Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA
EPIC GAMES, INC.,
Plaintiff,
vs.
APPLE INC.,
Defendant.
Case No. 4:20-cv-05640-YGR ORDER GRANTING IN PART AND DENYING IN PART MOTION FOR PRELIMINARY INJUNCTION
Re: Dkt. No. 61
Preliminary injunctive relief is an extraordinary measure rarely granted. Plaintiff Epic
Games, Inc.’s lawsuit against defendant Apple Inc. challenges the fundamental operation of digital
platforms affecting millions of users. To resolve it, the Court must apply the Sherman Act,
California’s Cartwright Act, and California’s Unfair Competition Law—statutes enacted more
than a century ago—to a technology context where lawyers and economists can merely
hypothesize about the future of the digital frontier. While courts are charged with adjudicating
cases of significant impact, they do so cautiously, and on full records, with the status quo intact.
In this motion for preliminary injunction, Epic Games asks the Court to force Apple to
reinstate Fortnite to the Apple App Store, despite its acknowledged breach of its licensing
agreements and operating guidelines, and to stop Apple from terminating its affiliates’ access to
developer tools for other applications, including Unreal Engine, while Epic Games litigates its
claims. Having carefully considered the parties’ arguments, and for the reasons set forth more
fully below, the Court maintains its findings from the temporary restraining order and hereby
GRANTS IN PART and DENIES IN PART Epic Games’ motion for a preliminary injunction.
Epic Games bears the burden in asking for such extraordinary relief. Given the novelty
and the magnitude of the issues, as well as the debate in both the academic community and society
at large,1 the Court is unwilling to tilt the playing field in favor of one party or the other with an
1 See e.g., “Investigation of Competition in Digital Marketplaces,” STAFF OF SUBCOMM.
ON ANTITRUST, COMMERCIAL, AND ADMIN. LAW OF THE COMM. ON THE JUDICIARY, UNITED
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 1 of 39
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
early ruling of likelihood of success on the merits. Epic Games has strong arguments regarding
Apple’s exclusive distribution through the iOS App Store, and the in-app purchase (“IAP”) system
through which Apple takes 30% of certain IAP payments. However, given the limited record,
Epic Games has not sufficiently addressed Apple’s counter arguments. The equities, addressed in
the temporary restraining order, remain the same.
I. BACKGROUND
The Court summarizes the limited record before it on this motion for preliminary
injunction. To expedite issuance of this Order, the Court cites to some of the basic background
from its prior order2 as the background is equally relevant here. The Court notes disputes in the
evidence where otherwise appropriate.
A. The Players
With respect to Epic Games: Epic Games is a United States-based tech-company that specializes in video games, including, as relevant here, the popular multi-platform3 game, Fortnite. Fortnite is structured around “seasons,” whereby narratives, themes, and events are introduced for a limited time. Cross-platform play is enabled for all users so long as those users remain on the same version of the game. . . .
Epic Games International, S.a.r.l (“Epic International”) is a related company based in Switzerland and hosts, among others, the Unreal Engine. The Unreal Engine is a graphics engine created by Epic International to assist in its development of video games that it later began licensing to other developers. The Unreal Engine 4, the current version of the engine on the market, is used by third-party developers for the development of video games for both console and mobile
STATES HOUSE OF REPRESENTATIVES, (Oct. 6, 2020). The Court finds it appropriate to take judicial notice of public documents generated by Congress, although the Court does not consider the content therein for purposes of this motion. See Vasserman v. Henry Mayo Newhall Mem’l Hosp., 65 F. Supp. 3d 932, 942-43 (C.D. Cal. 2014) (noting that court can take notice of ‘[o]fficial acts of legislative, executive, and judicial departments of the United States’”); Del Puerto Water Dist. v United States Bur. of Reclamation, 271 F. Supp. 2d 1224, 1234 (E.D. Cal. 2003) (taking judicial notice of House Reports).
2 See Epic Games, Inc. v. Apple Inc., 4:20-cv-05640-YGR, 2020 WL 5073937 (N.D. Cal. Aug. 24, 2020) (Dkt. No. 48).
3 These platforms include Android, iOS, macOS, Windows, Sony PlayStation 4, Microsoft Xbox One, Nintendo Switch. Fortnite is also available for download through the Epic Games Store, as discussed herein.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 2 of 39
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
platforms, including for games currently offered in the iPhone App Store. These third parties range from smaller game developers to larger corporations, such as Microsoft Corporation. The Unreal Engine has also been used by third parties for architecture projects, film and television production, and medical training.
Epic Games, 2020 WL 5073937 at *1 (Dkt. No. 48 at 2). Epic Games has released twenty-five
(25) updates to Unreal Engine since 2014, and anticipates releasing future updates to ensure that
Unreal Engine remains compatible with new versions of Apple’s software, such as the now
released iOS 14. Developers can use Unreal Engine commercially on a royalty model or
negotiated license, but it is otherwise free for non-commercial use. Although more applications on
the iOS platform are powered by a rival game engine, Unity, a significant number of iOS
applications are constructed based on Unreal Engine, including Fortnite competitor
PlayerUnknown’s Battlegrounds (“PUBG”).
Epic Games also maintains or controls other affiliates including: Epic International, Life
on Air, Inc. (both in San Francisco, California and Austin, Texas), KA-RA S.a.r.l., Psyonix LLC,
and Quixel AB (collectively, “Epic Affiliates”). The Epic Affiliates maintain control over certain
applications and software within the Epic Games business. These identified applications include:
Unreal Engine, Unreal Remote 2, Unreal Match 3, Action RPG Game Sample, Unreal Remote,
Live Link Face, and House Party, among others. Meanwhile, Epic Games itself controls Battle
Breakers, Infinity Blade Stickers, Spyjinx, and, as relevant here, Fortnite.
Beyond these games and applications, Epic Games also operates a digital marketplace to
sell game software called the Epic Games Store. As pled in the operative complaint: the Epic
Games Store was created to compete against the leading multi-publisher digital video game
marketplace on computer platforms, Steam, which is operated by Valve Corporation. The Epic
Games Store provides access to more than 250 games from more than 200 developers. Like other
video game digital distribution platforms, the Epic Games Store offers personalized features,
including friends list management and game matchmaking services. As alleged, absent Apple’s
alleged anti-competitive conduct, Epic Games would also create an analogous Epic Games Store
for the iOS platform independent of Apple’s digital marketplace.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 3 of 39
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
With respect to Apple: Apple is a ubiquitous tech-company that makes products ranging from hardware to software. Apple, as relevant here, maintains an App Store for the iOS platform that is geared for its mobile devices, the iPhones [and iPads]. The App Store allows third-party developers an opportunity to create and thereafter sell applications to iPhone [and iPad] users. Apple generally takes 30% of the sale of the application or of the IAP made within the third-party application itself. Apple’s agreements with developers and the App Store guidelines do not generally permit third-party developers to circumvent the IAP system.
Id. at *2 (Dkt. No. 48 at 2). In addition to preventing developers from circumventing the IAP
system, developers are also prohibited from distributing applications outside of the App Store on
the iOS platform.4 In short: Apple maintains the iOS platform as a walled garden or closed
platform model, whereby Apple has strict and exclusive control over the hardware, the operating
system, the digital distribution, and the IAP system.
In order to access the App Store and to obtain developer tools, developers are required to
comply with Apple’s rules and regulations through a web of agreements and guidelines: As relevant here, Apple maintains separate developer agreements and developer program licensing agreements between Epic Games, Epic International and four other affiliated entities. Apple also maintains a separate agreement, “Xcode and Apple SDKs Agreement,” regarding its developer tools (software development kits, or “SDKs”).
Epic Games, 2020 WL 5073937 at *1 (Dkt. No. 48 at 2-3). These agreements have broad
language including terminable at-will clauses.5
The relationship between Epic Games and Apple dates to at least 2011, when Epic Games
released its first Infinity Blade game on the iOS platform. Epic Games and Apple collaborated for
several Apple events, showcasing Epic Games’ iOS games and the earlier iterations of the Unreal
Engine running on the iOS and macOS platforms. Following the success of Fortnite on other
4 For purposes of this motion, the parties refer to the operating system for both iPhones
and iPads as iOS. (See Opp’n at 4, n.2 (Dkt. No. 73 at 10).) Moreover, Epic Games pleads that there are no differences between iOS and iPadOS to the allegations in the complaint. (Compl. ¶ 39, n.1 (Dkt. No. 1).) Similarly, this Order refers to iOS to refer to both the iPhone and iPad platforms, and references to iPhones generally also apply to iPads.
5 The record also contains two enterprise account agreements for Epic Games and YEVVO Entertainment, Inc. The parties do not otherwise discuss the significance of these agreements.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 4 of 39
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
video game platforms, Epic Games launched Fortnite on iOS in April 2018, where it remained on
the platform until, as discussed below, August 13, 2020. During this time period: (i) 116 million
iOS device users accessed Fortnite, spending more than 2.86 billion hours in the game; (ii) the
daily average users numbered approximately 2.5 million daily iOS players, representing nearly
10% of Fortnite’s total average daily players; and (iii) 63% of iOS players on Fortnite have only
accessed Fortnite from an iOS device. Finally, iOS users accounted for more IAPs within
Fortnite than those on the Android platform, but iOS users spend less on IAPs than those on the
console platforms, including the Sony PlayStation 4 and Microsoft Xbox One.
B. Relevant Background
On June 30, 2020, the developer program licensing agreements for the Epic Games
account, the Epic International account, KA-RA S.a.r.l. account, and the Epic Games enterprise
account were renewed by the payment of separate consideration.6 That same day, Epic Games
founder and Chief Executive Officer (“CEO”) Tim Sweeney sent an email to Apple executives,
including Apple CEO Tim Cook, requesting the ability to offer iOS consumers: (1) competing
payment processing options, “other than Apple payments, without Apple’s fees, in Fortnite and
other Epic Games software distributed through the iOS App Store”; and (2) a competing Epic
Games Store app “available through the iOS App Store and through direct installation that has
equal access to underlying operating system features for software installation and update as the
iOS App Store itself has, including the ability to install and update software as seamlessly as the
iOS App Store experience.” (Sweeney Decl. ¶ 14, Ex. A (Dkt. No. 65-1 at 2).) Mr. Sweeney
highlights that these two offerings would allow consumers to pay less for digital products, and
allow developers to earn more money. Mr. Sweeney also wrote that he “hope[d] that Apple
w[ould] also make these options equally available to all iOS developers in order to make software
sales and distribution on the iOS platform as open and competitive as it is on personal computers.”
(Id.) In this email, Mr. Sweeney does not provide any offer to pay Apple any portion of the 30
6 The renewal price for the enterprise accounts were each $299; the other agreements were
each renewed at a price of $99.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 5 of 39
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
percent it charges on either app distribution or for IAP.
On July 10, 2020, Apple Vice President and Associate General Counsel Douglas G. Vetter
responded to Mr. Sweeney’s email with a formal letter. In short, Apple’s response to Epic Games’
requests was no. Both requests were unequivocally refused. (Sweeney Decl. ¶ 15, Ex. B (Dkt.
No. 65-2).) As relevant here and with respect to the Epic Games Store request, Mr. Vetter wrote: Apple has never allowed this. Not when we launched the App Store in 2008. Not now. We understand this might be in Epic’s financial interests, but Apple strongly believes these rules are vital to the health of the Apple platform and carry enormous benefits for both consumers and developers. The guiding principle of the App Store is to provide a safe, secure and reliable experience for users and a great opportunity for all developers to be successful but, to be clear, when it comes to striking the balance, Apple errs on the side of the consumer.
(Id.) Mr. Vetter also reiterated that Epic Games’ request to establish a separate payment processor
would interfere with Apple’s own IAP system, the business model of which has been used in the
App Store since its inception. (Id.)
On July 17, 2020 Mr. Sweeney responded to what he described as a “self-righteous and
self-serving screed,” writing that he hoped “Apple someday chooses to return to its roots building
open platforms in which consumers have freedom to install software from sources of their
choosing, and developers can reach consumers and do business directly without intermediation.”
(Sweeney Decl. ¶ 16, Ex. C (Dkt. No. 65-3 at 2).) He stated that Epic Games “is in a state of
substantial disagreement with Apple's policy and practices,” and promised that it would “continue
to pursue this, as [it] ha[s] done in the past to address other injustices in [the] industry.” (Id.)
In fulfilling Mr. Sweeney’s promise to “pursue this” perceived “injustice,” Epic Games
covertly introduced a “hotfix” into the Fortnite version 13.40 update on August 3, 2020. Epic
Games did not disclose the full extent of this hotfix to Apple, namely that this hotfix would enable
a significant and substantive feature to Fortnite permitting a direct pay option to Epic Games that
would be activated when signaled by Epic Games’ servers. Until this signal was sent out, this
direct pay option would remain dormant. When activated, however, this direct pay option would
allow iOS Fortnite players to choose a direct pay option that would circumvent Apple’s IAP
system. Relying on the representations, that intentionally omitted the full extent and disclosure of
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 6 of 39
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
this hotfix, Apple approved of the Fortnite version 13.40 to the App Store.7
The hotfix remained inactive until the early morning of August 13, 2020, when Epic
Games made the calculated decision to breach its allegedly illegal agreements with Apple by
activating the undisclosed code in Fortnite, allowing Epic Games to collect IAPs directly.
Fortnite remained on the App Store until later that morning, when Apple removed Fortnite from
the App Store, where it remains unavailable. Later that same day, Epic Games filed this action
and began a pre-planned, and blistering, marketing campaign against Apple. This marketing
campaign included: a large-scale twitter campaign, a releasing of a parody video of the iconic
Apple 1984 commercial, a Fortnite tournament in support of its lawsuit with in-game prizes, and a
releasing of a limited time skin in Fortnite called the Tart Tycoon,8 among other actions.
The following day, Apple responded sternly. It informed Epic Games that, based on its
breaches of the App Store guidelines, and the developer program license agreement, it would be
revoking all developer tools, which would preclude updates for its programs and software. Apple
gave two weeks to comply with the App Store guidelines and the agreements. Apple also
identified general consequences for any failure to comply, but specifically cited Unreal Engine as
potentially being subject to harm should Epic Games fail to comply within the two-week period.
Thereafter on August 17, 2020, Epic Games filed the request for a temporary restraining
order, requesting the reinstatement of Fortnite with its activated hotfix onto the App Store, and to
enjoin Apple from revoking the developer tools belonging to the Epic Affiliates. The Court
declined to reinstate Fortnite onto the App Store, but temporarily restrained Apple from taking any
action with respect to the Epic Affiliates’ developer tools and accounts.
7 Epic Games disputes that its use of the hotfix was deceptive where it is common practice
in the gaming and software industry. The deceptive conduct does not derive from Epic Games’ use of the hotfix specifically, but from using a hotfix to clandestinely add features in violation of the guidelines and its agreements with Apple, and then failing to disclose such code. Moreover, Epic Games did this despite receiving an unambiguous refusal from Apple only a few weeks prior to the introduction of its hotfix. The record further reflects that while hotfixes are commonly used in the industry, their uses are generally to fix or patch critical bugs or defects—not to enact substantive and significant new features. Epic Games’ adamant refusal to understand this basic distinction is not only baffling, but undermines its credibility with this Court.
8 Modeled presumably on Mr. Cook’s likeness.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 7 of 39
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
On August 27, 2020, as planned by Epic Games, an updated version containing season four
of Fortnite was released on all platforms except for the iOS platform, which Epic Games could no
longer update due to its breaches of the Apple agreements and guidelines. By design, Fortnite
users can only play amongst other users currently operating the same version. Because of this
release, iOS Fortnite players no longer had the ability to play cross-platform with other players
(unless these players chose not to update their version, forgoing playing the new season).
On August 28, 2020, on the expiration of the two-week deadline, Apple terminated Epic
Games’ developer program account, referenced as Team ID ’84, stating “Apple is exercising its
right in Apple’s sole discretion to terminate your status as a registered Apple Developer pursuant
to the Apple Developer Agreement and is terminating the Developer Agreement and the Program
License Agreement pursuant to their terms. . . . [W]e will deny your reapplication to the Apple
Developer Program for at least a year.” (Grant Decl. ¶ 35, Ex. H (Dkt. No. 63-8 at 2).)9
Following this, the parties engaged in briefing on the motion for preliminary injunction on
a slightly expedited basis. The Court heard oral argument on the motion on September 28, 2020.
II. LEGAL FRAMEWORK
Preliminary injunctive relief, whether in the form of a temporary restraining order or a
preliminary injunction, is an “extraordinary and drastic remedy,” that is never awarded as of right.
Munaf v. Geren, 553 U.S. 674, 689-90 (2008) (internal citations omitted). “It is so well settled as
not to require citation of authority that the usual function of a preliminary injunction is to preserve
the status quo ante litem pending a determination of the action on the merits.” Tanner Motor
Livery, Ltd. v. Avis, Inc., 316 F.2d 804, 808 (9th Cir. 1963). A preliminary injunction is “not a
preliminary adjudication on the merits but rather a device for preserving the status quo and
preventing the irreparable loss of rights before judgment.” Sierra On-Line, Inc. v. Phoenix
Software, Inc., 739 F.2d 1415, 1422 (9th Cir. 1984) (citation omitted).
9 The record also reflects that Apple made moves in early September to cancel Epic
Games’ ability to use the Sign in with Apple (“SIWA”) on the Fortnite game. Apple eventually relented to allowing its continued use without waiving any right to revoke SIWA in the future.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 8 of 39
9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
In order to obtain such relief, plaintiffs must establish four factors: (1) they are likely to
succeed on the merits; (2) they are likely to suffer irreparable harm in the absence of preliminary
relief; (3) the balance of equities tips in their favor; and (4) an injunction is in the public interest.
Winter v. Natural Res. Def. Council. Inc., 555 U.S. 7, 20 (2008). With respect to the success on
the merits and balance of harms factors, courts permit a strong showing on one factor to offset a
weaker showing on the other, so long as all four factors are established. Alliance for the Wild
Rockies v. Cottrell, 632 F.3d 1127, 1135 (9th Cir. 2011). In other words, “if a plaintiff can only
show that there are serious questions going to the merits—a lesser showing than likelihood of
success on the merits—then a preliminary injunction may still issue if the balance of hardships tips
sharply in the plaintiff’s favor, and the other two Winter factors are satisfied.” Shell Offshore, Inc.
v. Greenpeace, Inc., 709 F.3d 1281, 1291 (9th Cir. 2013) (citations and quotations omitted). Thus,
under the Ninth Circuit’s “‘sliding scale’ approach to these factors,” “when the balance of
hardships tips sharply in the plaintiff’s favor, the plaintiff need demonstrate only ‘serious
questions going to the merits.’” hiQ Labs, Inc. v. LinkedIn Corp., 938 F.3d 985, 992 (9th Cir.
2019) (quoting Alliance for the Wild Rockies, 632 F.3d at 1131). The Court addresses each.
III. ANALYSIS
The Court finds it appropriate to evaluate, once again, Apple’s actions with respect to
(i) Epic Games specifically, including the delisting of Fortnite and other games authorized under
Epic Games’ contract with Apple, and (ii) the attempt to suspend and terminate developer rights
authorized under other contracts, such as the one controlling Unreal Engine.
A. Likelihood of Success on the Merits
Epic Games brings ten claims for violations of Sherman Act, the California Cartwright
Act, and California Unfair Competition. For purposes of the motion for preliminary injunction,
Epic Games focuses on two: the monopoly maintenance claim under section 2 of the Sherman
Act, and the tying claim under section 1 of the Sherman Act. Accordingly, the Court cabins its
analysis with respect to these only. Having reviewed the limited record, while Epic Games raises
serious questions on the merits, the Court cannot conclude that Epic Games will likely succeed on
the merits of those claims. Too many unknowns remain.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 9 of 39
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
1. Preliminary Considerations
The current legal landscape cautions against preliminarily finding antitrust violations based
on less than a full record. As the parties acknowledge, this matter presents questions at the
frontier edges of antitrust law in the United States. Simply put, no analogous authority exists.
The questions and issues raised in this litigation concern novel and innovative business practices
in the technology market that have not otherwise been the subject of antitrust litigation.10
As the Ninth Circuit recently recognized in Federal Trade Commission v. Qualcomm Inc.,
“novel business practices—especially in technology markets—should not be ‘conclusively
presumed to be unreasonable and therefore illegal without elaborate inquiry as to the precise harm
they have cause or the business excuse for their use.’” 969 F.3d 974, 990-91 (9th Cir. 2020)
(emphasis in original) (quoting United States v. Microsoft Corp., 253 F.3d 34, 91 (D.C. Cir. 2001)
(en banc)). This is “[b]ecause innovation involves new products and business practices, courts[’]
and economists[’] initial understanding of these practices will skew initial likelihoods that
innovation is anticompetitive and the proper subject of antitrust scrutiny.” Id. at 991 (internal
quotation marks omitted) (quoting Geoffrey A. Manne & Joshua D. Wright, Innovation and the
Limits of Antitrust, 6 J. COMP. L. & ECON. 153, 167 (2010)); see also Rachel S. Tennis &
Alexander Baier Schwab, Business Model Innovation and Antitrust Law, 29 YALE J. ON REG. 307,
319 (2012) (explaining how “antitrust economists, and in turn lawyers and judges, tend to treat
novel products or business practices as anticompetitive” and “are likely to decide cases wrongly in
rapidly changing dynamic markets,” which can have long-lasting effects particularly in
technological markets, where innovation “is essential to economic growth and social welfare” and
“an erroneous decision will deny large consumer benefits”). The Court therefore has an even
greater obligation to conduct an “elaborate inquiry” before determining that the alleged practices
10 The exceptions involve the related In re Apple Antitrust, 4:11-cv-06714-YGR (N.D.
Cal.) (Pepper), and Donald Cameron v. Apple Inc., 4:19-cv-03074-YGR (N.D. Cal.), matters that are currently before this Court. Both Pepper and Cameron are in the middle of discovery, with motions for class certification anticipated in early 2021. No substantive rulings as to the merits of the claims have otherwise been made in those cases. Similar issues arise in Epic Games, Inc. v. Google LLC, 3:20-cv-05671-JD (N.D. Cal.), filed at the same time but which does not have similar motions for preliminary injunctive relief.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 10 of 39
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
violate antitrust law.
Second, the record remains insufficient to conclude that Epic Games will likely succeed on
the merits of its claims. As discussed below, the record includes conflicting evidence in support
of both Epic Games and Apple; a lack of crucial evidence without which the merits cannot be
determined; and fundamental disagreement by expert witnesses that is not resolvable at this stage
of the case. With respect to the last, the Court highlights that the parties’ retained expert witnesses
are all accomplished and distinguished individuals. Epic Games submits declarations from Dr.
David S. Evans, an economist with degrees from the University of Chicago, whose scholarly work
has been widely read and cited, including by the Supreme Court in Ohio v. American Express Co.,
138 S. Ct. 2274 (2018) (Amex). Apple submits declarations from Dr. Richard Schmalensee—an
economist with degrees from the Massachusetts Institute of Technology (“MIT”), whose work is
also widely read and cited, including in Amex and Microsoft—and Dr. Lorin Hitt—an academic
with a business management background and degrees from MIT and Brown University, who has
background in electrical engineering and technology. These expert reports reflect fundamental
disagreements from luminaries in the field as to the foundational questions of this matter. While
ultimately one view will likely prevail, at this juncture, the Court concludes that reasonable minds
differ.
With these considerations in mind, the Court turns to the merits of the claims.
2. Monopoly Maintenance under Section 2 of the Sherman Act11
a. Legal Framework
In order to prevail on its theory that Apple engaged in unlawful monopolization under
section 2 of the Sherman Act, Epic Games must show: “(a) the possession of monopoly power in
the relevant market; (b) the willful acquisition or maintenance of that power; and (c) causal
antitrust injury.” Qualcomm, 969 F.3d at 990 (internal quotation marks omitted); see also United
States v. Grinnell Corp., 384 U.S. 563, 570-71 (1966) (stating that a section 2 claim requires “(1)
11 The Court’s discussion of the section 2 claim before the section 1 claim mirrors the
parties’ briefing.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 11 of 39
12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
the possession of monopoly power in the relevant market and (2) the willful acquisition or
maintenance of that power as distinguished from growth or development as a consequence of a
superior product, business acumen, or historic accident”).
“A threshold step in any antitrust case is to accurately define the relevant market, which
refers to ‘the area of effective competition.’” Qualcomm, 969 F.3d at 992 (quoting Amex, 138 S.
Ct. at 2285); see also Image Tech. Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195, 1202 (9th
Cir. 1997) (“The relevant market is the field in which meaningful competition is said to exist.”).
Monopoly power under the first element can be defined as “the power to control prices or exclude
competition”12 and may be inferred from defendant’s predominant market share in the relevant
market. Grinnell, 384 U.S. at 571. In addition, “courts usually cannot properly apply the rule of
reason without an accurate definition of the relevant market.” Amex, 138 S. Ct. at 2285. Without
a relevant market definition, “there is no way to measure the defendant's ability to lessen or
destroy competition.” Id. (brackets and citation omitted).
“The relevant market must include both a geographic market and a product market.” Hicks
v. PGA Tour, Inc., 897 F.3d 1109, 1120 (9th Cir. 2018). The latter “must encompass the product
at issue as well as all economic substitutes for the product.” Newcal Indus., Inc. v. Ikon Office
Sol., 513 F.3d 1038, 1045 (9th Cir. 2008); see also id. (“The consumers do not define the
boundaries of the market; the products or producers do [and] the market must encompass the
product at issue as well as all economic substitutes for the product.”). “Economic substitutes have
a ‘reasonable interchangeability of use’ or sufficient ‘cross-elasticity of demand’ with the relevant
product.” Hicks, 897 F.3d at 1120 (quoting Newcal, 513 F.3d at 1045); see also United States v.
E.I. du Pont de Nemours & Co., 351 U.S. 377, 404 (1956). “Including economic substitutes
ensures that the relevant product market encompasses ‘the group or groups of sellers or producers
who have actual or potential ability to deprive each other of significant levels of business.’”
Hicks, 897 F.3d at 1120 (quoting Newcal, 513 F.3d at 1045); see also Du Pont, 351 U.S. at 393
12 More precisely, “a firm is a monopolist if it can profitably raise prices substantially
above the competitive level.” Microsoft, 253 F.3d at 51.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 12 of 39
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
(“Illegal power must be appraised in terms of the competitive market for the product.”).
“[I]n some instances one brand of a product can constitute a separate market.” See
Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 482 (1992); see also Newcal, 513
F.3d at 1048 (“[T]he law permits an antitrust claimant to restrict the relevant market to a single
brand of the product at issue . . . .”). However, such “[s]ingle-brand markets are, at a minimum,
extremely rare” and courts have rejected such market definitions “[e]ven where brand loyalty is
intense.” Apple, Inc. v. Psystar Corp., 586 F. Supp. 2d 1190, 1198 (N.D. Cal. 2008) (internal
quotation marks omitted). But see id. (“Antitrust markets consisting of just a single brand,
however, are not per se prohibited . . . . In theory, it may be possible that, in rare and unforeseen
circumstances, a relevant market may consist of only one brand of a product.”)
Nevertheless, “it is legally permissible to premise antitrust allegations on a submarket” or
an aftermarket. Newcal, 513 F.3d at 1045. A submarket “is economically distinct from the
general product market.” Id. at 1045. There are “several ‘practical indicia’ of an economically
distinct submarket,” including: industry or public recognition of the submarket as a separate economic entity, the product's peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors.
Id. (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 325 (1962)).13 An aftermarket is
“wholly derivative from and dependent on the primary market.” Id. at 1049. An aftermarket may
constitute the relevant market where market imperfections, such as information and switching
costs, “prevent consumers from realizing that their choice in the initial market will impact their
freedom to shop in the aftermarket.” Id. at 1050. Thus, “[d]etermining the relevant market can
involve a complicated economic analysis, including concepts like cross-elasticity of demand, and
‘small but significant nontransitory increase in price’ (‘SSNIP’) analysis.” Theme Promotions,
Inc. v. News America Marketing FSI, 546 F.3d 991, 1002 (9th Cir. 2008); see also Psystar, 586. F.
13 Epic Games’ economic expert does not address these factors; instead, he principally
relies on the those that follow.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 13 of 39
14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
Supp. 2d at 1198.
The determination of a “relevant market” is a highly factual question. See Eastman Kodak,
504 U.S. at 482 (“The proper market definition in this case can be determined only after a factual
inquiry into the ‘commercial realities’ faced by consumers.”); see also Newcal, 513 F.3d at 1051
(“The actual existence of an aftermarket . . . is a factual question. The actual existence of a
separate economic entity (i.e. a submarket) . . . is a factual question. The actual existence of [a
party’s] market power within the alleged submarket is a factual question. . . . The initial market's
actual ability, through cross-elasticity of demand, to discipline anti-competitive conduct in the
aftermarket is a factual question.”); Teradata Corp. v. SAP SE, Case No. 18-cv-03670-WHO,
2018 WL 6528009, at *14 (N.D. Cal. Dec. 12, 2018) (“The definition of a ‘relevant market’ in
which defendant has market power is typically a factual rather than legal question.”).14
Even if a plaintiff establishes monopoly power in the relevant market under the first
element, courts will not condemn it unless “it is accompanied by an element of anticompetitive
conduct” under the second element. Qualcomm, 969 F.3d at 990 (emphasis in original) (quoting
Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004)). “The
mere possession of monopoly power, and the concomitant charging of monopoly prices, is . . . an
important element of the free market system.” Verizon Commc’ns, 540 U.S. at 407. Thus, courts
distinguish between “the willful acquisition or maintenance of [monopoly] power” from “growth
or development as a consequence of a superior product, business acumen, or historic accident.”
See Grinnell, 384 U.S. at 571. To demonstrate the former, plaintiff must show “anticompetitive
abuse or leverage of monopoly power, or a predatory or exclusionary means of attempting to
monopolize the relevant market.” Qualcomm, 969 F.3d at 990. “To be condemned as
exclusionary, a monopolist’s act must have an ‘anticompetitive effect’—that is, it must harm the
competitive process and thereby harm consumers[, i]n contrast [to] harm to one or more
14 The parties are reminded that Newcal was decided on a motion to dismiss and has
limited reach. The Ninth Circuit explicitly indicated that the case is not “guarantee[d]” to survive a motion for summary judgment because the “actual existence of a separate economic entity (i.e. a submarket) that includes only IKON’s customers is a factual question.” 513 F.3d at 1051. The same is true of Teradata and Psystar. The Court relies on each for those limited propositions.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 14 of 39
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
competitors[, which] will not suffice.” Id. (emphasis in original) (internal quotation marks and
alternations omitted) (quoting Microsoft, 253 F.3d at 58).
Anticompetitive conduct is evaluated under the “rule of reason.” Id. at 991. First, plaintiff
must show “diminished consumer choices and increased prices” as “the result of a less competitive
market due to either artificial restrains or predatory or exclusionary conduct” by the defendant. Id.
Then, “if a plaintiff successfully establishes a prima facie case . . . by demonstrating
anticompetitive effect, then the monopolist may offer a ‘procompetitive justification’ for its
conduct.” Id. (internal quotation marks omitted) (quoting Microsoft, 253 F.3d at 59). For
example, the monopolist may show “that its conduct is . . . a form of competition on the merits
because it involves, for example, greater efficiency or enhanced consumer appeal.” Id. (internal
quotation marks omitted) (quoting Microsoft, 253 F.3d at 59). Finally, if defendant offers a non-
pretextual procompetitive justification, the burden shifts back to the plaintiff to rebut defendant’s
claim or “demonstrate that the anticompetitive harm of the conduct outweighs the procompetitive
benefit.” Id. (internal quotation marks omitted) (quoting Microsoft, 253 F.3d at 59).
Last, if plaintiff satisfies the first and second elements of monopoly power and willful
maintenance or acquisition of that power in the relevant market, the last element of causation may
be inferred “when exclusionary conduct is aimed at producers of nascent competitive technologies
as well as when it is aimed at producers of established substitutes.” Microsoft, 253 F.3d at 79
(cited with approval by Qualcomm, 969 F.3d at 992).
b. Relevant Market Analysis
In summary, the record does not yet establish how the “relevant market” should be defined.
Without a definition of the relevant market, the existence of market power—the foundation of a
monopolization claim—cannot be assessed. Accordingly, Epic Games has not yet shown that it
will likely succeed on the merits of the monopolization claim.
The relevant market must include both a geographic market and a product market.
Unsurprisingly, the parties disagree on the product market.15 Epic Games avers that the relevant
15 Both Epic Games and Apple agree, however, that the “geographic market” is likely
global. (But see Evans Decl. at 10 n.37 (Dkt. No. 62 at 12) (reserving future opinion on whether
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 15 of 39
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
product market is the market for distribution of apps on the iOS software platform, which it refers
to as the “iOS App Distribution Market.” Thus, Epic Games narrows the relevant market to
consider only how iOS apps are distributed on the iOS platform. Apple meanwhile asserts that the
relevant market must include competing platforms on which Fortnite is distributed and monetized.
In other words, Apple argues that the Court must consider the wider video game market and
distribution on other platforms, including the Microsoft Xbox One, the Sony PlayStation 4, the
Nintendo Switch, computer platforms (Microsoft Windows PCs, macOS computers), and tablets
(Google Android and Microsoft Surface). Thus, Apple seeks a broader market definition that
includes the digital distribution of video games across all video game platforms. Ultimately, the
Court must discern where competition exists and whether such competition is sufficient to impact
price and discipline market players.
Epic Games’ relevant market definition that iOS App Distribution is an “aftermarket” of
the smartphone OS market is plausible.16 See Newcal, 513 F.3d at 1050. However, in some ways,
Epic Games offers a failsafe definition by restricting the market so narrowly. By definition, Epic
Games’ proposed market definition excludes other smartphone systems, including the Google
Android system, as well as video game platforms and their digital distribution markets. Courts
have expressly cautioned against such a narrowing of the relevant market definition. See Du Pont,
351 U.S. at 392-93 (“A retail seller may have in one sense a monopoly on certain trade because of
location . . . or because no one else makes a product of just the quality or attractiveness of his
product . . . . Thus one can theorize that we have monopolistic competition in every
nonstandardized commodity with each manufacturer having power over the price and production
of his own product. [However, i]llegal power must be appraised in terms of the competitive
market for the product.”); Psystar, 586 F. Supp. 2d at 1198 (“[M]anufacturer’s own products do
the Chinese mobile market should be included in the geographic market).)
16 Apple fails to respond adequately to the “aftermarket” theory, devoting a single paragraph to it and stating, in a conclusory fashion, that “this is not an aftermarket case.” Should Epic Games continue to assert this theory, Apple should explain why switching and information costs do not render the IOS app distribution market distinct. Silence can be interpreted as an admission.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 16 of 39
17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
not themselves comprise a relevant product market.”).17
Moreover, Apple avers that an “aftermarket” requires user lock-in in the primary market.
Given the lack of legal citation, the Court surmises that this theory has not been adopted by any
court, even if embraced by economists. The term “lock-in” appears to derive from the Supreme
Court mention that “[i]f the cost of switching is high, consumers who already have purchased the
equipment, and are thus ‘locked in,’ will tolerate some level of service-price increases before
changing equipment brands.” Eastman Kodak, 504 U.S. at 476. In evaluating Epic Games’
response, resolution of the issue is focused on timing: Apple argues that consumers are not
locked-in to the purchase of iPhones, while Epic Games assumes the purchase and argues that
after the purchases occurs, a consumer is locked-in and unlikely to switch to a different
smartphone in response to slightly more expensive IAPs. Under the latter perspective, app
developers who wish to reach iOS users have no choice but to tolerate Apple’s 30% rate.18
Thus, at this stage of the litigation, and with the record before the Court, Apple’s relevant
market definition is also plausible. As Apple correctly points out, alternative means exist to
distribute Fortnite.19 Indeed, Epic Games expressly advertised the multiplatform nature of its
product following its breach of the Apple terms and service. (See Hitt Decl. ¶ 39 (Dkt. No. 77)
17 Apple further avers that as intellectual property owner, even if it is a monopolist, Apple
is not required to allow unfettered and uncompensated use of its own technology. See Herbert Hovenkamp et al., IP and Antitrust: An Analysis of Antitrust Principles Applied to Intellectual Property Law § 13.03 (3rd ed., 2016 & Supp. 2019) (citing Microsoft Corp., 253 F.3d at 63-64). That said, “intellectual property rights do not confer a privilege to violate the antitrust laws.” Microsoft, 253 F.3d at 64. Moreover, the parties fail to brief whether Apple possesses “essential facilities,” which may require (compensated) access. The Court makes no express finding on these issues, but notes these as other potential hurdles.
18 The Court also leaves for another day the proper classification of the 30% at issue, that is, whether it is a commission, a licensing fee, a “tax”, or a “price.” Each may have legal ramifications which have not been fully briefed, and therefore carry with them unintended consequences of choosing a term too quickly.
19 However, the Court notes that Apple’s argument assumes a user who owns multiple
devices, pays attention to prices for in-app purchases, and switches devices in response to price increases. There is little evidence that the ordinary iOS consumer carries such characteristics. Cf. U.S. v. Engelhard Corp., 126 F.3d 1302, 1306 (11th Cir. 1997) (rejecting relevant market analysis based on customer interviews where proponent failed to show that the customers were representative).
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 17 of 39
18
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
(“[The] party continues on PlayStation 4, Xbox One, Nintendo Switch, PC, Mac, GeForce Now,
and through both the Epic Games app at epicgames.com and the Samsung Galaxy Store.”).) The
multiplatform nature of Fortnite suggests that these other platforms and their digital distributions
may be economic substitutes that should be considered in any “relevant market” definition
because they are “reasonably interchangeable” when used “for the same purposes.” Du Pont, 351
U.S. at 395; see also Hicks, 897 F.3d at 1120-21 (dismissing antitrust claim when alleged relevant
market ignored multiple ways of reaching consumers). “If competitors can reach the ultimate
consumers of the product by employing existing or potential alternative channels of distribution, it
is unclear whether such restrictions foreclose from competition any part of the relevant market.”
Omega Envtl., Inc. v. Gilbarco, Inc., 127 F.3d 1157, 1163 (9th Cir. 1997).
Epic Games’ arguments distinguishing these other platforms as potential economic
substitutes have not been sufficiently tested. First, Epic Games avers that the iOS market is
distinct from other video game platforms because Sony, Nintendo, and Microsoft do not make
much profit, if any, on the sale of the hardware or console—unlike Apple, which allegedly makes
significant profits from the sale of each iPhone. This distinction is without legal precedent under
section 2 of the Sherman Act. Indeed, Sony, Nintendo and Microsoft all operate similar walled
gardens or closed platform models as Apple, whereby the hardware, operating system, digital
marketplace, and IAPs are all exclusive to the platform owner. As such, a final decision should be
better informed regarding the impact of the walled garden model given the potential for significant
and serious ramifications for Sony, Nintendo and Microsoft and their video game platforms.
Second, Epic Games’ avers that the iOS platform is unique from other gaming devices.
Specifically, Epic Games argues that gaming consoles and computers require electrical outlets and
separate screens and thus lack capacity for mobile play, which demands portable, battery operated,
and cellularly connected devices with built-in screens. (See Sweeney Reply Decl. ¶ 14 (Dkt. No.
86).) Yet, Epic Games repeatedly ignored discussion of gaming laptops, tablets, and the Nintendo
Switch, all of which can be played in a mobile fashion. These devices could have significant
overlap with the iOS platform in terms of the ultimate consumer. Again, however, at this stage,
the record does not contain sufficient information to determine whether such other devices are
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 18 of 39
19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
economic substitutes or are merely complimentary to iOS devices.
Thus, and for other reasons, Apple’s market definition also faces hurdles. Antitrust law is
not concerned with individual consumers or producers, like Epic Games; it is concerned with
market aggregates. Substitutes may not deprive a monopolist of market power if they fail to affect
enough customers to make a price increase unprofitable. See Theme Promotions, 546 F.3d at 1002
(defining relevant market by whether a price increase would cause a “significant number” of
customers to substitute to make the price increase unprofitable). Alternatively, constraints among
some consumers may not render the market as a whole narrow. See Telecor Commc’ns, Inc. v.
Southwestern Bell Telephone Co., 305 F.3d 1124, 1131-32 (10th Cir. 2002) (rejecting relevant
market definition based on a finding that “some” consumers could not substitute products because
the record did not show they were “significant enough to render the market as a whole non-cross-
elastic”). But see Engelhard, 126 F.3d at 1306 (noting that “it is possible for only a few customers
who switch to alternatives to make the price increase unprofitable, thereby protecting a larger
number of customers who would have acquiesced in higher . . . prices.”).
Here, both parties cite factors impacting the elasticity of their proposed markets. A final
determination may depend on the magnitude of those effects. For instance, focusing on Fortnite
alone, the record shows that (i) more than 116 million (out of 350 million) Fortnite players have
accessed Fortnite through the iOS platform; (ii) iOS players constitute roughly 10% of the daily
active Fortnite users since its iOS launch in April 2018; and (iii) 63% of Fortnite players on iOS
only play on the iOS platform. (Sweeney Decl. ¶ 3 (Dkt. No. 65).) Notably, the record is silent on
how often these 116 million individuals play Fortnite and devoid of information on the
characteristics of 10% of daily active users or whether these users access Fortnite through other
platforms. More broadly, there is no evidence regarding the size of the game app market
compared to other apps and whether they constitute a separate submarket with unique
characteristics that do not apply to other app developers.
Thus, the market definition rests on factual questions regarding the nature of the iOS
market as a whole: how many iOS users own multiple devices; how many iOS users would switch
to another device in response to a price increase; and how many producers can afford to forego
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 19 of 39
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
iOS customers altogether. Neither party adequately addresses these factual questions. Epic
Games assumes all iOS customers are the same, and Apple assumes that only Epic Games
customers are relevant.
Moreover, underlying these questions is a significant and unresolved dispute over
clustering. Apple focuses narrowly on game distribution channels because of the nature of Epic
Games’ business. But courts have often combined different services together when “the product
package is significantly different from, and appeals to buyers on a different basis from, the
individual products considered separately.” Image Tech., 125 F.3d at 1204-05. For example, in
United States v. Phillipsburg National Bank and Trust Company, the Supreme Court grouped
multiple financial services together into a relevant market of “commercial banking”—even though
they differed in their availability of substitutes—because customers generally obtain all banking
services from one place. 399 U.S. 350, 360-61 & n.4 (1970). Here, Epic Games may establish
that app distribution generally should be considered separately from app distribution of individual
games, which could have a significant impact on how alternative distribution channels are
evaluated.
Finally, underlying each of these issues is the question of perspective. Interchangeability
for purposes of the relevant market may vary depending on perspective. See, e.g., Little Rock
Cardiology Clinic PA v. Baptist Health, 591 F.3d 591, 597 (8th Cir. 2009) (reversing relevant
market definition based on improper perspective); Flovac, Inc. v. Airvac, Inc., 817 F.3d 849, 854-
55 (1st Cir. 2016) (same); Telecor, 305 F.3d at 1132-33 (same). Here, there are at least three
possible perspectives on the relevant market: (1) the customer who purchases the apps or games,
(2) the developer who makes the apps or games, and (3) the competing app store or digital
marketplace that distributes the apps or games. The parties adopt different perspectives, but
neither justifies its choice. And as the parties’ briefing demonstrates, the resolution of this
question could lead to radically different analysis.
In short, without the record to define the relevant antitrust market, Epic Games has not
established likelihood of success as to monopoly maintenance, only serious questions. Further,
without such definition, the Court need not evaluate the second or third elements of the section 2
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 20 of 39
21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
claim. Additionally, even under a section 2 claim, plaintiff must show anticompetitive conduct.
One way to do so includes a rule a reason analysis. Given the overlap of this issue with a section
1 claim, the Court addresses it below. See, Qualcomm, 969 F.3d at 991 (“The similarity of the
burden-shifting tests under [sections] 1 and 2 means that courts often review claims under each
section simultaneously.”).
3. Tying under Section 1 of the Sherman Act
a. Legal Framework
Tying arrangements under section 1 of the Sherman Act20 may be evaluated under either
per se or rule of reason analysis. See Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 29
(1984), abrogated on other grounds by Ill. Tool Works Inc. v. Indep. Ink, Inc., 547 U.S. 28 (2006).
Per se analysis allows “condemnation without inquiry into actual market conditions” based
on precedent that deems certain contractual arrangements “unreasonable as a matter of law.” Id. at
9, 15. “For a tying claim to suffer per se condemnation, a plaintiff must prove: (1) that the
defendant tied together the sale of two distinct products or services; (2) that the defendant
possesses enough economic power in the tying product market to coerce its customers into
purchasing the tied product; and (3) that the tying arrangement affects a not insubstantial volume
of commerce in the tied product market.”21 Cascade Health Sols. v. PeaceHealth, 515 F.3d 883,
913 (9th Cir. 2008).
To assess the first element, courts apply the purchaser demand test, which “examines direct
and indirect evidence of consumer demand and whether [a] defendant[] foreclosed competition on
the merits in a product market distinct from the market for the tying item.” Teradata, 2018 WL
6528009, at *12 (internal quotation marks omitted). “Direct evidence of demand includes
20 Section 1 of the Sherman Act broadly prohibits “[e]very contract, combination . . . , or
conspiracy, in restraint of trade or commerce among the several States.” 15 U.S.C. § 1. The term “restraint of trade” has been limited to “undue” (unreasonable) restrains. Amex, 138 S.Ct. at 2283.
21 As explained in Jefferson Parish, “the essential characteristic of an invalid tying arrangement lies in the seller’s exploitation of its control over the tying product to force the buyer into the purchase of a tied product that the buyer either did not want at all, or might have preferred to purchase elsewhere on different terms.” 466 U.S. at 12. Per se condemnation is only appropriate where such forcing is “probable.” Id. at 15.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 21 of 39
22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
‘whether, when given a choice, consumers purchase the tied good from the tying good maker, or
from other firms.’” Id. (quoting Rick-Mik Enters., Inc. v. Equilon Enters. LLC, 532 F.3d 963, 975
(9th Cir. 2008)). “Indirect evidence includes firm behaviors, for instance a single product is
apparent if ‘competitive firms always bundle the tying and tied goods’ together.” Id. (quoting
Rick-Mik, 532 F.3d at 975). A tie requires a “condition linked to a sale.” Aerotec Int’l, Inc. v.
Honeywell Int’l, Inc., 836 F.3d 1171, 1178 (9th Cir. 2016).
The second element of “forcing (or coercion) is likely if the seller has power in the tying
product market.” CollegeNet, Inc. v. Common Application, Inc., 355 F. Supp. 3d 926, 955 (D. Or.
2018) (quoting Robert’s Waikiki U-Drive, Inc. v. Budget Rent-a-Car Sys., Inc., 732 F.2d 1403,
1407 (9th Cir. 1984)). The third element asks “simply whether a total amount of business,
substantial enough in terms of dollar-volume so as not to be merely de minimis, is foreclosed to
competitors by the tie . . . .” Fortner Enters., Inc. v. U.S. Steel Corp., 394 U.S. 495, 501 (1969);
see also Datagate, Inc. v. Hewlett-Packard Co., 60 F.3d 1421, 1425 (9th Cir. 1995) (foreclosure of
a single purchaser sufficient so long as the dollar volume of sales is “not insubstantial”).
If a plaintiff fails to establish per se liability, a plaintiff must demonstrate that a defendant
“violated the Sherman Act because it unreasonably restrained competition” under the rule of
reason. Jefferson Parish, 466 U.S. at 29. The rule of reason “requires courts to conduct a fact-
specific assessment of ‘market power and market structure . . . to assess the restraint’s actual
effect’ on competition.” Amex, 138 S.Ct. at 2284 (internal brackets omitted) (quoting Copperweld
Corp. v. Indep. Tube Corp., 467 U.S. 752, 768 (1984). Recent cases suggest that the rule of
reason applies to any tying claim that “involves software that serves as a platform for third-party
applications.” Microsoft, 253 F.3d 34 at 89; see also id. at 95 (no per se claim where “the tying
product is software whose major purpose is to serve as a platform for third-party applications and
the tied product is complementary software functionality”).
“[T]he three-part burden-shifting test under the rule of reason is essentially the same” for
section 1 as for section 2 claims. Qualcomm, 969 F.3d at 991. First, plaintiff has “the initial
burden to prove that the challenged restraint has a substantial anticompetitive effect that harms
consumers in the relevant market.” Amex, 138 S. Ct. at 2284. That said, the Court need not
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 22 of 39
23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
“consider whether competition was in fact unreasonably restrained.” See Digidyne Corp. v. Data
Gen. Corp., 734 F.2d 1336, 1338 (9th Cir. 1984). Second, if “the plaintiff carries its burden, then
the burden shifts to the defendant to show a procompetitive rationale for the restraint.” Amex, 138
S. Ct. at 2284. Finally, “[i]f the defendant makes this showing, then the burden shifts back to the
plaintiff to demonstrate that the procompetitive efficiencies could be reasonably achieved through
less anticompetitive means.” Id.
b. Per Se Tying Analysis
Epic Games avers that Apple ties the iOS app distribution “product,” over which Apple has
economic power, to a separate “product” of the IAP system. Based upon the current record, the
Court concludes that Epic Games has not yet shown that the IAP system is a separate and distinct
service from iOS app distribution sufficient to constitute a “tie” under antitrust law.
Where the allegedly tied product is an essential ingredient of the overall “method of
business” with customers, courts view them as one product not as two tied together. Rick-Mik,
532 F.3d at 974 (quoting Will v. Comprehensive Accounting Corp., 776 F.2d 665, 670 n.1 (7th Cir
1985)). That is especially true where the allegedly separate products have always been integrated.
See id. at 975. As the Ninth Circuit has recognized, payment processing can be part of a single
integrated product. See id. at 974 (“The franchise and the method of processing credit transactions
are not separate products, but part of a single product (the franchise).”).
Here, the IAP system appears to be integrated with the App Store and, historically, to have
never been a separate product. If so, the construct of the IAP appears to reinforce the notion that
the App Store is a digital marketplace where developers on the App Store are able to structure
their business models however they choose.22 Many of these developers, like Epic Games,
22 Epic Games’ Best Buy and QuickBooks analogy misses the mark. Epic Games stated:
[W]hat Apple wants to do is to have the consumer go in to Best Buy, buy the Quick[B]ooks . . . pay for it there, that's fine. That's the app distribution. But then take it home, and every time you do your taxes or every time you close your books using Quick[B]ooks, after you have the product, to keep paying Best Buy every single time another 30 percent. They are reaching into subsequent transactions.”
(Dkt. No. 50 at 50-51.) With respect to video games, however, at least two significant distinctions
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 23 of 39
24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
structure these models so that the game or app is free, presumably to entice customers to download
the game or app initially, and only monetize the subsequent IAPs. The IAP system does not
appear to be a payment processor in the same way that Visa, Mastercard, or PayPal is a payment
processor; it is more akin to a link back to the App Store whereby the transaction must occur
within the digital confines of the App Store.23 The IAP system appears to have been created, in
part, to capture the value of a developer being on the digital shelf of the App Store which is owed
to Apple—either on the initial download, or in subsequent IAPs.24
Nevertheless, Epic Games raises serious questions about the existence of separate demand
for IAP-type services. Payment processing markets are ubiquitous outside of IAPs. Epic Games
offers indirect evidence of separate demand through analogy to these markets, including the
markets for the sale of physical goods sold through apps on the iOS platform. See Rick-Mik, 532
F.3d at 975. The experts disagree over whether the distinctions between IAP and these payment
exist. First, at a brick-and-mortar store, games were not distributed for free; that is, free-to-play games like Fortnite did not exist. In the digital context, consumers can obtain some games for free, and, under the license, no payment from Epic Games to Apple is due in that transaction.
Second, an analogous pre-digital marketplace transaction exists: namely, the sale of expansion packs, which could unlock additional content for base version of games, including new gameplay mechanics and functions. Consumers would initially purchase the base game from a brick-and-mortar store. Assuming the expansion pack was not available at the time of purchase of the base game, consumers were thereafter required to return to a store to purchase in a separate transaction the expansion pack—thereby unlocking this additional content in the base game.
IAP appears to operate analogously: the base version of a game is required to play, but IAP similarly unlocks additional content including new gameplay mechanics and functions. These analogous pre-digital transactions suggest that IAP is not merely a payment processor, as Epic Games contends, but rather an integrated part of the digital marketplace, permitting a prior historical business model in the gaming industry. The Court highlights that neither party discusses these analogous transactions, but the Court discloses that this conceptual similarity further colors the Court’s analysis, including the need for a more complete record. See also Amex, 138 S. Ct. at 2286-87 (discussing a transaction platform like the App Store, noting that it “facilitate[s] a single, simultaneous transaction between” two parties).
23 Indeed, it is the Court’s understanding that all video game digital distribution marketplaces require a consumer to similarly return to the marketplace to complete an IAP.
24 The Court notes that the conceptualization of the IAP system as integrated within the App Store may generally defeat a per se analysis. Microsoft suggests that perhaps the appropriate lens to view a tying claim involving innovative technological business models is under the rule of reason analysis, not under a per se tying analysis. See Microsoft, 253 F.3d 34 at 89-95.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 24 of 39
25
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
processing services actually impact consumer demand.25 (Compare Schmalensee Decl. ¶ 49 (Dkt.
No. 78) with Evans Reply Decl. ¶ 36 (Dkt. No. 88).) Moreover, Epic Games provides evidence
that developers have demanded their own in-app purchase payment processing services. (See
Evans Reply Decl. ¶ 45 (Dkt. No. 88).)
Epic Games further points to evidence in the record demonstrating that some customers
chose to use Epic Games’ payment processing service when given the choice with IAP. The
trouble with this argument is that it conflates competition on the merits with Epic Games’ goal of
avoiding Apple’s 30%. It is not surprising that some customers would choose competing payment
services if they provided lower prices offered only because of this non-payment. This does not
evidence separate demand for payment processing services, as much as a demand for alternatives
to Apple’s “integrated services” of iOS app distribution. When framed in this way, Epic Games’
argument is no more than a collateral attack on Apple’s App Store model, not a demonstration of
separate demand. In this respect, Epic Games’ strongest argument—left woefully underexplored
in the record—lies with competition on other features provided by IAP, such as customer service,
parental controls, and security.26 This evidence suggests that a more fully developed record could
25 The question of perspective underlies the tying claim as much as the monopolization
claim. In Rick-Mik, the court found that “[t]he relevant ‘purchaser’ is the franchisee (not the general consumer)” for purposes of separate demand for credit card processing services. 532 F.3d at 975. Here, the equivalent of the franchisee is the developer, which may demonstrate stronger “separate demand” for payment processing services than the user who makes the purchases.
26 Epic Games shows that at least some developers have demanded separate payment processing services based on these features, independent of Apple’s 30%. For example, Epic Games claims the CEO of the company “Hey,” which provides email service, made the following statement months before Epic Games’ motion:
[A]s the owner of a business, this isn’t just about money. Money grabs the headlines, but there’s a far more elemental story here. It’s about the absence of choice, and how Apple forcibly inserts themselves between your company and your customer. . . .
When Apple forces companies to offer In App Purchases in order to be on their platform, they also dictate the limits to which you can help your customer. This has a detrimental impact on the customer experience, and your relationship with your customer. It can flat out ruin an interaction, damage your reputation, and it can literally cost you customers. It prevents us from providing exceptional customer service when someone who uses our product needs help.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 25 of 39
26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
plausibly show demand for a separate product.
Should Epic Games satisfy the “purchaser demand” test for finding distinct products, it
may prevail on the remaining elements under the per se tying analysis. While Apple claims that it
does not “tie” IAP to iOS app distribution because developers may choose other business models,
it does not dispute that its App Store Review Guidelines require the IAP system’s use for IAPs as
a condition of app distribution. (See Schiller Decl. ¶¶ 5, 7, 33, 41 (Dkt. No. 74).) This
requirement manifests the coercion, that is, developers who offer IAP must do so on Apple’s
terms. Apple also does not dispute that it holds market power in the iOS app distribution market
and that the alleged tie affects a substantial volume of commerce in in-app payment processing.
Accordingly, Epic Games raises serious questions with regard to per se tying, but fails to
demonstrate the likelihood of success due to lack of evidence of “purchaser demand” for IAP
processing service separate from the “integrated service” of app distribution. 27
c. Rule of Reason Analysis
The rule of reason analysis is more fact specific than the per se analysis. Here, the first
element focuses on the harm to competition and consumers. Epic Games errs by focusing on harm
to competitors, and for that reason has not sustained its burden at this juncture.28
(Evans Reply Decl. ¶ 45 & n.40 (Dkt. No. 88).) This statement suggests that the IAP dispute is not simply about Apple’s fee, but also about whether “the world’s largest company [gets] to decide how millions of other businesses can interact with their own customers.” See Jason Fried, “Our CEO’s take on Apple’s App Store payment policies, and their impact on our relationship with our customers,” HEY (June 19, 2020), available at https://hey.com/apple/iap/.
27 Commentators have suggested that separate demand is a “threshold requirement” for separate products, following which defendant may show that the products are nevertheless a single product due to their being an “integrated service.” See Philip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application ¶ 1743 (4th Ed. 2020). United States v. Microsoft Corp., 147 F.3d 935, 948-49 (D.C. Cir. 1998), lays out the general requirements for an integrated service.
28 Nevertheless, for the same reasons as described for separate demand under the per se analysis, the Court can envision a plausible case for anticompetitive effect given the serious questions referenced above regarding Apple’s IAP restrictions and whether they reduce consumer choice or increase price due to exclusionary conduct. See Qualcomm, 969 F.3d at 990. Competitors could conceivably provide equal or superior services than IAP— better security, better customer service, and better parental controls. Moreover, Epic Games may be able to prove anticompetitive effects even if it cannot show separate products: Microsoft suggests that the separate-products test “is a rough proxy for whether a tying arrangement may [be] . . . unsuited to per se condemnation,” not a determination of ultimate efficiency. 253 F.3d at 87.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 26 of 39
27
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
Even if it had, Apple, of course, offers a procompetitive justification consistent with step
two of the three-part burden shifting analysis. Apple claims that the IAP provides the business
mechanism for it to be paid for the App Store given its support of 1.8 million apps, of which 84
percent are free, to 1.5 billion Apple devices and 900 million iPhone users. Apple claims that IAP
also provides: (1) a “centralized, convenient way” to transact online, (2) security and fraud
protection, (3) refunds and customer support from Apple, (4) parental controls, and (5)
comprehensive list of purchases,” all of which are facially reasonable. (Schiller Decl. ¶ 36 (Dkt.
No. 74); see also Schmalensee Decl. ¶ 29 (Dkt. No. 78).) That developers may not want to pay a
commission or licensing fee does not necessarily translate to antitrust behavior.
Under a rule of reason analysis, the burden shifts back to Epic Games to demonstrate that
the “procompetitive efficiencies could be reasonably achieved through less anticompetitive
means.” Here, the record is not fully developed, and mixed, at best. Competitors could
conceivably provide equal or superior services. Indeed, it is entirely plausible that app developers
could provide better refunds and customer support for goods purchased through their own apps
than can Apple, or not. As noted, the sale of physical goods sold on the iOS already uses separate
payment processors or mechanisms outside of the Apple IAP system. On the other hand, Apple
has produced evidence that its security features are more effective than its competitors, a basis on
which it competes. Thus, the dispute likely comes down to whether these features and Apple’s
monetization can be achieved through less anti-competitive means. The record on these issues is
thin, as is any briefing on the method of proof given the frontier on which the questions sit.
4. Summary
For the reasons set forth above, Epic Games has shown that serious questions exist with
respect to its section 1 and section 2 claims against Apple but has not proven a likelihood of
success on the merits on this record.
B. Irreparable Harm
As the Court stated in the temporary restraining order: the issue of irreparable harm
focuses on the harm caused by not maintaining the status quo, as opposed to the separate and
distinct element of a remedy under the likelihood of success factor. Here, once again, the Court’s
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 27 of 39
28
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
evaluation is guided by the general notion that “self-inflicted wounds are not irreparable injury.”
Al Otro Lado v. Wolf, 952 F.3d 999, 1008 (9th Cir. 2020) (quoting Second City Music, Inc. v. City
of Chicago, 333 F.3d 846, 850 (7th Cir. 2003)). Further courts generally decline to find
irreparable harm that “results from the express terms of [the] contract.” See Salt Lake Tribune
Publ’g Co., LLC v. AT&T Corp., 320 F.3d 1081, 1106 (10th Cir. 2003) (no irreparable harm
where the alleged harm “results from the express terms of [the] contract”). Quite simply,
irreparable harm is harm or injury that cannot be repaired.
1. Fortnite
Epic Games contests the Court’s prior determination with respect to Fortnite, namely that
no irreparable harm exists where Epic Games chose to breach its agreements with Apple in
enacting its own direct IAP system. Epic Games cites to precedent involving affirmative defenses
to argue that the Court should not aid in the enforcement of contracts that are anti-competitive and
violative of antitrust laws. See generally McMullen v. Hoffman, 174 U.S. 639, 654 (1899) (“The
authorities from the earliest times to the present unanimously hold that no court will lend its
assistance in any way towards carrying out the terms of an illegal contract.”); Memorex Corp. v.
Int’l Bus. Mach. Corp., 555 F.2d 1379,1383 (9th Cir. 1977) (“[Courts should] continue to side
with the goal of vigorous enforcement of our antitrust laws.”); Perma Life Mufflers, Inc. v. Int’l
Parts Corp., 392 U.S. 134, 139 (1968) (“[T]he purposes of the antitrust laws are best served by
insuring that private action will be an ever-present threat to deter anyone contemplating business
behavior in violation of the antitrust laws.”); Kaiser Steel Corp. v. Mullins, 455 U.S. 72, 83-84
(1982) (enforcement of “private agreements” is subject to “the restrictions and limitations of the
public policy of the United States”). Epic Games concedes that these affirmative defenses are
procedurally not at issue. Instead, it claims it should “not be penalized for defying Apple’s
monopolistic edicts” (Mot. at 23 (Dkt. No. 61 at 30)), and that the Court should proactively extend
the principle in support of Epic Games’ proffer of irreparable injury.
Epic Games further avers that ongoing harm continues to its reputation, the Fortnite
gaming community, and its ongoing ambitions in the creation of a metaverse. In support, Epic
Games introduces declarations attesting to a 60% decline in the number of iOS users in Fortnite,
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 28 of 39
29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
and that those who continue to play are doing so for significantly fewer hours per week, given that
these players are stuck on earlier version of the game and unable to play with other individuals.
(Sweeney Decl. ¶ 22 (Dkt. No. 65).) Next, Epic Games provides records reflecting customer
complaints and online comments about the unavailability of Fortnite on the iOS platform. (Id. ¶¶
25-26, Exs. E-F (Dkt. Nos. 65-5, 65-6); Byars Decl., Ex. Q (Dkt. No. 61-18).) Indeed, Epic
Games argues that “[m]any of these customers blame Epic [Games] for being cut off from access
to Fortnite.” (Mot. at 27 (Dkt. No. 61 at 34).) Finally, Epic Games includes declarations attesting
to difficulty in creating and sustaining a metaverse in the Fortnite community given that it is no
longer on the iOS platform.
Epic Games does not persuade. The cited cases are singularly premised on the fact that the
consequences from a breach of contract in which the parties are seeking to escape are actually in
violation of antitrust laws in the United States.29 As discussed, the Court has made no such
finding in this Order as to Epic Games’ likelihood of success on the merits beyond only finding
serious questions as to the merits. This is especially so where the alleged monopolistic practices
and conduct concern innovative technology platforms without analogous prior precedent.
Moreover, the Court considers countervailing interests in ensuring that antitrust laws are not
otherwise stretched into areas that are beyond what is required or contemplated. See Kelly v.
Kosuga, 358 U.S. 516, 519 (1959) (“Obviously . . . federal courts should not be quick to create a
policy of nonenforcement of contracts beyond that which is clearly the requirement of the
Sherman Act.”); Germon v. Times Mirror Co., 520 F.2d 786, 788 (9th Cir. 1975) (“The purposes
of the antitrust laws deal with promoting competition, not with extending unsatisfactory
contractual relationships beyond their stipulated periods of effectiveness.”).
In short, Epic Games cannot simply exclaim “monopoly” to rewrite agreements giving
29 Epic Games’ citation to Acquaire v. Canada Dry Bottling Co., 24 F.3d 401 (2d Cir.
1994) is markedly distinguishable. As Apple correctly notes, Acquaire involved a defendant who made after-the-fact changes to its policies, did not even comply with its own stated policy, and the plaintiffs made a showing that they would be driven out of business absent an injunction. Id. at 412. None of these facts are present, where Apple has maintained the same policies since the inception of the App Store, and there is no evidence in the record that Epic Games will be driven out of business based on the unavailability of Fortnite on the iOS platform.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 29 of 39
30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
itself unilateral benefit. Its other identified bases: damage to its reputation30 and the Fortnite
gaming community cannot constitute irreparable harm where such harm flows from Epic Games’
own actions and its strategic decision to breach its agreements with Apple.31 While consumers are
feeling the impact of this litigation, the fact remains: these are business disputes. A putative class
action on behalf of all developers on these exact same issues was already in progress when Epic
Games breached the agreements. See Cameron, 4:19-cv-03074-YGR. Yet, Epic Games has never
adequately explained its rush, other than its disdain for the situation. The current predicament is
of its own making. See Second City Music, 333 F.3d at 850 (“Only the injury inflicted by one’s
adversary counts for this purpose.”).32
Epic Games remains free to maintain its agreements with Apple in breach status as this
litigation continues and ignore what the Seventh Circuit recognized in Second City Music: “[t]he
sensible way to proceed is for [Epic Games to comply with the agreements and guidelines] and
continue to operate while it builds a record.” Id. There is no loss of face if one’s goal is to protect
its consumers, the Fortnite player base. To assist, the Court even offered to require the 30% be
placed in escrow pending resolution of the trial which Epic Games flatly rejected.33 The refusal to
30 Even reviewing the record before the Court, the Court is not persuaded that Epic Games
has suffered reputational harm. Epic Games unleashed a pre-planned and scorching marketing campaign against Apple following its breach of the operating agreements and guidelines. As the Court noted at oral arguments, if anything, it appears Epic Games’ actions have only increased its reputation in the wider community.
31 It is further difficult to conceive how Epic Games’ own ongoing ambitions in the creation of a metaverse would create a basis for a finding of irreparable injury.
32 Indeed, Second City Music is illustrative for why Epic Games’ actions cannot constitute irreparable harm. In Second City Music, the plaintiff-appellant challenged a city ordinance as unconstitutionally vague—not merely violative of a statutory regime. 333 F.3d at 847. The Seventh Circuit found that “some real injury may lurk beneath the surface . . . but evaluating this possibility requires evidence so far missing from the record.” Id. at 850. Likewise, as discussed, the record here is inadequate for the Court to conclude that the agreements and guidelines are violative of antitrust laws such that Epic Games truly has irreparable harm as to Fortnite.
33 As made apparent at the oral argument, Apple’s form letter purportedly banning Epic Games from the iOS platform for a one-year period is no barrier to Epic Games’ return to the iOS platform during the pendency of this litigation. (Dkt. No. 111 at 83-84.) As Apple stated at the hearing, Epic Games is able to return to the App Store under the Court’s supervision provided that Epic Games complies with the relevant agreements and guidelines and further pays Apple its
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 30 of 39
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
do so suggests Epic Games is not principally concerned with iOS consumers, but rather, harbors
other tactical motives. Certainly, no technical issue exists. Epic Games admits that the
technology exists to “fix” the problem easily by deactivating the “hotfix.” Thus, given the totality
of these circumstances, the Court can easily find that the injury Epic Games “incurs by following a
different course is of its own choosing.” Second City Music, 333 F.3d at 850. It is self-harm
caused by self-help.34 Accordingly, Epic Games has failed to demonstrate irreparable harm as to
Fortnite and the games under the Epic Games developer account.
2. Epic Affiliates
By contrast, with respect to Unreal Engine and the Epic Affiliates, the Court concludes that
Epic Games has made a sufficient showing as to the irreparable harm. As the Court previously
found: Apple is hard-pressed to dispute that even if Epic Games succeeded on the merits, it could be too late to save all the projects by third-party developers relying on the engine that were shelved while support was unavailable. Indeed, such a scenario would likely lead to nebulous, hard-to-quantify questions, such as, how successful these other projects might have been, and how much in royalties would have been generated, much less the collateral damage to the third-party developers themselves
Epic Games, 2020 WL 5073937, at 4 (Dkt. No. 48 at 6). Apple does not challenge these prior
findings. Indeed, there is ample evidence in the record demonstrating: (1) that the removal of
developer tools could have significant irreparable harm to Unreal Engine and to Epic Games and
its affiliates; and (2) that Apple’s threat to revoke developer tools (SDKs) from Unreal Engine is
already having a negative impact on Unreal Engine. (See Sweeney Decl. ¶¶ 38, 40 (Dkt. No. 65);
Penwarden Decl. ¶¶ 8-13 (Dkt. No. 64); Byars Decl., Exs. R (Dkt. No. 61-19), S ¶¶ 3-4 (Dkt. No.
61-20).) In this regard, Epic Games could not otherwise be made whole even if victorious at trial.
commission on the IAP that occurred after the breach on August 13, 2020. (Id.)
34 Epic Games’ argument that people are incorrectly blaming Epic Games for the unavailability of Fortnite on the iOS platform ignores the record in this matter. The decision of whether to return Fortnite to the iOS platform during the pendency of this litigation rests with Epic Games—not Apple. Indeed, Epic Games has a choice, and it has exercised this choice by weighing its own beliefs and principles as to the alleged illegality of the Apple agreements and guidelines above its interest in continuing to provide iOS users with access to Fortnite. As noted, Epic Games is free to make that choice; but it is Epic Games’ choice nonetheless.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 31 of 39
32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
See trueEX, LLC v. MarkitSERV Ltd., 266 F.Supp.3d 705, 728 (S.D.N.Y. 2017) (“trueEX is likely
also to suffer irreparable harm . . . [as] some [customers] have threatened to stop doing business
with trueEX . . . . Another client sought to accelerate a number of planned trades . . . suggesting
that the client did not believe it could do business with trueEX in the future”).
Instead, Apple advances three arguments: (1) Apple has a well-established practice of
removing affiliated developer accounts and developer tools (SDKs) in similar circumstances based
on broad language in the agreements and guidelines (Schiller Decl. ¶¶ 54-55, Ex. C at 2); (2) the
harm to Unreal Engine is also self-inflicted harm and cannot be irreparable harm; and (3) Epic
Games and/or its affiliates could insert and distribute secret code in Unreal Engine and the other
applications remaining on the iOS and macOS platforms.
Apple does not persuade. It is clear from the record that Apple’s long-standing practice of
removing affiliated accounts based on broad language regarding termination in the relevant
agreements and guidelines would generally be permissible. However, as applied to the specific
facts, the Court concludes that this matter presents an exception to the ordinary practices. The
Court notes that the totality of facts is not overwhelming for either side, but leans towards Epic
Games on this topic. On the one hand, facts weighing in favor of Apple include: the agreements
are at-will; the developer accounts for both Epic International and Epic Games list the same
taxpayer identification number; a single individual is listed as the registered account holder for
both accounts; a single credit card paid for both accounts; share the same test devices; the accounts
were renewed within a minute of each other; and Epic International receives customer payments
made by iOS Fortnite users who are playing outside the United States. On the other hand, facts
weighing in favor of Epic Games, Epic International, and other Epic Affiliates include: each have
separate agreements with Apple; each of the Epic Affiliates pays separate consideration (i.e.
annual developer fees); all agreements were renewed at separate times; the Epic Affiliates’
agreements have not otherwise been breached; and Epic International has been represented by
Epic Games to be a different legal entity despite overlapping financial accounts (e.g. credit cards,
taxpayer identification number, etc.). Additionally, despite the inclusion of broad termination
language in the agreements, the relevant agreement governing developer tools (SDKs), the Apple
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 32 of 39
33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
Xcode and Apple SDKs Agreement, is a fully integrated document that explicitly excludes the
developer program license agreement.
Although it is a close question, the Court finds that, with respect to access to the developer
tools (SDKs), Apple’s reaching into separate agreements with separate entities appears to be
retaliatory, especially where these agreements have not been otherwise breached. Indeed, the form
letter first issued by Apple in response to Epic Games’ breach does not mention Unreal Engine or
the possibility of Apple revoking the developer tools (SDKs). However, after the commencement
of this lawsuit, Apple apparently sent a more personalized letter specifically identifying and
targeting Unreal Engine as a consequence of Epic Games’ breach. Significantly, the letter does
not otherwise identify or list any other specific application or software at risk from Epic Games or
any of the Epic Affiliates. (See Grant Decl. ¶ 27, Ex. C (Dkt. No. 63-3 at 3-4) (“You will also lose
access to the following programs, technologies, and capabilities: . . . Engineering efforts to
improve hardware and software performance of Unreal Engine on Mac and iOS hardware;
optimize Unreal Engine on the Mac for creative workflows, virtual sets and their CI/Build
Systems; and adoption and support of ARKit features and future VR features into Unreal Engine
by their XR team.”).) The subtext of the letter where one, and only one, significant product is
mentioned evidences that Apple was impermissibly pressuring and retaliating against Epic Games
and the Epic Affiliates on Unreal Engine product.
Apple’s remaining two arguments are also without merit. Apple has not shown that Epic
Games’ breach with respect to Fortnite results in a breach of agreements with Epic International
or the Epic Affiliates. In the normal course of business, parties can terminate such at will
agreements pursuant to their express terms. Here though, Apple reaches beyond these separate
agreements to inflict harm, or pressure, upon Epic Games and the Epic Affiliates. In this regard,
the injury cannot be said to be self-inflicted.
Finally, the Court is not persuaded by Apple’s exaggerated claims that Epic Games would
insert hidden or malicious code into Unreal Engine or its products to damage the iOS platform.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 33 of 39
34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
The record is devoid of any evidence to support such a finding or inference.35 To the extent any
valid concern exists, however, it is easily remedied by narrowing the scope of the injunction to
permit Epic Affiliates’ continued access to the developer tools (SDKs) and to the App Store only
so long as such applications and the Epic Affiliates remain in continued compliance with the terms
of the relevant agreements and guidelines.
The Court notes that expanded briefing by Apple on the agreements and its historical
practice has made this a closer question than was presented earlier. On balance, however, and in
light of the foregoing analysis, the Court concludes that ongoing irreparable harm and significant
potential irreparable harm to Unreal Engine exist absent a preliminary injunction.
C. Balance of Equities
1. Fortnite
As the Court stated in its prior order: The battle between Epic Games and Apple has apparently been brewing for some time. It is not clear why now became so urgent. The Cameron case which addresses the same issues has been pending for over a year, and yet, both Epic Games and Apple remain successful market players. If plaintiffs there, or here, prevail, monetary damages will be available and injunctive relief requiring a change in practice will likely be required. Epic Games moves this Court to allow it to access Apple’s platform for free while it makes money on each purchase made on the same platform. While the Court anticipates experts will opine that Apple’s 30 percent take is anti-competitive, the Court doubts that an expert would suggest a zero percent alternative. Not even Epic Games gives away its products for free.
35 Further, to do so would be tactically disastrous for Epic Games and its affiliates as it
would prove Apple’s point with respect to its need to maintain its walled garden or closed platform to protect iOS consumers against security attacks.
Moreover, Apple’s arguments—that Mr. Sweeney’s statements to Apple announcing the breach reflect a risk to the iOS platform—do not persuade. Mr. Sweeney states that should Apple reject its demands for the ability to introduce a separate app market and use a different payment processor, then Epic Games will be in conflict with Apple on “a multitude of fronts - creative, technical, business, and legal - for so long as it takes to bring about change.” (Sweeney Decl., Ex. D (Dkt. No. 65-4).) These statements appear hyperbolic, but even a generous reading in Apple’s favor does not reflect any intent to harm the iOS platform with respect to the Unreal Engine or the other applications that are under other affiliates’ developer accounts. Indeed, Unreal Engine does not even utilize the App Store or itself offer IAP, as it is a graphics engine available to developers on computer platforms. It is hard to determine how the Unreal Engine would or could be used to try to affect such changes as described in the above cited correspondence.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 34 of 39
35
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
Thus, in focusing on the status quo, the Court observes that Epic Games strategically chose to breach its agreements with Apple which changed the status quo. No equities have been identified suggesting that the Court should impose a new status quo in favor of Epic Games.
Epic Games, 2020 WL 5073937, at 4 (Dkt. No. 48 at 6).
The Court’s prior findings remain instructive in addressing the parties’ arguments: Epic
Games advances two arguments for why the balance of equities tilt sharply in its favor with
respect to Fortnite. First, Epic Games dismisses Apple’s concern that an injunction with respect
to Fortnite would set off a rash of other developers breaching their agreements, and asserts that
any harm to Apple would be limited to loss of commissions for a short time, “which is harm easily
compensable by damages.” (Mot. at 30 (Dkt. No. 61 at 37).) Second, the balance tilts towards
Epic Games where the injunction seeks to ensure that Apple complies with antitrust laws.
Epic Games does not persuade on either of these two bases. As discussed, Epic Games has
not made a preliminary showing of the likelihood of success on its claim. Therefore, it is not yet
established that such an injunction reinstating Fortnite would issue in compliance with antitrust
laws. Moreover, Epic Games’ argument with respect to damages only demonstrates that the harm
to Fortnite is not irreparable. As Epic Games states, the loss of commissions to Apple would be
for a short duration and would be “easily compensable . . . .” The converse is also true.
Finally, the Court finds that the balance of equities weighs toward Apple where Epic
Games breached both its agreements and the guidelines, and an injunction would potentially
incentivize similar breaches among developers. Epic Games does not dispute that it breached its
agreements. Nor is a breach required to maintain or even commence this lawsuit as reflected by
the fact that the named plaintiffs in Cameron did not breach their agreements. As explained
herein, Epic Games can similarly proceed. The Court declines to incentivize breaches of contracts
where the legality of those provisions has not yet been conclusively or presumptively determined
to be illegal.
In sum, no equities have been identified suggesting that the Court should impose a new
status quo in favor of Epic Games. The balance of equities tilts sharply toward Apple on the issue
of Fortnite.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 35 of 39
36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
2. Epic Affiliates
By contrast, the Court finds that the balance of equities weighs in favor of Epic Games and
the Epic Affiliates, including as to Unreal Engine, on the issue of continued access to developer
tools and the App Store for the Epic Affiliates. As the Court previously found: [W]ith respect to the Unreal Engine and the developer tools, the Court finds the opposite result. In this regard, the contracts related to those applications were not breached. Apple does not persuade that it will be harmed based on any restraint on removing the developer tools. The parties’ dispute is easily cabined on the antitrust allegations with respect to the App Store. It need not go farther. Apple has chosen to act severely, and by doing so, has impacted non-parties, and a third-party developer ecosystem. In this regard, the equities do weigh against Apple.
Epic Games, 2020 WL 5073937, at 4 (Dkt. No. 48 at 6). The Court finds that this analysis
remains unchanged. The only equity that Apple has identified concerns an alleged potential
“trojan horse” or insertion of malicious code by Epic Games or the Epic Affiliates. The Court
rejects this argument for the same reasons discussed under the irreparable harm factor. The
modification in the preliminary injunction to ensure continued compliance with the operating
agreements and guidelines addresses this issue.
Apple’s aggressive targeting of separate contracts in an attempt to eradicate Epic Games
and its affiliates fully from the iOS platform was unnecessary and imperiled a thriving third-party
developer ecosystem. Providing continued access for the Epic Affiliates to developer tools and
the App Store preserves the status quo. Accordingly, the balance of equities tilts sharply toward
Epic Games on the issue of continued access to developer tools and the App Store for the Epic
Affiliates.
D. Public Interest
“[T]he public interest inquiry primarily addresses the impact on non-parties rather than
parties.” HiQ Labs, 938 F.3d at 1004 (internal quotation marks omitted). “The plaintiffs bear the
initial burden of showing that the injunction is in the public interest.” Stormans, Inc. v. Selecky,
586 F.3d 1109, 1139 (9th Cir. 2009). The current briefing does not change significantly the
parties’ initial submissions or the Court’s findings, which it reaffirms:
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 36 of 39
37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
With respect to the gaming requests, the Court recognizes based on the numerous internet postings and comments submitted in the record that Fortnite players are passionate supporters of the game, and eagerly anticipate its return to the iOS platform. The Court further recognizes that during these coronavirus pandemic (COVID-19) times, virtual escapes may assist in connecting people and providing a space that is otherwise unavailable. However, the showing is not sufficient to conclude that these considerations outweigh the general public interest in requiring private parties to adhere to their contractual agreements or in resolving business disputes through normal, albeit expedited, proceedings. See S. Glazer’s Distrib. of Ohio, LLC v. Great Lakes Brewing Co., 860 F.3d 844, 853 (6th Cir. 2017) (declining to enjoin termination of contract according to its terms because the “public has a strong interest in holding private parties to their agreements”). With respect to the Unreal Engine and the developer tools, the calculus changes. The record shows potential significant damage to both the Unreal Engine platform itself, and to the gaming industry generally, including on both third-party developers and gamers. The public context in which this injury arises differs significantly: not only has the underlying agreement not been breached, but the economy is in dire need of increasing avenues for creativity and innovation, not eliminating them. Epic Games and Apple are at liberty to litigate against each other, but their dispute should not create havoc to bystanders. . . .
Epic Games, 2020 WL 5073937, at 4 (Dkt. No. 48 at 7).
As to Fortnite, nothing has changed in the prior analysis. The Court has empathy for
Fortnite players regarding the continued unavailability of the game on the iOS platform. This is
especially so during these continued difficult times that is the COVID-19 pandemic era, where
gaming and virtual worlds are both social and safe. However, there is significant public interest in
requiring parties to adhere to their contractual agreements or in resolving business disputes
through the normal course.36 Thus, the public interest factor weighs in favor of Apple as to
Fortnite.
The record has also remained the same as to the Epic Affiliate accounts and Unreal Engine.
The record demonstrates potential significant damage to both developers and gamers absent the
issuance of a preliminary injunction. Indeed, many games on iOS and on other platforms,
36 Epic Games cites authority that it is not in the public interest to enforce illegal contracts.
Of course, these cases presuppose a showing on the illegality of the contract, which Epic Games has not yet done, and are therefore inapposite.
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 37 of 39
38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
including Fortnite competitor PUBG, are built using Unreal Engine and rely on the engine
remaining compatible with future Apple software updates. Without the ability to update the
underlying engine for these and other games, the gaming industry built upon developers and
fervent consumers, including iOS consumers, will be unnecessarily impacted. Moreover, the need
for increasing avenues for creativity and innovation has not abated since the prior order. If
anything, the continued ongoing pandemic has demonstrated the imperative for substantial digital
and virtual innovation. Epic Games and Apple are at liberty to litigate this action for the future of
the digital frontier, but their dispute should not create havoc to bystanders. Thus, the public
interest weighs overwhelmingly in favor of Unreal Engine and the Epic Affiliates.
E. Weighing of Factors
In sum, the Court finds that based upon the record before it, the Winter factors weigh
against granting a preliminary injunction based on Epic Games’ requests as to Fortnite and other
games and in favor of granting a preliminary injunction order as the Epic Affiliates effected
developer tools, including as to Unreal Engine.
IV. CONCLUSION
Accordingly, for the foregoing reasons, the Court GRANTS IN PART and DENIES IN PART
the motion for preliminary injunction.
THEREFORE, APPLE AND ALL PERSONS IN ACTIVE CONCERT OR PARTICIPATION WITH
APPLE, ARE PRELIMINARILY ENJOINED from taking adverse action against the Epic Affiliates
with respect to restricting, suspending or terminating the Epic Affiliates from the Apple’s
Developer Program, on the basis that Epic Games enabled IAP direct processing in Fortnite
through means other than the Apple IAP system, or on the basis of the steps Epic Games took to
do so. This preliminary injunction shall remain in effect during the pendency of this litigation
unless the Epic Affiliates breach: (1) any of their governing agreements with Apple, or (2) the
operative App Store guidelines. This preliminary injunction SUPERSEDES the prior temporary
restraining order.
For the reasons set forth above, this preliminary injunction is EFFECTIVE IMMEDIATELY
and will remain in force until the disposition of this case. Neither party has requested a security
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 38 of 39
39
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Uni
ted
Sta
tes
Dis
tric
t Cou
rt
Nor
ther
n D
istr
ict o
f C
alif
orni
a
bond and the Court finds that none is necessary as contemplated under Fed. R. Civ. P. 65(c). See
Connecticut Gen. Life Ins. Co. v. New Images of Beverly Hills, 321 F.3d 878, 882 (9th Cir. 2003)
(“The district court is afforded wide discretion in setting the amount of the bond, . . . and the bond
amount may be zero if there is no evidence the party will suffer damages from the injunction.”).
This Order terminates Docket Number 61.
IT IS SO ORDERED.
Dated: October 9, 2020 YVONNE GONZALEZ ROGERS UNITED STATES DISTRICT JUDGE
Case 4:20-cv-05640-YGR Document 118 Filed 10/09/20 Page 39 of 39
IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA
UNITED STATES OF AMERICA U.S. Department of Justice 950 Pennsylvania Avenue NW Washington, DC 20530
STATE OF ARKANSAS 323 Center Street, Suite 200 Little Rock, AR 72201 STATE OF FLORIDA PL-01, The Capitol Tallahassee, FL 32399 STATE OF GEORGIA 40 Capitol Square SW Atlanta, GA 30334 STATE OF INDIANA 302 West Washington Street IGCS – 5th Floor Indianapolis, IN 46204 COMMONWEALTH OF KENTUCKY 1024 Capital Center Drive, Suite 200 Frankfort, KY 40601 STATE OF LOUISIANA 1885 North Third Street Baton Rouge, LA 70802 STATE OF MISSISSIPPI P.O. Box 220 Jackson, MS 39205 STATE OF MISSOURI P.O. Box 899 Jefferson City, MO 65102 STATE OF MONTANA P.O. Box 200151 Helena, MT 59620
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 1 of 64
2
STATE OF SOUTH CAROLINA 1000 Assembly Street Rembert C. Dennis Building P.O. Box 11549 Columbia, SC 29211-1549
and
STATE OF TEXAS P.O. Box 12548 Austin, TX 78711
Plaintiffs,
v.
GOOGLE LLC 1600 Amphitheatre Parkway Mountain View, CA 94043
Defendant.
COMPLAINT
The United States of America, acting under the direction of the Attorney General of the
United States, and the States of Arkansas, Florida, Georgia, Indiana, Kentucky, Louisiana,
Mississippi, Missouri, Montana, South Carolina, and Texas, acting through their respective
Attorneys General, bring this action under Section 2 of the Sherman Act, 15 U.S.C. § 2, to
restrain Google LLC (Google) from unlawfully maintaining monopolies in the markets for
general search services, search advertising, and general search text advertising in the United
States through anticompetitive and exclusionary practices, and to remedy the effects of this
conduct.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 2 of 64
3
I. NATURE OF THIS ACTION
1. Two decades ago, Google became the darling of Silicon Valley as a scrappy
startup with an innovative way to search the emerging internet. That Google is long gone. The
Google of today is a monopoly gatekeeper for the internet, and one of the wealthiest companies
on the planet, with a market value of $1 trillion and annual revenue exceeding $160 billion. For
many years, Google has used anticompetitive tactics to maintain and extend its monopolies in the
markets for general search services, search advertising, and general search text advertising—the
cornerstones of its empire.
2. As in many other businesses, a general search engine must find an effective path
to consumers for it to be successful. Today, general search engines are distributed primarily on
mobile devices (smartphones and tablets) and computers (desktops and laptops). These devices
contain web browsers (software applications for accessing information on the internet) and other
“search access points” that call on a general search engine to respond to a user’s query. Over the
last ten years, internet searches on mobile devices have grown rapidly, eclipsing searches on
computers and making mobile devices the most important avenue for search distribution in the
United States.
3. For a general search engine, by far the most effective means of distribution is to
be the preset default general search engine for mobile and computer search access points. Even
where users can change the default, they rarely do. This leaves the preset default general search
engine with de facto exclusivity. As Google itself has recognized, this is particularly true on
mobile devices, where defaults are especially sticky.
4. For years, Google has entered into exclusionary agreements, including tying
arrangements, and engaged in anticompetitive conduct to lock up distribution channels and block
rivals. Google pays billions of dollars each year to distributors—including popular-device
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 3 of 64
4
manufacturers such as Apple, LG, Motorola, and Samsung; major U.S. wireless carriers such as
AT&T, T-Mobile, and Verizon; and browser developers such as Mozilla, Opera, and UCWeb—
to secure default status for its general search engine and, in many cases, to specifically prohibit
Google’s counterparties from dealing with Google’s competitors. Some of these agreements also
require distributors to take a bundle of Google apps, including its search apps, and feature them
on devices in prime positions where consumers are most likely to start their internet searches.
5. Google’s exclusionary agreements cover just under 60 percent of all general
search queries. Nearly half the remaining queries are funneled through Google owned-and-
operated properties (e.g., Google’s browser, Chrome). Between its exclusionary contracts and
owned-and-operated properties, Google effectively owns or controls search distribution channels
accounting for roughly 80 percent of the general search queries in the United States. Largely as a
result of Google’s exclusionary agreements and anticompetitive conduct, Google in recent years
has accounted for nearly 90 percent of all general-search-engine queries in the United States, and
almost 95 percent of queries on mobile devices.
6. Google has thus foreclosed competition for internet search. General search engine
competitors are denied vital distribution, scale, and product recognition—ensuring they have no
real chance to challenge Google. Google is so dominant that “Google” is not only a noun to
identify the company and the Google search engine but also a verb that means to search the
internet.
7. Google monetizes this search monopoly in the markets for search advertising and
general search text advertising, both of which Google has also monopolized for many years.
Google uses consumer search queries and consumer information to sell advertising. In the United
States, advertisers pay about $40 billion annually to place ads on Google’s search engine results
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 4 of 64
5
page (SERP). It is these search advertising monopoly revenues that Google “shares” with
distributors in return for commitments to favor Google’s search engine. These enormous
payments create a strong disincentive for distributors to switch. The payments also raise barriers
to entry for rivals—particularly for small, innovative search companies that cannot afford to pay
a multi-billion-dollar entry fee. Through these exclusionary payoffs, and the other
anticompetitive conduct described below, Google has created continuous and self-reinforcing
monopolies in multiple markets.
8. Google’s anticompetitive practices are especially pernicious because they deny
rivals scale to compete effectively. General search services, search advertising, and general
search text advertising require complex algorithms that are constantly learning which organic
results and ads best respond to user queries; the volume, variety, and velocity of data accelerates
the automated learning of search and search advertising algorithms. When asked to name
Google’s biggest strength in search, Google’s former CEO explained: “Scale is the key. We just
have so much scale in terms of the data we can bring to bear.” By using distribution agreements
to lock up scale for itself and deny it to others, Google unlawfully maintains its monopolies.
9. Google’s grip over distribution also thwarts potential innovation. For example,
one company recently started a subscription-based general search engine that does not rely on
advertising profits derived from monetizing user information. Another, DuckDuckGo,
differentiates itself from Google through its privacy-protective policies. But Google’s control of
search access points means that these new search models are denied the tools to become true
rivals: effective paths to market and access, at scale, to consumers, advertisers, or data.
10. Google’s practices are anticompetitive under long-established antitrust law.
Almost 20 years ago, the D.C. Circuit in United States v. Microsoft recognized that
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 5 of 64
6
anticompetitive agreements by a high-tech monopolist shutting off effective distribution channels
for rivals, such as by requiring preset default status (as Google does) and making software
undeletable (as Google also does), were exclusionary and unlawful under Section 2 of the
Sherman Act.
11. Back then, Google claimed Microsoft’s practices were anticompetitive, and yet,
now, Google deploys the same playbook to sustain its own monopolies. But Google did learn
one thing from Microsoft—to choose its words carefully to avoid antitrust scrutiny. Referring to
a notorious line from the Microsoft case, Google’s Chief Economist wrote: “We should be
careful about what we say in both public and private. ‘Cutting off the air supply’ and similar
phrases should be avoided.” Moreover, as has been publicly reported, Google’s employees
received specific instructions on what language to use (and not use) in emails because “Words
matter. Especially in antitrust law.” In particular, Google employees were instructed to avoid
using terms such as “bundle,” “tie,” “crush,” “kill,” “hurt,” or “block” competition, and to avoid
observing that Google has “market power” in any market.
12. Google has refused to diverge from its anticompetitive path. Earlier this year,
while the United States was investigating Google’s anticompetitive conduct, Google entered into
agreements with distributors that are even more exclusionary than the agreements they replaced.
Also, Google has turned its sights to emerging search access points, such as voice assistants,
ensuring that they too are covered by the same anticompetitive scheme. And Google is now
positioning itself to dominate search access points on the next generation of search platforms:
internet-enabled devices such as smart speakers, home appliances, and automobiles (so-called
internet-of-things, or IoT, devices).
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 6 of 64
7
13. Absent a court order, Google will continue executing its anticompetitive strategy,
crippling the competitive process, reducing consumer choice, and stifling innovation. Google is
now the unchallenged gateway to the internet for billions of users worldwide. As a consequence,
countless advertisers must pay a toll to Google’s search advertising and general search text
advertising monopolies; American consumers are forced to accept Google’s policies, privacy
practices, and use of personal data; and new companies with innovative business models cannot
emerge from Google’s long shadow. For the sake of American consumers, advertisers, and all
companies now reliant on the internet economy, the time has come to stop Google’s
anticompetitive conduct and restore competition.
II. JURISDICTION, VENUE, AND COMMERCE
14. The United States brings this action pursuant to Section 4 of the Sherman Act,
15 U.S.C. § 4, to prevent and restrain Google’s violations of Section 2 of the Sherman Act,
15 U.S.C. § 2.
15. Plaintiffs Arkansas, Florida, Georgia, Indiana, Kentucky, Louisiana, Mississippi,
Missouri, Montana, South Carolina, and Texas by and through their respective Attorneys
General, bring this action in their respective sovereign capacities and as parens patriae on behalf
of the citizens, general welfare, and economy of their respective States under their statutory,
equitable, or common law powers, and pursuant to Section 16 of the Clayton Act,
15 U.S.C. § 26, to prevent and restrain Google’s violations of Section 2 of the Sherman Act,
15 U.S.C. § 2.
16. This Court has subject matter jurisdiction over this action under Section 4 of the
Sherman Act, 15 U.S.C. § 4, and 28 U.S.C. §§ 1331, 1337(a), and 1345.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 7 of 64
8
17. The Court has personal jurisdiction over Google; venue is proper in this District
under Section 12 of the Clayton Act, 15 U.S.C. § 22, and under 28 U.S.C. § 1391 because
Google transacts business and is found within this District.
18. Google is a limited liability company organized and existing under the laws of the
State of Delaware, and is headquartered in Mountain View, California. Google is owned by
Alphabet Inc., a publicly traded company incorporated and existing under the laws of the State of
Delaware and headquartered in Mountain View, California. Google engages in, and its activities
substantially affect, interstate trade and commerce. Google provides a range of products and
services that are marketed, distributed, and offered to consumers throughout the United States, in
the plaintiff States, across state lines, and internationally.
III. INDUSTRY BACKGROUND
A. Search Engines, Search Advertising, and General Search Text Advertising
19. In the early 1990s, computer scientists and entrepreneurs explored different ways
to search and index the growing number of internet sites. The first computer program or general
“search engine” that could perform this task was designed in 1990 by a student at McGill
University in Montreal and called “Archie.” Other early general search engines emerged, with
different methods of gathering, organizing, and presenting information about internet sites.
Google’s founders launched their research project “Backrub” on Stanford University’s network
in 1996.
20. Most modern general search engines use software to “crawl” the internet,
indexing webpages and the information within them. As Google explains, “The web is like an
ever-growing library with billions of books and no central filing system. We use software known
as web crawlers to discover publicly available webpages. Crawlers look at webpages and follow
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 8 of 64
9
links on those pages, much like you would if you were browsing content on the web. They go
from link to link and bring data about those webpages back to Google’s servers.”
21. When a search user enters a query into a general search engine, the software uses
algorithms to evaluate the relevance of information on any given webpage to the user’s query.
Depending on the query, some general search engines may also search selected proprietary
databases for pertinent information to offer additional “specialized” search results. The general
search engine then delivers the results on the SERP, with links to, and short descriptions of,
webpages the algorithm has curated and ranked. Sometimes, the general search engine will serve
ads with the search results.
22. Given the internet’s enormous breadth and constant evolution, establishing and
maintaining a commercially viable general search engine is an expensive process. Google’s
search index contains hundreds of billions of webpages and is well over 100,000,000 gigabytes
in size. Developing a general search index of this scale, as well as viable search algorithms,
would require an upfront investment of billions of dollars. The costs for maintaining a scaled
general search business can reach hundreds of millions of dollars a year.
23. General search engines are “one-stop shops” consumers can use to search the
internet for answers to a wide range of queries. The United States has only three general search
engines that crawl the internet: Google, Bing, and, to a lesser extent, privacy-focused search
provider DuckDuckGo. DuckDuckGo combines search results from different sources (including
Bing) depending on the search query. A fourth general search engine, Yahoo!, does not currently
crawl the internet and instead purchases search results from Bing.
24. Consumers can find certain specialized information online using sources other
than general search engines. For example, consumers can search retail marketplaces such as
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 9 of 64
10
Amazon or eBay to shop for products, or go to Expedia or Priceline to compare airfares. Search
sites that offer users a narrower, focused set of answers to queries are “specialized search
engines.” Specialized search engines are often able to give users deeper topical results than
general search engines by using specialized data or information gathered from users or supplied
by third parties.
25. Most general search engines do not charge a cash price to consumers. At least
one, Bing, even offers to pay consumers rewards for using its general search engine. That does
not mean, however, that these general search engines are free. When a consumer uses Google,
the consumer provides personal information and attention in exchange for search results. Google
then monetizes the consumer’s information and attention by selling ads.
26. Search advertising first appeared on Google in 2000. During that same year
Google launched AdWords, its buying platform for search ads. Two years ago, Google rebranded
AdWords as Google Ads.
27. To sell ads on its SERP, in 2002, Google adopted auctions for keywords;
advertisers would bid on selected keywords, and when those keywords arose in a query, the
winning bidder’s ad was shown. At that time, Google also started using a compensation scheme
where advertisers pay only when the user clicks on the ad, known as cost-per-click pricing. Some
SERP displayed multiple ads. Eventually, Google discovered that it could increase the number of
clicks—and its own profits—by ranking ads to promote those with greater relevance and
therefore higher expected click-through rates. To help determine placement of ads, Google still
uses a “quality score” based on various factors.
28. Advertisers use various types of ads to achieve different objectives. Marketers and
advertisers typically refer to a “purchase funnel” or “customer acquisition funnel” to describe the
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 10 of 64
11
average consumer’s various states of mind leading up to a potential purchase, and the type of
advertising most effective at each state. The following is an illustration of the purchase funnel:
Figure 1
29. Search ads enable advertisers to target potential customers based on keywords
entered by these users, at the exact moment users express interest in the topic of the queries. For
this reason, search ads are lower in the purchase funnel—closer to the consumer’s ultimate intent
to make a purchase—than other types of ads that are primarily intended to drive brand
awareness. The ability of search ads to provide advertising based on a consumer’s self-disclosed
interests, when the consumer is actively seeking information, makes search ads uniquely
valuable to advertisers.
30. Historically, general search engines such as Google sold only general search text
ads. General search text ads resemble the organic search results that appear on a SERP—what
Google refers to as the “10 blue links”—but with a subtle notation that they are “ads” or
“sponsored.” Google describes its text ads as follows:
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 11 of 64
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 12 of 64
13
32. Some specialized search providers also sell search ads. For example, advertisers
can buy specialized search ads for goods sold on Amazon, hotels presented on Expedia, and local
services listed on Yelp.
33. As the number of users of a general search engine grows, advertisers benefit
because they want their marketing campaigns to reach large groups of consumers. But users do
not benefit from indirect network effects in an equivalent way. As Google’s Chief Economist has
explained, “users do not decide which search engine to use based on the number of advertisers.”
34. Today, the search advertising business in the United States is enormous—over
$50 billion per year—and dominated by Google. Because of Google’s user base and scale, the
company’s search ads have become a “must have” for many advertisers. Advertising agencies
and larger companies often have entire groups that manage search advertising, mostly focused on
Google.
B. Importance of Scale
35. Scale is of critical importance to competition among general search engines for
consumers and search advertisers. Google has long recognized that without adequate scale its
rivals cannot compete. Greater scale improves the quality of a general search engine’s
algorithms, expands the audience reach of a search advertising business, and generates greater
revenue and profits.
36. The additional data from scale allows improved automated learning for algorithms
to deliver more relevant results, particularly on “fresh” queries (queries seeking recent
information), location-based queries (queries asking about something in the searcher’s vicinity),
and “long-tail” queries (queries used infrequently).
37. Scale is also important for search advertising because advertisers pay more to buy
ads from a search provider with a large audience of potentially interested customers. Google can
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 13 of 64
14
deliver enormous audiences, especially in mobile, which its competitors cannot. Google’s scale
also enables it to better discern which ads are most relevant for which queries.
38. Further, to recoup the large investment in creating and maintaining a general
search engine, scale is critical to generating the necessary revenues and profits. Even a
competitor that syndicates its search results from other general search engines must make
substantial investments to compete. The most effective way to achieve scale is for the general
search engine to be the preset default on mobile devices, computers, and other devices, as
described in more detail below.
C. General Search Engine Distribution and Default Status
39. Search is like many other businesses in that the owners of general search engines
can benefit greatly from a network of distributors to get their products to consumers. Distribution
of general search engines takes place primarily through search access points, such as browsers
and search apps, typically located on mobile devices and computers. More recently, searches
have become available on IoT devices.
40. General search service providers can enter into agreements with various
distributors, including computer and mobile-device manufacturers, cell phone carriers, and
browser developers, to secure preset default status on computer and mobile-device search access
points.
41. New computers and new mobile devices generally come with a number of
preinstalled apps and out-of-the-box settings. Computers and mobile devices generally have apps
preinstalled that include search access points, such as browsers, search apps and widgets, and
voice assistants. Mobile devices may also have hardware features—such as a home button
triggering a voice assistant—that a consumer can use to invoke apps with search functionality.
Each of these search access points can and almost always does have a preset default general
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 14 of 64
15
search engine. Being the preset default general search engine is particularly valuable because
consumers rarely change the preset default.
1. The Mobile Search Distribution Channel
42. With roughly 60 percent of searches, mobile devices represent the largest and,
over the last five years, fastest growing search distribution channel.
43. In the United States, Apple iOS devices—those running on Apple’s proprietary
mobile operating system—account for roughly 60 percent of mobile-device usage. Apple’s iOS
is a closed ecosystem; Apple does not license iOS to third-party mobile-device manufacturers.
Another roughly 40 percent of mobile-device usage comes from devices that use Android, an
open-source mobile operating system controlled by Google. Unlike iOS, Android is licensable,
which means third-party mobile-device manufacturers can use it as the operating system for their
devices. All other mobile operating systems, combined, account for less than one percent of
mobile-device usage in the United States.
44. General search services can be delivered to mobile-device users through a variety
of search access points, including: (1) a browser, (2) a static search bar (search widget, referred
to in Figure 4 as the QSB or quick search box) on the device’s home screen, (3) a search app,
(4) artificial intelligence software (voice assistants) accessed by a button or voice command and
designed to answer voice-initiated queries, and (5) other apps that link to general search engines,
such as smart keyboards. Figure 4, from a 2018 Google strategy deck, provides a more specific
breakdown of how Google delivers its general search service on Android devices.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 15 of 64
16
Figure 4
45. In the United States, both cell phone carriers and manufacturers sell mobile
devices. As discussed above, these phones or tablets typically have search access points preset
with a general search engine as the default. These preset defaults are usually governed by a
distribution or licensing agreement. For instance, Google has contracted with Apple for many
years to preset Google’s search engine as the default for Apple’s Safari browser and, more
recently, other search access points on Apple’s mobile devices. When a consumer takes a new
iPhone or iPad out of its box, all the significant access points default to Google as their general
search provider. Indeed, Google has preset default status for an overwhelming share of the search
access points on mobile devices sold in the United States.
46. For mobile browsers, Google is the default search provider for both Apple Safari
(approximately 55 percent share) and Google Chrome (over 35 percent share), which together
account for over 90 percent of the browser usage on mobile devices in the United States.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 16 of 64
17
47. Consumers typically do not change their mobile device’s default search functions,
making securing preset default status for search access points important for effective distribution
of general search engines (and delivery of search ads). As one search competitor noted in 2019,
“For the most part, despite the simplicity of changing a default setting to enable customer choice,
experience shows us that users accept the default search experience that comes with their device
or the browser.” This fact is especially true on mobile devices; as Google observed in a 2018
strategy document, “People are much less likely to change [the] default search engine on
mobile.” Alternative methods of obtaining search access points or encouraging general search
engine usage—such as direct marketing to consumers—are not nearly as effective as
preinstalling search access points on mobile devices and computers.
2. The Computer Search Distribution Channel
48. When using a computer, most consumers access a general search engine through a
browser, either by (1) typing a query directly into the address bar at the top of the browser, or
(2) visiting a general search engine web page and entering a query. Many browsers default to a
general-search-engine web page as the home or start screen each time a user activates the
browser; this offers users a convenient way to start their search experience.
49. In the United States, Google Chrome is the leading computer browser, with
almost 60 percent market share. Apple’s Safari browser has approximately 16 percent share on
computers. Mozilla’s Firefox has approximately 7 percent share, and Microsoft’s Edge and
Internet Explorer together have approximately 15 percent share. Other small browsers have a
combined share of less than 4 percent. With the exception of Microsoft, most browser developers
have agreed with Google to preset its search engine as the default search provider.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 17 of 64
18
50. Preset default settings are important for computers. Consumers may not
understand that they can change the browser’s preset default general search engine, or consumers
may not bother to invest the time to make such a switch.
51. For both mobile and computer search access points, being preset as the default is
the most effective way for general search engines to reach users, develop scale, and become or
remain competitive.
D. Distribution Agreements in Mobile and Computer Channels
52. General search services providers typically enter into licensing and distribution
agreements with manufacturers and carriers that distribute mobile devices with search access
points. In the United States, roughly 60 percent of all search queries are covered by Google’s
exclusionary agreements. On mobile devices, Google’s exclusionary agreements cover more than
80 percent of all U.S. search queries.
53. Of the remaining search queries not covered by Google’s exclusionary contracts,
almost half take place on search access points owned by Google. Google is a vertically integrated
search provider and distributes search in part through several of its own properties, including for
example its browser (Chrome) and phone (Pixel). Between its exclusionary contracts and owned-
and-operated properties, Google effectively owns or controls search distribution channels
accounting for roughly 80 percent of the general search queries in the United States.
54. Google’s distribution agreements come in three basic types, with the specific
terms of each agreement depending upon the counterparty and the search access points at issue.
First, Google requires Android device manufacturers that want to preinstall Google’s proprietary
apps to sign an anti-forking agreement; these agreements set strict limits on the manufacturers’
ability to sell Android devices that do not comply with Google’s technical and design standards.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 18 of 64
19
55. Next, for Android device manufacturers that sign an anti-forking agreement,
Google provides access to its vital proprietary apps and application program interfaces (APIs) for
preinstallation, but only if the manufacturers contractually agree to (1) take a bundle of other
Google apps, (2) make certain apps undeletable, and (3) give Google the most valuable and
important real estate on the default home screen.
56. Finally, Google provides a share of its search advertising revenue to Android
device manufacturers, mobile phone carriers, competing browsers, and Apple; in exchange,
Google becomes the preset default general search engine for the most important search access
points on a computer or mobile device. As a practical matter, users rarely switch the preset
default general search engine. In many cases, the agreements relating to mobile devices go even
further, expressly prohibiting (1) the preinstallation of any rival general search services, and (2)
the setting of other defaults to rival general search engines. This means that Google is the only
preset default search provider preinstalled on the device.
57. These agreements work exactly as Google designed them—to foreclose
distribution to Google’s search rivals, weakening them as competitive alternatives for consumers
and advertisers by denying them scale.
1. Background on Mobile Strategy and Development of Android Ecosystem
58. Google’s anticompetitive agreements must be understood against the backdrop of
Google’s overall business strategy. When Google was formed and achieved initial success in the
late 1990s and early 2000s, internet searches were almost exclusively performed through
browsers on computers. But as Google told investors in its 2007 Form 10-K: “More individuals
are using non-desktop devices to access the internet. If users of these devices do not widely
adopt versions of our web search technology, products or operating systems developed for these
devices, our business could be adversely affected.”
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 19 of 64
20
59. In a mobile world, Google had to deal with mobile device manufacturers (such as
LG, Motorola, and Samsung), and carriers (such as AT&T, T-Mobile/Sprint, and Verizon) that
would hold sway over distribution of search and search ads. Google thus asked internally, “How
can we conquer the world’s major wireless markets simultaneously?”
60. The answer started with Android, a mobile operating system that Google
purchased in 2005. In 2007, Google released the Android code for free under an open-source
license. Being “open source” means that anyone can access the source code and use it to make
their own, modified operating system—a “fork.” This was key to Android’s adoption.
61. First, Google’s apparent lack of control over an open-source operating system
attracted skeptical manufacturers and carriers of mobile phones to use Android instead of the
other choices then available. As the Android team leader observed to Google’s board of
directors, “Google was historically seen as a threat” to these distributors. But an open-source
model suggested that they—and not Google—would ultimately retain control over their devices
and the app ecosystem on those devices.
62. Second, once enough major distributors agreed to use Android, the operating
system attracted developers looking for wide distribution of their apps. As more app developers
focused their efforts on designing Android apps, Android became more attractive to consumers,
which in turn led even more developers to design for Android. The result was a must-have
ecosystem of Android apps.
63. Third, to help the Android ecosystem achieve critical mass and to advance the
network effects, Google “shared” its search advertising and app store revenues with distributors
as further inducement to give up control. As one senior executive explained about Android
Market, an earlier name for Google’s app store, “Android Market is a bitter pill for carriers, and
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 20 of 64
21
a generous revenue share is the sugar that makes it go down smoother.” In other words,
beginning over ten years ago, Google used revenue sharing to attract partners to Android; as
discussed below, Google uses revenue sharing to keep them locked in today.
64. By 2010, the Android team leader noted that “Android is poised for world
domination—the success story of the decade.” He was right; the strategy worked. The “Google
Play” app store has a massive library of apps, making it essential for Android distributors to have
on their devices. As for the operating system itself, it quickly became the dominant licensable
mobile operating system in the United States. In the four years between 2009 and 2012,
Android’s share of licensable mobile operating systems on smartphones in the United States
more than tripled, reaching about 80 percent. Today, Android represents over 95 percent of
licensable mobile operating systems for smartphones and tablets in the United States and
accounts for over 70 percent of all mobile device usage worldwide. The only other mobile
operating system with significant market share in the United States is Apple’s iOS, which is not
licensable.
65. Control over Android has always been a critical issue. As Google’s Android team
leader asked at the time: “How do we retain control of something we gave away?” Google’s
answer is the set of contractual “carrots” and “sticks” discussed below that empower Google to
“[o]wn the ecosystem” and help thwart any alternative mobile ecosystem from developing that
could support a different search provider.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 21 of 64
22
Figure 5
2. Anti-Fragmentation Agreements and Compatibility Commitments (Android Mobile Devices)
66. The Android operating system is open source; Google updates the Android code
periodically and makes it publicly available. But Google takes steps to minimize the risk that a
developer creates an Android fork to compete with the Android ecosystem controlled by Google.
By limiting the existence of devices running Android forks, Google limits possible distribution
channels available to its search rivals.
67. One way Google retains control of the Android ecosystem is through anti-forking
agreements. These agreements broadly prohibit manufacturers from taking “any actions that may
cause or result in the fragmentation of Android.” Notably, “fragmentation” is left undefined,
giving Google wide latitude in practice.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 22 of 64
23
68. Google’s anti-forking agreements specifically forbid manufacturers from
developing or distributing versions of Android that do not comply with Google-controlled
technical standards, as defined in its Android Compatibility Definition Document (CDD).
69. Two types of anti-forking agreements exist. Before 2017, Google required
distributors to sign Anti-Fragmentation Agreements (AFAs). In 2017, while being investigated
by the European Commission (and long after Google had locked up its monopoly status), Google
began shifting its anti-forking restrictions from AFAs to new Android Compatibility
Commitments (ACCs). Today, Google has an AFA or ACC with the leading Android device
manufacturers, including LG, Motorola, and Samsung.
70. ACCs are marginally less onerous than AFAs because they allow manufacturers
to build devices or components for third parties to sell to consumers, even if those devices or
components do not comply with Google’s technical standards. But ACCs, like AFAs, prohibit
signatories from manufacturing Android forks of their own, distributing devices with Android
forks, or using their powerful brands to market forks on behalf of third parties. Most well-known
Android manufacturers are bound by AFAs or ACCs.
71. The AFAs and ACCs do not just restrict manufacturers’ ability to build and
distribute innovative versions of mobile phones. Over time, Google has extended the Android
CDD such that its specifications apply to tablets and emerging technologies such as smart TVs,
watches, and automotive devices. Manufacturers that hope to release Android-based versions of
these products must comply with Google’s standards as well.
3. Mobile Application Distribution Agreements (Android Mobile Devices)
72. Manufacturers agree to anti-forking agreements in part because they are a
precondition to receiving a license to distribute devices with must-have proprietary Google apps
and APIs (the set of technical specifications that enable software applications to communicate
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 23 of 64
24
with each other, operating systems, and hardware). This license is provided only through
preinstallation agreements—called Mobile Application Distribution Agreements, or MADAs.
Leading Android device manufacturers, such as LG, Motorola, and Samsung, are MADA
licensees.
73. Over time, Google has chosen to include important features and functionality in
Google’s own ecosystem of proprietary apps and APIs, rather than the open-source Android
code. Google refers to this proprietary layer as “Google Mobile Services” (GMS). GMS includes
many popular apps, such as Google’s search app, Chrome, YouTube, and Google Maps. GMS
also includes Google Play, Google’s app store. An app store is one of the most valuable features
of a mobile device because it offers access to compatible apps that do not come preinstalled on
the device. Google Play offers about three million apps, more than any other app store (including
Apple’s App Store, which is compatible only with Apple devices). More than 90 percent of apps
on Android devices are downloaded through Google Play. For years, Google Play has been the
only commercially significant app store option for Android manufacturers.
74. Another key part of GMS is the set of APIs that allow developers to access certain
important features. The APIs available within GMS are part of “Google Play Services” (GPS).
GPS allows apps, including third-party apps, to perform functions that are not possible using the
open-source version of Android. For example, using the open-source Android system, third-party
apps cannot provide basic “push notifications,” enable in-app purchases through Google Play, or
use data from Google Maps; to have these functionalities, third-party apps must use GPS.
75. The integration of key functions with GPS makes it more difficult for third-party
Android developers to port their apps to Android forks because the apps are designed to interact
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 24 of 64
25
with Google’s proprietary APIs. And as the functionality gap between open-source Android apps
and Google’s proprietary apps grows, developers are more dependent on GPS.
76. Signing a preinstallation agreement is the only way for an Android device
manufacturer to preinstall any Google app, including Google Play. It is also the only way an
Android device manufacturer can gain access to GPS and the APIs many developers need for
their apps to work properly, at least without expensive and time-consuming reprogramming. But
any manufacturer installing Google Play or GPS must preinstall a full suite of apps identified by
Google, including the search access points most frequently used by consumers: Chrome, Google
search app, Google search widget, and Google Assistant. Google’s search engine is the default
on all these search access points. Indeed, Google uses the MADAs to control the appearance of
Android devices, requiring the manufacturer to place the Google search widget on the home
screen, and to preinstall Chrome, the Google search app, and other apps in a way that makes
them undeletable by the user.
77. Moreover, before 2017, most MADAs also required manufacturers to set Google
as the default general search engine for all key search access points on any device with
preinstalled Google apps—these requirements are now found in the revenue sharing agreements
discussed below.
4. Revenue Sharing Agreements (Android Mobile Devices)
78. Google enters into search revenue sharing agreements (RSAs) with Android
manufacturers and carriers. Google generally requires exclusive distribution as the sole preset
default general search service on an ever-expanding list of search access points; in exchange,
Google remits to these companies a percentage of search advertising revenue. Google offers
revenue share to Android device manufacturers only if they are MADA licensees, and Google
offers revenue share to carriers only for devices built by manufacturers that are MADA
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 25 of 64
26
licensees. The leading U.S. carriers (AT&T, T-Mobile, and Verizon) and the leading Android
device manufacturer (Samsung) have RSAs with Google.
79. Some of Google’s revenue sharing agreements require blanket coverage for all
Android devices sold by Google’s counterparty. Under this version of the revenue sharing
agreements, the distributor receives a payment from Google only if all the distributor’s Android
devices comply with the exclusivity requirements. Other revenue sharing agreements provide for
a model-by-model choice. Under this version of the agreements, for the distributor to receive a
cut of the advertising revenue from any units of a model, every unit of that model must comply
with the exclusivity requirements.
80. As innovation has increased the number of search access points on mobile
devices—including smart keyboards and voice assistants—Google has expanded its RSAs to
close off these avenues to search rivals.
5. Mobile Incentive Agreements (Android Mobile Devices)
81. In Google’s latest round of negotiations with some Android manufacturers,
Google has replaced RSAs with mobile incentive agreements (MIAs), under which Google pays
manufacturers to (1) forego preinstalling rival general search services on their Android devices
and (2) comply with a significant number of “incentive implementation requirements”—
including preloading up to fourteen additional Google apps. LG and Motorola have MIAs with
Google.
82. To maximize payments under the MIAs, the manufacturers must also set Google
as the default for all search access points on nearly all of their devices. Moreover, Google
generally retains “sole discretion” to determine what constitutes a “search access point,” and thus
controls the coverage of its exclusive contracts. Although the MIAs change the payment
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 26 of 64
27
structure for certain manufacturers, the agreements achieve the same end as their predecessors:
search exclusivity for Google.
83. Today, Google has revenue sharing agreements (RSAs or MIAs) with all major
U.S. carriers and Android device manufacturers, as well as a number of smaller carriers and
manufacturers. Google’s revenue sharing agreements (and preinstallation agreements) with
Android device manufacturers, together, account for more than 30 percent of mobile device
usage in the United States.
6. Revenue Sharing Agreements (Apple and Others)
84. Google’s revenue sharing agreements are not limited to its Android partners.
Google has entered into revenue sharing agreements with rival browsers and other device
manufacturers, further blocking off search access points from competition.
85. Most significantly, Google has had a series of search distribution agreements with
Apple, effectively locking up one of the most significant distribution channels for general search
engines. Apple operates a tightly controlled ecosystem and produces both the hardware and the
operating system for its popular products. Apple does not license its operating systems to third-
party manufacturers and controls preinstallation of all apps on its products. The Safari browser is
the preinstalled default browser on Apple computer and mobile devices. Apple devices account
for roughly 60 percent of mobile device usage in the United States. Apple’s Mac OS accounts for
approximately 25 percent of the computer usage in the United States.
86. In 2005, Apple began using Google as the preset default general search engine for
Apple’s Safari browser. In return, Google gave Apple a significant percentage of Google’s
advertising revenue derived from the search queries on Apple devices. Two years later, Google
extended this agreement to cover Apple’s iPhones. In 2016, the agreement expanded further to
cover additional search access points—Siri (Apple’s voice-activated assistant) and Spotlight
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 27 of 64
28
(Apple’s system-wide search feature)—making Google the preset default general search engine
for both services. Today, Google’s distribution agreement with Apple gives Google the coveted,
preset default position on all significant search access points for Apple computers and mobile
devices.
87. Today, Google has RSAs with nearly every significant, non-Google browser other
than those distributed by Microsoft, including Mozilla’s Firefox, Opera, and UCWeb. These
agreements generally require the browsers to make Google the preset default general search
engine for each search access point on both their web and mobile versions.
IV. RELEVANT MARKETS
A. General Search Services in the United States
1. General Search Services in the United States Is a Relevant Antitrust Market
88. General search services in the United States is a relevant antitrust market. General
search services allow consumers to find responsive information on the internet by entering
keyword queries in a general search engine such as Google, Bing, or DuckDuckGo.
89. General search services are unique because they offer consumers the convenience
of a “one-stop shop” to access an extremely large and diverse volume of information across the
internet. Consumers use general search services to perform several types of searches, including
navigational queries (seeking a specific website), informational queries (seeking knowledge or
answers to questions), and commercial queries (seeking to make a purchase).
90. Other search tools, platforms, and sources of information are not reasonable
substitutes for general search services. Offline and online resources, such as books, publisher
websites, social media platforms, and specialized search providers such as Amazon, Expedia, or
Yelp, do not offer consumers the same breadth of information or convenience. These resources
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 28 of 64
29
are not “one-stop shops” and cannot respond to all types of consumer queries, particularly
navigational queries. Few consumers would find alternative sources a suitable substitute for
general search services. Thus, there are no reasonable substitutes for general search services, and
a general search service monopolist would be able to maintain quality below the level that would
prevail in a competitive market.
91. The United States is a relevant geographic market for general search services.
Google offers users in the United States a local domain website with search results optimized
based on the user’s location in the United States. General search services available in other
countries are not reasonable substitutes for general search services offered in the United States.
Google analyzes search market shares by country, including the United States. Therefore, the
United States is a relevant geographic market.
2. Google Has Monopoly Power in the General Search Services Market in the United States
92. Google has monopoly power in the United States general search services market.
There are currently only four meaningful general search providers in this market: Google, Bing,
Yahoo!, and DuckDuckGo. According to public data sources, Google today dominates the
market with approximately 88 percent market share, followed far behind by Bing with about
seven percent, Yahoo! with less than four percent, and DuckDuckGo with less than two percent.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 29 of 64
30
Figure 6
93. Over the years, Google has steadily increased its dominant position in general
search services. In July 2007, Google estimated its general search services market share at
68 percent. By June 2013, Google estimated that its share in the United States had already
increased to 77 percent on computers. By April 2018, Google estimated that its share was
79 percent on computers and 93.5 percent on mobile. More recently, Google has accounted for
almost 90 percent of all general search engine queries in the United States, and almost 95 percent
of queries on mobile devices. Recent share estimates are in Figures 7 and 8.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 30 of 64
31
Figure 7 Figure 8
94. There are significant barriers to entry in general search services. The creation,
maintenance, and growth of a general search engine requires a significant capital investment,
highly complex technology, access to effective distribution, and adequate scale. For that reason,
only two U.S. firms—Google and Microsoft—maintain a comprehensive search index, which is
just a single, albeit fundamental, component of a general search engine.
95. Scale is also a significant barrier to entry. Scale affects a general search engine’s
ability to deliver a quality search experience. The scale needed to successfully compete today is
greater than ever. Google’s anticompetitive conduct effectively eliminates rivals’ ability to build
the scale necessary to compete.
96. Google’s large and durable market share and the significant barriers to entry in
general search services demonstrate Google’s monopoly power in the United States.
B. Search Advertising in the United States and General Search Text Advertising in the United States Are Relevant Antitrust Markets
1. Search Advertising Is a Relevant Product Market
97. Search advertising in the United States is a relevant antitrust market. The search
advertising market consists of all types of ads generated in response to online search queries,
including general search text ads (offered by general search engines such as Google and Bing)
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 31 of 64
32
and other, specialized search ads (offered by general search engines and specialized search
providers such as Amazon, Expedia, or Yelp).
98. Search ads enable advertisers to target marketing messages in real time in
response to queries entered by a consumer. Thus, a user’s general search query has the important
function to an advertiser of revealing the searcher’s intent. The ability of search ads to respond to
consumer inquiries, at the moment the consumer is investigating a subject relevant to an
advertiser’s product or service, makes these ads highly valuable to advertisers and distinguishes
them from other types of advertising that cannot be similarly targeted, whether online or offline.
99. Other forms of advertising are not reasonably substitutable for search ads. For
example, “offline” ads such as newspaper, billboard, TV, and radio ads cannot be targeted at a
specific consumer based on the consumer’s real-time, self-disclosed interests. Similarly, other
forms of online ads, such as display ads or social media ads, do not enable advertisers to target
customers based on specific queries and are generally aimed at consumers who are further from
the point of purchase. As Google’s Chief Economist explained: “One way to think about the
difference between search and display/brand advertising is to say that ‘search ads help satisfy
demand’ while ‘brand advertising helps to create demand,’” and “[d]isplay and search
advertising are complementary tools, not competing ones.”
100. Few advertisers would find alternative sources a suitable substitute for search
advertising. Thus, there are no reasonable substitutes for search advertising, and a search
advertising monopolist would be able to maintain prices above the level that would prevail in a
competitive market.
2. General Search Text Advertising Is a Relevant Product Market
101. There also is a relevant product market for general search text advertising that is
wholly contained within the broader search advertising market. General search text ads are sold
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 32 of 64
33
by general search engines, typically placed just above or below the organic search results on a
SERP, and resemble the organic results that appear on a general search engine’s SERP, with a
subtle notation that they are “ads” or “sponsored.” In contrast, other types of search ads—
specifically, specialized search ads—typically are visually different from general search text ads
and convey different types of information. For example, a Google Shopping Ad normally
includes an image of the product, its price, and star-based ratings (see Figure 3). In 2018, general
search text ads accounted for close to 85 percent of Google’s search ad revenue.
102. General search text ads are distinct from specialized search ads in ways that limit
their substitutability for most purposes, including their scope of coverage, purpose, format, and
sales process. Indeed, for many advertisers that purchase general search text ads, there are no
reasonable alternatives for these ads, which renders these advertisers particularly vulnerable to
targeted price increases. General search text ads on Google are vitally important for many
different types of advertisers, including companies that prefer to sell directly to consumers from
their own websites and companies that want to protect their brand names on Google.
103. General search text ads can be delivered in response to search queries related to
any subject that users explore on the internet. General search text ads are offered predominantly
by the two companies that operate general search engines: Google and Microsoft (Bing); Bing
also syndicates general search text ads for Yahoo! and DuckDuckGo. By contrast, other kinds of
search ads are provided by specialized search providers in response to narrower and deeper
searches in their areas of specialization, such as retail (e.g., Amazon), travel (e.g., Kayak), or
local (e.g., Yelp).
104. General search text ads often target consumers further from an actual sale or
“conversion” than specialized search ads. Users often rely on a general search engine such as
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 33 of 64
34
Google to start a search of the entire web to explore an interest, consider options, and form a
preference, often about a purchase. An advertiser often will buy a general search text ad to drive
these searchers down the purchase funnel to the advertiser’s website to shop for a product or
service. In part for this reason, specialized search providers, such as Amazon, Expedia, and
eBay, are among Google’s largest customers for general search text ads—i.e., they buy general
search text ads to drive consumers to their specialized search sites, where they then sell
specialized search ads to advertisers who want to reach those interested consumers at or near the
point of purchase. Because general search text ads and specialized ads serve different functions,
advertisers often view these ads as complements.
105. General search text ads link to the advertiser’s website, so the user can “click out”
to that site. By contrast, ads by specialized search providers often link to webpages on that
specialized search provider’s own website. For example, if a company sells a product on
Amazon and buys an ad on Amazon to promote its product, the ad links to the Amazon page on
which the advertised product can be purchased—not the seller’s own website. This kind of
search ad is called a “click in” ad. Thus, general search text ads can be purchased by advertisers
that do not sell their products or services on specialized search sites (such as Amazon) as well as
advertisers that prefer to sell their products or services directly to consumers.
106. Few general search text advertisers would find alternative sources a suitable
substitute for general search text advertising. Thus, there are no reasonable substitutes for
general search text advertising, and a general search text advertising monopolist would be able to
maintain prices above the level that would prevail in a competitive market.
3. The United States Is a Relevant Geographic Market
107. The United States is a relevant geographic market for both the search advertising
and the general search text advertising markets. Market participants recognize this in the
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 34 of 64
35
ordinary course of business. For example, Google offers advertisers the ability to target and
deliver ads based on the location of consumers in the United States, and Google search is
customized for particular countries. Google also separately tracks revenue for the United States.
4. Google Has Monopoly Power in the Search Advertising and General Search Text Advertising Markets in the United States
108. Google has monopoly power in the search advertising market. Based on public
estimates of total search advertising spending in the United States, Google’s share of the U.S.
search advertising market is over 70 percent. This market share understates Google’s market
power in search advertising because many search-advertising competitors offer only specialized
search ads and thus compete with Google only in a limited portion of the market.
109. Google also has monopoly power in the general search text advertising market.
Google’s market share of the U.S. general search text advertising market also exceeds
70 percent. Google’s share of the general search text advertising market well exceeds its share of
the search advertising market.
110. There are barriers to entry in these advertising markets that protect Google’s
advertising monopolies. Most critically, search advertising of any kind requires a search engine
with sufficient scale to make advertising an efficient proposition for businesses. Specialized
search engines require significant investment, including the cost of populating and indexing
relevant data, distribution, developing and maintaining a search algorithm, and attracting users.
Search advertising of any kind also requires (1) a user interface through which advertisers can
buy ads, (2) software to facilitate the sales process, and (3) a sales and technical support staff.
The same barriers to entry that apply to general search services also protect Google’s general
search text advertising monopoly.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 35 of 64
36
V. ANTICOMPETITIVE CONDUCT
111. Google is a monopolist in the general search services, search advertising, and
general search text advertising markets. Google aggressively uses its monopoly positions, and
the money that flows from them, to continuously foreclose rivals and protect its monopolies.
112. Google has unlawfully maintained its monopolies by implementing and enforcing
a series of exclusionary agreements with distributors over at least the last decade. Particularly
when taken together, Google’s exclusionary agreements have denied rivals access to the most
important distribution channels. In fact, Google’s exclusionary contracts cover almost 60 percent
of U.S. search queries. Almost half the remaining searches are funneled through properties
owned and operated directly by Google. As a result, the large majority of searches are covered
by Google’s exclusionary contracts and own properties, leaving only a small fraction for
competitors.
113. Google’s continued use of the exclusionary agreements over many years, long
after there was any real competition in general search, has denied its rivals access to the scale
that would allow rivals to increase quality. By depriving them of scale, Google also hinders its
rivals’ ability to secure distribution going forward, insulating Google from competition.
114. Google’s exclusionary motives influence its negotiations with distributors. Some
of these exclusionary agreements have been described by Google as an “[i]nsurance policy that
preserves our search and assistant usage.” To preserve its dominance, Google has developed
economic models to measure the “defensive value” of foreclosing search rivals from effective
distribution, search access points, and ultimately competition. Google recognized it could pay
search distributors to “protect [its] market share from erosion.” Google continues to focus on the
exclusionary defensive value of its distribution contracts as it tries to expand its search
dominance into new distribution channels, such as smart home speakers. Here, Google’s
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 36 of 64
37
defensive value “is attributable to protecting access to Search and other Google services that may
otherwise be blocked in a given household” if a user chooses a rival.
115. In sum, Google deprives rivals of the quality, reach, and financial position
necessary to mount any meaningful competition to Google’s longstanding monopolies. By
foreclosing competition from rivals, Google harms consumers and advertisers.
A. Google’s Agreements Lock Up Mobile Distribution of Search
116. Launched in the infancy of mobile smartphones, Google’s strategy to ward off
competition for mobile search distribution had two parts. First, Google expanded its existing
search deal with Apple to cover mobile. Second, for other mobile distributors, Google offered its
Android operating system for “free” but with a series of interlocking distribution agreements to
ensure it search-engine dominance in the Android ecosystem.
117. Google’s strategy worked. Google has almost completely shut out its competitors
from mobile distribution. As one executive for a competing search product recognized in
frustration last year: “Google essentially [has] locked up ALL DISTRIBUTION” with its Apple
deal and restrictive Android licensing terms, leaving the competitor’s product with “no mobile
volume.”
1. Distribution on Apple iOS Devices
118. Apple has not developed and does not offer its own general search engine. Under
the current agreement between Apple and Google, which has a multi-year term, Apple must
make Google’s search engine the default for Safari, and use Google for Siri and Spotlight in
response to general search queries. In exchange for this privileged access to Apple’s massive
consumer base, Google pays Apple billions of dollars in advertising revenue each year, with
public estimates ranging around $8–12 billion. The revenues Google shares with Apple make up
approximately 15–20 percent of Apple’s worldwide net income.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 37 of 64
38
119. Although it is possible to change the search default on Safari from Google to a
competing general search engine, few people do, making Google the de facto exclusive general
search engine. That is why Google pays Apple billions on a yearly basis for default status.
Indeed, Google’s documents recognize that “Safari default is a significant revenue channel” and
that losing the deal would fundamentally harm Google’s bottom line. Thus, Google views the
prospect of losing default status on Apple devices as a “Code Red” scenario. In short, Google
pays Apple billions to be the default search provider, in part, because Google knows the
agreement increases the company’s valuable scale; this simultaneously denies that scale to rivals.
120. Apple’s RSA incentivizes Apple to push more and more search traffic to Google
and accommodate Google’s strategy of denying scale to rivals. For example, in 2018, Apple’s
and Google’s CEOs met to discuss how the companies could work together to drive search
revenue growth. After the 2018 meeting, a senior Apple employee wrote to a Google
counterpart: “Our vision is that we work as if we are one company.”
121. The current version of the Google–Apple agreement substantially forecloses
Google’s search rivals from an important distribution channel for a significant, multi-year term.
This agreement covers roughly 36 percent of all general search queries in the United States,
including mobile devices and computers. Google estimates that, in 2019, almost 50 percent of its
search traffic originated on Apple devices.
122. Particularly when considered with the other exclusionary distribution agreements
discussed below, Google’s hold on Apple’s distribution channel is self-reinforcing, impairing
rival general search engines’ ability to offer competitive products and making Google’s
monopolies impenetrable to competitive discipline. By paying Apple a portion of the monopoly
rents extracted from advertisers, Google has aligned Apple’s financial incentives with its own
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 38 of 64
39
and set the price of bidding for distribution extraordinarily high—in the billions. And, even if a
rival was willing to make no money from a distribution relationship or could afford to lose
money indefinitely, the rival would likely still fall short because the existing distribution
agreements have for more than a decade denied rivals the benefits of scale, thus limiting (1) the
quality of their general search and search advertising products, as well as (2) the audience to
attract advertisers. In other words, because of the longtime deprivation of scale, no other search
engine can offer Apple (or any other partner) the mix of quality, brand recognition, and
economics that market-dominant Google can.
2. Distribution on Android Devices
123. Google controls the Android mobile distribution channel with its distributor
agreements and owned-and-operated distribution properties.
124. Even though Android is open source, Google has used Android to protect
Google’s lucrative general search and search advertising monopolies. Google sets the rules
through anti-forking agreements, preinstallation agreements, and revenue sharing agreements.
Notably, each of these agreements builds on the others to preserve control. Thus, Google will not
pay a revenue share or financial incentive payment on a mobile device unless it is covered by
(1) an anti-forking agreement, (2) a preinstallation agreement ensuring that Google’s search
access points are preinstalled and given prominent placement, and (3) a revenue sharing or
mobile incentive agreement that entitles Google to preset default status and, in most cases,
prohibits preinstallation of search access points with rival general search providers.
125. Through these interlocking, anticompetitive agreements, Google insulates and
protects its monopoly profits. One internal Google analysis of these restrictive agreements
concluded that only one percent of Google’s worldwide Android search revenue was currently at
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 39 of 64
40
risk to competitors. This analysis noted that the growth in Google’s search advertising revenue
from Android distribution was “driven by increased platform protection efforts and agreements.”
a) Anti-Forking Agreements
126. An alternative operating system could serve as a pathway for distribution of
general search services other than Google. However, Google’s anti-forking agreements inhibit
the development of an operating system based on an Android fork that could serve as a viable
path to market for a search competitor.
127. Developing an operating system from scratch is extremely expensive, but a
manufacturer could start with existing Android open-source code for a fraction of the cost.
Moreover, the costs to app developers of “porting” GMS-compatible Android apps to an Android
fork are substantially less than developing apps for an entirely new operating system.
128. Google’s anti-forking agreements, however, have inhibited operating system
innovation through forking, ensuring that manufacturers and distributors are beholden to
Google’s version of Android. Distributors know that any violation of an anti-forking agreement
could mean excommunication from Google’s Android ecosystem, loss of access to Google’s
must-have GPS and Google Play, and millions or even billions of dollars in lost revenue sharing.
Thus, distributors avoid anything that Google might deem “fragmentation”—a term that Google
“purposely leave[s] . . . very vague” and interprets broadly.
129. Pursuant to the preinstallation agreements discussed below, Google also has final
say over whether a device is found to be compatible with the technical specifications Google
requires manufacturers to meet before they can preinstall GMS. As a Google engineer noted, it
must be “obvious to the [manufacturers] that we are using compatibility as a club to make them
do things we want.” Google views its anti-fragmentation mandate, and its final approval of
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 40 of 64
41
devices before they launch, as a “poison pill” to prevent deviation from the Google-controlled
Android ecosystem.
130. Google’s broad interpretation of the anti-forking agreements, and the reluctance it
creates among Android distributors to support alternative versions of Android, presents barriers
to entry. These were on display when Amazon developed its Fire OS operating system, a
competing fork of Android. Rather than preinstall Google’s search engine, GPS, Google Play, or
other Google apps on Fire devices, Amazon preinstalled its own proprietary apps and agreed to
make Microsoft’s Bing the preset default general search engine. Amazon originally sold only
Fire OS tablets, but in 2014 it launched a phone that ran on Fire OS. The phone was not a
commercial success and Amazon quickly exited the phone business. Amazon continues to sell
Fire tablets, which account for less than two percent of mobile device usage in the United States.
131. Google’s anti-forking provisions and policies limited the growth of Amazon’s
mobile phone, and of Fire OS, because major manufacturers declined to support Amazon’s
phone out of fear doing so would risk their lucrative deals with Google. Manufacturers willing to
work with Amazon did not have the same marketing and logistics capabilities as top
manufacturers. Despite hundreds of millions of dollars in investment over nearly ten years across
tablets and phones, Fire OS still has not reached sufficient critical mass to challenge Google’s
version of Android and provide a significant alternative path to market for search rivals.
132. No Android fork has made significant inroads to challenge Google for mobile
devices, and there is no meaningful operating system alternative for manufacturers and carriers
to license. These manufacturers and carriers are beholden to Google’s Android ecosystem, which
Google uses to preserve its monopolies in general search, search advertising, and general search
text advertising. Google’s anti-forking agreements further inhibit the development of alternative
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 41 of 64
42
Android operating systems for the next generation of search distribution channels, such as smart
watches, smart speakers, smart TVs, and connected automobiles.
b) Preinstallation Agreements
133. Google uses preinstallation agreements—MADAs—to ensure that its entire suite
of search-related products is given premium placement on Android GMS devices. Consumers
naturally and regularly turn to these prominently placed search access points to conduct searches.
Preinstallation agreements also reinforce Google’s anti-forking requirements, either by including
an anti-forking clause of their own or, more commonly, requiring device manufacturers to be
signatories to an anti-forking agreement.
134. If a manufacturer wants even one of Google’s key apps and APIs, the device must
be preloaded with a bundle of other Google apps selected by Google. The six “core” apps are
Google Play, Chrome, Google’s search app, Gmail, Maps, and YouTube. Manufacturers must
preinstall the core apps in a manner that prevents the consumer from deleting them, regardless of
whether the consumer wants them. These preinstallation agreements cover almost all Android
devices sold in the United States.
135. Google’s preinstallation agreements effectuate a tie, that is, they condition the
distribution of Google Play and GPS to the distribution of these other apps. This tie reinforces
Google’s monopolies. The preinstallation agreements provide Android device manufacturers an
all-or-nothing choice: if a manufacturer wants Google Play or GPS, then the manufacturer must
also preinstall, and in some cases give premium placement to, an entire suite of Google apps,
including Google’s search products. The forced preinstallation of Google’s apps deters
manufacturers from preinstalling those of competitors. This forecloses distribution opportunities
to rival general search engines, protecting Google’s monopolies.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 42 of 64
43
136. Google recognizes it could “make [the] phone experience better for user[s] by
ensuring . . . preloaded apps are deletable.” In large part, this is because “[u]sers can free up
space by deleting apps they don’t want.” Consumers desiring to use non-Google search access
points thus suffer because they cannot save storage space on their devices by deleting unwanted
Google apps. In this way, manufacturers must agree to make their phones less attractive to
consumers to accommodate Google’s efforts to lock up search distribution.
137. Once the manufacturer adopts the necessary suite of Google apps, the search
access points of those apps are preset to default to Google’s search engine. For example, the
preinstalled version of Chrome is preset to default to Google search. A senior executive at
Google referred to changing Chrome’s preset search default as “totally off the table” and insisted
that if a manufacturer “values their MADA, they cannot modify Chrome’s settings.” The result is
that Google locks up the access points to general search on Android phones, as shown in
Figure 9:
Figure 9
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 43 of 64
44
138. The preinstallation agreements are even more pernicious than basic ties because
these agreements force distributors to configure the appearance of their phones to Google’s
specifications. For example, they require manufacturers to put the Google search widget on the
device’s default home screen. Google considers the search widget “an essential part of the
Google brand” and rejects requests by manufacturers to waive the preinstallation agreement’s
search-widget requirement. This locks up another search access point, as it would be impractical
for a manufacturer to preinstall two search widgets on the same home screen.
139. Google’s preinstallation agreements also impose voice-search preferencing. In
addition to requiring the preinstallation of Google Assistant, preinstallation agreements require
manufacturers to (1) implement a Google hotword, which activates Google Assistant, and
(2) ensure certain touch actions on the device’s home button directly access Google Assistant or
Google. Google’s agreements with most manufacturers also (3) set Google Assistant as the
default assistant app.
140. Rivals to Google Assistant are deprived of the same opportunities. Most of
Google’s preinstallation agreements prevent rival assistants from being the preset default or
using a home button. Google also handicaps rival assistants by limiting the APIs that non-Google
apps can use, ensuring that the useful features, such as “always on” microphone access that
would enable the use of a hotword or the initiation of phone calls, are available only to Google
Assistant. Even Google Assistant’s chief rival—Amazon’s Alexa—is unable to navigate these
disincentives to get significant preinstallation or functional integration on Android devices.
141. Voice search is an important, emerging access point. Internal Google documents
have recognized that the “[v]oice platform will become the future of search” and financial
projections for the assistant category recognize “search defensive value.”
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 44 of 64
45
142. Partners that depart from the preinstallation agreements risk discipline from
Google. For example, in 2011, one major electronics manufacturer considered giving a group of
consumers outside the United States a choice between two home screen experiences for their
device: one home screen with the Google search widget and a second home screen with a rival
search widget. Discussing this proposal with colleagues, one Google employee noted “[a]llowing
a mode that does not have Google as the default search provider and completely changes the
home screen” would violate Google’s terms and risk breach.
143. In 2015, Google was concerned that a major United States carrier would ask
manufacturers to install a search widget powered by the carrier’s in-house search engine.
Google’s Vice President of Partnerships wrote to a colleague that Google needed to make clear
to manufacturers that “[these] customization requests will not go far” and replacing the Google
search widget with a different search box would violate the preinstallation agreement.
Termination of this default agreement would, in turn, prohibit access to the entire GMS suite,
including Google Play and GPS, and forfeit any potential cut of Google’s search advertising
revenue under a revenue sharing agreement. In short, as the above examples illustrate, Google’s
documents show its efforts to discipline its counterparties, including major electronics companies
and carriers.
c) Revenue Sharing Agreements
144. In exchange for a substantial portion of Google’s search advertising revenues,
Android distributors agree to make Google the preset default general search engine for all
significant search access points on the device. In addition, these agreements typically contain an
exclusivity provision prohibiting the preinstallation of a competing general search service.
145. Google has recognized for some time that its revenue sharing agreements with
Android device manufacturers and carriers provide exclusivity for its general search service on
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 45 of 64
46
those devices. As stated explicitly in a draft 2014 Google strategy deck in Figures 10 and 11
below, Google’s revenue sharing arrangements with Android manufacturers or OEMs “provide
exclusivity of Search” and its deals with carriers similarly “prevent[] the pre-installation of other
Search engines or browsers,” thus enabling Google “to protect Search exclusivity on the device
as it makes its way to the user.”
Figure 10 (Android manufacturers)
Figure 11 (Android carriers)
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 46 of 64
47
146. Similarly, one Google executive acknowledged that exclusivity is “the general
philosophy of the RSA or one of the tenets of the value exchanged in the RSA.” Another Google
executive noted, “our philosophy is that we are paying revenue share *in return for* exclusivity.”
These agreements are, as that executive further explained, “really important” because “otherwise
Bing or Yahoo can come and steal away our Android search distribution at any time.”
147. As Google’s documents recognize, the preinstallation agreements and revenue
sharing agreements work together as a belt-and-suspenders strategy for driving searches to
Google (and therefore away from competitors) on Android devices. As one Google executive
explained in 2017, Google uses revenue sharing agreements “as a lever for motivating partner
behavior that is consistent with our goals for Google and the ecosystem,” and to “drive
incremental revenue (securing search defaults not covered by MADA).” By using its monopoly
profits, Google is able to secure even “more stringent requirements” on manufacturers and
carriers to obtain the preset default position on search access points not covered by the
preinstallation agreements. The combined result of Google’s preinstallation and revenue sharing
agreements is to lock up all the main pathways through which consumers access search on
Android devices, thus foreclosing rivals and protecting Google’s monopoly positions.
148. The size of Google’s payments to Android distributors demonstrates the
enormous value of default status and exclusivity provided by the agreements. Last year, Google
paid major U.S. carriers, collectively, more than a billion dollars.
149. Other channels of distribution left for competitors are far inferior to those paid for
by Google and protected by its agreements. For example, a consumer can in theory download a
competing search app on his or her own. But as one of Google’s executives bluntly put it, “most
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 47 of 64
48
users just use what comes on the device” and do not attempt to download or use other general
search services.
150. Google’s revenue-sharing partners turn down opportunities to preinstall or
otherwise enable innovative, search-related apps because those new partnerships could violate
Google’s demand for exclusivity.
151. Google also uses its agreements to ensure that new search access points are not
available to competitors. For example, Google developed a smart keyboard—a mobile app that
can be used as an alternative for the standard-issued keyboards on smart phones—with the
recognition that such keyboards might be “the next big search access point.” Google relies on its
preinstallation and default restrictions in its revenue sharing agreements as a “strategic defense”
against rival keyboards that might provide a “[b]ridge” to rival general search engines. Thus,
search queries cannot leak out to Google’s rivals even in niche areas.
152. Google likewise structures its agreements to penalize any distributor that might
walk away, tying them to Google. The typical term of the carrier and manufacturer revenue
sharing agreement is two to three years. If a carrier or manufacturer does not renew its revenue
sharing agreement with Google, the distributor loses out on revenue share not only for new
mobile devices but also for the phones and tablets previously sold and in the hands of consumers.
This provision is punitive to the carrier or manufacturer and helps to ensure that carriers and
manufacturers will not stray from Google.
153. To be attractive to a carrier or manufacturer, a rival search provider’s offer for
preset default status would need to cover not only the revenue the carrier or manufacturer would
have earned from Google for new devices, but also the revenue that the carrier or manufacturer
would have earned on all the devices that are currently in the hands of consumers. Google will
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 48 of 64
49
continue to benefit from those devices with defaults previously set to Google. A rival search
provider is left with no practical way to ensure that it will generate revenues from those devices,
regardless of how competitive its general search service might otherwise be.
154. In roughly a decade, no other general search provider has secured preset default
search status on any preinstalled search access point on GMS Android devices. As with the
Apple distribution agreements, the Android distribution agreements—taken together—are self-
reinforcing, depriving rivals of the quality, audience, and financial benefits of scale that would
allow them to mount an effective challenge to Google.
155. Particularly for newer entrants, the revenue sharing agreements present a
substantial barrier to entry. These entrants cannot pay the billions of dollars that Google does for
the most effective forms of distribution—premium placement and default status. Instead, they are
relegated to inferior forms of distribution that do not allow them to build scale, gain brand
recognition, and generate momentum to challenge Google.
B. Google Agreements Lock Up Browser Distribution
156. Beyond its agreements locking up distribution on Android and Apple devices,
Google also has entered into exclusive revenue sharing agreements with browsers. Google has
recognized it is “crucial to retain web browser partnerships.” Google’s agreements with browsers
generally require the browsers to make Google the preset default general search engine for
search access points in both the browser’s computer and mobile versions.
157. In exchange for being the preset default general search engine, Google shares up
to 40 percent of the advertising revenue it generates from these search access points with
Google’s browser rivals. Browser revenue sharing agreements typically last at least two years
and are routinely extended.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 49 of 64
50
158. Browsers are one of the most important distribution channels for general search
services because they are the gateway to the internet for most consumers. Many search queries
on mobile devices and computers are performed through the device’s browser. Today, Google
has revenue sharing agreements with the most widely used browsers in the United States, such as
Apple’s Safari browser and Mozilla’s Firefox browser; Microsoft’s browsers are the only notable
exceptions. Over 85 percent of all browser usage in the United States occurs on Google’s own
Chrome browser or on one of the browsers covered by these revenue sharing agreements.
159. In a competitive market, rivals could compete to be the preset default general
search engine on a browser. The general search services market has not, however, been
competitive for many years. When considered with Google’s other exclusionary agreements and
its monopoly power, Google’s conduct forecloses a critical avenue for search competitors to
enter the market or increase distribution. In the absence of these agreements, rival browsers
would have the ability to consider making other general search engines the preset default for
some or all search access points, spurring greater competition in the general search services
market and offering additional choices to consumers. As a Google employee once noted,
Google’s browser agreements can be “a good way to keep [a browser] away from Bing.”
C. Google Is Positioning Itself to Control the Next Generation of Search Distribution Channels
160. Although mobile phones and computers account for the vast majority of general
searches on the internet today, in the future, an increasing number of searches will likely be
conducted on next generation devices such as smart watches, smart speakers, smart TVs, and
connected automobiles. Google is positioning itself to control these emerging channels for search
distribution, excluding new and established rivals.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 50 of 64
51
161. As noted above, Google has interpreted its anti-forking agreements (AFAs and
ACCs) with Android mobile partners to cover next generation devices. Additionally, Google
uses other points of leverage in the mobile channel to discourage mobile partners from working
with rival operating systems on next generation devices. The result is that Google is positioned to
retain control over the operating systems that power next generation devices manufactured by
mobile partners and to inhibit adoption of alternative search services on those devices.
162. Google also requires connected-device manufacturers that do not sell Android
mobile phones to agree to restrictive contract terms that mirror the effects of the mobile
distribution agreements. For instance, Google partners with automobile manufacturers on the
condition that they not preinstall rival search-related apps. Google has similarly restrictive
agreements with smart watch manufacturers: its agreements to license Google’s “free” smart
watch operating system (Wear OS) prohibit manufacturers from preinstalling any third-party
software, including any rival search services.
163. Additionally, Google refuses to license its Google Assistant to IoT device
manufacturers that would host another voice assistant simultaneously—a feature commonly
known as “concurrency.” Through concurrency, a rival voice assistant could grow in popularity
to challenge Google for control over the way that consumers access the internet generally, even
on more established devices such as mobile phones. Google recognizes that concurrency is a
feature that consumers would value, but it sees too great a competitive risk from allowing
consumers to decide which voice assistant to use on a case-by-case basis.
164. Finally, Google uses its control over hardware products—including smart
speakers and Google Nest smart home products—to protect its general search monopoly. Google
recognizes that its “[h]ardware products also have HUGE defensive value in virtual assistant
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 51 of 64
52
space AND combatting query erosion in core Search business.” Looking ahead to the future of
search, Google sees that “Alexa and others may increasingly be a substitute for Search and
browsers with additional sophistication and push into screen devices.”
165. Google therefore aims to control emerging search access points to protect its
monopolies in the general search services, search advertising, and general search text advertising
markets in the present and the future. Google is poised to ensure that history repeats itself, and
that all search access points funnel users in one direction: toward Google.
VI. ANTICOMPETITIVE EFFECTS
166. Google has maintained unlawful monopolies in the general search services, search
advertising, and general search text advertising markets through its many exclusionary
agreements and other conduct that have separately and collectively harmed competition by:
a. Substantially foreclosing competition in general search services and
protecting a large majority of search queries in the United States against
any meaningful competition;
b. Excluding general search services rivals from effective distribution
channels, thereby denying rivals the necessary scale to compete effectively
in the general search services, search advertising, and general search text
advertising markets;
c. Impeding other potential distribution paths for general search services
rivals;
d. Increasing barriers to entry and excluding competition at emerging search
access points from nascent competitors on both computers and mobile
devices;
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 52 of 64
53
e. Stunting innovation in new products that could serve as alternative search
access points or disruptors to the traditional Google search model; and
f. Insulating Google from significant competitive pressure to improve its
general search, search advertising, and general search text advertising
products and services.
167. By restricting competition in general search services, Google’s conduct has
harmed consumers by reducing the quality of general search services (including dimensions such
as privacy, data protection, and use of consumer data), lessening choice in general search
services, and impeding innovation.
168. Google’s exclusionary conduct also substantially forecloses competition in the
search advertising and general search text advertising markets, harming advertisers. By
suppressing competition, Google has more power to manipulate the quantity of ad inventory and
auction dynamics in ways that allow it to charge advertisers more than it could in a competitive
market. Google can also reduce the quality of the services it provides to advertisers, including by
restricting the information it offers to advertisers about their marketing campaigns.
169. Google’s conduct also has harmed competition by impeding the distribution of
innovative apps that offer search features that would otherwise challenge Google. Google has
also harmed competition by raising rivals’ costs and foreclosing them from effective distribution
channels, such as distribution through voice assistant providers, preventing them from
meaningfully challenging Google’s monopoly in general search services.
170. Google’s monopoly in general search services also has given the company
extraordinary power as the gateway to the internet, which it uses to promote its own web content
and increase its profits. Google originally prided itself as being the “turnstile” to the internet,
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 53 of 64
54
sending users off its results pages through organic links designed to connect the user with a third-
party website that would best “answer” a user query. Over time, however, Google has pushed the
organic links further and further down the results page and featured more search advertising
results and Google’s own vertical or specialized search offerings. This, in turn, has demoted
organic links of third-party verticals, pushing these links “below-the-fold” (i.e., on the portion of
the SERP that is visible only if the user scrolls down) and requiring them to buy more search
advertising from Google to remain relevant. This raises their costs, reduces their
competitiveness, and limits their incentive and ability to invest in innovations that could be
attractive to users. Not surprisingly, investors also report being unwilling to provide funding to
vertical startups with business models similar to or potentially competitive with Google’s search
advertising monopoly.
171. Absent Google’s exclusionary agreements and other conduct, dynamic
competition for general search services would lead to higher quality search, increased consumer
choice, and a more beneficial user experience. In addition, more competitive search advertising
and general search text advertising markets would allow advertisers to purchase ads at more
attractive terms, with better quality and service. Finally, the incentives and abilities for
companies to develop and distribute innovative search products would be restored, resulting in
more options, better products, and higher consumer welfare overall.
172. The anticompetitive effects flowing from Google’s distribution agreements,
particularly when considered collectively, have allowed Google to develop and maintain
monopolies in the markets for general search services, search advertising, and general search text
advertising; these anticompetitive effects outweigh any benefits from those agreements, or those
benefits could be accomplished by less restrictive means.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 54 of 64
55
VII. VIOLATIONS ALLEGED
First Claim for Relief: Maintaining Monopoly of General Search Services in Violation of Sherman Act § 2
173. Plaintiffs incorporate the allegations of paragraphs 1 through 172 above.
174. General search services in the United States is a relevant antitrust market and
Google has monopoly power in that market.
175. Google has willfully maintained and abused its monopoly power in general search
services through anticompetitive and exclusionary distribution agreements that lock up the preset
default positions for search access points on browsers, mobile devices, computers, and other
devices; require preinstallation and prominent placement of Google’s apps; tie Google’s search
access points to Google Play and Google APIs; and other restrictions that drive queries to
Google at the expense of search rivals.
176. Google’s exclusionary conduct has foreclosed a substantial share of the general
search services market.
177. Google’s anticompetitive acts have had harmful effects on competition and
consumers.
178. The anticompetitive effects of Google’s exclusionary agreements outweigh any
procompetitive benefits in this market, or can be achieved through less restrictive means.
179. Google’s anticompetitive and exclusionary practices violate Section 2 of the
Sherman Act, 15 U.S.C. § 2.
Second Claim for Relief: Maintaining Monopoly of Search Advertising in Violation of Sherman Act § 2
180. Plaintiffs incorporate the allegations of paragraphs 1 through 172 above.
181. Search advertising in the United States is a relevant antitrust market and Google
has monopoly power in that market.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 55 of 64
56
182. Google has willfully maintained and abused its monopoly power in search
advertising through anticompetitive and exclusionary distribution agreements that lock up the
preset default positions for search access points on browsers, mobile devices, computers, and
other devices; require preinstallation and prominent placement of Google’s apps; tie Google’s
search access points to Google Play and Google APIs; and other restrictions that benefit Google
at the expense of search advertising rivals.
183. Google’s exclusionary conduct has foreclosed a substantial share of the search
advertising market.
184. Google’s anticompetitive acts have had harmful effects on competition,
advertisers, and consumers.
185. The anticompetitive effects of Google’s exclusionary agreements outweigh any
procompetitive benefits in this market, or can be achieved through less restrictive means.
186. Google’s anticompetitive and exclusionary practices violate Section 2 of the
Sherman Act, 15 U.S.C. § 2.
Third Claim for Relief: Maintaining Monopoly of General Search Text Advertising in Violation of Sherman Act § 2
187. Plaintiffs incorporate the allegations of paragraphs 1 through 172 above.
188. General search text advertising in the United States is a relevant antitrust market,
and Google has monopoly power in that market.
189. Google has willfully maintained and abused its monopoly power in general search
text advertising through anticompetitive and exclusionary distribution agreements that lock up
the preset default positions for search access points on browsers, mobile devices, computers, and
other devices; require preinstallation and prominent placement of Google’s apps; tie Google’s
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 56 of 64
57
search access points to Google Play and Google APIs; and other restrictions that benefit Google
at the expense of general search text advertising rivals.
190. Google’s exclusionary conduct has foreclosed a substantial share of the general
search text advertising market.
191. Google’s anticompetitive acts have had harmful effects on competition,
advertisers, and consumers.
192. The anticompetitive effects of Google’s exclusionary agreements outweigh any
procompetitive benefits in this market, or can be achieved through less restrictive means.
193. Google’s anticompetitive and exclusionary practices violate Section 2 of the
Sherman Act, 15 U.S.C. § 2.
VIII. REQUEST FOR RELIEF
194. To remedy these illegal acts, Plaintiffs request that the Court:
a. Adjudge and decree that Google acted unlawfully to maintain general
search services, search advertising, and general search text advertising
monopolies in violation of Section 2 of the Sherman Act, 15 U.S.C. § 2;
b. Enter structural relief as needed to cure any anticompetitive harm;
c. Enjoin Google from continuing to engage in the anticompetitive practices
described herein and from engaging in any other practices with the same
purpose and effect as the challenged practices;
d. Enter any other preliminary or permanent relief necessary and appropriate
to restore competitive conditions in the markets affected by Google’s
unlawful conduct;
e. Enter any additional relief the Court finds just and proper; and
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 57 of 64
58
f. Award each Plaintiff an amount equal to its costs incurred in bringing this
action on behalf of its citizens.
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 58 of 64
59
Respectfully submitted,
October 20, 2020
FOR PLAINTIFF UNITED STATES OF AMERICA,
/s/ Ryan A. Shores RYAN A. SHORES (D.C. Bar #500031) Associate Deputy Attorney General, Senior Advisor for Technology Industries
/s/ Alexander P. Okuliar ALEXANDER P. OKULIAR (D.C. Bar #481103) Deputy Assistant Attorney General
/s/ Michael F. Murray MICHAEL F. MURRAY (D.C. Bar #1001680) Deputy Assistant Attorney General
/s/ Kathleen S. O’Neill KATHLEEN S. O’NEILL Senior Director of Investigations and Litigation
/s/ Craig W. Conrath CRAIG W. CONRATH Director of Civil Litigation
/s/ Aaron D. Hoag AARON D. HOAG Chief, Technology & Financial Services Section
/s/ Adam T. Severt ADAM T. SEVERT Assistant Chief, Technology & Financial Services Section
/s/ Kenneth M. Dintzer KENNETH M. DINTZER* DIANA A. AGUILAR ALDAPE JESÚS M. ALVARADO-RIVERA BRENDAN BALLOU (D.C. BAR#241592) MICHAEL E. BLAISDELL (D.C. BAR#998354) ALEX COHEN THOMAS P. DEMATTEO DANIEL S. GUARNERA (D.C. BAR#1034844) JEREMY M. P. GOLDSTEIN R. CAMERON GOWER THOMAS GREENE KARL E. HERRMANN (D.C. BAR#1022464) ELIZABETH S. JENSEN MICHAEL G. MCLELLAN (D.C. BAR#489217) TAYLOR M. OWINGS (D.C. BAR#1031064) DANIEL A. PRINCIPATO RYAN M. SANDROCK DAVID J. SHAW (D.C. BAR#996525) CHRISTINE SOMMER (D.C. BAR#470763) LAUREN S. WILLARD (D.C. BAR#1024636) Attorneys for the United States U.S. Department of Justice, Antitrust Division Technology & Financial Services Section 450 Fifth St, NW, Suite 7100 Washington, D.C. 20530 Phone: 202-227-1967 Fax: 202-616-8544 Email: [email protected] * LEAD ATTORNEY TO BE NOTICED
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 59 of 64
60
FOR PLAINTIFF STATE OF ARKANSAS:
LESLIE RUTLEDGE, Attorney General
/s/ Johnathan R. Carter
JOHNATHAN R. CARTER, Assistant Attorney General
Office of the Attorney General, State of Arkansas
323 Center Street, Suite 200
Little Rock, Arkansas 72201
Phone: 501-682-8063
Email: [email protected]
FOR PLAINTIFF STATE OF FLORIDA:
ASHLEY MOODY, Attorney General
/s/ R. Scott Palmer
R. SCOTT PALMER, Special Counsel for Antitrust Enforcement
PATRICIA A. CONNERS, Chief Associate Deputy Attorney General
NICHOLAS D. NIEMIEC, Assistant Attorney General
LEE ISTRAIL, Assistant Attorney General
Office of the Attorney General, State of Florida
PL-01 The Capitol
Tallahassee, Florida 32399
Phone: 850-414-3300
Email: [email protected]
FOR PLAINTIFF STATE OF GEORGIA:
CHRISTOPHER CARR, Attorney General
/s/ Christopher Carr
CHRISTOPHER CARR, Attorney General
MARGARET ECKROTE, Deputy Attorney General
DANIEL WALSH, Senior Assistant Attorney General
DALE MARGOLIN CECKA, Assistant Attorney General
Office of the Attorney General, State of Georgia
40 Capitol Square, SW
Atlanta, Georgia 30334-1300
Phone: 404-651-7675
Email: [email protected]
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 60 of 64
61
FOR PLAINTIFF STATE OF INDIANA:
CURTIS HILL, Attorney General
/s/ Scott L. Barnhart
SCOTT L. BARNHART, Chief Counsel and Director, Consumer Protection Division
MATTHEW MICHALOSKI, Deputy Attorney General
ERICA SULLIVAN, Deputy Attorney General
Office of the Attorney General, State of Indiana
Indiana Government Center South, Fifth Floor
302 West Washington Street
Indianapolis, Indiana 46204
Phone: 317-232-6309
Email: [email protected]
FOR PLAINTIFF COMMONWEALTH OF KENTUCKY:
DANIEL CAMERON, ATTORNEY GENERAL
/s/ Justin D. Clark
JUSTIN D. CLARK, Deputy Director of Consumer Protection
J. CHRISTIAN LEWIS, Executive Director of Consumer Protection
PHILIP R. HELERINGER, Assistant Attorney General
JONATHAN E. FARMER, Assistant Attorney General
Office of the Attorney General, Commonwealth of Kentucky
1024 Capital Center Drive, Suite 200
Frankfort, Kentucky 40601
Phone: 502-696-5300
Email: [email protected]
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 61 of 64
62
FOR PLAINTIFF STATE OF LOUISIANA:
JEFF LANDRY, Attorney General
/s/ Stacie L. Deblieux
STACIE L. DEBLIEUX, Assistant Attorney General
Office of the Attorney General, State of Louisiana
Public Protection Division
1885 North Third St.
Baton Rouge, Louisiana 70802
Phone: 225-326-6400
Email: [email protected]
FOR PLAINTIFF STATE OF MISSOURI:
ERIC SCHMITT, Attorney General
/s/ Amy Haywood
AMY HAYWOOD, Chief Counsel, Consumer Protection
KIMBERLEY BIAGIOLI, Assistant Attorney General
Office of the Attorney General, State of Missouri
P.O. Box 899
Jefferson City, Missouri 65102
Phone: 573-571-3321
Email: [email protected]
FOR PLAINTIFF STATE OF MISSISSIPPI:
LYNN FITCH, Attorney General
/s/ Hart Martin
HART MARTIN, Consumer Protection Division
CRYSTAL UTLEY SECOY, Consumer Protection Division
Office of the Attorney General, State of Mississippi
P.O.Box 220
Jackson, Mississippi 39205
Phone: 601-359-4223
Email: [email protected]
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 62 of 64
63
FOR PLAINTIFF STATE OF MONTANA:
TIMOTHY C. FOX, Attorney General
/s/ Mark Mattioli
MARK MATTIOLI, Chief, Office of Consumer Protection
Office of the Attorney General, State of Montana
P.O. Box 200151
555 Fuller Avenue, 2nd Floor
Helena, MT 59620-0151
Phone: 406-444-2026
Email: [email protected]
FOR PLAINTIFF STATE OF SOUTH CAROLINA:
ALAN WILSON, Attorney General
/s/ Rebecca M. Hartner
REBECCA M. HARTNER, Assistant Attorney General
W. JEFFREY YOUNG, Chief Deputy Attorney General
C. HAVIRD JONES, JR., Senior Assistant Deputy Attorney General
MARY FRANCES JOWERS, Assistant Deputy Attorney General
Office of the Attorney General, State of South Carolina
1000 Assembly Street
Rembert C. Dennis Building
P.O.Box 11549
Columbia, South Carolina 29211-1549
Phone: 803-734-3970
Email: [email protected]
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 63 of 64
64
FOR PLAINTIFF STATE OF TEXAS:
KEN PAXTON, Attorney General
/s/ Bret Fulkerson
BRENT WEBSTER, First Assistant Attorney General
RYAN BANGERT, Deputy First Assistant Attorney General
DARREN MCCARTY, Deputy Attorney General for Civil Litigation
KIM VAN WINKLE, Chief, Antitrust Division
BRET FULKERSON, Deputy Chief, Antitrust Division
KELSEY PAINE, Assistant Attorney General
Office of the Attorney General, State of Texas
300 West 15th Street
Austin, Texas 78701
Phone: 512-936-1674
Email: [email protected]
Case 1:20-cv-03010 Document 1 Filed 10/20/20 Page 64 of 64
376
These practices have recently come under scrutiny by antitrust authorities. In March 2019,
Spotify filed a complaint against Apple before the European Commission, reportedly alleging, among
other things, that Apple is restricting Spotify’s access to Siri.2402 In July 2020, the European
Commission’s antitrust authority announced that it had opened an inquiry into the use of digital
assistants and smart home products by Apple, Google, and Amazon, among other companies.2403 In her
statement accompanying the announcement, Margrethe Vestager, the Commission’s Executive Vice
President, identified interoperability and self-preferencing as areas of concern.2404
RECOMMENDATIONS
As part of its top-to-bottom review of competition in digital markets, the Subcommittee
examined whether current laws and enforcement levels are adequate to address the market power
concerns identified through this investigation. In pursuit of this goal, on March 13, 2020, the
Subcommittee requested submissions from antitrust and competition policy experts. These experts
were chosen on a careful, bipartisan basis to ensure the representation of a full range of views.
Throughout the investigation, the Subcommittee received additional submissions and written
statements from antitrust enforcers and other leading experts, including Margrethe Vestager, the
Executive Vice President of the European Commission, and Rod Sims, the Chair of the Australian
Competition and Consumer Commission. Most recently, the Subcommittee held an oversight hearing
on October 1, 2020 regarding “Proposals to Strengthen the Antitrust Laws and Restore Competition
Online,” its seventh and final hearing as part of the investigation.
2402 Thomas Ricker, Apple to be formally investigated over Spotify’s antitrust complaint, says report, THE VERGE (MAY 6,
2019), https://www.theverge.com/2019/5/6/18530894/apple-music-monopoly-spotify-app-store-europe.
2403 Margrethe Vestager, Exec. Vice Pres., Eur. Comm’n, Statement by Executive Vice-President Margrethe Vestager on
the launch of a Sector Inquiry on the Consumer Internet of Things (July 16, 2020),
https://ec.europa.eu/commission/presscorner/detail/en/speech_20_1367.
2404 Id.
377
Subcommittee Chairman David N. Cicilline (D-RI) requested that staff provide Members of the
Subcommittee with a series of recommendations, informed by this investigation, on how to strengthen
the antitrust laws and restore competition online. As he noted in remarks to the American Antitrust
Institute in June 2019:
No doubt, other branches of government have a key role to play in the development of
antitrust law. But Congress—not the courts, agencies, or private companies—enacted
the antitrust laws, and Congress ultimately decides what the law should be and whether
the law is working for the American people. As such, it is Congress’ responsibility to
conduct oversight of our antitrust laws and competition system to ensure that they are
properly working and to enact changes when they are not. While I do not have any
preconceived ideas about what the right answer is, as Chairman of the Antitrust
Subcommittee, I intend to carry out that responsibility with the sense of urgency and
serious deliberation that it demands.2405
In response to this request, Subcommittee staff identified a broad set of reforms for further
examination by the Members of the Subcommittee for purposes of crafting legislative and oversight
responses to the findings of this Report. These reforms include proposals to: (1) promote fair
competition in digital markets; (2) strengthen laws relating to mergers and monopolization; and (3)
restore vigorous oversight and enforcement of the antitrust laws.
Subcommittee staff intends for these recommendations to serve as a complement, not a
substitute, to strong enforcement of the antitrust laws. This is particularly true for acquisitions by
dominant firms that may have substantially lessened competition or tended to create a monopoly in
violation of the Clayton Act. In these cases, Subcommittee staff supports as a policy matter the
examination of the full range of remedies—including unwinding consummated acquisitions or
divesting business lines—to fully restore competition that was harmed as a result of these acquisitions
and to prevent future violations of the antitrust laws.2406
A. Restoring Competition in the Digital Economy
For more than a century, Congress has addressed the market power of dominant intermediaries
using a robust antitrust and antimonopoly toolkit.2407 The antitrust laws prohibit anticompetitive
2405 Hon. David N. Cicilline, Chairman, Subcomm. on Antitrust, Commercial and Admin. Law of the H. Comm. on the
Judiciary, Keynote Address at American Antitrust Institute’s 20th Annual Policy Conference (June 20, 2019),
https://cicilline.house.gov/press-release/cicilline-delivers-keynote-address-american-antitrust-institute%E2%80%99s-20th-
annual-policy.
2406 Due to separation of powers concerns and other relevant considerations, we do not take a position on the outcome of
any individual matter before the Justice Department or Federal Trade Commission.
2407 See, e.g., Subcomm. on Study of Monopoly Power of the H. Comm. on the Judiciary, 81st Cong. 2d Sess., The Antitrust
Laws: A Basis for Economic Freedom iii (1950) (identifying an extensive list of statutes “dealing directly with the
378
mergers and monopolistic conduct in order to promote open markets and prevent undue concentration
of economic power. In many critical sectors of the economy—including financial services,
telecommunications, and transportation—Congress has also relied on a broad set of policies to create
the conditions necessary for fair competition, even when economies of scale may favor concentration.
In a similar vein, the remedies identified in this section seek to restore competition online by
addressing harmful business practices as well as certain features of digital markets that tend to tip the
market towards concentration.
Reduce Conflicts of Interest Thorough Structural Separations and Line of Business Restrictions
In addition to controlling one or multiple key channels of distribution, the dominant firms
investigated by the Subcommittee are integrated across lines of business. When operating in adjacent
markets, these platforms compete directly with companies that depend on them to access users, giving
rise to a conflict of interest. As discussed earlier in this Report, the Subcommittee’s investigation
uncovered several ways in which Amazon, Apple, Facebook, and Google use their dominance in one
or more markets to advantage their other lines of business, reducing dynamism and innovation.
First, the investigation revealed that the dominant platforms have misappropriated the data of
third parties that rely on their platforms, effectively collecting information from customers only to
weaponize it against them as rivals. For example, the investigation produced documents showing that
Google used the Android operating system to closely track usage trends and growth patterns of third-
party apps—near-perfect market intelligence that Google can use to gain an edge over those same
apps. Facebook used its platform tools to identify and then acquire fast-growing third-party apps,
thwarting competitive threats at key moments. A former Amazon employee told the Subcommittee that
Amazon has used the data of third-party merchants to inform Amazon’s own private label strategy,
identifying which third-party products were selling well and then introducing copycat versions. These
and other examples detailed in this Report demonstrate a dangerous pattern of predatory conduct that,
if left unchecked, risk further concentrating wealth and power.
Some have suggested that there is little difference between the dominant platforms’ access to
and use of this data and the way that brick-and-mortar retailers track popular products. The
Subcommittee’s investigation, however, produced evidence that the platforms’ access to competitively
significant market data is unique. Specifically, the dominant platforms collect real-time data which,
given the scale of their user-base, is akin to near-perfect market intelligence. Whereas firms with a
choice among business partners might seek to protect their proprietary data, the platforms’ market
power lets them compel the collection of this data in the first place.
preservation of the American competitive economy” and reflecting the legislative policy that “under no circumstances
should [laws] foster the growth of monopoly”).
379
Second, dominant platforms can exploit their integration by using their dominance in one
market as leverage in negotiations in an unrelated line of business. For example, evidence produced
during the investigation showed that Amazon has leveraged its dominance in online commerce as
pressure during negotiations with firms in a separate line of business. Market participants that depend
on Amazon’s retail platform are effectively forced to accept its demands—even in markets where
Amazon would otherwise lack the power to set the terms of commerce.
Third, dominant platforms have used their integration to tie products and services in ways that
can lock in users and insulate the platform from competition. Google, for example, required that
smartphone manufacturers seeking to use Android also pre-install and give default status to certain
Google apps—enabling Google to maintain its search monopoly and crowd out opportunities for third-
party developers.
And fourth, these firms can use supra-competitive profits from the markets they dominate to
subsidize their entry into other markets. Documents uncovered during the Subcommittee’s
investigation indicate that the dominant platforms have relied on this strategy to capture markets, as
startups and non-platform businesses tend to lack the resources and capacity to bleed billions of dollars
over multiple years in order to drive out rivals. For dominant platforms, meanwhile, this strategy
appears to be a race to capture ecosystems and control interlocking products that funnel data back to
the platforms, further reinforcing their dominance.
By using market power in one area to advantage a separate line of business, dominant firms
undermine competition on the merits. By functioning as critical intermediaries that are also integrated
across lines of business, the dominant platforms face a core conflict of interest. The surveillance data
they collect through their intermediary role, meanwhile, lets them exploit that conflict with unrivaled
precision. Their ability both to use their dominance in one market as negotiating leverage in another,
and to subsidize entry to capture unrelated markets, have the effect of spreading concentration from
one market into others, threatening greater and greater portions of the digital economy.
To address this underlying conflict of interest, Subcommittee staff recommends that Congress
consider legislation that draws on two mainstay tools of the antimonopoly toolkit: structural separation
and line of business restrictions.2408 Structural separations prohibit a dominant intermediary from
operating in markets that place the intermediary in competition with the firms dependent on its
infrastructure. Line of business restrictions, meanwhile, generally limit the markets in which a
dominant firm can engage.
2408 See Submission from Sally Hubbard, Dir. of Enforcement Strategy, Open Mkts. Inst. et al., to H. Comm. on the
Judiciary, 7–8 (Apr. 17, 2020) (on file with Comm.) [hereinafter Hubbard Submission]; Submission from Stacy Mitchell,
Co-Dir., Inst. for Local Self-Reliance, to H. Comm. on the Judiciary, 4 (May 4, 2020) (on file with Comm.) [hereinafter
Mitchell Submission]; Submission from Zephyr Teachout, Assoc. Prof. of Law, Fordham Univ. Sch. of Law, to H. Comm.
on the Judiciary, 6 (Apr. 23, 2020) (on file with Comm.) [hereinafter Teachout Submission]; Submission from Ams. for
Fin. Reform, to H. Comm. on the Judiciary, 3–4 (Apr. 17, 2020) (on file with Comm.).
380
Congress has relied on both policy tools as part of a standard remedy for dominant
intermediaries in other network industries, including railroads and telecommunications services.2409 In
the railroad industry, for example, a congressional investigation found that the expansion of common
carrier railroads’ into the coal market undermined independent coal producers, whose wares the
railroads would deprioritize in order to give themselves superior access to markets. In 1893, the
Committee on Interstate and Foreign Commerce wrote that “[n]o competition can exist between two
producers of a commodity when one of them has the power to prescribe both the price and output of
the other.”2410
Congress subsequently enacted a provision to prohibit railroads from transporting any goods
that they had produced or in which they held an interest.2411 Congress has legislated similar
prohibitions in other markets. The Bank Holding Company Act of 1956 broadly prohibited bank
holding companies from acquiring nonbanking companies.2412 Vertically integrated television
networks, meanwhile, were subject to “fin-syn” rules, which prohibited networks from entering
production and syndication markets.2413
Both structural separations and line of business restrictions seek to eliminate the conflict of
interest faced by a dominant intermediary when it enters markets that place it in competition with
dependent businesses. In certain cases, structural separations have also been used to prevent
monopolistic firms from subsidizing entry into competitive markets and to promote media
diversity.2414
At a general level, there are two forms of structural separation: (1) ownership separations,
which require divestiture and separate ownership of each business; and (2) functional separations,
which permit a single corporate entity to engage in multiple lines of business but prescribe the
particular organizational form it must take.2415 Importantly, both forms of structural limits apply on a
market-wide basis, while divestitures in antitrust enforcement generally apply to a single firm or
merging party.
2409 Mitchell Submission at 4.
2410 H.R. REP. NO. 52-2278, vii–viii (1893).
2411 Hepburn Act, Pub. L. No. 59-337, § 1, 34 Stat. 584, 585 (1906).
2412 Bank Holding Company Act of 1956, Pub. L. No. 84-511, § 2(a), 70 Stat. 133, 133 (codified as amended at 12 U.S.C. §
1841(a) (2012)).
2413 Competition & Responsibility in Network Television Broad., 23 F.C.C.2d 382, 398, para. 30 (1970) (report and order).
2414 Mitchell Submission at 4.
2415 John Kwoka & Tommaso Valletti, Scrambled Eggs and Paralyzed Policy: Breaking Up Consummated Mergers and
Dominant Firms 22 (forthcoming Oct. 2020) (on file with Comm.).
381
A benefit of these proposals is their administrability. By setting rules for the underlying
structure of the market—rather than policing anticompetitive conduct on an ad hoc basis—structural
rules are easier to administer than conduct remedies, which can require close and continuous
monitoring.2416
The challenges of crafting and implementing structural solutions vary by market and market
participants. In response to the Subcommittee’s requests for comments on potential reforms, some
antitrust experts have cautioned that crafting separations can pose a major cost and challenge,
especially in dynamic markets.2417 Others have responded by identifying certain principles that can
make identifying the fault lines easier. In the case of separations undoing vertical mergers, the fault
lines designating the separate companies are likely to still be apparent, even in the new structure.2418 In
cases where a firm grew through internal expansion or when the constituent parts are no longer clearly
distinguishable, scholars have suggested identifying distinct business operations.2419 Experts have also
noted that business-initiated corporate restructuring and divestitures may in some cases also provide a
guide to designing and implementing successful break-ups.2420
Several enforcement bodies around the world are exploring the use of structural separations in
digital markets. In July 2020, the United Kingdom’s Competition and Markets Authority
recommended that its digital regulatory body have powers to “implement ownership separation or
operational separation,” concluding that “there could be significant benefits if there were more formal
separation between businesses with market power” in digital advertising markets in particular.2421
Meanwhile, the OECD in 2001 adopted recommendations to structurally separate vertically integrated
regulated firms that operate in concentrated markets.2422 In its 15-year overview, the OECD concluded
that “structural separation remains a relevant remedy” and identified other market areas where it might
be adopted.2423
2416 OECD, STRUCTURAL SEPARATION IN REGULATED INDUSTRIES: REPORT ON IMPLEMENTING THE OECD
RECOMMENDATION 9 (2016), https://www.oecd.org/daf/competition/Structural-separation-in-regulated-industries-
2016report-en.pdf (“[S]eparation limits the need for regulation that is difficult and costly to devise and implement, and may
be only partly effective; it improves information; and it eliminates the risk of cross-subsidies by the incumbent from its
non-competitive to its competitive segments.”).
2417 See, e.g., Submission from Maureen K. Ohlhausen, Partner, Baker Botts L.L.P., to H. Comm. on the Judiciary, 5 (Apr.
17, 2020) (on file with Comm.).
2418 John Kwoka & Tommaso Valletti, Scrambled Eggs and Paralyzed Policy: Breaking Up Consummated Mergers and
Dominant Firms 11 (forthcoming Oct. 2020) (on file with Comm.).
2419 Id. at 15.
2420 Id.; Rory Van Loo, In Defense of Breakups: Administering a ‘Radical’ Remedy, 105 CORNELL L. REV (forthcoming
2020), https://ssrn.com/abstract=3646630.
2421 Competition & Mkts. Auth. Report at 405–06.
2422 OECD, STRUCTURAL SEPARATION IN REGULATED INDUSTRIES: REPORT ON IMPLEMENTING THE OECD
RECOMMENDATION 9 (2016), https://www.oecd.org/daf/competition/Structural-separation-in-regulated-industries-
2016report-en.pdf.
2423 Id. at 3.
382
Implement Rules to Prevent Discrimination, Favoritism, and Self-Preferencing
As discussed throughout this Report, the Subcommittee identified numerous instances in which
dominant platforms engaged in preferential or discriminatory treatment. In some cases, the dominant
platform privileged its own products or services. In others, a dominant platform gave preferential
treatment to one business partner over others. Because the dominant platform was, in most instances,
the only viable path to market, its discriminatory treatment had the effect of picking winners and losers
in the marketplace.
Google, for example, engaged in self-preferencing by systematically ranking its own content
above third-party content, even when its content was inferior or less relevant for users. Web publishers
of content that Google demoted suffered economic losses and had no way of competing on the merits.
Over the course of the investigation, numerous third parties also told the Subcommittee that self-
preferencing and discriminatory treatment by the dominant platforms forced businesses to lay off
employees and divert resources away from developing new products and towards paying a dominant
platform for advertisements or other ancillary services. They added that some of the harmful business
practices of the platforms discouraged investors from supporting their business and made it
challenging to grow and sustain a business even with highly popular products. Without the opportunity
to compete fairly, businesses and entrepreneurs are dissuaded from investing and, over the long term,
innovation suffers.
In response to these concerns, the Subcommittee recommends that Congress consider
establishing nondiscrimination rules to ensure fair competition and to promote innovation online.
Nondiscrimination rules would require dominant platforms to offer equal terms for equal service and
would apply to price as well as to terms of access. As several experts noted, nondiscrimination has
been a mainstay principle for governing network intermediaries, especially those that play essential
roles in facilitating transportation and communications.2424
The 1887 Interstate Commerce Act, for example, prohibited discriminatory treatment by
railroads.2425 In the century years since, Congress and policymakers have continued to apply
nondiscrimination principles to network monopolies, even as technologies have rapidly evolved. Most
recently, the Open Internet Order written by the Federal Communications Commission (FCC) in 2015
was effectively a nondiscrimination regime, prohibiting internet service providers from picking
2424 See, e.g., Submission from Harry First, Charles L. Denison Prof. of Law, N.Y.U. Sch. of Law & Eleanor Fox, Walter J.
Derenberg Prof. of Trade Reg., N.Y.U. Sch. of Law, to H. Comm. on the Judiciary (Aug. 6, 2020) [hereinafter First & Fox
Submission] (“[Google, Amazon, Facebook, and Apple] are akin to essential facilities for many smaller businesses. Many
businesses, to do business, must use the platform. They have almost no choice. The GAFA compete with the businesses on
their platforms.”) (on file with Comm.); Submission from Albert A. Foer, Founder & Sr. Fellow, Am. Antitrust Inst., to H.
Comm. on the Judiciary, 1–2 (Apr. 14, 2020) (on file with Comm.) [hereinafter Foer Submission]; Hubbard Submission at
5–7; Remedies Hearing 6–7 (statement of K. Sabeel Rahman, Pres., Demos).
2425 Hubbard Submission at 4–5.
383
winners and losers among content providers and other users.2426 Other jurisdictions have begun to
apply nondiscrimination principles to digital markets. For example, after determining that Google had
engaged in illegal self-preferencing, the European Commission required that Google follow “the
simple principle of equal treatment.”2427
Historically, Congress has implemented nondiscrimination requirements in a variety of
markets. With railroads, the Interstate Commerce Commission oversaw obligations and prohibitions
applied to railroads designated as common carriers.2428 More recently, the Cable Act of 1992 included
a provision requiring the Federal Communications Commission to oversee a nondiscrimination
requirement for cable operators.2429 Some experts have proposed establishing a similar venue to
adjudicate discrimination disputes between dominant platforms and the third parties that depend on
them.2430 Others note that the Federal Trade Commission could also use its existing competition
rulemaking authority to “require dominant gatekeepers to apply a rule of neutrality in operating their
platforms.”2431
Finally, on several occasions, nondiscrimination rules have been treated as an important
complement to divestitures in antitrust enforcement. For example, the Justice Department combined
AT&T’s divestiture of the Regional Bell Operating Companies with an equal access obligation,
requiring AT&T to offer independent long-distance providers access to its network on equal terms of
quality and price.2432 The DOJ argued that requiring equal access without mandating divestiture would
be insufficient due to AT&T’s incentive and ability to discriminate against local carriers.2433
2426 Protecting and Promoting the Open Internet, 30 FCC Rcd. 5601, 5603, para. 4 (2015) (“[C]arefully-tailored rules that
would prevent specific practices we know are harmful to Internet openness—blocking, throttling, and paid prioritization—
as well as a strong standard of conduct designed to prevent the deployment of new practices that would harm Internet
openness.”).
2427 Press Release, Eur. Comm’n, Antitrust: Commission Fines Google €2.42 Billion for Abusing Dominance as Search
Engine by Giving Illegal Advantage to Own Comparison Shopping Service (June 27, 2017),
https://ec.europa.eu/commission/presscorner/detail/en/MEMO_17_1785.
2428 Hubbard Submission at 5.
2429 See, e.g., Submission from Hal Singer, Managing Dir., Econ One Research, to H. Comm. on the Judiciary, 4–5 (Mar.
30, 2020) (on file with Comm.) [hereinafter Singer Submission].
2430 Id.
2431 First & Fox Submission at 12.
2432 See United States v. AT&T Co., 552 F. Supp. 131 (D.D.C. 1982).
2433 Mitchell Submission at 4 (“It’s important to note here that applying this kind of [nondiscrimination-based] regulatory
oversight to the big tech firms will not be effective unless it’s done in conjunction with breakups. In the case of Amazon,
it’s my view that several factors make it virtually impossible to establish a system of oversight and adjudication that would
be robust enough to protect competition and fair market access, absent spinning off its shopping platform from its other
divisions. These factors include the enormous number of sellers and transactions, the low dollar value of most transactions,
and the many subtle and hard-to-detect ways that Amazon can skew outcomes to favor its own interests. Therefore,
oversight must be combined with structural separation, which would do much of the work by removing the underlying
conflicts of interest, thus allowing for an effective and less bureaucratic system of oversight.”).
384
Promote Innovation Through Interoperability and Open Access
As discussed elsewhere in the Report, digital markets have certain characteristics—such as
network effects, switching costs, and other entry barriers—that make them prone to tipping in favor of
a single dominant firm. As a result, these markets are no longer contestable by new entrants,2434 and
the competitive process shifts from “competition in the market to competition for the market.”2435
This dynamic is particularly evident in the social networking market. As discussed earlier in the
Report, Facebook’s internal documents and communications indicate that due to strong network effects
and market tipping, the most significant competitive pressure to Facebook is from within its own
family of products—Facebook, Instagram, Messenger, and WhatsApp—rather than from other social
apps in the market, such as Snapchat or Twitter. In the case of messaging apps, Facebook’s documents
show that network effects can be even more extreme. And because Facebook is not interoperable with
other social networks, its users face high costs to switch to other platforms, locking them into
Facebook’s platform.
High switching costs are also present in other markets. In the smartphone market, switching
costs include learning a new operating system, which can discourage users from leaving Google or
Apple due to familiarity with their distinct operating systems, as well as the inability to easily port all
of their data, such as messages, call history, and photos. In online commerce, sellers have high
switching costs associated with their reputation. Sellers can be locked into an incumbent platform for
online commerce if they are unable to transfer their reputation—ratings and customer reviews accrued
over a long period of time—to a different platform. Switching costs involving data for other services,
such as email, can also contribute to user lock-in.2436 In response to these concerns, Subcommittee staff
recommends that Congress consider data interoperability and portability to encourage competition by
lowering entry barriers for competitors and switching costs for consumers. These reforms would
complement vigorous antitrust enforcement by spurring competitive entry.
a. Interoperability
Interoperability is fundamental to the open internet.2437 It is present in email, which is an open,
interoperable protocol for communicating online regardless of a person’s email service or the type of
2434 Competition & Mkts. Auth. Report at 10–11.
2435 See Stigler Report at 29; Michael Kades & Fiona Scott Morton, Interoperability as a Competition Remedy for Digital
Networks, WASH. CTR. FOR EQUITABLE GROWTH 1 (Sept. 2020) (on file with Comm.) (“The monopolist operates in a
market with significant network effects, scale and scope economies, and low distribution costs. Therefore, the competition
that matters most is often for the market not within the market. Anticompetitive conduct is more likely to succeed. And, the
harm to consumers greater because the market tends to be winner-take-all, or most.”).
2436 Chris Riley, A Framework for Forward-Looking Tech Competition Policy, MOZILLA 10 (2019),
https://blog.mozilla.org/netpolicy/files/2019/09/Mozilla-Competition-Working-Paper.pdf.
2437 See generally id. at 18–24.
385
device they use to send the email.2438 It has also been built into numerous other services online2439 and
is a “core technical structure of the Internet.”2440 Interoperability standards are also present in other
communications systems, from telephones to telegraphs.2441 Telecommunications would not work
without the ability of users on one carrier’s network to interconnect with other carriers.2442 And in the
absence of interoperability, dominant carriers could foreclose new entrants from offering lower prices
or better services, reinforcing their monopoly power while harming consumers and competition.2443
An interoperability requirement would allow competing social networking platforms to
interconnect with dominant firms to ensure that users can communicate across services.2444 Foremost,
interoperability “breaks the power of network effects” by allowing new entrants to take advantage of
existing network effects “at the level of the market, not the level of the company.”2445 It would also
lower switching costs for users by ensuring that they do not lose access to their network as a result of
switching.
The implementation cost of requiring interoperability by dominant firms would be relatively
low. Unlike interconnecting in traditional communications markets, there is little direct cost associated
with interoperating with dominant platforms.2446
2438 Michael Kades & Fiona Scott Morton, Interoperability as a Competition Remedy for Digital Networks 14 (Sept. 2020)
(on file with Comm.).
2439 Becky Chao & Russ Schulman, Promoting Platform Interoperability, NEW AM. FOUND. (May 13, 2020),
https://www.newamerica.org/oti/reports/promoting-platform-interoperability/.
2440 Chris Riley, A Framework for Forward-Looking Tech Competition Policy, MOZILLA 18 (2019),
https://blog.mozilla.org/netpolicy/files/2019/09/Mozilla-Competition-Working-Paper.pdf.
2441 Becky Chao & Russ Schulman, Promoting Platform interoperability, NEW AM. FOUND. (May 13, 2020),
https://www.newamerica.org/oti/reports/promoting-platform-interoperability/.
2442 Michael Kades & Fiona Scott Morton, Interoperability as a Competition Remedy for Digital Networks 13–14 (Sept.
2020) (on file with Comm.).
2443 Id.
2444 Competition in Digital Technology Markets: Examining Self-Preferencing by Digital Platforms: Hearing Before
Subcomm. on Antitrust, Competition Policy and Consumer Rights of the S. Comm. on the Judiciary, 116th Cong. 21 (2020)
(statement of Sally Hubbard, Dir. of Enforcement Strategy, Open Mkts. Inst.) (“Interoperability is an anti-monopoly tool
that has been used successfully many times to promote innovation by reducing barriers to entering markets.”).
2445 Michael Kades & Fiona Scott Morton, Interoperability as a Competition Remedy for Digital Networks 13–14 (Sept.
2020) (on file with Comm.).
2446 Id. at 15 (“Unlike the familiar AT&T example, there would be no cost to interconnection in the digital platform context.
The standard is simply a way to present and transfer information that is already being presented and transferred. No wire
needs to be connected to achieve it, nor do machines need to be co-located, or special workers employed. Transferring
digital files has almost zero cost, but regardless of that cost, Facebook would be transferring those files to serve its users in
any case. Facebook might need to pay some costs to redesign the format in which it transfers text and images, but if it has
been found liable for monopolization by a court, it is expected that a remedy will have costs. The real cost of ongoing
interoperability to Facebook.com is the possibility that it loses customers once the barriers to entry fall. But that risk is what
every firm faces in a competitive market and represents a benefit to consumers.”).
386
Finally, interoperability is an important complement, not substitute, to vigorous antitrust
enforcement. As discussed in this Report, Facebook has tipped the social network toward a monopoly,
and due to its strong network effects, does not face competitive pressure. On its own, interoperability is
unlikely to fully restore competition in the social networking market due to the lack of meaningful
competition in the market today. On the other hand, in the absence of pro-competitive policies like
interoperability, it is also possible that enforcement alone may provide incomplete relief due to future
market tipping.2447
b. Data Portability
Data portability is also a remedy for high costs associated with leaving a dominant platform.
These costs present another barrier to entry for competitors and a barrier to exit for consumers.
Dominant platforms can maintain market power in part because consumers experience significant
frictions when moving to a new product.2448 Users contribute data to a platform, for example, but can
find it hard to migrate that data to a rival platform.2449 The difficulty of switching tends to keep users
on incumbent platforms.2450 Providing consumers and businesses with tools to easily port or rebuild
their social graph, profile, or other relevant data on a competing platform would help address these
concerns.2451 Although complementary to interoperability, data portability alone would not fully
address concerns related to network effects since consumers would still need to recreate their networks
on a new platform and would not be able to communicate with their network on the incumbent
platform.2452
2447 Id. at 10. (“A divestiture may reduce the existing market power of the dominant network but not eliminate the market
power due to network effects that was achieved through anticompetitive conduct. And, alone, divestiture may not prevent
future tipping. Thus, on their own, they risk being insufficient to fully restore the lost competition.”).
2448 See JOSHUA GANS, THE HAMILTON PROJECT, ENHANCING COMPETITION WITH DATA AND IDENTITY PORTABILITY 5
(June 2018), http://www.hamiltonproject.org/assets/files/Gans_20180611.pdf.
2449 See id.
2450 See Josh Constine, Friend Portability Is the Must-Have Facebook Regulation, TECHCRUNCH (May 12, 2019),
https://technologycrunch.com/2019/05/12/friends-wherever; Chris Dixon, The Interoperability of Social Networks, BUS.
INSIDER (Nov. 10, 2010), https://www.businessinsider.com/the-interoperability-of-social-networks-2011-2; Data and
Privacy Hearing at 2 (statement of Dina Srinivasan, Fellow, Yale Thurman Arnold Project).
2451 Submission from Charlotte Slaiman, Competition Policy Dir., Public Knowledge, to H. Comm. on the Judiciary (May
14, 2020) (on file with Comm.); Appendix I at 3–4 (statement of Gene Kimmelman, Sr. Advisor, Public Knowledge)
[hereinafter Slaiman Submission].
2452 Competition in Digital Technology Markets: Examining Self-Preferencing by Digital Platforms: Hearing Before
Subcomm. on Antitrust, Competition Policy and Consumer Rights of the S. Comm. on the Judiciary, 116th Cong. 21 (2020)
(statement of Sally Hubbard, Dir. of Enforcement Strategy, Open Mkts. Inst.) (on file with Comm.). Last year, Senators
Mark R. Warner (D-VA), Josh Hawley (R-MO), and Richard Blumenthal (D-CT) introduced S.2648, the “Augmenting
Compatibility and Competition by Enabling Service Switching (ACCESS) Act,” bipartisan legislation to require that
dominant platforms make user data portable and their services interoperable. Additionally, this proposal would also allow
users to delegate management of their privacy preferences to a third-party service. Press Release, Sen. Mark R. Warner,
Senators Introduce Bipartisan Bill to Encourage Competition in Social Media (Oct. 22, 2019),
https://www.warner.senate.gov/public/index.cfm/2019/10/senators-introduce-bipartisan-bill-to-encourage-competition-in-
social-media.
387
Reduce Market Power Through Merger Presumptions
The firms investigated by the Subcommittee owe part of their dominance to mergers and
acquisitions. Several of the platforms built entire lines of business through acquisitions, while others
used acquisitions at key moments to neutralize competitive threats. Although the dominant platforms
collectively engaged in several hundred mergers and acquisitions between 2000-2019, antitrust
enforcers did not block a single one of these transactions. The Subcommittee’s investigation revealed
that several of these acquisitions enabled the dominant platforms to block emerging rivals and
undermine competition.
Despite a significant number of ongoing antitrust investigations, the dominant platforms have
continued to pursue significant deal-making. Over the last year, for example, Google purchased Fitbit
for $2.1 billion and Looker for $2.6 billion; Amazon purchased Zoox for $1.3 billion; and Facebook
acquired Giphy for an undisclosed amount.2453 Meanwhile, all four of the firms investigated by the
Subcommittee have recently focused on acquiring startups in the artificial intelligence and virtual
reality space.2454
Ongoing acquisitions by the dominant platforms raise several concerns. Insofar as any
transaction entrenches their existing position, or eliminates a nascent competitor, it strengthens their
market power and can close off market entry. Furthermore, by pursuing additional deals in artificial
intelligence and in other emerging markets, the dominant firms of today could position themselves to
control the technology of tomorrow.
It is unclear whether the antitrust agencies are presently equipped to block anticompetitive
mergers in digital markets. The record of the Federal Trade Commission and the Justice Department in
this area shows significant missteps and repeat enforcement failures. While both agencies are currently
pursuing reviews of pending transactions, it is not yet clear whether they have developed the analytical
tools to challenge anticompetitive deals in digital markets. For example, the Justice Department in
February permitted Google’s acquisition of Looker, a data analytics and business intelligence startup,
despite serious risks that the deal would eliminate an independent rival and could allow Google to cut
2453 Chaim Gartenberg, Google buys Fitbit for $2.1 billion, THE VERGE (Nov. 1, 2019),
https://www.theverge.com/2019/11/1/20943318/google-fitbit-acquisition-fitness-tracker-announcement; Lauren Feiner &
Jordan Novet, Google cloud boss Thomas Kurian makes his first big move — buys Looker for $2.6 billion, CNBC (June 6,
2019), https://www.cnbc.com/2019/06/06/google-buys-cloud-company-looker-for-2point6-billion.html; Karen Weise &
Erin Griffith, Amazon to Buy Zoox, in a Move Toward Self-Driving Cars, N.Y. TIMES (June 26, 2020),
https://www.nytimes.com/2020/06/26/business/amazon-zoox.html; Kurt Wagner & Sarah Frier, Facebook Buys Animated
Image Library Giphy for $400 Million, BLOOMBERG (May 15, 2020), https://www.bloomberg.com/news/articles/2020-05-
15/facebook-buys-animated-image-library-giphy-to-boost-messaging.
2454 See infra Appendix.
388
off access to rivals.2455 These concerns are especially acute today, given the combined national health
and economic crises, which have widened the gap between the dominant platforms and businesses
across the rest of the economy.
To address this concern, Subcommittee staff recommends that Congress consider shifting
presumptions for future acquisitions by the dominant platforms. Under this change, any acquisition by
a dominant platform would be presumed anticompetitive unless the merging parties could show that
the transaction was necessary for serving the public interest and that similar benefits could not be
achieved through internal growth and expansion. This process would occur outside the current Hart-
Scott-Rodino Act (HSR) process, such that the dominant platforms would be required to report all
transactions and no HSR deadlines would be triggered. Establishing this presumption would better
reflect Congress’s preference for growth through ingenuity and investment rather than through
acquisition.
Create an Even Playing Field for the Free and Diverse Press
The free and diverse press—particularly local press—is the backbone of a healthy and vibrant
democracy. But as discussed in this Report, the rise of market power online has corresponded with a
significant decline in the availability of trustworthy sources of news.2456 Through dominating both
digital advertising and key communication platforms, Google and Facebook have outsized power over
the distribution and monetization of trustworthy sources of news online,2457 creating an uneven playing
field in which news publishers are beholden to their decisions.2458
To address this imbalance of bargaining power, we recommend that the Subcommittee consider
legislation to provide news publishers and broadcasters with a narrowly tailored and temporary safe
harbor to collectively negotiate with dominant online platforms.
In April 2019, Subcommittee Chairman Cicilline and Doug Collins (R-GA), the former-
Ranking Member of the Committee on the Judiciary, introduced H.R. 2054, the “Journalism
2455 Letter from Diana L. Moss, Pres., Am. Antitrust Inst., to Hon. Makan Delrahim, Assistant Att’y Gen., U.S. Dep’t of
Justice, Antitrust Div. (July 8, 2019), https://www.antitrustinstitute.org/wp-content/uploads/2019/07/AAI-Ltr-to-
DOJ_Google-Looker_7.8.19.pdf.
2456 Free and Diverse Press Hearing at 3 (statement of David Chavern, Pres. & CEO, News Media Alliance) (“In effect, a
couple of dominant tech platforms are acting as regulators of the digital news industry.”).
2457 Submission of Source 52, to H. Comm. on the Judiciary, 12 (Oct. 30, 2019) (on file with Comm.).
2458 Submission from Source 53, to H. Comm. on the Judiciary, 7 (Oct. 14, 2019) (on file with Comm.). Although Apple
News and Apple News Plus are increasingly popular news aggregators, most market participants that the Subcommittee
received evidence from during the investigation do not view it as a critical intermediary for online news at this time. Some
publishers raised competition concerns about the tying of payment inside Apple’s news product.
389
Competition and Preservation Act of 2019.”2459 H.R. 2054 would allow coordination by news
publishers under the antitrust laws if it: (1) directly relates to the quality, accuracy, attribution or
branding, or interoperability of news; (2) benefits the entire industry, rather than just a few publishers,
and is non-discriminatory to other news publishers; and (3) directly relates to and is reasonably
necessary for these negotiations, instead of being used for other purposes. As Subcommittee Chairman
Cicilline noted at the time of the bill’s introduction:
The free press is a cornerstone of our democracy. Journalists keep the public informed,
root out corruption, and hold the powerful accountable. This bill will provide a much-
needed lifeline to local publishers who have been crushed by Google and Facebook. It’s
about time we take a stand on this issue.2460
Mr. Collins added that the proposed legislation would allow “community newspapers to more fairly
negotiate with large tech platforms that are operating in an increasingly anti-competitive space,” which
would “help protect journalism, promote competition and allow communities to stay informed.”2461
We recommend the consideration of this legislation as part of a broader set of reforms to
address the rise of market power online. This proposed legislation follows a long congressional
tradition of allocating coordination rights to individuals or entities that lack bargaining power in a
marketplace.2462 Although antitrust exemptions have been disfavored, at various times lawmakers have
created exemptions in order to rectify imbalances of power or to promote non-competition values.2463
In this instance, the risk associated with antitrust exemptions to preserve the free and diverse press—a
bedrock constitutional value—is low, while the benefits of preserving access to high-quality
journalism are difficult to overstate. As discussed earlier in the Report, the bill would follow steps that
other jurisdictions are similarly taking to rebalance the power between news publishers and the
dominant platforms.
Prohibit Abuse of Superior Bargaining Power and Require Due Process
By virtue of functioning as the only viable path to market, dominant platforms enjoy superior
bargaining power over the third parties that depend on their platforms to access users and markets.
2459 Press Release, Rep. David N. Cicilline, Collins Introduce Bill to Provide Lifeline to Local News (Apr. 3, 2019),
https://cicilline.house.gov/press-release/cicilline-collins-introduce-bill-provide-lifeline-local-news.
2460 Id.
2461 Id.
2462 See generally Submission from Sanjukta Paul, Ass’t Prof. of Law, Wayne State Univ., to H. Comm. on the Judiciary,
2–4 (Apr. 21, 2020) (on file with Comm.) [hereinafter Paul Submission].
2463 See, e.g., Clayton Act, 15 U.S.C. § 17 (1914); Capper-Volstead Act, ch. 57, 42 Stat. 388–89 (1922) (codified as
amended at 7 U.S.C. §§ 291, 292 (2012)).
390
Their bargaining leverage is a form of market power,2464 which the dominant platforms routinely use to
protect and expand their dominance.
Through its investigation, the Subcommittee identified numerous instances in which the
dominant platforms abused this power. In several cases, dominant platforms used their leverage to
extract greater money or data than users would be willing to provide in a competitive market. While a
firm in a competitive market would lose business if it charged excessive prices for its goods or services
because the customer would switch to a competitor, dominant platforms have been able to charge
excessive prices or ratchet up their prices without a significant loss of business. Similarly, certain
dominant platforms have been able to extort an ever-increasing amount of data from their customers
and users, ranging from a user’s personal data to a business’s trade secrets and proprietary content. In
the absence of an alternative platform, users effectively have no choice but to accede to the platform’s
demands for payment whether in the form of dollars or data.
The Subcommittee’s investigation found that dominant platforms have also leveraged their
market power in negotiations with businesses and individuals to dictate the terms of the relationship.
The dominant platforms frequently impose oppressive contractual provisions or offer “take-it-or-leave-
it” terms in contract negotiations—even when dealing with relatively large companies represented by
sophisticated counsel.2465 Lacking bargaining power, dependent third parties often find themselves at
the whims of the platform’s arbitrary decisions. Subcommittee staff encountered numerous instances in
which a third party had been abruptly delisted or demoted from a platform, without notice or
explanation, and often without a clear avenue for recourse.
The dominant platforms’ ability to abuse their superior bargaining power in these ways can
cause long-term and far-reaching harm. To address these issues, the Subcommittee recommends that
Congress consider prohibiting the abuse of superior bargaining power, including through potentially
targeting anticompetitive contracts, and introducing due process protections for individuals and
businesses dependent on the dominant platforms.2466
2464 Aviv Nevo, Deputy Assistant Att’y Gen. for Econ., U.S. Dep’t of Justice, Antitrust Div., Mergers that Increase
Bargaining Leverage, Remarks at the Stanford Institute for Economic Policy Research, 7 (Jan. 22, 2014),
https://www.justice.gov/atr/file/517781/download (“[A]s a matter of economic theory and case law bargaining leverage is a
source of market power.”).
2465 See, e.g., Dig. Competition Expert Panel Report at 45 (noting how a report commissioned by the UK’s Department for
Digital, Culture, Media & Sport found that as “a consequence of their high market share, ownership of key technologies
and strong user data assets, Google and Facebook are, to some extent, able to set their own terms to advertisers and
publishers”).
2466 Foer Submission at 2–3; Submission from Marshall Steinbaum, Assistant Prof. of Econ., Univ. of Utah, to H. Comm.
on the Judiciary, 8 (Apr. 2020) (on file with Comm.) [hereinafter Steinbaum Submission]. See generally Austl. Competition
& Consumer Comm’n Report at 205–79; Competition & Mkts. Auth. Report at 328–49.
391
B. Strengthening the Antitrust Laws
Restore the Antimonopoly Goals of the Antitrust Laws
The antitrust laws that Congress enacted in 1890 and 1914—the Sherman Act, the Clayton Act,
and the Federal Trade Commission Act—reflected a recognition that unchecked monopoly power
poses a threat to our economy as well as to our democracy.2467 Congress reasserted this vision through
subsequent antitrust laws, including the Robinson-Patman Act of 1936, the Celler-Kefauver Act of
1950, and the Hart-Scott-Rodino Act of 1976.2468
In the decades since Congress enacted these foundational statutes, the courts have significantly
weakened these laws and made it increasingly difficult for federal antitrust enforcers and private
plaintiffs to successfully challenge anticompetitive conduct and mergers.2469 By adopting a narrow
construction of “consumer welfare” as the sole goal of the antitrust laws, the Supreme Court has
limited the analysis of competitive harm to focus primarily on price and output rather than the
competitive process2470—contravening legislative history and legislative intent.2471 Simultaneously,
courts have adopted the view that underenforcement of the antitrust laws is preferable to
overenforcement, a position at odds with the clear legislative intent of the antitrust laws, as well as the
view of Congress that private monopolies are a “menace to republican institutions.”2472 In recent
decades, the Justice Department and the Federal Trade Commission have contributed to this problem
by taking a narrow view of their legal authorities and issuing guidelines that are highly permissive of
market power and its abuse. The overall result is an approach to antitrust that has significantly
diverged from the laws that Congress enacted.
2467 See generally First & Fox Submission at 10–11; Steinbaum Submission; Submission from Robert H. Lande, Venable
Prof. of Law, Univ. of Balt. Sch. of Law, to H. Comm. on the Judiciary (Apr. 16, 2020) (on file with Comm.) [hereinafter
Lande Submission]; Paul Submission at 2–4; Submission from Maurice Stucke, Douglas A. Blaze Distinguished Prof. of
Law, Univ. of Tennessee, to H. Comm. on the Judiciary, 2 (Mar. 13, 2020) (on file with Comm.) [hereinafter Stucke
Submission].
2468 Thomas J. Horton, Rediscovering Antitrust’s Lost Values, 16 U.N.H. L. REV. 179 (2018).
2469 See generally Submission from Tim Wu, Julius Silver Prof.of Law, Columbia Law Sch., to H. Comm. on the Judiciary
(Apr. 25, 2020) (on file with Comm.) [hereinafter Wu Submission]; Submission from Spencer Weber Waller, John Paul
Stevens Chair in Competition Law, Loyola Univ. Chicago Sch. of Law, to H. Comm. on the Judiciary (Apr. 28, 2020) (on
file with Comm.) [hereinafter Waller Submission].
2470 Jonathan Sallet, Protecting the “Competitive Process”—The Evolution of Antitrust Enforcement in the United States,
WASH. CTR. FOR EQUITABLE GROWTH (Oct. 31, 2018), https://equitablegrowth.org/competitive-edge-protecting-the-
competitive-process-the-evolution-of-antitrust-enforcement-in-the-united-states/.
2471 Submission from John Newman, Assoc. Prof. of Law, Univ. of Miami Sch. of Law, to the Subcomm. on Antitrust,
Commercial and Admin. Law of the H. Comm. on the Judiciary, 2 (Apr. 1, 2020) (on file with Comm.) [hereinafter
Newman Submission]; Stucke Submission at 2.
2472 21 CONG. REC. 3146 (1890) (statement of Sen. Hoar).
392
In part due to this narrowing, some of the anticompetitive business practices that the
Subcommittee’s investigation uncovered could be difficult to challenge under current law.2473 In
response to this concern, this section identifies specific legislative reforms that would help renew and
rehabilitate the antitrust laws in the context of digital markets. In addition to these specific reforms, the
Subcommittee recommends that Congress consider reasserting the original intent and broad goals of
the antitrust laws by clarifying that they are designed to protect not just consumers, but also workers,
entrepreneurs, independent businesses, open markets, a fair economy, and democratic ideals.2474
Invigorate Merger Enforcement
Section 7 of the Clayton Act of 1914 prohibits any transaction where “the effect of such
acquisition may be substantially to lessen competition, or to tend to create a monopoly.”2475 In 1950,
Congress passed the Celler-Kevauver Anti-Merger Act to broaden the types of transactions covered by
the Clayton Act, specifically to include vertical mergers, conglomerate mergers, and purchases of
assets.2476
As noted above, since 1998, Amazon, Apple, Facebook, and Google collectively have
purchased more than 500 companies.2477 The antitrust agencies did not block a single acquisition. In
one instance—Google’s purchase of ITA—the Justice Department required Google to agree to certain
terms in a consent decree before proceeding with the transaction.2478
The Subcommittee’s review of the relevant documents revealed that several of these
acquisitions lessened competition and increased market power. In several cases, antitrust enforcers
permitted dominant platforms to acquire a competitive threat. For example, documents produced
during the investigation demonstrate that Facebook acquired Instagram to neutralize an emerging rival,
while Google purchased Waze to eliminate an independent provider of mapping data. In other
instances, the platform engaged in a series of acquisitions that enabled it to gain a controlling position
across an entire supply chain or ecosystem. Google’s acquisitions of DoubleClick, AdMeld, and
AdMob, for example, let Google achieve a commanding position across the digital ad tech market.
2473 See Wu Submission at 2 (“If read broadly, the prohibitions on ‘monopolization,’ ‘unfair means of competition,’ and
‘restraints on trade’ could be used to handle the challenges of our time. But ‘broadly’ is manifestly not how the laws are
read by the judiciary at this point. For the courts have grafted onto these laws burdens of proof, special requirements and
defenses that are found nowhere in the statutes, and that have rendered the laws applicable only to the narrowest of
scenarios, usually those involving blatant price effects. And it is this that makes the laws inadequate for the challenges
presented by digital markets.”).
2474 See generally First & Fox Submission at 10–11; Stucke Submission at 2; Wu Submission; Waller Submission.
2475 Clayton Act, 15 U.S.C. § 18 (1914).
2476 Celler-Kefauver Anti-Merger Act, 64 Stat. 1125 (1950).
2477 See infra Appendix.
2478 Stipulation and Order, United States v. Google Inc. & ITA Software Inc., No. 1:11-cv-00688 (D.D.C. 2011).
393
In light of this, Subcommittee staff recommends that Congress considers a series of reforms to
strengthen merger enforcement.
a. Codify Bright-Line Rules and Structural Presumptions in Concentrated Markets
A major change in antitrust enforcement over the last few decades has been the shift away from
bright-line rules in favor of “rule of reason” case-by-case analysis. Although the rule of reason
approach is said to reduce errors in enforcement through fact-specific analysis, in practice the standard
tilts heavily in favor of defendants.2479 The departure from bright-line rules and presumptions has
especially affected merger enforcement, where enforcers seeking to challenge a merger must fully
prove that it will have anticompetitive effects, even in cases where the merging parties are dominant
firms in highly concentrated markets. Scholarship by Professor John Kwoka of Northeastern
University shows that the antitrust agencies acted in only 38% of all mergers that led to price increases,
suggesting that the current approach to merger review is resulting in significant underenforcement.2480
To respond to this concern, the Subcommittee recommends that Members consider codifying
bright-line rules for merger enforcement, including structural presumptions.2481 Under a structural
presumption, mergers resulting in a single firm controlling an outsized market share, or resulting in a
significant increase in concentration, would be presumptively prohibited under Section 7 of the
Clayton Act.2482 This structural presumption would place the burden of proof upon the merging parties
to show that the merger would not reduce competition. A showing that the merger would result in
efficiencies should not be sufficient to overcome the presumption that it is anticompetitive. It is the
view of Subcommittee staff that the 30% threshold established by the Supreme Court in Philadelphia
National Bank is appropriate, although a lower standard for monopsony or buyer power claims may
deserve consideration by the Subcommittee.
By shifting the burden of proof to the merging parties in cases involving concentrated markets
and high market shares, codifying the structural presumption would help promote the efficient
allocation of agency resources and increase the likelihood that anticompetitive mergers are blocked.
2479 Michael A. Carrier, The Rule of Reason: An Empirical Update for the 21st Century, 16 GEO. MASON L. REV. 827
(2009).
2480 JOHN KWOKA, MERGERS, MERGER CONTROL, AND REMEDIES 155 (2014).
2481 For support of codifying the structural presumption, see Submission from John Kwoka, Finnegan Prof. of Econ.,
Northeastern Univ., to H. Comm. on the Judiciary, 3 (Apr. 17, 2020) (on file with Comm.) [hereinafter Kwoka
Submission]; Submission from Michael Kades, Dir., Mkts. & Competition Pol’y, Wash. Ctr. for Equitable Growth et al., to
H. Comm. on the Judiciary, 9 (Apr. 30, 2020) (on file with Comm.) [hereinafter Kades Submission]; Lande Submission at
5; Slaiman Submission at 3; Foer Submission at 9. See also Herbert Hovenkamp & Carl Shapiro, Horizontal Mergers,
Market Structure, and Burdens of Proof, 127 YALE L.J. 1996 (2018); Steven C. Salop, The Evolution and Vitality of
Merger Presumptions: A Decision-Theoretic Approach, 80 ANTITRUST L.J. 269 (2015).
2482 Although some courts still follow the structural presumption adopted by the Supreme Court in Philadelphia National
Bank, it is not universally followed, especially given the D.C. Circuit’s decision in United States v. Baker Hughes Inc., 908
F.2d 981 (D.C. Cir. 1990).
394
b. Protect Potential Rivals, Nascent Competitors, and Startups
The Subcommittee’s investigation produced evidence that several of the dominant platforms
acquired potential rivals and nascent competitors. Potential rivals are firms that are planning to enter or
could plausibly enter the acquirer’s market. Nascent competitors are firms whose “prospective
innovation represents a serious future threat to an incumbent.”2483 In digital markets, potential rivals
and nascent competitors play a critical role in driving innovation, as their prospective entry may
dislodge incumbents or spur competition. For this reason, incumbents may view potential rivals and
nascent competitors as a significant threat, especially as their success could render the incumbent’s
technologies obsolete.
To strengthen the law relating to potential rivals and nascent competitors, Subcommittee staff
recommends strengthening the Clayton Act to prohibit acquisitions of potential rivals and nascent
competitors. This could be achieved by clarifying that proving harm on potential competition or
nascent competition grounds does not require proving that the potential or nascent competitor would
have been a successful entrant in a but-for world.2484 Given the patchwork of cases that are unfavorable
to potential and nascent competition-based theories of harm, this amendment should also make clear
that Congress intends to override this case law.2485
Since startups can be an important source of potential and nascent competition, the antitrust
laws should also look unfavorably upon incumbents purchasing innovative startups. One way that
Congress could do so is by codifying a presumption against acquisitions of startups by dominant firms,
particularly those that serve as direct competitors, as well as those operating in adjacent or related
markets.2486
Lastly, Subcommittee staff’s review of relevant documents produced by the Federal Trade
Commission and Justice Department demonstrated that the antitrust agencies consistently
underestimated—by a significant margin—the degree to which an acquisition would undermine
competition and impede entry. In light of this tendency, Subcommittee staff recommends that
Congress consider strengthening the incipiency standard by amending the Clayton Act to prohibit
acquisitions that “may lessen competition or tend to increase market power.”2487 Revising the law
2483 Wu Submission at 4–5; see also C. Scott Hemphill & Tim Wu, Nascent Competitors, 168 U. PA. L. REV. (forthcoming
2020); Kades Submission at 14.
2484 Wu Submission at 6; Kwoka Submission at 6.
2485 See, e.g., United States v. Marine Bancorporation, Inc., 418 U.S. 602 (1974).
2486 Submission from Mark Lemley, William H. Neukom Prof. of Law, Stanford Law Sch., to H. Comm. on the Judiciary,
7–8 (Apr. 8, 2020) (on file with Comm.) [hereinafter Lemley Submission].
2487 Submission from Consumer Reports, to H. Comm. on the Judiciary, 5 (Apr. 17, 2020) (on file with Comm.) [hereinafter
Consumer Reports Submission]; Submission from Richard M. Steuer, Adjunct Prof., Fordham Univ. Sch. of Law, to H.
Comm. on the Judiciary (Apr. 8, 2020) (on file with Comm.) [hereinafter Steuer Submission]; Peter C. Carstensen &
395
would “arrest the creation of trusts, conspiracies, and monopolies in their incipiency and before
consummation.” 2488
c. Strengthen Vertical Merger Doctrine
The Subcommittee’s investigation identified several ways in which vertical integration of
dominant platforms enabled anticompetitive conduct. For this reason, the Subcommittee recommends
that Congress examine proposals to strengthen the law relating to vertical mergers. The current case
law disfavors challenges to vertical mergers. Specifically, courts tend to defer to claims from the
merging parties that the transaction will yield efficiencies through the “elimination of double
marginalization” and are skeptical about claims that the merger will result in foreclosure.
To address this concern, the Subcommittee recommends that Congress explore presumptions
involving vertical mergers, such as a presumption that vertical mergers are anticompetitive when either
of the merging parties is a dominant firm operating in a concentrated market, or presumptions relating
to input foreclosure and customer foreclosure.2489
Rehabilitate Monopolization Law
Section 2 of the Sherman Act makes it illegal to “monopolize, or attempt to monopolize, or
combine or conspire with any other person or persons, to monopolize any part of the trade or
commerce among the several States.”2490 Over recent decades, courts have significantly heightened the
legal standards that plaintiffs must overcome in order to prove monopolization. Several of the business
practices the Subcommittee’s investigation uncovered should be illegal under Section 2. This section
briefly identifies the relevant business practices and the case law that impedes effective enforcement of
Section 2 of the Sherman Act.
a. Abuse of Dominance
Robert H. Lande, The Merger Incipiency Doctrine and the Importance of ‘Redundant’ Competitors, 2018 WIS. L. REV. 783
(2018).
2488 S. REP. NO. 698 (1914) in EARL W. KINTNER, THE LEGISLATIVE HISTORY OF THE FEDERAL ANTITRUST LAWS AND
RELATED STATUTES 1744–52 (1978) (noting that the Senate Judiciary Committee report stated that the purpose of the bill
was to supplement the Sherman Act “by making these practices illegal, to arrest the creation of trusts, conspiracies, and
monopolies in their incipiency and before consummation”).
2489 Kades Submission at 5; Jonathan Baker et al., Five Principles for Vertical Merger Enforcement Policy, 33 ANTITRUST 3
(2019).
2490 Sherman Act, 15 U.S.C. § 2 (1890).
396
The Subcommittee’s investigation found that the dominant platforms have the incentive and
ability to abuse their dominant position against third-party suppliers, workers, and consumers. Some of
these business practices are a detriment to fair competition, but they do not easily fit the existing
categories identified by the Sherman Act, namely “monopolization” or “restraint of trade.” Since
courts have shifted their interpretation of the antitrust law to focus primarily on the formation or
entrenchment of market power, and not on its exploitation or exercise, many of the business practices
that Subcommittee staff identified as undermining competition in digital markets could be difficult to
reach under the prevailing judicial approach.
To address this concern, Subcommittee staff recommends that Congress consider extending the
Sherman Act to prohibit abuses of dominance.2491 Furthermore, the Subcommittee should examine the
creation of a statutory presumption that a market share of 30% or more constitutes a rebuttable
presumption of dominance by a seller, and a market share of 25% or more constitute a rebuttable
presumption of dominance by a buyer.2492
b. Monopoly Leveraging
The Subcommittee’s investigation found that the dominant platforms have engaged in
“monopoly leveraging,” where a dominant firm uses its monopoly power in one market to boost or
privilege its position in another market. For example, Google’s use of its horizontal search monopoly
to advantage its vertical search offerings is a form of monopoly leveraging. Although monopoly
leveraging was previously a widely cognizable theory of harm under antitrust law, courts now require
that use of monopoly power in the first market “actually monopolize” the secondary market or
“dangerously threaten[] to do so.”2493 The Subcommittee’s investigation identified several instances in
which use of monopoly power in one market to privilege the monopolist’s position in the second
market injured competition, even if the conduct did not result in monopolization of the second market.
For this reason, Subcommittee staff recommends overriding the legal requirement that monopoly
leveraging “actually monopolize” the second market, as set out in Spectrum Sports, Inc. v.
McQuillan.2494
c. Predatory Pricing
2491 First & Fox Submission at 2; Foer Submission at 2–4; Newman Submission at 7–8; Stucke Submission at 14; Waller
Submission at 13.
2492 Waller Submission at 12.
2493 506 U.S. 447 (1993).
2494 Id. See also Alaska Airlines, Inc. v. United Airlines, Inc., 948 F.2d 536 (9th Cir. 1991).
397
The Subcommittee’s investigation identified several instances in which a dominant platform
was pricing goods or services below-cost in order to drive out rivals and capture the market. For
example, documents produced during the investigation revealed that Amazon had been willing to lose
$200 million in a single quarter in order to pressure Diapers.com, a firm it had recognized as its most
significant rival in the category. Amazon cut prices and introduced steep promotions, prompting a
pricing war that eventually weakened Diapers.com. Amazon then purchased the company, eliminating
its competitor and subsequently cutting back the discounts and promotions it had introduced.
Predatory pricing is a particular risk in digital markets, where winner-take-all dynamics
incentivize the pursuit of growth over profits, and where the dominant digital platforms can cross-
subsidize between lines of business. Courts, however, have introduced a “recoupment” requirement,
necessitating that plaintiffs prove that the losses incurred through below-cost pricing subsequently
were or could be recouped. Although dominant digital markets can recoup these losses through various
means over the long term, recoupment is difficult for plaintiffs to prove in the short term. Since the
recoupment requirement was introduced, successful predatory pricing cases have plummeted.2495
The Subcommittee recommends clarifying that proof of recoupment is not necessary to prove
predatory pricing or predatory buying, overriding the Supreme Court’s decisions in Matsushita v.
Zenith Ratio Corp.,2496 Brooke Group Ltd. v. Brown & Williamson Tobacco Corp.,2497 and
Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co.2498
d. Essential Facilities and Refusals to Deal
The Subcommittee’s investigation uncovered several instances in which a dominant platform
used the threat of delisting or refusing service to a third party as leverage to extract greater value or
more data or to secure an advantage in a distinct market. Because the dominant platforms do not face
meaningful competition in their primary markets, their threat to refuse business with a third party is the
equivalent of depriving a market participant of an essential input. This denial of access in one market
can undermine competition across adjacent markets, undermining the ability of market participants to
compete on the merits.
To address this concern, the Subcommittee recommends that Congress consider revitalizing the
“essential facilities” doctrine, the legal requirement that dominant firms provide access to their
2495 Hubbard Submission at 20; Stucke Submission at 7; Teachout Submission at 12; Christopher R. Leslie, Predatory
Pricing and Recoupment, 113 COLUM. L. REV. 1695 (2013).
2496 475 U.S. 574 (1986).
2497 509 U.S. 209 (1993).
2498 549 U.S. 312 (2007).
398
infrastructural services or facilities on a nondiscriminatory basis.2499 To clarify the law, Congress
should consider overriding judicial decisions that have treated unfavorably essential facilities- and
refusal to deal-based theories of harm.2500
e. Tying
The Subcommittee’s investigation identified several instances in which a dominant platform
conditioned access to a good or service that the dominant platform controlled on the purchase or use of
a separate product or service. This business practice undermines competition on the merits by enabling
a firm with market power in one market to privilege products or services in a distinct market.
Although antitrust law has long treated tying by a monopolist as anticompetitive, in recent
decades, courts have moved away from this position. Subcommittee staff recommends that Congress
consider clarifying that conditioning access to a product or service in which a firm has market power to
the purchase or use of a separate product or service is anticompetitive under Section 2, as held by the
Supreme Court in Jefferson Parish Hosp. Dist. v. Hyde.2501
f. Self-Preferencing and Anticompetitive Product Design
The Subcommittee’s investigation uncovered several instances in which a dominant platform
used the design of its platform or service to privilege its own services or to disfavor competitors. This
practice undermines competition by enabling a firm that controls an essential input to distort
competition in separate markets. The Subcommittee recommends that Congress consider whether
making a design change that excludes competitors or otherwise undermines competition should be a
violation of Section 2, regardless of whether the design change can be justified as an improvement for
consumers.2502
Additional Measures to Strengthen the Antitrust Laws
In response to the Subcommittee’s requests for submissions, experts identified other proposals
that Subcommittee staff believes warrant review by Congress. These include:
2499 Submission from the Am. Antitrust Inst., to H. Comm. on the Judiciary, 4 (Apr. 17, 2020) (on file with Comm.)
[hereinafter AAI Submission]; Waller Submission at 13.
2500 Verizon Commc’ns Inc. v. Law Offices of Curtis v. Trinko, LLP, 540 U.S. 398 (2004); Pacific Bell Telephone Co. v.
LinkLine Commc’ns, Inc., 555 U.S. 438 (2009).
2501 466 U.S. 2 (1984).
2502 This would require overriding Allied Orthopedic Appliances, Inc. v. Tyco Health Care Grp. LP, 592 F.3d 991 (9th Cir.
2010).
399
• Overriding Ohio v. American Express by clarifying that cases involving platforms do not
require plaintiffs to establish harm to both sets of customers;2503
• Overriding United States v. Sabre Corp., clarifying that platforms that are “two-sided,” or serve
multiple sets of customers, can compete with firms that are “one-sided”;2504
• Clarifying that market definition is not required for proving an antitrust violation, especially in
the presence of direct evidence of market power;2505 and
• Clarifying that “false positives”—or erroneous enforcement—are not more costly than “false
negatives”—or erroneous non-enforcement—and that, in relation to conduct or mergers
involving dominant firms, “false negatives” are costlier.2506
C. Strengthening Antitrust Enforcement
Congressional Oversight
As discussed earlier in the Report, Congress has a strong tradition of performing vigorous
oversight of the enforcement and adequacy of the antitrust laws. Over the last century, Congress at key
moments responded forcefully to the courts’ narrowing of antitrust laws, the rising tide of economic
concentration, or other challenges to the sound and effective administration of the antitrust laws.2507
This tradition includes the creation of the Federal Trade Commission and concurrent enactment
of the Clayton Antitrust Act in 1914, as both a response to the Supreme Court’s narrow construction of
the Sherman Act in 1911 and an effort to limit the discretion of the courts.2508 It also includes
Congress’s broadening of merger enforcement to cover non-horizontal acquisitions and other
transactions in the Celler-Kefauver Anti-Merger Act of 1950 as well as establishing a mechanism for
judicial oversight of consent decrees in response to political interference in merger enforcement with
2503 AAI Submission at 4; Submission from Herbert Hovenkamp, James G. Dinan Univ. Prof., Univ. of Pa. Law Sch., to H.
Comm. on the Judiciary, 3 (Apr. 17, 2020) (on file with Comm.) [hereinafter Hovenkamp Submission]; Hubbard
Submission at 20; Kades Submission at 8.
2504 United States v. Sabre Corp., 452 F. Supp. 3d 97 (D. Del. 2020). See also Kades Submission at 10.
2505 Hovenkamp Submission at 3–4; Newman Submission at 5–6.
2506 Subcommittee staff believes that Congress could clarify that the views set out by then-Professor Frank Easterbrook in
The Limits of Antitrust, 63 TEX. L. REV. 1 (1984) do not reflect the views of the Congress in enacting the antitrust laws. See
also Submission from Bill Baer, Visiting Fellow, Brookings Inst., to H. Comm. on the Judiciary, 3 (May 19, 2020) (on file
with Comm.) [hereinafter Baer Submission] (“That is my fundamental concern with the state of antitrust enforcement
today. It is too cautious, too worried about adverse effects of “over enforcement” (so called Type I errors).”).
2507 See generally, Marc Winerman, The Origins of the FTC: Concentration, Cooperation, Control, and Competition, 71
ANTITRUST L.J. 1 (2003).
2508 Clayton Act, 15 U.S.C. § 12; Fed. Trade Comm’n Act, 15 U.S.C. § 41.
400
the Tunney Act of 1974.2509 Additionally, Congress has regularly investigated the rise and abuse of
market power in important markets.2510 In support of these efforts, Congress dedicated substantial
congressional and agency resources to perform the task of identifying and responding to
anticompetitive conduct.2511
In recent decades, Congress has departed from this tradition, deferring largely to the courts and
to the antitrust agencies in the crafting of substantive antitrust policy.2512 Its inaction has been read as
acquiescence to the narrowing of the antitrust laws and has contributed to antitrust becoming “overly
technical and primarily dependent on economics.”2513
In other cases, congressional attention has fallen short as lawmakers tried to address
competition problems without sustained efforts to implement enforcement changes, leading some
reform efforts in recent decades to misfire.2514 Responding to these concerns, Congress has increased
appropriations and provided modest improvements to the Federal Trade Commission’s budget and
remedial authority during this period. But these efforts were insufficient without sustained support in
the face of “ferocious opposition” from large defendants and businesses lobbying Congress.2515
To remedy these broader trends, Subcommittee staff recommends that Congress revive its long
tradition of robust and vigorous oversight of the antitrust laws and enforcement, along with its
2509 5 U.S.C. § 16. See also Consent Decree Program of the Dep’t of Justice: Hearings Before the Subcomm. on Antitrust of
the H. Comm. on the Judiciary, 85th Cong. (1957); REPORT OF THE SUBCOMM. ON ANTITRUST OF THE H. COMM. ON THE
JUDICIARY, CONSENT DECREE PROGRAM OF THE DEP’T OF JUSTICE, 86TH CONG., 1ST SESS. (1959).
2510 In the 1990s, the Committee on the Judiciary conducted significant oversight of competition in the telecommunications
market in the wake of the breakup of Ma Bell and through oversight of the 1982 consent decree. These efforts culminated
in the passage of H.R. 3626, the “Antitrust and Communications Reform Act,” by the House of Representatives in 1994 by
a vote of 423 to 5. Chairman Jack B. Brooks introduced this bill—a precursor to the Telecommunications Act of 1996—to
address monopolization in the telecommunications market. See generally H. REP. NO. 103-559 (1994); Robert M. Frieden,
The Telecommunications Act of 1996: Predicting the Winners and Losers, 20 HASTINGS COMM. & ENT. L.J. 11, 57 n.8
(1997).
2511 Submission from Alison Jones & William E. Kovacic, to H. Comm. on the Judiciary, 4 (Apr. 17, 2020) (on file with
Comm.) [hereinafter Jones & Kovacic Submission].
2512 Harry First & Spencer Weber Waller, Antitrust’s Democracy Deficit, 81 FORDHAM L. REV. 2543, 2556 (2013)
(“[D]espite a history of bipartisan congressional support for the importance of the antitrust laws and their enforcement, of
late Congress has done little. And when it has done something, it has focused on the micro rather than the macro changes
that have occurred in the field.”).
2513 Id. at 2559.
2514 Jones & Kovacic Submission at 4 (“The miscalculation of Congress (and the agencies) about the magnitude of
implementation tasks in this earlier period came at a high price. Implementation weaknesses undermined many
investigations and cases that the federal agencies launched in response to congressional guidance. The litigation failures
raised questions about the competence of the federal agencies, particularly their ability to manage large cases dealing with
misconduct by dominant firms and oligopolists. The wariness of the federal agencies since the late 1970s to bring cases in
this area—a wariness that many observers today criticize as unwarranted—is in major part the residue of bitter litigation
experiences from this earlier period.”).
2515 Id. at 6.
401
commitment to ongoing market investigations and legislative activity. Additionally, greater attention to
implementation challenges will enable Congress to better see its reform efforts through.
Agency Enforcement
Over the course of the investigation, the Subcommittee uncovered evidence that the antitrust
agencies consistently failed to block monopolists from establishing or maintaining their dominance
through anticompetitive conduct or acquisitions. This institutional failure follows a multi-decade trend
whereby the antitrust agencies have constrained their own authorities and advanced narrow readings of
the law. In the case of the Federal Trade Commission, the agency has been reluctant to use the
expansive set of tools with which Congress provided it, neglecting to fulfill its broad legislative
mandate. Restoring the agencies to full strength will require overcoming these trends.
As a general matter, Congress created the FTC to police and prohibit “unfair methods of
competition,”2516 and to serve as an “administrative tribunal” that carefully studied ongoing business
practices and economic conditions.2517 To enable the agency to carry out these functions, Congress
assigned the Commission powers to “make rules and regulations for the purpose of carrying out the
[FTC Act’s] provisions,” as well as broad investigative authority to compel business information and
conduct market studies.2518 Notably, Congress established the provision prohibiting “unfair methods of
competition” to reach beyond the other antitrust statutes, “to fill in the gaps in the other antitrust laws,
to round them out and make their coverage complete.”2519 Lawmakers delegated to the FTC the task of
defining what constituted an “unfair method of competition,” recognizing that an expert agency
equipped to continuously monitor business practices would be best positioned to ensure the legal
definition kept pace with business realities.
2516 See S. REP. NO. 63-597, 13 (1914) (“The committee gave careful consideration to the question as to whether it would
attempt to define the many and variable unfair practices which prevail in commerce and to forbid [them] . . . or whether it
would, by a general declaration condemning unfair practices, leave it to the commission to determine what practices were
unfair. It concluded that the latter course would be better, for the reason . . . that there were too many unfair practices to
define, and after writing 20 of them into the law it would be quite possible to invent others.”).
2517 Neil W. Averitt, The Meaning of “Unfair Methods of Competition” in Section 5 of the Federal Trade Commission Act,
21 B.C. L. REV. 227 (1980); see also Marc Winerman, The Origins of the FTC: Concentration, Cooperation, Control, and
Competition, 71 ANTITRUST L.J. 1 (2003).
2518 15 U.S.C. § 46.
2519 Neil W. Averitt, The Meaning of “Unfair Methods of Competition” in Section 5 of the Federal Trade Commission Act,
21 B.C. L. REV. 227, 251 (1980) (“Section 5 is not confined to conduct that actually violates, or that threatens to violate,
one of the other antitrust statutes. If it were limited to this extent it would be a largely duplicative provision. The legislative
purpose instead assigned to Section 5 a broader role. It was to be an interstitial statute: it was to fill in the gaps in the other
antitrust laws, to round them out and make their coverage complete. In addition to overt violations, therefore, Section 5
would reach closely similar conduct that violates the policy or ‘spirit’ of the antitrust laws, even though it may not come
technically within its terms.”).
402
In practice, however, the Commission has neglected to play this role. In its first hundred years,
the FTC promulgated only one rule defining an “unfair method of competition.”2520 In 2015 the
Commission adopted a set of “Enforcement Principles,” stating that the FTC’s targeting of “unfair
methods of competition” would be guided by the “promotion of consumer welfare,” a policy goal
absent from any legislative directive given to the Commission.2521 Since the adoption of this
framework, the FTC has brought only one case under its standalone Section 5 authority.2522 The
agency has also failed to regularly produce market-wide studies, having halted regular data collection
in the 1980s.2523
Together with the DOJ, the FTC has also chosen to stop enforcing certain antitrust laws
entirely. For two decades, neither agency has filed a suit under the Robinson-Patman Act, which
Congress passed in order to limit the power of large chain retailers to extract concessions from
independent suppliers.2524 In 2008, the Justice Department issued a report recommending that Section
2 of the Sherman Act be curbed dramatically.2525 Although the report was subsequently rescinded, the
Justice Department has not filed a significant monopolization case in two decades. Meanwhile, both
agencies have targeted their enforcement efforts on relatively small players—including ice skating
teachers and organists—raising questions about their enforcement priorities.2526
The agencies have also been hamstrung by inadequate budgets. In 1981, FTC Chairman Jim
Miller won steep budget cuts at the Commission, a drastic rollback from which the agency has not yet
recovered. Prior to this Congress, appropriations for both agencies have reached historic lows.2527 To
2520 Discriminatory Practices in Men’s and Boys’ Tailored Clothing Industry, 16 C.F.R. pt. 412 (1968).
2521 Fed. Trade Comm’n, Statement of Enforcement Principles Regarding “Unfair Methods of Competition” Under Section
5 of the FTC Act (Aug. 13, 2015),
https://www.ftc.gov/system/files/documents/public_statements/735201/150813section5enforcement.pdf.
2522 The one exception is FTC’s recent suit against Qualcomm. Fed. Trade Comm’n v. Qualcomm Inc., 411 F. Supp. 3d 658
(N.D. Cal. 2019) (5:17-cv-00220).
2523 FED. TRADE COMM’N, BUR. OF ECON., ANNUAL LINE OF BUSINESS REPORT 1977 (1985), https://www.ftc.gov/reports/us-
federal-trade-commission-bureau-economics-annual-line-business-report-1977-statistical.
2524 In a memo submitted on behalf of the United States to the OECD, the Justice Department stated that “a shift in
emphasis based on economic analysis resulted in a significant reduction in enforcement actions brought by the Agencies
under the Robinson-Patman Act. As a result, current enforcement of the Act occurs mainly through private treble damages
actions.” Note by the United States, Roundtable on “Price Discrimination,” OECD (Nov. 2016),
https://www.justice.gov/atr/case-document/file/979211/download.
2525 Thomas O. Barnett & Hill B. Wellford, The DOJ’s Single-Firm Conduct Report: Promoting Consumer Welfare
Through Clearer Standards for Section 2 of the Sherman Act (Sept. 8, 2008),
https://www.justice.gov/sites/default/files/atr/legacy/2009/05/11/238599.pdf.
2526 Sandeep Vaheesan, Accommodating Capital and Policing Labor: Antitrust in the Two Gilded Ages, 78 MD. L. REV. 766
(2019). See also Brief for the United States and the Fed. Trade Comm’n as Amicus Curiae in Support of Appellant and in
Favor of Reversal, Chamber of Commerce of the United States of America and Rasier, LLC, v. City of Seattle, 890 F.3d
769 (9th Cir. 2018) (No. 17-35640).
2527 MICHAEL KADES, WASH. CTR. FOR EQUITABLE GROWTH, THE STATE OF U.S. FEDERAL ANTITRUST ENFORCEMENT
(2019), https://equitablegrowth.org/wp-content/uploads/2019/09/091719-antitrust-enforcement-report.pdf.
403
restore the antitrust agencies to full strength, Subcommittee staff recommends that Congress consider
the following:
• Triggering civil penalties and other relief for violations of “unfair methods of competition”
rules, creating symmetry with violations of “unfair or deceptive acts or practices” rules;
• Requiring the Commission to regularly collect data and report on economic concentration and
competition in sectors across the economy, as permitted under Section 6 of the FTC Act;
• Enhancing the public transparency and accountability of the antitrust agencies, by requiring the
agencies to solicit and respond to public comments for merger reviews, and by requiring the
agencies to publish written explanations for all enforcement decisions;2528
• Requiring the agencies to conduct and make publicly available merger retrospectives on
significant transactions consummated over the last three decades;
• Codifying stricter prohibitions on the revolving door between the agencies and the companies
that they investigate, especially with regards to senior officials;2529 and
• Increasing the budgets of the Federal Trade Commission and the Antitrust Division.2530
Private Enforcement
Private enforcement plays a critical role in the nation’s antitrust system. The Sherman Act and
Clayton Act both include a private right of action. This reflected lawmakers’ desire to ensure that those
abused by monopoly power have an opportunity for direct recourse.2531 It also reflected a recognition
that public enforcers would be susceptible to capture by the very monopolists that they were supposed
to investigate, necessitating other means of enforcement.
Empirical surveys of trends in antitrust enforcement indicate that private enforcement deters
anticompetitive conduct and strengthens enforcement overall.2532 In recent decades, however, courts
2528 Mitchell Submission at 9–10.
2529 See submission from Source 17.
2530 See Baer Submission at 7–8; Kades Submission at 12–13.
2531 See, e.g., 51 CONG. REC. 9073 (1914) (remarks of Rep. Webb) (stating that private Section 7 remedies “open the door
of justice to every man, whenever he may be injured by those who violate the antitrust laws, and give the injured party
ample damages for the wrong suffered”).
2532 Joshua P. Davis & Robert H. Lande, Toward an Empirical and Theoretical Assessment of Private Antitrust
Enforcement, 36 SEATTLE U. L. REV. 1269, 1276 (2013).
404
have erected significant obstacles for private antitrust plaintiffs, both through procedural decisions and
substantive doctrine.
One major obstacle is the rise of forced arbitration clauses, which undermine private
enforcement of the antitrust laws by allowing companies to avoid legal accountability for their
actions.2533 These clauses allow firms to evade the public justice system—where plaintiffs have far
greater legal protections—and hide behind a one-sided process that is tilted in their favor.2534 For
example, although Amazon has over two million sellers in the United States, Amazon’s records reflect
that only 163 sellers initiated arbitration proceedings between 2014 and 2019.2535 This data seems to
confirm studies showing that forced arbitration clauses often fail to provide a meaningful forum for
resolving disputes and instead tend to suppress valid claims and shield wrongdoing.2536
Several other trends in judicial decisions have hampered private antitrust plaintiffs, including in
cases involving dominant platforms. To address these concerns, the Subcommittee recommends that
Congress consider:
• Eliminating court-created standards for “antitrust injury”2537 and “antitrust standing,”2538 which
undermine Congress’s grant of enforcement authority to “any person . . . injured . . . by reason
of anything forbidden in the antitrust laws;”2539
• Reducing procedural obstacles to litigation, including through eliminating forced arbitration
clauses2540 and undue limits on class action formation;2541 and
2533 Justice Denied: Forced Arbitration and the Erosion of our Legal System: Hearing Before the Subcomm. on Antitrust,
Commercial and Admin. Law of the H. Comm. on the Judiciary, 116th Cong. 2 (2019) (statement of Myriam Gilles, Paul R.
Verkuil Research Chair in Public Law & Prof. of Law, Benjamin N. Cardozo Sch. of Law).
2534 Justice Denied: Forced Arbitration and the Erosion of our Legal System: Hearing Before the Subcomm. on Antitrust,
Commercial and Admin. Law of the H. Comm on the Judiciary, 116th Cong. 2 (2019) (statement of Deepak Gupta,
Founding Principal, Gupta Wessler PLLC).
2535 Innovation and Entrepreneurship Hearing at 49 (response to Questions for the Record of Nate Sutton, Assoc. Gen.
Counsel, Competition, Amazon.com, Inc.).
2536 Judith Resnik, Diffusing Disputes: The Public in the Private of Arbitration, the Private in Courts, and the Erasure of
Rights, 124 YALE L. J. 2804 (2015).
2537 Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477 (1977).
2538 Assoc. Gen. Contractors v. California State Council of Carpenters, 459 U.S. 519 (1983).
2539 Clayton Act, 15 U.S.C. § 15 (1914).
2540 American Express v. Italian Colors, 570 U.S. 228 (2013); AT&T Mobility v. Concepcion, 563 U.S. 333 (2011).
2541 Comcast v. Behrend, 569 U.S. 27 (2013).
405
• Lowering the heightened pleading requirement introduced in Bell Atlantic Corp. v.
Twombly.2542
* * *
2542 550 U.S. 544 (2007).