`
`Nos. 21-16506 & 21-16695
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IN THE UNITED STATES COURT OF APPEALS
`FOR THE NINTH CIRCUIT
`____________________
`EPIC GAMES, INC.,
`Plaintiff, Counter-Defendant–Appellant, Cross-Appellee
`v.
`APPLE, INC.,
`Defendant, Counterclaimant–Appellee, Cross-Appellant
`____________________
`On Appeal from the United States District Court
`for the Northern District of California
`No. 4:20-cv-05640-YGR-TSH
`The Honorable Yvonne Gonzalez Rogers
`
` ____________________
`BRIEF OF THE AMERICAN ANTITRUST INSTITUTE AS AMICUS
`CURIAE IN SUPPORT OF PLAINTIFF, COUNTER-DEFENDANT–
`APPELLANT
`____________________
`
`
`LAURA M. ALEXANDER
`
`
` Counsel of Record
`
`
`RANDY M. STUTZ
`AMERICAN ANTITRUST INSTITUTE
`1025 Connecticut Avenue, NW
`Suite 1000
`
`Washington, DC 20036
`
`
`(202) 276-4050
`
`lalexander@antitrustinstitute.org
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`January 27, 2022
`
`Counsel for Amicus Curiae
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 2 of 39
`
`
`
`CORPORATE DISCLOSURE STATEMENT
`
`Pursuant to Appellate Rule 26.1(a), the American Antitrust Institute states
`
`that it is a nonprofit, non-stock corporation. It has no parent corporations, and no
`
`publicly traded corporations have an ownership interest in it.
`
`
`
`
`
`
`
`i
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 3 of 39
`
`
`
`
`CORPORATE DISCLOSURE STATEMENT ..................................................... i
`
`TABLE OF CONTENTS
`
`TABLE OF AUTHORITIES ................................................................................ iii
`
`INTEREST OF AMICUS CURIAE ...................................................................... 1
`
`INTRODUCTION AND SUMMARY OF ARGUMENT ................................... 2
`
`ARGUMENT ........................................................................................................... 5
`
`I. COURTS MUST DEFINE MARKETS CONSISTENT
`WITH FIRST PRINCIPLES OF ANTITRUST LAW ............................. 5
`
`A. The District Court’s Opinion Cannot be Squared with
`First Principles of Antitrust Law Reflected in Kodak .......................... 5
`
`B. The District Court Did Not Properly Assess Apple’s Conduct
`as It Relates to the Source of Apple’s Market Power ........................ 10
`
`C. A Common Error Underlies the District Court’s Failure to
`Recognize Distinct Markets in at Least Two Instances ..................... 14
`
`D. AmEx Applies Only to the Narrow Economic Circumstances
`that Underly Its Departure from Well-established Market
`Definition Principles .............................................................................. 21
`
`II. UPHOLDING THE DECISION BELOW WOULD PROVIDE
`A ROADMAP FOR PLATFORMS TO EVADE FEDERAL
`ANTITRUST LAWS WITH ENORMOUS IMPLICATIONS
`FOR CONSUMERS, DEVELOPERS, AND MARKETS ...................... 29
`
`CONCLUSION ..................................................................................................... 31
`
`
`
`ii
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 4 of 39
`
`
`
`
`
`CASES
`
`
`
`TABLE OF AUTHORITIES
`
`
`American Needle, Inc. v. National Football League,
`560 U.S. 183 (2010) .............................................................................................. 3
`
`Apple, Inc. v. Pepper,
`139 S.Ct. 1514 (2019) ......................................................................................... 29
`
`Digidyne Corp. v. Data Gen. Corp.,
`734 F.2d 1336 (9th Cir. 1984) ............................................................................. 19
`
`Eastman Kodak Co. v. Image Tech. Servs.,
`540 U.S. 451 (1992) ..................................................................................... passim
`
`Eastman Kodak Co. v. Image Tech. Servs.,
`903 F.2d 612 (9th Cir. 1990) ............................................................................... 19
`
`Image Tech. Servs. v. Eastman Kodak Co.,
`125 F.3d 1195 (9th Cir. 1997) ............................................................................. 11
`
`Jefferson Parish Hosp. Dist. No. 2 v. Hyde,
`466 U.S. 2 (1984) ...................................................................................... 2, 17, 19
`
`Leegin Creative Leather Prods., Inc. v. PSKS, Inc.,
`551 U.S. 877 (2007) ............................................................................................ 23
`
`Nat’l Soc’y of Prof’l Eng’rs v. United States,
`435 U.S. 679 (1978) .............................................................................................. 2
`
`Newcal Industries, Inc. v. IKON Office Solutions,
`513 F.3d 1038 (9th Cir. 2008) ............................................................................... 8
`
`Ohio v. American Express,
`138 S. Ct. 2274 (2018) ................................................................................. passim
`
`Rick-Mik Enters., Inc. v. Equilon Enters. LLC,
`532 F.3d 963 (9th Cir. 2008) ......................................................................... 17, 18
`
`
`
`iii
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 5 of 39
`
`
`
`Times-Picayune Publishing Co. v. United States,
`345 U.S. 594 (1953) ...................................................................................... 11, 25
`
`United States v. Microsoft Corp.,
`253 F.3d 34 (D.C. Cir. 2001) (en banc) .............................................................. 19
`
`United States v. Sabre Corp.,
`452 F. Supp. 3d 97 (D. Del. 2020) ...................................................................... 24
`
`OTHER AUTHORITIES
`
`Brief for Petitioner Apple, Apple, Inc. v. Pepper, No. 17-204,
`139 S. Ct. 1514 (filed Aug. 10, 2018) ..................................................... 26, 27, 28
`
`Daniel J. Solove, The Myth of the Privacy Paradox,
`89 GEO. WASH. L. REV. 1 (Jan. 2021) ................................................................. 30
`
`Dennis W. Carlton, The Anticompetitive Effects of Vertical
`Most-Favored-Nation Restraints and the Error of Amex,
`2019 Columbia Bus. L. Rev. 93 (2019) ........................................................ 22, 24
`
`Donald Turner, Antitrust Policy and the Cellophane Case,
`70 Harv. L. Rev. 281 (1956) ............................................................................... 16
`
`Fed. Trade Comm’n & U.S. Dep’t of Just.,
`Antitrust Div., Horizontal Merger Guidelines (2010) ................................... 16, 28
`
`Federal Trade Comm’n & U.S. Dep’t of Justice,
`Commentary on the Horizontal Merger Guidelines (2006) ................................ 28
`
`Herbert Hovenkamp, Platforms and the Rule of Reason:
`The American Express Case, 2019 COLUMBIA BUS. L. REV. 35 (2019) ............. 22
`
`Jack Kirkwood, Antitrust and Two-Sided Platforms:
`The Failure of American Express, 41 CARDOZO L. REV. 1805 (2020) ............... 22
`
`Jeffrey Harrison, Ohio v. American Express: Misunderstanding
`Two-Sided Platforms; the Charge Card “Market;” and the Need for
`Procompetitive Justifications, 70 MERCER L. REV. 437 (2019) .......................... 22
`
`
`
`iv
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 6 of 39
`
`
`
`Lapo Filistrucchi et al., Market Definition in Two-Sided Markets:
`Theory and Practice, 10 J. Comp. L. & Econ. 293 (2014) ................................. 28
`
`Michael Katz & Jonathan Sallet, Multisided Platforms
`and Antitrust Enforcement, 127 YALE L.J. 2142 (2018) ......................... 22, 24, 25
`
`Michael L. Katz & A. Douglas Melamed, Competition Law
`as Common Law: American Express and the Evolution of Antitrust,
`168 U. PENN. L. REV. 2061 (2000) ................................................................ 22, 23
`
`Steven C. Salop et al., Rebuilding Platform Antitrust: Moving on from
`Ohio v. American Express (forthcoming), .................................................... 22, 24
`
`Steven C. Salop, The First Principles Approach to Antitrust, Kodak,
`and Antitrust at the Millennium, 68 ANTITRUST L.J. 187 (2000) .......... 2, 3, 13, 16
`
`TREATISES
`
`Carl Shapiro & Hal R. Varian, Information Rules: A Strategic Guide to the
`Network Economy (1999) ................................................................................... 29
`
`Phillip Areeda & Herbert Hovenkamp,
`Antitrust Law (4th and 5th eds. 2013-2020) ..................................... 14, 17, 19, 20
`
`
`
`v
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 7 of 39
`
`
`
`INTEREST OF AMICUS CURIAE1
`The American Antitrust Institute (“AAI”) is an independent nonprofit organ-
`
`ization devoted to promoting competition that protects consumers, businesses, and
`
`society. It serves the public through research, education, and advocacy on the ben-
`
`efits of competition and the use of antitrust enforcement as a vital component of
`
`national and international competition policy. AAI enjoys the input of an Advisory
`
`Board that consists of over 130 prominent antitrust lawyers, law professors, econo-
`
`mists, and business leaders. See http://www.antitrustinstitute.org.2
`
`
`
`
`
`
`1 The parties have lodged blanket consents with the clerk. No counsel for a party
`has authored this brief in whole or in part, and no party, party’s counsel, or any
`other person—other than amicus curiae or its counsel—has contributed money that
`was intended to fund preparing or submitting this brief.
`2 Individual views of members of AAI’s Board of Directors or Advisory Board
`may differ from AAI’s positions.
`
`
`
`1
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 8 of 39
`
`
`
`INTRODUCTION AND SUMMARY OF ARGUMENT
`
`Despite radical technological and business innovation over the last century,
`
`the Sherman Act remains as relevant to markets today as ever. Moreover, while
`
`the Act’s application has evolved with these developments, the core legal and eco-
`
`nomic principles underlying the Sherman Act have remained remarkably constant.
`
`For good reason: faithful adherence to these first principles of antitrust law means
`
`“courts can maintain logic and consistency while avoiding analytic traps and fac-
`
`tual errors.” Steven C. Salop, The First Principles Approach to Antitrust, Kodak,
`
`and Antitrust at the Millennium, 68 Antitrust L.J. 187, 3 (2000) (hereinafter
`
`“Salop”).
`
`Central among them are the legislative judgment that consumers and society
`
`are best served by competitive markets and that the ultimate test under any antitrust
`
`analysis should be the effect of the defendant’s conduct on competition. Nat’l
`
`Soc’y of Prof’l Eng’rs v. United States, 435 U.S. 679, 695 (1978); see also Jeffer-
`
`son Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 12 (1984).
`
`Antitrust law has also steadfastly emphasized the need to confront markets
`
`as they are and to eschew formalism in favor of market realities. Eastman Kodak
`
`Co. v. Image Tech. Servs., 540 U.S. 451, 462 (1992) (“Legal presumptions that rest
`
`on formalistic distinctions rather than actual market realities are generally disfa-
`
`vored in antitrust law.”); see also American Needle, Inc. v. Nat’l Football League,
`
`
`
`2
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 9 of 39
`
`
`
`560 U.S. 183, 191 (2010) (modern antitrust law “eschew[s]…formalistic distinc-
`
`tions…in favor of a functional consideration of how the parties involved in the al-
`
`leged anticompetitive conduct actually operate”). Where the anticompetitive
`
`impact of conduct is readily apparent, tests and proxies must yield to the directly
`
`observable market realities. Salop at 4 (“Market definition and market power
`
`should be evaluated in the context of the alleged anticompetitive conduct and ef-
`
`fect, not as a flawed filter carried out in a vacuum divorced from these factors.”).
`
`This case provides a prime example of the pitfalls that come with applying
`
`multi-pronged tests and proxies without tying them back to first principles or tak-
`
`ing account of “the context of the alleged anticompetitive conduct and effect.”
`
`Salop at 191. The district court found direct anticompetitive effects from Apple’s
`
`conduct yet held that Apple lacks market power. Its legal conclusion cannot fol-
`
`low from its finding of fact. The analysis is economically incoherent.
`
` The district court failed to appreciate the significance of its finding of direct
`
`anticompetitive effects. Fundamentally, courts must define markets “in reference
`
`to the economic analysis of the alleged anticompetitive conduct and its market ef-
`
`fect.” Id. Here, the court failed to do that in at least three ways:
`
`First, despite factual findings that directly evidence iPhone purchasers are
`
`locked-in by both information costs that prevent accurate lifecycle pricing and
`
`switching costs that weigh against switching from iOS to Android to access
`
`
`
`3
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 10 of 39
`
`
`
`alternative app-distribution services and in-app payment processors, the district
`
`court held there could be no relevant aftermarkets in Apple app-distribution and in-
`
`app payment processing. Had the district court properly understood the economic
`
`and legal significance of its factual findings in the context of Apple’s alleged anti-
`
`competitive conduct and its effects, it could not have arrived at this result.
`
`Second, in defining markets, the district court at least twice fell into the ana-
`
`lytical trap of assuming that current, competitively-constrained market conditions
`
`reflect the contours of consumer demand. In at least two instances, it found a sin-
`
`gle product where, had it properly focused on consumer demand in the context of
`
`Epic’s claims, it would have found two. As a result, it failed to credit plaintiffs’
`
`foremarket in mobile operating systems and mistook the results of Apple’s leverag-
`
`ing as evidence that no leveraging was possible.
`
`Third, the district court applied the Amex decision without attention to the
`
`economic justifications for that decision’s limited application and narrow scope.
`
`As a result, the district court treated one-sided app distribution and in-app payment
`
`markets as transaction services markets, despite the fact that, by Apple’s own
`
`
`
`4
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 11 of 39
`
`
`
`admission, it does not sell app-distribution services and in-app payment services to
`
`consumers and these services are not jointly consumed.
`
`ARGUMENT
`
`I.
`
`COURTS MUST DEFINE MARKETS CONSISTENT WITH FIRST
`PRINCIPLES OF ANTITRUST LAW
`
`Legal tests and frameworks developed in the context of tangible goods can
`
`be difficult to apply in new and evolving digital markets. But such tests are only a
`
`proxy for the ultimate goal of any antitrust inquiry: to determine the impact of
`
`conduct on competition. Courts can correctly apply old tests to new contexts only
`
`by adhering to the first principles underlying this ultimate goal.
`
`A.
`
`The District Court’s Opinion Cannot be Squared with First Prin-
`ciples of Antitrust Law Reflected in Kodak
`
`The district court correctly identified the issue at the heart of Epic’s after-
`
`market monopolization argument: “whether competition in the initial market [for
`
`operating systems] suffices to discipline anticompetitive practices in the aftermar-
`
`kets” for iOS in-app payments and iOS app distribution. Slip op. at 130; see Ko-
`
`dak, 504 U.S. at 469 n.15 (“Whether considered in the conceptual category of
`
`‘market definition’ or ‘market power,’ the ultimate inquiry is the same—whether
`
`competition in the equipment market will significantly restrain power in the service
`
`and parts markets.”). And, the court rightly recognized that “[i]ssues of lock-in or
`
`switching costs, and notice or consumer knowledge fall under th[is] analysis.” Id.
`
`
`
`5
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 12 of 39
`
`
`
`But the district court abandoned first principles by failing to connect its assessment
`
`of switching and information costs to the ultimate competition issue Epic has
`
`raised.
`
`In Kodak, the plaintiffs alleged that Kodak had market power in replacement
`
`parts for Kodak copiers and used that market power to unlawfully tie Kodak parts
`
`to Kodak repair services. Id. at 459. Kodak’s near-100% market share in replace-
`
`ment parts for Kodak copiers was undisputed; the question was whether Kodak ac-
`
`tually possessed market power over parts purchases. Kodak argued, as Apple did
`
`here, that it had no meaningful market power in the Kodak parts market, because if
`
`it tried to charge supracompetitive prices or impose anticompetitive terms on parts
`
`purchases, customers would either switch to a different brand of copier or would
`
`avoid buying Kodak copiers in the first place. On summary judgment, the Su-
`
`preme Court disagreed, explaining that the theory was “perhaps intuitively appeal-
`
`ing [but] may not accurately explain the behavior of the primary and derivative
`
`markets for complex durable goods.” Id. at 473.
`
`“The extent to which one market prevents exploitation of another market,”
`
`the Court reasoned, “depends on the extent to which consumers will change their
`
`consumption of one product in response to a price change in another, i.e., the
`
`cross-elasticity of demand.” Id. at 469 (internal quotations omitted). Where con-
`
`sumers in the aftermarket can readily change their consumption in the foremarket
`
`
`
`6
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 13 of 39
`
`
`
`in response to exploitation in the aftermarket, there is no market power. Where
`
`consumers subject to exploitation in the aftermarket cannot readily avoid or change
`
`their consumption in the foremarket in response, they are “locked-in.”
`
`The Court recognized that this lock-in is a source of the market power in the
`
`aftermarket, even if the lock-in is not complete. Id. at 471 (“The fact that the
`
`[fore]market imposes a restraint on prices in the after-markets by no means dis-
`
`proves the existence of power in those markets.”). Consequently, a finding of
`
`lock-in and market power in this context necessarily suggests that the aftermarket
`
`is a relevant antitrust market. Id. at 469 n.15 (“Whether considered in the concep-
`
`tual category of ‘market definition’ or ‘market power,’ the ultimate inquiry is the
`
`same—whether competition in the equipment market will significantly restrain
`
`power in the service and parts markets.”) (emphasis added).
`
`The Supreme Court identified two causal factors that can generate lock-in:
`
`information costs and switching costs. Id. at 473. Information costs prevent
`
`“lifecycle pricing.” Id. High information costs matter because consumers who
`
`cannot readily assess the full lifecycle costs when they purchase a product cannot
`
`account for those costs when making their initial purchasing decision. Id. Switch-
`
`ing costs prevent switching to a different product after the initial purchasing deci-
`
`sion. High switching costs matter because consumers who cannot easily switch
`
`
`
`7
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 14 of 39
`
`
`
`products in the foremarket cannot escape the exercise of market power in the after-
`
`market.
`
`Information costs lead consumers into a vulnerable position, and switching
`
`costs keeps them there. Together, they generate a relevant antitrust market: an af-
`
`termarket where consumers are locked-in. Absent competition in the aftermarket,
`
`market power can be exercised in the aftermarket with anticompetitive effect.
`
`Here, the district court found that both switching costs and information costs
`
`were present,3 yet found no aftermarket for iOS app distribution or in-app payment.
`
`Again, the legal conclusion does not coherently follow from the fact-finding. To
`
`find such switching and information costs without finding an aftermarket is funda-
`
`mentally incompatible with the market definition principles underlying Kodak and
`
`Newcal Industries, Inc. v. IKON Office Solutions, 513 F.3d 1038 (9th Cir. 2008).
`
`The presence of these information and switching costs necessarily suggests that
`
`competition in the foremarket will not “significantly restrain” market power in the
`
`aftermarket. Kodak, 504 U.S. at 469 n.15.
`
`
`3 The district court found switching costs when it held: “The Court can agree that it
`takes time to find and reinstall apps or find substitute apps; to learn a new operat-
`ing system; and to reconfigure app settings. It is further apparent that one may
`need to repurchase phone accessories.” Slip op. at 50. It found information costs
`when it held: “Apple’s anti-steering provisions hid critical information from con-
`sumers,” Slip op. at 2, and that “Apple used other provisions to hide information
`on those commission rates from the consumers…. Without information, consum-
`ers cannot have a full understanding of the costs.” Slip op. at 50-51.
`
`
`
`8
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 15 of 39
`
`
`
`Moreover, the district court is wrong to assert that the popularity of Apple’s
`
`products reduces the magnitude of switching costs. Slip op. at 51 (“evidence
`
`strongly suggests that low switching between operating systems stems from overall
`
`satisfaction with existing devices, rather any [sic] ‘lock-in’”). To be sure, demand
`
`for Apple’s devices and operating system has significant implications for competi-
`
`tion in the markets for devices and operating systems. But the only relevant ques-
`
`tion here is whether that foremarket competition disciplines the aftermarkets. If it
`
`does not, the foremarket competition says nothing about market power in the after-
`
`markets.
`
`Similarly, the district court is wrong to assert that consumer knowledge that
`
`the iPhone is a walled garden reduces the magnitude of information costs. Slip op.
`
`at 131 (“In short, there is no evidence in the record demonstrating that consumers
`
`are unaware that the App Store is the sole means of digital distribution on the iOS
`
`platform.”). The only relevant question is whether customers have a way to assess
`
`the costs of Apple’s walled-garden approach at the time they purchase an iPhone.
`
`It is consumers’ inability to assess Apple’s aftermarket pricing constraints at the
`
`time of purchase that generates the information costs and prevents foremarket com-
`
`petition from disciplining aftermarket conduct.
`
`The district court held that streaming services and cross-platform gaming
`
`have also reduced switching costs, Slip op. at 131, but again the court cannot
`
`
`
`9
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 16 of 39
`
`
`
`connect its dots. This holding is fundamentally incompatible with its holding that
`
`these services are not included in the mobile gaming market. Slip op. at 85. When
`
`the district court found that streaming services and cross-platform gaming are not
`
`part of the relevant market in this case, it necessarily found that switching between
`
`mobile gaming and these products is not sufficient to constrain prices in mobile
`
`gaming. To simultaneously hold that these products are outside the relevant mar-
`
`ket yet meaningfully reduce switching costs is pure nonsense. The holdings cannot
`
`coexist as a matter of law and economics.
`
`B.
`
`The District Court Did Not Properly Assess Apple’s Conduct as It
`Relates to the Source of Apple’s Market Power
`
`Where high information and switching costs lock consumers into an after-
`
`market, the lock-in can be illegally exploited in the aftermarket in at least two key
`
`ways that Epic raises: (1) leveraging and (2) raising entry barriers. First, where
`
`lock-in creates market power in an aftermarket, a firm can use its power in the af-
`
`termarket as a means to secure market power in a second aftermarket. That is lev-
`
`eraging. Epic alleges that Apple did so by exploiting customers locked into an
`
`app-distribution aftermarket to force them to use Apple’s IAP services.
`
`Second, a company can monopolize by taking steps to strengthen or deepen
`
`the existing lock-in of its customers in any or all of its aftermarkets. Anticompeti-
`
`tive acts that increase lock-in raise entry barriers. Epic alleges that Apple
`
`
`
`10
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 17 of 39
`
`
`
`increased lock-in of aftermarket App Store and IAP customers by anticompeti-
`
`tively increasing or maintaining information and switching costs.
`
`These two methods of exploiting lock-in can be analyzed differently depend-
`
`ing on the source of market power in the aftermarket. In a tying or monopoly lever-
`
`aging case like Kodak, the source of market power in the market being leveraged
`
`(e.g., the tying market) is irrelevant. The anticompetitive conduct is the leveraging
`
`of that monopoly in one (after)market into monopoly power in another (after)mar-
`
`ket (e.g., leveraging parts to get power in service). Monopoly leveraging is illegal
`
`and anticompetitive even if the monopoly power being leveraged is completely
`
`lawful.
`
`Accordingly, the fact that the original power comes from producing a supe-
`
`rior product that customers love, Slip op. at 50-51 (customer satisfaction), or from
`
`some other legal conduct, Slip op. at 48 (“not necessarily nefarious”), does not ex-
`
`cuse leveraging that market power into another market where competition would
`
`otherwise exist. Kodak, 504 U.S. at 479 n.29 (“The Court has held many times
`
`that power gained through some natural and legal advantage such as a patent, cop-
`
`yright, or business acumen can give rise to liability if ‘a seller exploits his domi-
`
`nant position in one market to expand his empire into the next.’”) (quoting Times-
`
`Picayune Publishing Co. v. United States, 345 U.S. 594, 611 (1953)); Image Tech.
`
`Servs. v. Eastman Kodak Co., 125 F.3d 1195, 1216 (9th Cir. 1997) (“Section 2 of
`
`
`
`11
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 18 of 39
`
`
`
`the Sherman Act condemns exclusionary conduct that extends natural monopolies
`
`into separate markets.”).
`
`On the other hand, when it comes to increasing lock-in to raise entry barri-
`
`ers, the source of the information and switching costs that create lock-in does mat-
`
`ter. When lock-in is increased, the allegation in a monopolization case is that the
`
`defendant took actions with the purpose and design to acquire, enhance or maintain
`
`monopoly power. Merely possessing market or monopoly power is not illegal, and
`
`firms with monopoly power are permitted and encouraged to compete like any oth-
`
`ers, including in an aftermarket. Only the acquisition or maintenance of monopoly
`
`power (or attempts) by means other than competition on the merits is illegal. Thus,
`
`where the source of market power is lock-in created by switching and information
`
`costs, the question of whether those costs have been raised or maintained by un-
`
`lawful actions of the defendant, or instead by competition on the merits, is at the
`
`heart of the inquiry.
`
`In undertaking that inquiry, though, courts must bear in mind that market
`
`power arising from sources other than defendants allegedly anticompetitive actions
`
`is irrelevant. Monopolistic conduct that raises entry barriers almost always arises
`
`in a context where the defendant has some degree of legitimate and legal market
`
`power, be it from a first-mover advantage, a high-quality product, or for structural
`
`reasons. The source of that market power is generally not relevant, because
`
`
`
`12
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 19 of 39
`
`
`
`monopolization is concerned not with the background market power of the defend-
`
`ant, but with the defendant’s allegedly illegal actions to increase or maintain that
`
`market power. Salop, at 4 (“Market definition and market power should be evalu-
`
`ated in the context of the alleged anticompetitive conduct and effect….”).
`
`So too here. Apple has a certain degree of lawfully earned market power
`
`from the fact that the iPhone is a unique product that consumers like and is pro-
`
`tected from competition by intellectual property rights. As discussed above, that
`
`market power relates to the foremarket. Likewise, aftermarkets for the App Store
`
`and in-app payment systems exist because of switching and information costs that
`
`prevent foremarket competition from effectively cabining the exercise of market
`
`power in these aftermarkets.
`
`All of this is immaterial to Epic’s monopolization claims. It is not the exist-
`
`ence of aftermarkets where Apple has market power that Epic alleges constitutes
`
`illegal monopolization. It is that Apple took steps to increase or maintain the in-
`
`formation and switching costs, and thus its power in these aftermarkets, that Epic
`
`alleges is illegal.
`
`Properly understood against the context of Epic’s allegations, it is clear that
`
`the district court misunderstood the significance of its finding that people are reluc-
`
`tant to switch away from the iPhone in response to Apple’s conduct in the app dis-
`
`tribution and in-app payment markets because of “overall satisfaction with existing
`
`
`
`13
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 20 of 39
`
`
`
`devices.” Slip op. at 51. It is because consumers are overall satisfied with Apple’s
`
`iPhone and iOS that Apple’s monopolization scheme was effective. For consum-
`
`ers who are “overall happy with their devices,” switching is a high cost, indeed.
`
`At its core, Epic’s monopolization claim is that Apple engaged in conduct to
`
`maintain or increase its market power in app distribution and in-app payment mar-
`
`kets by increasing or maintaining the high switching and information costs that
`
`create lock-in. How? By imposing contractual restrictions on developers that (a)
`
`prevented them from reducing information costs by the use of links and other
`
`forms of communication through their apps and (b) prevented them from reducing
`
`the costs of switching away from Apple’s App Store or iOS IAP by allowing cus-
`
`tomers to keep the iPhone and iOS they like while blocking them from accessing
`
`rival app stores or in-app payment methods. That consumers genuinely like the
`
`overall iPhone product is exactly what makes the second restriction so effective; it
`
`means the cost of switching away from the App Store or iOS IAP is the cost of giv-
`
`ing up the iPhone altogether. See Phillip Areeda & Herbert Hovenkamp, Antitrust
`
`Law ¶ 564 (4th and 5th eds. 2013-2020) (markups in aftermarkets are particularly
`
`effective when small relative to the overall cost of the product).
`
`C. A Common Error Underlies the District Court’s Failure to Recog-
`nize Distinct Markets in at Least Two Instances
`
`Proceeding from first principles rooted in economics also helps identify a
`
`common error behind the district court’s failure to recognize distinct markets in at
`
`
`
`14
`
`
`
`Case: 21-16506, 01/27/2022, ID: 12353961, DktEntry: 57, Page 21 of 39
`
`
`
`least two instances. The district court rejected plaintiffs’ claim that a foremarket
`
`for mobile operating systems exists that is distinct from the market for
`
`smartphones. Slip op. at 45 (“[I]t is illogical to argue that there is a market for
`
`something that is not licensed or sold to anyone. Competition exists for
`
`smartphones which are more than just operating systems.”). Accordingly, it con-
`
`cluded there could be no aftermarkets for the App Store and iOS IAP.4 Slip op. at
`
`45–46 (“Given the Court’s rejection of the foremarket theory, the aftermarket the-
`
`ory fails as it is tethered to the foremarket.”). The district court also rejected
`
`Epic’s claim that in-app payment processing and app-distribution are distinct prod-
`
`uct markets. Slip op. at 65–67 (finding “[t]he [IAP] system is not something that is
`
`bought or sold” and, therefore, not “a separate and distinct product” from the App
`
`Store). As a result, it held the App Store and IAP could not be anticompetitively
`
`tied together. Both of these holdings are rooted in a common and fundamental er-
`
`ror: assuming that the way markets currently operate reflects the way competitive
`
`markets would operate. This same fundamental error underlies the infamous “cel-
`
`lophane fallacy” and many other intellectual traps for the unwary in antitrust law.
`
`
`4 There is no logical reason why the conclusion that there can be no aftermarket
`must follow from the district court’s conclusion that no separate market for mobile
`operating systems exists. If the district court concluded that mobile operat