Opinion

Epic Games, Inc. v. Apple Inc.

Court
District Court, N.D. California
Filed
Dec 31, 2020
Cited by
0 cases
Authority
More cited than 18.8%

The opinion

1

2

3

4 UNITED STATES DISTRICT COURT

5 NORTHERN DISTRICT OF CALIFORNIA

6

7 EPIC GAMES, INC., Case No. 20-cv-05640-YGR (TSH)

8 Plaintiff and Counter-

DISCOVERY ORDER

9 defendant,

Re: Dkt. No. 213

10 v.

11 APPLE INC.,

12 Defendant and

13 Counterclaimant.

14

15 We are here on a joint discovery letter brief concerning Apple’s responses to Epic Games’

16 requests for production (“RFPs”). ECF No. 213. The Court held a hearing on December 30,

17 2020, and now issues this order.

18 A. Non-U.S. Documents

19 The first dispute is over Apple’s general refusal to produce documents concerning its

20 activities outside the United States. Apple has agreed to produce documents that reference its

21 activities both within and outside of the U.S., as well as documents relating to Epic’s own dealings

22 with Apple outside of the U.S. But it will not agree to produce documents that reference only

23 extraterritorial conduct and that do not relate to Epic. Epic says this geographic limitation is

24 unjustified, and it moves to compel documents relating to foreign activities on all 70 RFPs in its

25 first set of RFPs.1

26

1 Epic has provided its first set of RFPs as Exhibit 1 to the joint discovery letter brief. Apple

27

reports in its section of the letter brief that Epic has served a total of 83 RFPs. At the hearing Epic

1 Notwithstanding its assertion that it has alleged global markets, Epic is suing under the

2 federal Sherman Act, the California Cartwright Act, and California Business and Professions Code

3 section 17200. See ECF No. 1 (Complaint). Therefore, wholly extraterritorial conduct not

4 directed at the U.S. cannot be a basis for liability in this case. Having said that, foreign conduct

5 can sometimes be relevant evidence of domestic conduct. The clearest example of this is an

6 international price-fixing conspiracy where you have to see the whole conspiracy to know how

7 broad it is, the role the various executives played, how the conspiracy was enforced and concealed,

8 and so on, before you can really understand what happened in the U.S. See, e.g., In re Aspartame

9 Antitrust Litig., 2008 WL 2275531, *2 (E.D. Pa. May 13, 2008) (citing cases). However, in other

10 cases, documents about purely foreign conduct may not be relevant. Rule 26 limits discovery to

11 what is relevant and proportional, after all, and the Foreign Trade Antitrust Improvements Act

12 generally removes from antitrust liability commercial activities abroad, subject to a few

13 exceptions. See U.S. v. Hui Hsuing, 778 F.3d 738, 751 (9th Cir. 2015).

14 So, the Court cannot endorse a simplistic holding that documents about foreign conduct are

15 always relevant or never relevant because neither proposition is true. Instead, the analysis comes

16 down to having a good theory of relevance. The moving party needs to explain why documents

17 concerning foreign activities are relevant to U.S. claims or defenses, and the Court must conduct a

18 careful analysis to determine if the foreign documents actually would be relevant. See, e.g., In re

19 eBay Seller Antitrust Litig., 2008 WL 3925350, *1-2 (N.D. Cal. Aug. 22, 2008) (“relevance does

20 not necessarily stop at the shores of the United States,” so “at least some of the agreements with

21 the third parties, including those connected to activities overseas, may reflect upon plaintiffs’

22 claims,” but “[t]hat said, to require production of all third party agreements and backup materials

23 at this junction would be premature in light of the significant probability that a number of these

24 contracts and agreements may have nothing whatsoever to do with the issues in this litigation”).

Here, Epic has explained nothing. Epic’s assertions that it alleges worldwide markets and

25

that Apple also refers to its worldwide presence as part of its business justification defense do not

26

27

1 even begin to explain how documents about purely foreign conduct that are responsive to any of

2 these RFPs are relevant.2 The key legal principle that Epic misunderstands is that relevance is

3 measured against “any party’s claim or defense,” Fed. R. Civ. Proc. 26(b)(1). All of the claims

4 and defenses in this case arise under U.S. or California law, not some non-existent worldwide

5 antitrust law. To show relevance, Epic must explain – as the plaintiffs did in In re Aspartame

6 Antitrust Litig. and In re eBay Seller Antitrust Litig. – how the foreign documents it seeks would

7 tend to prove or disprove claims under U.S. or California law, claims that by definition have a

8 limited geographic reach. But here, Epic abjures that task entirely, insisting that because its

9 Complaint alleges global markets, it has no obligation to explain how the documents are relevant

10 within the meaning of Rule 26 to claims or defenses under U.S. domestic law. In Epic’s view, the

11 word “global” has magical power when used in a Complaint, wiping away the requirement of

12 relevance in discovery. The Court disagrees.

13 Consider RFP 59, which seeks “All Documents Concerning Customers’ awareness of,

14 familiarity with, lack of awareness of, and/or lack of familiarity with (a) the fact that Apple does

15 not permit any Software Store on iOS devices other than the iOS App Store; (b) the fact that

16 Apple does not allow Developers to use any method other than Apple’s IAP for accepting

17 payments from Customers for certain types of transactions; or (c) Apple’s fee or commission on

18 the purchase of Apps and Apple’s IAP transactions.” This RFP seems to be getting at a Kodak-

19 style “lock in” argument, suggesting that maybe customers don’t know what they’re getting into

20 when they buy an iPhone and then later it’s too expensive to switch. But why should we care

21 what foreign customers are aware of when they buy an iPhone? When the Court raised this

22 example at the hearing, Epic just repeated that it is alleging worldwide markets, but it did not

23 actually explain how the awareness or lack of awareness that people in foreign countries might

24 have could be relevant to the Sherman Act and California law, which don’t regulate Apple’s

transactions with foreign customers.

25

Or consider RFP 28. It requests: “Documents sufficient to show the number and

26

27

1 percentage of iPhone, iPad or iPod touch Customers, respectively, who own at least one iPhone,

2 iPad or iPod touch and used any of the following in the last 30, 90, 180 or 365 days, respectively:

3 (a) Apple Music; (b) Apple TV+; (c) Apple News; (d) Apple Arcade; (e) Apple Pay; (f) Apple

4 Card; (g) iMessage; (h) FaceTime; (i) Find My; (j) AirDrop; (k) iCloud Photos; (l) iCloud Drive;

5 (m) iTunes; (n) Apple Books; (o) Family Sharing; (p) Apple One; and (p) none of the above.” The

6 Court has a hard time understanding why we need to know how many people in Mongolia tried to

7 find their iPhone in the last month, or what percentage of iPad users in Sri Lanka use Apple pay,

8 or how popular FaceTime is in Brazil. How would such evidence be relevant to claims and

9 defenses under U.S. and California law? Epic doesn’t say. At the hearing Epic did not dispute

10 that RFP 28 asks for these things and did not present argument for why that information is relevant

11 to the U.S. and California claims and defenses in this case. Instead Epic argued that it did not

12 demand document custodians who are in those foreign countries. In other words, Epic argued that

13 it did not go out of its way to seek out documents that relate exclusively to foreign conduct. Well,

14 that’s good, but it still doesn’t answer the Court’s question about relevance. Epic says that if a

15 document is in the custodial collection of one of Apple’s document custodians, Apple should not

16 code it non-responsive merely because it relates to exclusively foreign conduct. However, that

17 appears to be an argument about burden and leaves unanswered the Court’s skepticism about the

18 relevance of such documents to claims and defenses under U.S. domestic law.

19 For a lot of the RFPs at issue, the Court can on its own dream up theories of how foreign

20 conduct might indeed be relevant to claims and defenses under U.S. or California law. But the

21 Court is concerned that the Court is the one dreaming up those theories of relevance. Epic’s

22 argument is that if an antitrust plaintiff says the words “global market” in the Complaint, then the

23 Court should forget that the Sherman Act and California law do not apply to foreign conduct not

24 directed at the U.S. Epic has not advanced any arguments that the particular foreign conduct at

issue in these RFPs actually is relevant to claims and defenses under U.S. law; Epic thinks it

25

doesn’t have to make that showing. In the adversarial system, we normally leave it to the litigants

26

to advocate for themselves rather than helping one side or the other. Here, where Epic has done

27

1 determine that foreign documents responsive to any particular RFP are relevant to claims and

2 defenses under U.S. domestic law would require the Court to write the motion to compel that Epic

3 didn’t write. That doesn’t seem like something the Court ought to do. And it would be grossly

4 unfair to Apple, which didn’t have an opportunity to respond to the arguments Epic didn’t make.

5 To be clear, the Court is not saying that each RFP had to be specifically discussed one by

6 one. It is common for litigants to group RFPs into related subjects and then discuss them in

7 groups. A common form of that argument is that RFPs 1-5 seek information about subjects A and

8 B; documents that are responsive will likely show X, Y or Z; and they are relevant to the

9 plaintiff’s claims for reasons 1, 2 and 3. And then the other side, having seen the moving party’s

10 arguments, can respond. Another popular approach is to use illustrative examples. In that

11 approach, the moving party selects a few RFPs that are representative of a number of issues in the

12 case, and the parties brief those examples. This allows the parties to obtain a ruling that they can

13 then apply to other RFPs without further judicial involvement. The Court’s experience is that a

14 well-constructed five-page discovery letter brief can effectively cover a lot of ground. But in any

15 event, the problem here is not that Epic’s discussion of why foreign documents responsive to any

16 particular RFP are relevant to claims or defenses under U.S. law was insufficiently detailed. The

17 problem is that Epic did not even attempt that showing.

18 Epic’s motion to compel Apple to produce documents concerning non-U.S. activities is

19 denied because Epic has not explained how the foreign documents responsive to these 70 RFPs are

20 relevant to the U.S. and California claims and defenses in this case.

21 B. RFP 3

22 Epic’s RFP 3 seeks: “Documents sufficient to show actual and projected revenue, costs,

23 expenses, and profits, by country, by year, incurred by, earned by and/or attributed to, sales of

24 each of the following, respectively: (a) iPhone; (b) iPad; (c) iPod touch; (d) Apple Watch; and (e)

25 Apple AirPods.”

Epic argues that “Apple has market power in the market for mobile operating systems, and

26

that this market power in turn supports Apple’s market power in aftermarkets for app distribution

27

1 evidence market power. Apple’s revenue from iOS comes primarily from selling devices that run

2 iOS (iPhone, iPad and iPod touch) and accessories that depend on ownership of iOS devices

3 (Apple Watch and Apple AirPods).”

4 The Court agrees, and Apple does not dispute, that this financial information is relevant for

5 devices that access the App Store, since Epic alleges that Apple has market power in the

6 aftermarkets for app distribution and in-app payment processing on iOS. However, the Court is

7 unable to discern the relevance of this information for Apple Watch or AirPods, which are just

8 accessories to those devices. It is true that “the consistent extraction of supracompetitive profits

9 may be an indication of anticompetitive market power,” Bailey v. Allgas, Inc., 284 F.3d 1237,

10 1252 (9th Cir. 2002), but the way Epic describes the alleged markets in the Complaint does not

11 make it sound like Apple Watch or AirPods are in or access either market. The app distribution

12 market refers only to smartphones and tablets. Complaint ¶¶ 35-50. Paragraph 40 of the

13 Complaint, which is part of the description of the app distribution market, alleges that “for mobile

14 device users, there are effectively only two mobile operating systems to choose from: Google’s

15 Android OS or Apple’s iOS. As of July 2020, these two operating systems accounted for nearly

16 100% of the worldwide mobile OSs.” If the referenced “mobile devices” included wearables such

17 as Apple Watch, then paragraph 40 would be false because Garmin’s and Fitbit’s products do not

18 use Android OS or Apple’s iOS. Similarly, paragraph 183 alleges that “nearly 100% of all mobile

19 devices run either Apple’s iOS or Google’s Android OS.” That allegation would also be false if

20 “mobile devices” included wearables such as Apple Watch. Paragraphs 40 and 183 therefore

21 make clear that the “mobile devices” that access the app distribution market are limited to

22 smartphones and tablets. And the in-app payment processing market seems to refer to financial

23 transactions that occur within the app distribution market. See id. ¶ 109 (“There is a relevant

24 market for the processing of payments for the purchase of digital content, including in-game

content, that is consumed within iOS apps, the iOS In-App Payment Processing Market.”). Thus,

25

as pleaded, the mobile devices that access the alleged relevant markets are Apple’s smartphones

26

and tablets, not Apple Watch – an interpretation that Epic confirmed at the hearing. And, of

27

1 The most that can be said is that Apple Watch and AirPods are part of an ecosystem of

2 products that are designed to be used with iPhones and iPads, along with cases, chargers, speakers,

3 and so on. At this point we’re not talking about the relevant markets anymore (app distribution

4 and in-app payment), or even products that allegedly access the relevant markets (smartphones and

5 tablets); we’re talking about something that can be used with something that accesses a relevant

6 market. The sole theory of relevance that Epic cites in the letter brief is that the consistent

7 extraction of supracompetitive profits may be an indication of anticompetitive market power. But

8 if the Court ordered Apple to produce profit information for Apple Watch and AirPods, how

9 would Epic know if those profits were supracompetitive? Epic would have to subpoena Garmin,

10 Fitbit and others in the market for wearables for their profit information, as well as the major

11 players in the market for headphones – a sprawling and unjustified expansion of discovery into

12 two entirely new markets that, at the hearing, Epic said it had no intention of undertaking.

13 Epic’s theory of relevance seems to be that Apple makes a lot of money off of iPhone

14 accessories. That is surely true, but the distance between that fact and evidence of market power

15 in the app distribution and in-app purchase markets is too great for this discovery to be either

16 relevant or proportional. Ask yourself this: What if Apple’s profit margins on Apple Watch or

17 AirPods are similar to the profit margins earned by competitors in the wearables or headphone

18 markets? Then Apple’s profit margins on those products would seem to mean nothing.

19 Alternatively, if Apple’s profit margins on Apple Watch and AirPods are huge compared to its

20 competitors’ profit margins in those markets, then maybe those profit margins do mean something,

21 although we would still have to figure out what. This is one of those times where unless we burn

22 down the entire forest in discovery, we won’t know what meaning to attach to the information the

23 moving party is seeking. At the hearing Epic’s counsel acknowledged that this theory of relevance

24 would technically extend to every single iPhone accessory in existence, including cases, chargers

and speakers – as well as accessories to the accessories, such as wristbands for Apple Watch – but

25

said Epic was not asking for all that. However, the logical reach of this theory of relevance

26

underscores just how sweeping and disproportional it is. Discovery into the profit margins of

27

1 Accordingly, the Court grants Epic’s motion in part and denies it in part and orders Apple

2 to produce documents responsive to RFP 33 for the iPhone, iPad and iPod touch.

3 C. RFP 54

4 Epic’s RFP 5 seeks “Documents sufficient to show actual and projected revenue, costs,

5 expenses, investments (Including research and development) and profits, by year, incurred by,

6 earned by and/or attributed to, Apple’s IAP.”

7 This information seems relevant because it concerns Apple’s profits in one of the relevant

8 markets. Apple argues that “IAP is a functionality of the App Store, and not a separate product

9 that ‘earns’ or is ‘attributed’ any costs or revenue.” Apple also states that it “is working to

10 produce data underlying App Store P&L calculations. This will include data on the expenses and

11 revenues associated with the App Store generally. The documents Epic cites show only that

12 Apple tracks revenues from the App Store.”

13 Apple is losing credibility by continuing to assert that it does not have data in the teeth of

14 documents proving that it does. What’s more remarkable is how this is playing out. It’s not the

15 case that Apple makes an incautious statement to the Court and then the Plaintiffs rummage

16 through Apple’s document production to try to find a document that undermines Apple’s

17 representation. Rather, in both this and the prior filing, Apple denied the existence of information

18 in the very same joint discovery letter brief in which the opposing party cited by Bates number a

19 document proving that Apple does have the requested information. Here, Epic cited and has now

20 provided to the Court APL-APPSTORE_00227526-27, which indicates that Apple tracks the

21 revenue associated specifically with in-app purchases, and is not limited to determining what

22 revenue is associated with the App Store generally. At the hearing Apple stated that it has likely

23 produced hundreds of iterations of that email report. While the Court appreciates the clarification

24 Apple provided at the hearing, in the letter brief Apple should not have said that revenue is not

25

26

3 As to the worldwide reach of RFP 3, the Court’s analysis in section A applies. Epic has not

explained why Apple should have to produce this information for every country in the world. The

27

parties did not discuss whether Apple actually has this information for each country in the world.

1 assigned or attributable to in-app purchases because Apple clearly does track that. Accordingly,

2 || the Court orders Apple to produce this revenue information for in-app purchases for the relevant

3 timeframe.

4 As for everything else requested by RFP 5 (costs, expenses, investments and profits

5 associated with in-app purchases), the Court does not know if that exists or not. Apple says it

6 || likely doesn’t. As to costs, Epic cites page 15 of the Fischer deposition, but in context that

7 testimony does not say that Apple is able to identify credit card fees that are specific to in-app

8 || purchases as opposed to credit card fees more generally associated with the App Store.

9 Accordingly, for these items, the Court orders Apple to produce whatever it has.

10 IT IS SO ORDERED.

11

12 Dated: December 31, 2020

□

13 TAA. |} □□

THOMAS S. HIXSON

14 United States Magistrate Judge

15

16

17

Z 18

19

20

21

22

23

24

25

26

27

28

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.

A word about cookies

We need a few to keep you signed in and the library working. The rest help us see which pages people use and where they get stuck. They stay off unless you say yes.