Epic Games, Inc v Apple Inc

PERRAM J: 1 There are currently two separate cases pending in the Court's docket: Epic Games, Inc v Apple Inc (NSD 1236 of 2020) and Epic Games, Inc v Google LLC (NSD 190 of 2021). Both cases arise out of the removal of Epic Games, Inc's ('Epic') game, Fortnite, from the iOS and Android platforms in August 2020. In both cases, Epic alleges that the manner in which Apple Inc ('Apple') and Google LLC ('Google') (together 'the Respondents') have conducted their respective platforms is, in various ways, a contravention of Pt IV of the Competition and Consumer Act 2010 (Cth) ('CCA'). It is also alleged that the Respondents have engaged in unconscionable conduct within the meaning of s 21 of the Australian Consumer Law, being Sch 2 to the CCA. 2 The question which now arises concerns the relationship between the two cases and the question of when, and in what way, they should be listed for hearing. The Apple case has already been listed for a six week hearing commencing in November 2022. At the joint case management hearing on 24 March 2022, both Epic and Apple indicated that this was no longer viable and that a trial could not realistically take place until the middle of 2023. Accordingly, I will vacate that trial. They also agreed that the current six week estimate was too short and should be increased to eight weeks. 3 No trial date has yet been set for the Google case but Epic and Google both agreed that it would not be ready for a hearing before late 2023 or early 2024. I will proceed on the more pessimistic basis that it will not be ready until early 2024. 4 The cases are potentially interrelated in various ways which although distinct are also, to a degree, overlapping. In both, Epic contends for the existence of two separate markets in which it alleges that each respective Respondent has engaged in conduct in contravention of the CCA Pt IV. In the Apple case, the two markets are the iOS App Distribution Market and the iOS In-App Payment Solutions Market. In the Google case, they are the Android App Distribution Market and the Android In-App Payment Processing Market. 5 To illustrate the problem which is looming on the horizon, it is sufficient to consider Epic's allegations concerning the iOS App Distribution Market. At s90 of its Further Amended Statement of Claim, Epic alleges: 90. At all relevant times, in the circumstances described … above, there is a market or markets for the supply of services for the distribution of iOS apps to iOS Device users (iOS App Distribution Market). Particulars i. The services consist of the provision of services to app developers enabling and/or facilitating app developers to reach, offer and provide iOS Device users with iOS apps and associated updates and/or the provision of services to iOS Device users enabling and/or facilitating iOS Device users to be presented with and/or find, obtain and utilise iOS apps and associated updates. ii. … 6 At the corresponding s90 in Apple's Defence this is said: 90. In answer to paragraph 90 of the [Further Amended Statement of Claim], the Respondents: (a) say that the allegation is vague and embarrassing because: (i) it fails to properly articulate a precise market(s) or a relevant demand said to give rise to the alleged market(s); and (ii) the reference to "app developers" in general terms in particular (i) to the paragraph exacerbates the lack of specificity in the definition of the alleged market(s); (b) under cover of that objection, and on the basis that the alleged contraventions pleaded in Parts VI through VIII of the [Further Amended Statement of Claim] are confined to the period from 6 November 2017, say that: (i) the App Store is a two-sided transaction platform that connects app developers with consumers by: A. creating a platform through which developers can publish their apps and from which a consumer can download the application; B. facilitating observable transactions that simultaneously connect developers with consumers; and C. adopting pricing strategies, service provision strategies, and rules of behaviour to attract these two distinct groups of users and to facilitate productive interactions between them; (ii) a successful interaction - a download, an update, or an in-app purchase - will result in a transaction simultaneously provided to the developer and the user, such that the single product supplied via the App Store is app transactions; (iii) the App Store is one of a number of substitute online platforms available on multiple devices, including tablets, smartphones, laptops, PCs, Macs, and game consoles, that facilitate two-sided transactions; (iv) the product and services supplied through the platforms described in 90(b)(i) through 90(b)(iii) above are dynamic in nature and continually evolving including as a result of the introduction of new and different online platforms; and (c) otherwise deny the allegations in the paragraph. Particulars i. The App Store competes with at least the following distributors of game apps: A. Game app transaction platforms on mobile devices (e.g., Google Play, launched as Android Market in 2008 becoming Google Play in 2012; Samsung Galaxy App Store, launched 2009); B. Game app transaction platforms on PCs/Macs (e.g., Steam, launched 2003; EGS launched 2018; Ubisoft Connect in 2012; Bethesda net 2016); C. game app transaction platforms on consoles (e.g., Sony PlayStation Store, launched 2006 and Microsoft Xbox Live Marketplace, launched 2005; Nintendo eShop, launched 2011; and, D. web-based game app transaction platforms including most recently, streaming game services (e.g., Nvidia GeForce Now; launched early 2020; Google Stadia launched November 2019; Microsoft Xbox Cloud Gaming; Amazon Luna available for web streaming of games to connected devices including iOS devices). ii. Fortnite is available through the EGS, the Nintendo eShop, the Xbox Marketplace, the PlayStation Store, the Samsung Galaxy Store, and GeForce Now, and was previously available on the App Store and Google Play. 7 It will be seen that Apple proposes to prove that the Apple App Store is a two-sided transaction platform that connects app developers with consumers and that, in relation to games, it competes with a number of other such platforms including, importantly, Google Play. 8 The differences between the parties are significant. On Epic's view, the market consists of the supply of services for the distribution of iOS apps to iOS device users, and iOS developers but that supply does not take place in a broader market in which other platforms provide similar services (it does plead a wider market at s91 but this is in the alternative). On Apple's view, the market is broader and consists of two-sided transaction platforms (of which iOS and Android are but examples) which provide services both to developers and to consumers and which compete against each other. It will be apparent that the ability of the closed nature of the Apple App Store substantially to lessen competition in these two differently defined markets is potentially quite different. The question of market definition will therefore be an important issue in the case. 9 It is explicitly part of Apple's response to Epic's allegations that the market it says exists includes, in relation to games, the Google Play Store on Android devices. There will therefore at least be an issue about the status of the Google Play Store on the Android platform in Epic's case against Apple about its App Store on the iOS platform. 10 It is foreseeable that Google will make the same allegation in the Google case in response to Epic's allegations about the Android App Distribution Market. This is not presently known because the pleadings are not yet closed but it seems likely that Google will seek to counter the market allegation which Epic makes against it in a similar way, viz, by alleging that the relevant market is actually a market defined by the activities of substitutable two-sided transaction platforms including, relevantly, the Apple App Store. This means that there is a real risk that both cases involve the same large economic question of how the market for the distribution of apps is to be characterised. The risks that this carries are significant. For example, if in the Apple case, Epic is found to be correct and that the iOS App Distribution Market consists of the provision of distribution services to the developers of apps on the iOS platform, then this necessarily implies that the App Store does not compete in relation to gaming with the Google Play Store. Having made that finding, could it then be concluded in the Google case that the Android App Distribution Market is in fact a market of two-sided transaction platforms providing services to both consumers and developers and that the platforms compete with each other? It may be that this is possible, but it is not difficult to see that both Google and Apple have a distinct interest in what is taking place in the case brought by Epic against the other. 11 This then gives rise to practical and procedural problems. If the cases are heard separately, what right should the parties in the Apple case have to be heard on the same question in the Google case (and vice versa)? I have chosen the example of the iOS App Distribution Market as an illustration only. It is not difficult, however, to see that a case about the economic qualities of the iOS platform and a case about the economic qualities of the Android platform are going to run into each other given the size of both platforms. Particularly where the cases have their origin in the same event, namely, the exclusion of the game Fortnite from both platforms, it does not require much imagination to see the two cases having much in common, at least from an economist's perspective. This is not to gainsay their significant differences. However, those differences cannot erase their similarities. 12 The position of Epic was straightforward: the Court should proceed to list the Apple case for trial in June 2023 and, if minded to list the Google matter at this stage, should do so in late 2023 or early 2024. It should not consolidate the two cases but could, if it wished, hear them sequentially. 13 The position of Google and Apple was different. They foresaw procedural difficulties about the relationship between the two cases. It may be that the difficulties would require them to be heard together, it may be that they would not, or it may be that some intermediate position would turn out to be necessary. The point was that it was impossible at this stage to know precisely what those problems were going to be or how they would be resolved. The Court should, therefore, manage the two cases in such a fashion so as to forestall the premature cutting off of any option concerning how they would eventually be conducted. In particular, the Court should not accede to the proposition that one case should be set down in advance of the other, at least at this stage. 14 At present, the Google case is but in an early stage of preparation. There are complex issues of discovery to be worked through. The Apple parties are working through these at the moment but they are much further along the path. Neither side had a firm view on how long the Google case might take. On the face of it, the Apple and Google cases are similar so eight weeks might be regarded as sensible. 15 On the other hand, the conduct complained of in the Google case is more complex than that in Apple case. There is no question that the iOS platform is a self-contained ecosystem (to select language that is endeavouring to be neutral). Epic alleges that the Android platform is, in substance, also a self-contained ecosystem but, as is well-known, the extent of the control of that platform is less total than it is in the iOS platform. But Epic wishes to say that in its practical effect the apparent openness of the Android platform turns out to be more illusory than real. Google has understandably indicated that this is a proposition it proposes most probably to deny. The trial of the Google case is likely therefore to be attended with additional baggage. This is the baggage of resolving what the practical restraints imposed by the Android platform are and what the effect of those practical restraints might be. 16 Mr Moore SC, for Google, submitted that one might perhaps be a little sceptical that the Google case presented the prospect of an eight week trial. In contradistinction to a price fixing case, for example, there was unlikely to be conflicting testimony about the existence of contested arrangements or understandings as there might be if the suit concerned the rigging of the corn syrup market at a trade conference in the Bahamas. Mr Young QC, for Epic, pointed out, however, that there were still allegations about Google's purposes in entering contracts, arrangements or understandings and that it was likely that testimonial evidence would become involved. 17 I accept both of these submissions. To them I would add this consideration: unlike most of the competition cases in this Court, a large quantity of data about actual consumer behaviour will be available in this litigation. For example, in the Apple case, Apple has been granted discovery from Epic about the behaviour of the players of Fortnite at the time the game was excluded from the iOS platform. The point of this discovery is to lay the groundwork for a contention that players of Fortnite, following its removal from the Android and iOS platforms, are likely simply to have switched to playing the game on another platform such as the PlayStation or Xbox. 18 The data about what these users actually did exists and will be examined by Apple (there are alleged to be around 400 million users). This opens up a new vista in competition cases for the economic witnesses. 19 For management purposes, I propose to assume that the Google case will take eight weeks. 20 Between them, the two cases therefore require 16 weeks of hearing time regardless of the manner in which they are heard. To this I think it would be wise to add two weeks by way of a buffer. If the cases are heard sequentially it will permit a break between the cases. If not, it can serve as a reserve in the event that the Google case turns out to be more complex than is presently anticipated. If not used for either of those purposes, it can be used to provide for a break after the close of evidence to permit the preparation of submissions. 21 Mr Young urged me to set down the Apple trial for hearing in the middle of 2023 when, I accept, it will be ready. To embrace that approach, however, would appear to involve determining at this stage the outcome of the procedural problems to which I have already adverted. The possibilities for how the Apple and Google cases will be heard are as follows: (a) The two cases will be consolidated into one case and heard together with evidence in one being evidence in the other; (b) The two cases will be heard together but they will maintain their separate identity; (c) The two cases will be heard sequentially with the parties in each case not being permitted to participate in the other; (d) The two cases will be heard sequentially with the parties in each case being be permitted to participate in the other to the extent of any common issue; (e) The two cases will be heard sequentially but by some mechanism any common issues will be tried together; or (f) One of the cases will reallocated to another judge and heard separately with, or without, a right of participation in the other case. 22 Each of (a)-(e) has its complexities. Option (c) will potentially put the Court in the untenable position of trying the same issue quite possibly on different evidence. It is not realistic to expect that the opinions of one set of economic experts on the market issues in the first case can be put out of mind when embarking upon the hearing of the same issue on different evidence in the second case. Consequently, this may well give rise to an apprehension of bias even before it is embarked upon as it was in GetSwift Limited v Webb [2021] FCAFC 26; 283 FCR 328 at [55]-[57] ('GetSwift'). 23 In relation to (c)-(e) it is apparent that no judgment could be delivered in the first case before judgment in the second case. This is necessary in order to avoid the risk of a recusal application on the grounds of apprehended bias arising from pre-judgment (for example, on the issue of the nature of the market for the distribution of apps). Further, given the certainty of an appeal, it would be much better if both cases went to the Full Court at the same time. In relation to (d) and (e), it is quite unclear at this stage what participation rights would be involved. In relation to (e), it is unclear how such an issue might be carved out cleanly enough to permit this to occur safely. 24 This is clear in relation to (a)-(e). There are real difficulties with going down the path of (f) at this stage. Whilst (f) will resolve the GetSwift problem immediately, it will leave unresolved the difficult issue of how the participation rights in the other case will be handled. Indeed, whereas that question will presently be resolved by one judge, adopting the GetSwift approach at this stage would create the spectacle of two judges deciding the participation rights, potentially inconsistently. Further, (f) may never arise depending on how (a)-(e) fall out. 25 All in all, it is best not to cross any of these one-way bridges at this time. 26 In relation to all of these matters, at this stage nothing is clear. For that reason, I accept the submission of Mr Free SC for Apple and Mr Moore for Google that the best approach is to manage the two cases in tandem towards a trial and, once more is known, to square up to the many procedural problems at that time. In other words, let us wait and see. 27 As I have said, I accept Mr Young's submission that Epic's case against Apple will be ready for trial by the middle of 2023 and that Apple would ordinarily be entitled, all other things being equal, to have it tried promptly. However, it seems to me that to do so at this stage would involve choosing from the basket of imponderables set out above, at a time when I do not have sufficient information to make any of the decisions which would be necessary in a just and fair way. Consequently, I do not think I should determine by a side wind that the matters are to be heard sequentially by now fixing the Apple case for trial in mid-2023. In any event, the delay involved in not doing so is only about eight months. Whilst this is not insignificant, when weighed against the other procedural perils involved, it seems to be one which will need to be tolerated. 28 The options therefore are either to list the two cases for hearing at a fixed date in some kind of way, yet to be determined. If this course is taken, the timing of the Google matter indicates that this would need to be in 2024. Alternatively, it might be best to do nothing at all. Under this scenario, no trial dates would be allocated at this stage. 29 I consider this second option inferior for a number of reasons. First, cases progress more rapidly to trial with a trial date set. Secondly, continuing uncertainty about when these cases will be heard imposes a significant strain on the management of the Court's docket. For example, in practical terms, if the manner in which the two cases are to be tried in relation to each other is resolved, say, in June 2023 then it is very likely that no hearing dates at the beginning of 2024 will remain available. Indeed, given the current flow of litigation in this Court and the number of judges in a position to try larger competition cases, it is quite possible that hearing dates might not be available until later in 2024 or possibly even 2025. This is undesirable. 30 Finally, it seems likely that in the preceding month before the trial there will be a significant increase in interlocutory applications as there often is in this kind of litigation. It would be unfortunate for that to happen in January 2024 when most of the profession are on vacation. Accordingly, I will list the two matters for trial on Monday, 4 March 2024 for a period of 18 weeks. They will be listed to start at the same time but that should not be understood as indicating any preference for how the 18 weeks will eventually be utilised. 31 The only orders I will make are orders vacating the hearing dates for the Apple trial and an order fixing both the Apple and Google cases for trial on 4 March 2024 with an estimate of 18 weeks. A date for a further case management hearing in both cases will be determined in consultation with my Associates. It is clearly to be understood that this order in no way determines questions such as whether they will be heard sequentially or together or in some other fashion. At this stage, all that is clear is that by whatever means, the two cases will be tried in that period. I will make arrangements to ensure that a judgment in both matters can be given towards the end of 2024. I certify that the preceding thirty-one (31) numbered paragraphs are a true copy of the Reasons for Judgment of the Honourable Justice Perram.