability to have a meaningful effect on competition, Crèmer et al. argue that data must be made available in real time, on a continuous basis.323 Data must also be formatted in a common standard, to
312 Coyle, above n 18, at 16.
313Biglaiser, Calvano and Crèmer, above n 24, at 43. For a broader discussion of data in this paper, see page 9.
314 At 43. The idea of the ‘incumbency advantage’ is discussed above at page 8.
315 Furman, above n 11, at 23.
316 ACCC, above n 6, at 115.
317 At 116.
318 Options for establishing a consumer data right in New Zealand (Ministry of Business, Innovation & Employment, Discussion Document, August 2020
319 At 39.
320 Regulation 679/2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) [2016] OJ L 119/1
321 Regulation 679/2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) [2016] OJ L 119/1
322 see Furman, above n 11, at 68. and Crèmer, de Montjoye and Schweitzer, above n 21, at 81.
323 Currently, under the GDPR, only past data must be made available to consumers, however many services rely on real time data to be effective. Competing services will likely struggle to use this older data to improve their offering, and consumers will be less likely to multi- home between platforms, as all their real time data (and its concomitant benefits) remain on the original platform. See this argument in Crèmer, de Montjoye and Schweitzer, above n 21, at 81
facilitate its easy transmission between suppliers.324 Neither requirements are secured by the GDPR, but ought to be considered as essential for New Zealand’s CDR.
It is very important to consider the potential harms of a CDR designed in this way. Compliance is likely to be costly, and therefore when imposed on non-dominant firms in particular, it could diminish competition and ultimately harm consumers.325 To this end, Crèmer et al. suggest a more stringent regime should apply only to dominant firms, something that is welcomed.326 Another unavoidable cost is that a company that has invested in collecting and accumulating personal data may have their investment devalued, by allowing users to easily migrate this data to other firms.327 In addition, mandating common data standards to allow for easy data mobility can impede a business’ ability to implement a new idea which goes beyond this standard, potentially stifling innovation.328 When considering exactly how broad the scope of the CDR should be, these costs will be important to balance against the benefits of enabling data portability with respect to any given type of data.
F Conclusion
A pro-active approach to tackling competition issues in the market appears promising. New Zealand’s Commerce Commission currently lacks the power to implement most of the ex ante regulation that is being adopted overseas. This needs to change. Whether in the form of a new digital platforms team within the Commerce Commission, or as an entirely collaborative exercise with Australia, New Zealand needs the ability to deliver ex ante digital platforms regulation. Codes of conduct, open standards and data portability may help stimulate competition in the digital platforms market. New Zealand needs the ability to implement these, to ensure the relationship between large platforms and our local businesses, news media and consumers is as productive as possible.
324 Although requiring a “commonly used format”, the GDPR does not require a common standard to facilitate the transmission of data between
suppliers. Currently, this means that businesses must convert data to their required format, a process that can be complex and time-consuming. See Furman, above n 11, at 68.
325 At 82.
326 At 82.
327 Furman, above n 11, at 71.
328 At 71.
V Conclusion
Jurisdictions around the world are fitted with an analogue system of competition law in an increasingly digital world. It is a system that has not served us well in the digital platform revolution. Digital platform markets are highly concentrated, competition is largely absent, and firms are of immense size and market power. However, the system can adapt, and indeed it must, as we can rely on no other discipline to control the excesses of dominant players.329
Issues stemming from a lack of competition can in part be blamed on the nature of the digital platforms market itself. But they can also be blamed on weak merger regimes that have driven concentration, and allowed incumbents to circumvent the competitive process altogether. That has left us in a position where inaction is no longer an option.
The mistakes of the past have created a market that is unlikely to correct itself. Intervention to stimulate competition ex ante is now a necessity, and that requires global co-operation in order to be effective.
As part of the international community, New Zealand needs to respond. The time is here for New Zealand to assess its own merger laws, to determine whether they ought to be strengthened in a manner similar to other jurisdictions. Far more importantly, we must address the relationship between large dominant platforms and New Zealand’s businesses, media and the general public. A new regulator, or branch of the Commerce Commission will likely be required to protect New Zealand’s best interests in these relationships.
They have been intentionally omitted from this paper, but the actions we take with respect to some of the world’s largest companies will necessarily have immense social and political implications. The economic choices we make are part of a broader social bargain between consumers, small businesses and big businesses. It is important that we get them right.
329 As was eloquently put by Laila Harré, a member of the Alliance political party in 1999 “Competition is the only alternative to
either State ownership and/or stringent industry regulation. We can rely on no other discipline to control the excesses of dominant players.” (Second Reading of the Commerce (Control of Dominant Position) Amendment Bill: (1999) 576 NZ Parl Deb 15814. as cited in Adhar, above n 199).
VI Bibliography
A Books
Francesco Ducci Natural Monopolies in Digital Platform Markets (Cambridge University Press, Cambridge, 2020).
Geoffrey Parker, Marshall W. Van Alstyne and Sangeet Choudary Platform Revolution: How Networked Markets Are Transforming the Economy and How to Make Them Work for You (W.W.Norton, United States, 2016).
Jose Alberro and Rainer Schwabe The Theory of Contestable Markets and its Legacy in Antitrust (Volume 16, number 1, Cornerstone Research, United States of America, 2016) Joseph Schumpeter Capitalism, Socialism and Democracy (George Allen & Unwin Ltd, UK, 1943).
Martin Moore and Damian Tambini Digital Dominance: The Power of Google, Amazon, Facebook, and Apple (Oxford University Press, New York, 2018).
Rex Adhar The Evolution of Competition Law in New Zealand (Oxford University Press, Oxford, 2020).
Richard Bork The Antitrust Paradox: A Policy at War with Itself (New York, Basic Books, 1978).
B Cases
European Union
Case AT.39740 European Commission v Google Search (Shopping)
ECLI:EU:C:2017:4444.
Case COMP/C-3/37.792 European Commission v Microsoft [2004] OJ L 32/23.
United States
United States v AT&T Co 552 F Supp 131 (D DC 1982).
United States v. Marine Bancorporation Inc. [1974] USSC 155; 418 US 602 (1974).
United States v. Microsoft Corporation [2001] USCADC 115; 253 F 3d 34 (DC Cir 2001).
C Internet Sources (including working papers)
Bill Roberts “Celebrating the first anniversary of Open Banking” (11 January 2019) GOV.UK < https://competitionandmarkets.blog.gov.uk>Damien Geradin “What Should EU Competition Policy do to Address the Concerns Raised by the Digital Platforms’ Market Power?” (12 December 2018) Social Science Research Network < https://www.ssrn.com>
David Evans “Multisided Platforms, Dynamic Competition, and the Assessment of Market Power for Internet-Based Firms” (10 March 2016) Social Science Research Network < https://www.ssrn.com/index>
Diane Coyle. "Making the most of platforms: a policy research agenda." (16 October 2016) Social Science Research Network <https://www.tse-fr.eu >
European Commission Principles for Better Self- and Co-Regulation (28 November 2018) European Commission Website < https://ec.europa.eu >
Mark Glick and Catherine Ruetschlin “Big Tech Acquisitions and the Potential Competition Doctrine: The Case of Facebook” (8 November 2019) Social Science Research Network < https://www.ssrn.com/index.cfm/en/>Nicolas Petit “Technology Giants , the “Moligopoly” Hypothesis and Holistic
Radio New Zealand “Facebook threatens to halt news sharing in Australia” (1 September 2020) RNZ < https://www.rnz.co.nz >
Sarah Frier and Nic Grant “Instagram Brings In More Than a Quarter of Facebook Sales” (February 5 2020) Bloomberg News < https://www.bloomberg.com>
Reuters Staff “TIMELINE-Google's decade-long antitrust battle in Europe” (21 March 2019) Reuters < https://www.reuters.com>
Moxie0 “Reflections: The ecosystem is moving” (10 May 2016) Signal
<https://signal.org>
Competition: A Primer” (20 October 2016) Social Science Research Network
<https://www.ssrn.com/index>
D Journals
A. Douglas Melamed and Nicholas Petit “The Misguided Assault on the Consumer Welfare Standard in the Age of Platform Markets” (2019) 54 Rev Ind Organ 741. Carl Shapiro “Antitrust in a Time of Populism” (2018) 61 IJIO 714.Cédric Argenton and Jens Prüfer “Search Engine Competition With Network Externalities” (2012) 8 JCL & E 73.
Chris Noonan, “Horizontal Mergers in New Zealand: Drifting Towards Structuralism” (2019) 26 CCLJ 263.
Darren Bush and Salvatore Massa “Rethinking the Potential Competition Doctrine” (2004) 2004 WIS L REV 1035.
David Evans “Attention Rivalry Among Online Platforms” (2013) 9(2) JCL & E 313. Diana Coyle, “Practical Competition Policy Implications of Digital Platforms" (2005) 82 ALJ 3.
Gary Biglaiser, Emilio Calvano and Jacques Cremer “Incumbency Advantage and its Value” (2019) 28 J Econ Manage Strat 41.
Geoffrey Manne and Joshua Wright “Innovation and the Limits of Antitrust” (2010) JCL & E 153.
Herbet Hovenkamp “Antitrust and Platform Monopoly” (2020) 130 Yale LJ 20. Howard Shelanski “Information, Innovation, and Competition Policy For the Internet” (2013) 161 U Pa L Rev 1663.
John Newman “Antitrust in Digital Markets” (2019) 72 VAND L REV 1497. Jonathan Baker “Beyond Schumpeter vs. Arrow: How Antitrust Fosters Innovation” (2007) 74 ALJ 575.
Kristina Shampanier, Nina Mazar and Dan Ariely “Zero as a Special Price: The True Value of Free Products” (2007) 26 Mark Sci 742.
Lina Khan “Amazon’s Antitrust Paradox” (2016) 126 Yale LJ 710.
Louis Kaplow “Why (Ever) Define Markets” (2010) 124 Harv L Rev 437. Louise Hill “Litigation Versus Regulation: How Successful Is Light-Handed Regulation?” [1999] AukULawRw 7; (1999) 8(4) Auckland U L Rev 1119.
Massimo Motta and Martin Peitz “Big Tech Mergers” (2020) Inf Econ Policy (in press).
Mike Walker "Competition policy and digital platforms: six uncontroversial propositions." (2020) 16 Eur Competition J 1.
Thibault Schrepel “Predatory Innovation: the Definite need for Legal Recognition” (2018) 21 SMU Sci & Tech L Rev 19.
Uğur Akgün Cristina Caffarra Federico Etro Robert Stillman “On the welfare impact of mergers of complements: Raising rivals’ costs versus elimination of double marginalization” (2020) 195 Econ Lett.
William Baumol "Contestable Markets: an Uprising in the Theory of Industry Structure" (1982) 72 American Economic Review 1.
E Legislation
NZCommerce Act 1986.
Merger and Acquisition Guidelines 2019.
AUS
Competition and Consumer Act 2010 (Cth).
Treasury Laws Amendment (News Media and Digital Platforms Mandatory Bargaining Code) Bill 2020 (Cth) (Exposure Draft).
European Union
Regulation 0112/2018 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L 186/57.
Regulation 679/2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) [2016] OJ L 119/1.
F Letters
Letter from Andrea Coscelli (Chief Eecutive of the Conduct Markets Authority) to Alex Chisholm (Permanent Secretary, Business, Energy and Industrial Strategy) regarding the CMA’s view of the recommendations of the Digital Competition Expert Panel (21 March 2019).
G Newspaper Sources
Alison Schrager “A Nobel-winning economist’s guide to taming tech monopolies”Quartz (New York City, June 27 2018).
Paul Rogers “Consumer watchdog chairman Lord Tyrie quits over ‘limits’ of powers”
The Times (London, June 18 2020).
H Press releases
Council of the European Union “EU introduces transparency obligations for online platforms” (press release, 14 June 2019).Facebook “Facebook Reports Second Quarter 2020 Results” (press release, July 30 2020).
Federal Trade Commission “FTC’s Bureau of Competition Launches Task Force to Monitor Technology Markets” (press release, 26 February 2019).
I Reports
Ania Thiemann and Pedro Gonzaga Big Data: Bringing Competition Policy to the Digital Era (Directorate for Financial and Enterprise Affairs Competition Committee, DAF/COMP14, October 2016).
Competition and Markets Authority Retail Banking market investigation: overview of the final report (August 2016).
Digital Platforms Inquiry: Final Report (Australian Competition and Consumer Commission, June 2019).
Elena Argentesi, Paolo Buccirossi, Emilio Calvano, Tomaso Duso, Alessia Marrazzo and Salvatore Nava Merger Policy in Digital Markets: An Ex-Post Assessment (LEAR, DIW Berlin Discussion Paper No. 1836, December 2019).
ICLE Final Report on Competition and Consumer Protection in the 21st Century: The Weakness of Interventionist Claims (ICLE, June 2019).
Jacques Crémer, Yves-Alexandre de Montjoye and Heike Schweitzer Competition Policy For the Digital Era (Publications Office of the European Union, 2019).
Jason Furman Unlocking digital competition: Report from the Digital Competition Expert Panel (UK government publication, HM Treasury, March 2019).
Jesse Norman BUDGET 2020 (Her Majesty’s Treasury, Copy of the Budget Report, March 2020).
Josh Frydenberg Competition and Consumer (Price Inquiry— Digital Platforms)
(Australian Government, Ministerial Direction, February 2020).
Options for establishing a consumer data right in New Zealand (Ministry of Business, Innovation & Employment, Discussion Document, August 2020).
Regulating in the digital age: Government Response and Implementation Roadmap for the Digital Platforms Inquiry (The Australian Government The Treasury, December 2019).
Stigler Committee on Digital Platforms Final Report (George J. Stigler Center for the Study of the Economy and the State, September 2019).
J Speeches
Carl Shapiro, Professor of Economics at the University of California at Berkeley “Hearing #5” (Federal Trade Commission Hearings, Washington, D.C. November 2018). Maurice Stucke, Professor of Law University of Tennessee-Knoxville “Hearing #5” (Federal Trade Commission Hearings, Washington, D.C. November 2018).Tim Wu, Julius Silver Professor of Law Columbia University “Hearing #5” (Federal Trade Commission Hearings, Washington, D.C. November 2018).
K Submissions to the report
Facebook Australia, “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).Google “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).
ICLE “Submission on the final report of the Australian Competition and Consumer Commission’s Digital Platforms Inquiry” (September 2019).
Media, Entertainment and Arts Alliance “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).
Microsoft “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).
Outdoor Media Association “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).
REA Group “Submission to the ACCC Digital Platforms Inquiry Preliminary Report” (February 2019).
L Unpublished papers
Henry Klimowicz “Reinvigorating the Perceived Potential Competition Theory: An Analysis of the Potential Competition Doctrine and FTC v. Steris Corp” (Law School Student Scholarship, Seton Hall University, 2019).
Jean Tirole Competition and the Industrial Challenge for the Digital Age (background paper for the IFS Deaton Review on “Inequalities in the Twenty- First Century”, April 2020).
Tim Wu “The “Protection of the Competitive Process” Standard” (Research Paper no 14- 612, Columbia Law School, 2018).
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/UOtaLawTD/2020/4.html