THE ANTITRUST CASE AGAINST FACEBOOK

A MONOPOLIST’S JOURNEY TOWARDS PERVASIVE SURVEILLANCE

IN SPITE OF CONSUMERS’ PREFERENCE FOR PRIVACY

Dina Srinivasan*

This version of The Antitrust Case Against Facebook is a draft version. The final version of this
article is forthcoming and will be published in January 2019 in the Berkeley Business Law
Journal Vol. 16, Issue 1.

*This article arose from my personal observations as an entrepreneur and executive in the digital
advertising industry. I am grateful to Fiona Scott Morton, Alvin Klevorick, Michael Kades,
Charles Reichmann, and Robert Litan for conversation and feedback at various stages. I extend a
warm thank you to George Priest for inspiring an interest in antitrust years ago. I would also like
to thank the editors at the University of California Berkley Business Law Journal for their careful
edits, the journalists and researchers for uncovering and reporting on the facts relied on in this
paper, and the University of Chicago for hosting the 2018 Antitrust and Competition Conference:
Digital Platforms and Concentration. No party provided to me any direct or indirect financial
support during the writing of this article.

Electronic copy available at: https://ssrn.com/abstract=3247362

CONTENTS

INTRODUCTION……………………………………………………….………….……………….. 3

I. PRIVACY WAS ONCE A CRUCIAL FORM OF COMPETITION………….…………………….. 10

II. THE PARADOX OF SURVEILLANCE REFLECTS MONOPOLY POWER……………………...... 18

A. Pre Power: Failure of Beacon and Early Misrepresentations………….………………. 23

B. Pre Power: More Backtracking and Pattern of Misrepresentations………….…...……. 32

C. Post Power: Deterioration of the Promise Not to Track…………….………….............. 41

1. Facebook Initiates Commercial Surveillance………………………………...………39

2. Facebook Leverages Consumer Identity for Stronger Surveillance…………..…. 43

3. Facebook Circumvents Consumer Attempts to Opt-Out…………………………….46

III. INDIRECT EVIDENCE CONFIRMS FACEBOOK’S MONOPOLY POWER…………………........ 52

IV. FACEBOOK’S PATTERN OF CONDUCT RAISES ILLEGAL MONOPOLIZATION CONCERNS…. 63

A. Heightened Scrutiny in Markets with Direct Network Effects……….…….…………..... 64

B. Pattern of False Statements, Misleading & Deceptive Conduct……….………..…….... 67

C. Wider Pattern of False Statements & Misleading Conduct………………..….………… 69

D. Antitrust Harm: Monopoly Rents and Allocative Inefficiency………………….……….. 73

CONCLUSION………….………….………….………….……...….………….………….……....75

Electronic copy available at: https://ssrn.com/abstract=3247362

ABSTRACT

The Facebook, Inc. (“Facebook”) social network, this era’s new communications service, plays

an important role in the lives of 2+ billion people across the world. Though the market was

highly competitive in the beginning, it has since consolidated in Facebook’s favor. Today, using

Facebook means to accept a product linked to broad-scale commercial surveillance—a paradox

in a democracy. This Paper argues that Facebook’s ability to extract this qualitative exchange

from consumers is merely this titan’s form of monopoly rents. The history of early competition,

Facebook’s market entry, and Facebook’s subsequent rise tells the story of Facebook’s

monopoly power. However, the history which elucidates this firm’s dominance also presents a

story of anticompetitive conduct. Facebook’s pattern of false statements and misleading conduct

induced consumers to trust and choose Facebook, to the detriment of market competitors and

consumers' own welfare.

Electronic copy available at: https://ssrn.com/abstract=3247362

INTRODUCTION

Social networks, considered electronic communication service providers,1 have become

a primary way that American consumers communicate.2 Facebook is the reigning platform, not

only in the lives of Americans, but in the lives of 2.2 billion people worldwide.3 Though the

social network market was highly competitive in the beginning,4 the market has since

consolidated in Facebook’s favor. Consumers effectively face a singular choice—use Facebook

and submit to the quality and stipulations of Facebook’s product or forgo all use of the only

social network used by most of their friends, family, and acquaintances.

When Facebook entered the market, the consumer’s privacy was paramount. The

company prioritized privacy, as did its users—many of whom chose the platform over others due

to Facebook’s avowed commitment to preserving their privacy. Today, however, accepting

Facebook’s policies in order to use its service means accepting broad-scale commercial

surveillance.

1
Various federal laws regulate electronic communication service providers, including the Electronic
Communications Privacy Act, 18 U.S.C.§ 2510 (ECPA), Titles I and II of the ECPA specifically (the Wiretap Act,
18 U.S.C. § 2510-2521, the Stored Communications Act, 18 U.S.C. § 2701-2711 (SCA)), and the Foreign
Intelligence Surveillance Act, 50 U.S.C. § 1801 (2008). Several courts have considered Facebook’s duties and lack
thereof under these statutes. See, e.g., Crispin v. Christian Audigier, Inc., 717 F. Supp. 2d 965, 991 (C.D. Cal. 2010)
(quashing subpoena portions for private messages from Facebook and MySpace because the SCA prohibits
production). See also In re Facebook Privacy Litig., 791 F. Supp. 2d 705, 711-13 (N.D. Cal. 2011); In re Facebook
Internet Tracking Litig., 140 F. Supp. 3d 922 (2015). For an overview of relevant cases, see IAN C. BALLON, Cyber
Boot Camp: Data Security at the Intersection of Law and Business, in E-COMMERCE AND INTERNET LAW: A LEGAL
TREATISE WITH FORMS (2d ed. 2016), http://legacy.callawyer.com/wp-content/uploads/2016/05/2017-01-12-DJ-LA-
CyberSecurityBootcamp-aPrivacyLit.pdf.
2
See generally Frank Newport, The New Era of Communication Among Americans, GALLUP.COM (Nov. 10,
2014), https://news.gallup.com/poll/179288/new-era-communication-americans.aspx.
3
Company Info, NEWSROOM.FB.COM (July 20, 2018), https://newsroom.fb.com/company-info/.
4
In 2007, there were hundreds of social networks, including competitive offerings from Google, Yahoo, and of
course, MySpace. See generally Danah M. Boyd & Nicole B. Ellison, Social Network Sites: Definition, History, and
Scholarship, 13 J, OF COMPUTER-MEDIATED COMM. 210 (2007).

Electronic copy available at: https://ssrn.com/abstract=3247362

Facebook knows consumers’ identities by virtue of its position as this century’s new

communications network and leverages this knowledge to build dossiers on consumers that are

unrivaled in the private market. For Max Schrems, then a 23-year-old law student who used

European law to petition Facebook for a copy of his data, his Facebook file as of 2011 was 1,200

pages long.5 That was before Facebook started monitoring and recording users’ activities not just

on the singular Facebook domain, but on millions of independently owned websites and mobile

applications.6 Facebook’s surveillance apparatus captures what people read, shop for, and even

think, 7 online. For Americans that use an Android phone, Facebook apparently also scrapes a

record of a user’s cellular phone calls—effectively rendering use of the Facebook social network

5
Max Schrems, now a leading advocate at the intersection of user privacy and Big Tech, initially filed claims
against Facebook with the Irish Data Protection Commissioner (IDPC) for violations of European data protection
law. For the files Schrems originally obtained from Facebook, see Facebook’s Data Pool, FACEBOOK.COM,
http://europe-v-facebook.org/EN/Data_Pool/data_pool.html (last visited July 27, 2018). Following the revelations of
whistleblower Edward Snowden, and the specific leak of National Security Agency (NSA) Prism slides claiming
Facebook was supplying the NSA with copies of user communications and other social network data, Schrems filed
an additional complaint with the IDPC alleging that the transfer of his data from the E.U. to U.S. jurisdiction
violated the European Union Directive on Data Protection (Directive 95/46/EC), which as of May 25, 2018, has
been replaced with the General Data Protection Regulation (GDPR). The suit eventually caused an international
crisis when it resulted in the downfall of the Safe Harbour framework for E.U.-U.S. corporate data transfers. See
Maximillian Schrems v. Digital Rights Ireland Ltd. [2015] C-362/14 (H. Ct.) (Ir.). For coverage of the Snowden
leaks, Prism, and Facebook’s alleged participation in Prism, see Barton Gellman & Laura Poitras, U.S., British
Intelligence Mining Data From Nine U.S. Internet Companies in Broad Secret Program, WASH. POST (June 7,
2013), https://www.washingtonpost.com/investigations/us-intelligence-mining-data-from-nine-us-internet-
companies-in-broad-secret-program/2013/06/06/3a0c0da8-cebf-11e2-8845-
d970ccb04497_story.html?utm_term=.8e7f22acf1c0; Glen Greenwald & Ewen MacAskill, NSA Prism Program
Taps in to User Data of Apple, Google and Others, GUARDIAN (June 6, 2013),
http://www.theguardian.com/world/2013/jun/06/us-techgiants-nsa-data [http://perma.cc/G4AD-DA88]. For the
leaked NSA PowerPoint slides describing Prism and Facebook’s involvement with Prism, see PRISM Overview
Powerpoint Slides, ACLU.ORG (April 2013), https://www.aclu.org/other/prism-overview-powerpoint-slides.
6
Facebook can track consumer behavior on the millions of independent websites and apps that use any of
Facebook’s business products (Like buttons, Logins, conversion tracking pixels, software development kit, et al.).
The Facebook Like buttons alone appear on nearly 3 million websites. Websites using Facebook Like Button,
BUILTWITH.COM (2018), https://trends.builtwith.com/websitelist/Facebook-Like-Button.
7 Facebook has collected the text users type, but then delete, into status updates, timeline posts, and comments,

before hitting an enter button. See Sauvik Das & Adam Kramer, Self-Censorship on Facebook, PROCEEDINGS OF
THE SEVENTH INTERNATIONAL AAAI CONF. ON WEBLOGS AND SOCIAL MEDIA,
https://www.aaai.org/ocs/index.php/ICWSM/ICWSM13/paper/viewFile/6093/6350. It is unclear whether
Facebook’s privacy policy at the time permitted Facebook to do this. See generally Jennifer Golbeck, On Second
Thought …, SLATE (Dec. 13, 2013), https://slate.com/technology/2013/12/facebook-self-censorship-what-happens-
to-the-posts-you-dont-publish.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

app to the equivalent of having a pen register on one’s device.8 Surveillance translates to

influence,9 on many levels. As we learned recently, information gathered thereby can be

misappropriated to influence political elections.10

Facebook also sells advertising to marketers. In the digital advertising market, the ability

to conduct unprecedented commercial surveillance is a bedrock of Facebook’s current revenues

and profits. Facebook generates nearly all of its revenues from the sale of advertising,11 and the

prices of ads sold today directly correlate with data derived from tracking consumers.12 Facebook

8
Tom Warren, Facebook Has Been Collecting Call History and SMS Data From Android Devices, THE VERGE
(Mar. 25, 2018), https://www.theverge.com/2018/3/25/17160944/facebook-call-history-sms-data-collection-android;
Alex Hern, Facebook Logs SMS and Calls, Users Find as They Delete Accounts, THE GUARDIAN (Mar. 26, 2018),
https://www.theguardian.com/technology/2018/mar/25/facebook-logs-texts-and-calls-users-find-as-they-delete-
accounts-cambridge-analytica.
9
Data derived from surveillance is used to create psychographic profiles of individual consumers. Whereas
demographic data refers to the variables of age, income, and education, psychographic data refers to the broader
range of data sets that reflect a consumer’s personality, opinions, attitudes, interests, values, past activity, and
general lifestyle. Psychographic data is used by companies to determine which piece of content, advertisement, or
even price, to display to a particular consumer, in an attempt to influence consumer outcome. For a study of how
psychographically targeted advertising changes the outcome of consumer behavior, see, e.g., S. C. Matz, M.
Kosinski, G. Nave, & D. J. Stillwell, Psychological Targeting as an Effective Approach to Digital Mass Persuasion,
PNAS (Nov. 2017), http://www.pnas.org/content/early/2017/11/07/1710966114.short. For background on
psychographics, see generally, William D. Wells, Psychographics: A Critical Review, 12 J. OF MARKETING RES. 196
(1975), www.jstor.org/stable/3150443.
10
After the conclusion of the 2016 U.S. presidential election, Facebook disclosed to congressional investigators
that it had sold to a Russian company, Internet Research Agency (IRA), ads which psychographically targeted
American voters, and that up to approximately 126 million people could have been exposed to such ads. See Dylan
Byers, Facebook Estimates 126 Million People Were Served Content from Russia-linked Pages, CNN.COM (Oct. 31,
2017), https://money.cnn.com/2017/10/30/media/russia-facebook-126-million-users/index.html; Facebook, Google
and Twitter Executives on Russian Disinformation, C-SPAN.ORG (Oct. 31, 2017), https://www.c-
span.org/video/?436454-1/facebook-google-twitter-executives-testify-russia-election-ads. In March 2018, it was
revealed that Facebook data, including Like data, on more than 87 million users was misappropriated and used by
political consulting firm Cambridge Analytica to psychographically target voters during the presidential race. For
The New York Times story that cascaded response from U.S. and British lawmakers (in Britain, for alleged
interference in the “Brexit” campaign) and resulted eventually in Facebook founder and chief executive officer Mark
Zuckerberg’s live testimony before Congress, see Matthew Rosenberg, Nicholas Confessore & Carole Cadwalladr,
How Trump Consultants Exploited the Facebook Data of Millions, N.Y. TIMES (March 17, 2018),
https://www.nytimes.com/2018/03/17/us/politics/cambridge-analytica-trump-campaign.html.
11
Advertising revenues contributed 98+% of Facebook’s 40+ billion in revenues in 2017. FACEBOOK INC.,
Annual Report Pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934 (Form 10-K) (2017).
12
Consider comments of Dave Wehner, the chief financial officer of Facebook, on Facebook’s Q2 earnings
call, with respect to how privacy (or, a reduction in user data) is inversely correlated with Facebook revenues.
Facebook Inc. Q2 2018 Earnings Conference Call, INVESTOR.FACEBOOK.COM (July 25, 2018),
https://s21.q4cdn.com/399680738/files/doc_financials/2018/Q2/Q218-earnings-call-transcript [hereinafter
“Facebook Q2 Earnings”].

Electronic copy available at: https://ssrn.com/abstract=3247362

leverages information it knows about users to sell more impression-targeted ads and more action-

based ads.13 Facebook then uses this capability to sell advertising on behalf of other market

participants—like Hearst, The Washington Post, or Time, Inc.14 Facebook’s power here is so

absolute that the duopoly of Facebook and Google accounts for 90-99% of year-over-year

growth in the U.S. digital advertising market.15

But in a country anchored by democratic values and fear of tyranny derived from

breaching individual civil liberties, why does the free market today offer no real alternative to the

exchange of free use of social media for pervasive surveillance? Why is it that thousands of

Facebook competitors on the advertising side—traditional publishers of content, such as

magazines and newspapers—also coordinate with Facebook to allow Facebook to watch and

monitor their own customers? Why must consumers consent to identical Facebook terms in the

privacy policies of other sellers of digital advertising? Does the free market today reflect

consumer welfare, or does it enable a monopolist to offer inferior quality products by

coordinating with other market participants?

13
For example, if one advertising company has 5,000 data points per consumer, including the data points for
age and propensity for depression, then that company can fulfill more orders for a campaign targeted to 20-year-olds
that read about suicide than can another company that reaches the same audience but does not possess said user data
points. I acknowledge that tailored advertising can reflect either a positive or negative product evolution for
consumers, and I do not wish to enter this parallel debate. Rather, I wish to focus on the successful operation, or lack
thereof, of consumer preference and choice in the market—a paramount concern in economics.
14
Facebook sells the advertising inventory of companies that compete with Facebook on the advertising side of
the market via a program called the Facebook Audience Network (FAN). Facebook for Developers,
FACEBOOK.COM (2018), https://developers.facebook.com/products/audience-network.
15
In a note to clients on 4/26/17, respected industry analyst Brian Wieser calculated that Facebook and Google
accounted for 77% of the industry’s gross revenues in 2016, 99% of the growth in the U.S. digital ad market, and
that “the average growth rate for every other company in the sector was close to 0.” Facebook alone accounted for
77% of the industry’s year-over-year growth. See Alex Heath, Facebook and Google Completely Dominate the
Digital Ad Industry, BUSINESSINSIDER.COM (April 26, 2017), http://www.businessinsider.com/facebook-and-google-
dominate-ad-industry-with-a-combined-99-of-growth-2017-4. Wieser estimated that Facebook and Google
contributed 90% to the U.S. digital ad market’s growth in 2018. Sarah Sluis, Digital Ad Market Soars to $88 Billion,
Facebook and Google Contribute 90% of Growth, ADEXCHANGER.COM (May 10, 2018),
https://adexchanger.com/online-advertising/digital-ad-market-soars-to-88-billion-facebook-and-google-contribute-
90-of-growth/.

Electronic copy available at: https://ssrn.com/abstract=3247362

Colloquially, and in the press, Facebook is a monopoly.16 Members of Congress,

reporters, academics, and even initial founders of Facebook are speaking of Facebook’s

monopoly power and questioning the need for regulation.17 However, from an academic

standpoint, the intellectual and legal case of monopoly has not yet been made,18 perhaps due to

the nature of Facebook’s product being still “free” to the consumer. The fact that the product is

free falsely diverts attention from what antitrust policymakers and economists are most

comfortable paying attention to: price. Nevertheless, antitrust law and economics concern quality

inasmuch as they do price. As I will argue in this Paper, Facebook is a monopolist, and what

Facebook extracts overtly from consumers today, from a quality perspective, is a direct function

of Facebook’s monopoly power.

16
See, e.g., recent dialogue between Lindsey Graham (R-SC) and Facebook founder and chief executive officer
Mark Zuckerberg (Graham pressed Zuckerberg to identity competitors in the private market), and between House
judiciary committee chairman Bob Goodlatte (R-Va.) and Facebook vice president of global policy management
Monika Bickert (Goodlatte asked Bickert about the lack of competition Facebook faces). Hearing on Facebook
Congressional Hearing Before the Committee on Energy and Commerce, 115th Cong. (April 2018). For a transcript
of the hearing, see Transcript of Zuckerberg’s Appearance Before House Committee, WASH. POST (April 11, 2018),
https://www.washingtonpost.com/news/the-switch/wp/2018/04/11/transcript-of-zuckerbergs-appearance-before-
house-committee/?utm_term=.c12ab1475612. Also consider recent coverage in leading news publications,
including: David J. Lynch, Big Tech and Amazon: Too Powerful to Break Up? FINANCIAL TIMES (Oct. 29, 2017),
https://www.ft.com/content/e5bf87b4-b3e5-11e7-aa26-bb002965bce8; Christopher Mims, Tech’s Titans Tiptoe
Toward Monopoly, WALL STREET J. (May 31, 2018), https://www.wsj.com/articles/techs-titans-tiptoe-toward-
monopoly-1527783845; Jonathan Taplin, Forget AT&T. The Real Monopolies Are Google and Facebook, N.Y.
TIMES (Dec. 13, 2016), https://www.nytimes.com/2016/12/13/opinion/forget-att-the-real-monopolies-are-google-
and-facebook.html; Micah L. Sifry, In Facebook We Antitrust, THENATION.COM (Oct. 12, 2017),
https://www.thenation.com/article/in-facebook-we-antitrust/; Barry C. Lynn, America’s Monopolies Are Holding
Back the Economy, THE ATLANTIC (Feb. 22, 2017),
https://www.theatlantic.com/business/archive/2017/02/antimonopoly-big-business/514358/.
17 Hearing on Facebook, supra note 16.
18
In the absence of clear rationale for how to interpret the paradoxes of Big Tech, there emerges a parallel push
to reconsider antitrust law’s coupling with economic rationale. See, e.g., Lina Khan, The New Brandeis Movement:
America’s Antimonopoly Debate, 9 J. OF EUR. COMPETITION L. & PRACTICE 131 (2018),
https://doi.org/10.1093/jeclap/lpy020 (referring to the movement as the “New Brandeis School,” which urges that
antitrust regulation should focus not only on consumer welfare but also the structure of markets to avoid mere
concentrations of economic power). For coverage of the debate in the wider media, see Lynch, supra note 16. Others
argue that economic thinking merely needs to be stretched to understand the 21st century’s modern economy market
problems. See generally Jonathan B. Baker, Jonathan Sallet & Fiona Scott Morton, Introduction: Unlocking
Antitrust Enforcement, 127 YALE L.J. 1916 (2018).

Electronic copy available at: https://ssrn.com/abstract=3247362

The fact that the dominant player in the market conducts wide-scale commercial

surveillance begs a question—how did the market get here? This Paper first explores this

question by revisiting the rich record of the market’s history. In so doing, I conclude that

Facebook’s ability to monitor and record consumers’ digital activity reflects Facebook’s ability

to extract monopoly rents in the social media market. In Part I, I revisit the history of social

networking and discuss how privacy was once a crucial promise made to secure a competitive

advantage. Facebook foreclosed competition in a contested market with superior representations

of protecting consumer privacy, including the specific promise not to track and monitor

consumers’ digital footprints.19

For a decade, the competitive market then enjoined Facebook’s ability to initiate

commercial surveillance. Facebook tried to renege on its promise not to track users in 2007, and

again in 2010, but the market was competitive enough with adequate consumer choice to thwart

Facebook’s attempts.20 Only after an historic public offering, the acquisition of over a billion

users, and the exit of competitors, was Facebook finally able to add the condition of surveillance

to its mandatory terms.21 In Part II, I argue that this pattern of events reflects Facebook’s

monopoly power.

Part III argues that the circumstantial, or structural, evidence of Facebook’s market

dominance also explains Facebook’s monopolistic power exhibited today. On the one hand,

Facebook is a market in and of itself. Ninety-nine percent of adults in the U.S. that use social

media use Facebook. 22 The reason for this is that Facebook is a “closed” communications

network and users must join simply to access the network. Not all social networks are closed—

LinkedIn, for example, disseminates user communications across its own network but also

19 See infra note 42, FACEBOOK PRIVACY POLICY (2004).

20 Infra Parts II A, II B.
21 See infra Part II C.
22 See infra Part III.

Electronic copy available at: https://ssrn.com/abstract=3247362

Twitter. But, Facebook, with a hold on the market and greatest number of users, declines the path

of interconnection, or inter-operability. Consumers cannot forgo Facebook, in much the same

way an earlier generation was beholden in the early 20th century to another communications

behemoth, AT&T— another company that leveraged a closed-network approach to protect

market share and foreclose competition in the telephone network market.23 However, even if one

does include other social networks as part of a relevant market, Facebook still dominates with

over 80% of total consumer time spent across various social networks.

The story of Facebook’s market entry, and the evolution of its commercial surveillance,

raises another relevant question under antitrust law. Under Section 2 of the Sherman Act, 24 this

country’s antitrust statute, it is illegal for a company to acquire monopoly power by engaging in

conduct outside the bounds of “competition on the merits.”25 Facebook engaged in a decade-long

pattern of false statements and misleading conduct that may have induced users to trust and

choose Facebook over alternatives in the market. Facebook also thereby secured the coordination

of independent publishers and other businesses while perpetuating the belief that Facebook did

not and would not leverage their coordination for surveillance. Facebook’s conduct may have

artificially shifted the demand curve for the Facebook social network to the right, effectively

accelerating the direct network effects that would lock in Facebook and foreclose competition.

Part IV explores Facebook’s pattern of conduct and argues that it was anticompetitive.26

23 Richard Gabel, The Early Competitive Era in Telephone Communication, 1893-1920, 34 LAW AND
CONTEMPORARY PROBLEMS 340, 341 (1969)
https://www.jstor.org/stable/pdf/1191094.pdf?refreqid=excelsior%3Ae9908dd1ed4a18092d06284ea77baf42.
24 15 U.S.C. § 2 (2000).
25 United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (finding Microsoft conduct outside the scope

of “competition on the merits” to be anticompetitive).

26
Misleading conduct is of particular concern in markets with direct network effects. See generally PHILLIP E.
AREEDA & HERBERT HOVENKAMP, FUNDAMENTALS OF ANTITRUST LAW § 350 (4th Ed. 2015) (Areeda and
Hovenkamp explain that a “monopolist’s misrepresentations encouraging the purchase of its product can fit [the]
general test for an exclusionary practice when the impact on rivals is significant; deception of buyers can impede the
opportunities of rivals.”). See also Speech delivered by Carl Shapiro before the A.L.I. and A.B.A., 6 (Jan. 25, 1996),
http://www.justice.gov/atr/public/speeches/0593.pdf (calling for heightened scrutiny of company behavior in

Electronic copy available at: https://ssrn.com/abstract=3247362

In digital markets where consumers do not pay a price, antitrust enforcement must

become comfortable with a paradigm that focuses on quality. Never before have we had to

grapple with one of the most valuable companies in the world, a half trillion-dollar market cap

company,27 that provides important communications services to over 2 billion consumers but

charges no price. Policymakers and antitrust enforcers today can refer to the record in order to

understand the consumer valuing privacy while still ceding control over their personal data, the

flawed market structure that allows for the perpetuation of present circumstances, and the

remedy which can restore competition in communications markets.

I. PRIVACY WAS ONCE A CRUCIAL FORM OF COMPETITION

To appreciate Facebook’s monopoly power in the social network market today, one must

begin by revisiting the story of Facebook’s rise. In the beginning, the social media market

reflected the type of competitive markets one reads about in textbooks. Dozens of companies

competed furiously in an attempt to win market share. In a market where all competing products

were priced at zero, startups competed not on price, but on quality. Privacy levels quickly

emerged as an important quality attribute.28

Facebook was not the world’s first major social network—MySpace was. In fact, in the

market’s early years, MySpace dominated. Founded in 2003, MySpace quickly became an

markets with strong direct network effects); see also Microsoft Corp., 253 F.3d at 62 (finding Microsoft’s
misrepresentations in a market with direct network effects to be anticompetitive conduct under Section 2 of the
Sherman Act).
27 Market cap fluctuates with daily stock price changes. YAHOO FINANCE, https://finance.yahoo.com/quote/FB/

(last visited Jan. 21, 2019).

28
For other discussions regarding competition based on quality rather than price, see Maurice E. Stucke &
Ariel Ezrachi, When Competition Fails to Optimize Quality: A Look at Search Engines, 18 YALE J. OF L. & TECH. 70
(2016).

Electronic copy available at: https://ssrn.com/abstract=3247362

internet darling, especially in the wake of the dot-com bust of 2001. Within two years, in a

heated bidding war between two media conglomerates, Rupert Murdoch’s News Corp. and

Sumner Redstone’s Viacom, MySpace was acquired by News Corp. for $580 million dollars.29

By 2006, MySpace overtook Google to become the most visited website in the U.S.30 With a

hundred million users, MySpace then signed a $900 million advertising deal with Google.31 By

2007, dozens of competitors existed, but none matched the scale and growth of MySpace.32

Journalists referred to the next generation, not as “millennials,” but as the “MySpace

generation.”33

Though it looked as though MySpace had tipped the scales, and that it would inevitably

become the de facto social media platform, it would not.34 Mark Zuckerberg, the founder and

chief executive officer of Facebook, and other enterprising entrepreneurs, had an opportunity for

market entry to subvert MySpace’s early gains. When Facebook rolled into the market, users,

29
See Richard Siklos, News Corp. to Acquire Owner of MySpace.com, N.Y. TIMES (July 18, 2005),
https://www.nytimes.com/2005/07/18/business/news-corp-to-acquire-owner-of-myspacecom.html.
30
See Hanging with the In-Crowd, THE ECONOMIST (Sept. 14, 2006),
https://www.economist.com/node/7918729.
31
Fox Interactive Media Enters into Landmark Agreement with Google Inc., NEWS FROM GOOGLE (Aug. 7,
2006), http://googlepress.blogspot.com/2006/08/fox-interactive-media-enters-into_07.html.
32
Other social networking sites included SixDegrees.com, which launched 1997 and shut down 2001;
Friendster, which launched 2002 and at one point, Friendster had over 115 million users and was the most used
social network in Asia; Orkut, which was founded by Google in 2004 and shut down 2014; Flip.com, which was
founded by Conde Nast in 2006 and shut down December 2008; Bebo, which was founded in 2005 and at one point
was the most used social network in the U.K. and was purchased by AOL in 2008 for $850 million, but in 2010
AOL announced it would shut down or sell the site; Mixi, which was founded in 2004 and at one point was the
largest social networking site in Japan; CyWorld, which was founded in 1999 and was once the primary social
network in Korea, but declined after Facebook entered the market in 2009; hi5, which was launched 2004 and was a
leading social network site in Spanish-speaking countries, including Mexico, Argentina, and Venezuela and is still
in business; BlackPlanet , which was founded in 1999; a social network for African-Americans and is still in
business; MiGente, which was founded 2001 for Latinos and is still operating; and Yahoo 360, which was Yahoo’s
effort at a social networking site with a soft-launch in 2005 that did not result in a full product launch.
33
See, e.g., Chris DeWolfe, The MySpace Generation, FORBES MAGAZINE (April 20, 2007),
https://www.forbes.com/forbes/2007/0507/072.html#60433a9b569e; The MySpace Generation, BLOOMBERG.COM
(Dec. 12, 2005), https://www.bloomberg.com/news/articles/2005-12-11/the-myspace-generation; Rhymer Rigbey,
Employers Target the MySpace Generation, FINANCIAL TIMES (Aug. 7, 2006),
https://www.ft.com/content/efce5b2a-2636-11db-afa1-0000779e2340.
34
See, e.g., John Barrett, MySpace is a Natural Monopoly, TECH NEWS WORLD (Jan. 17, 2007),
https://www.technewsworld.com/story/55185.html; Victor Keegan, Will MySpace ever Lose its Monopoly?, THE
GUARDIAN (New York), Feb. 8, 2007, https://www.theguardian.com/technology/2007/feb/08/business.comment.

Electronic copy available at: https://ssrn.com/abstract=3247362

parents, and critics were simultaneously gripped by an apprehension with social media, caused

primarily by MySpace.35 One could not ignore the barrage of negative MySpace headlines in the

media, permeating radio stations, television broadcasts, and newspaper reporting, across the

country.36 MySpace was blamed for sexual assaults, suicides, and murders, allegedly triggered

by open communications on the platform. Parents were concerned that MySpace was not safe for

children,37 and national and local news blamed the platform’s openness and lack of privacy.

Academia echoed this sentiment with consumer surveys demonstrating that social media users

were concerned about privacy and that “MySpace, the largest social networking site in the world,

has a poor reputation in terms of trust.”38

Indeed, Facebook’s disintermediation of MySpace’s rise is the story of just how

important Facebook’s qualitative differentiation was. Since MySpace was already free, Facebook

had to entice users with something unique. Thus, Facebook entered the market presenting itself,

among other things, as a privacy-centered alternative. MySpace made little to no effort to address

concerns around privacy, while Facebook appeared to take privacy seriously. Whereas MySpace

35
See CHARLES KRINSKY, THE ASHGATE RESEARCH COMPANION TO MORAL PANICS (2016).
36
See, e.g., Sue Downes, Teens Who Tell Too Much, N.Y. TIMES (Jan. 15, 2006),
http://www.nytimes.com/2006/01/15/opinion/nyregionopinions/teens-who-tell-too-much.html; Anna Bahney, Don’t
Talk to Invisible Strangers, N.Y. TIMES (May 9, 2006),
https://www.nytimes.com/2006/03/09/fashion/thursdaystyles/dont-talk-to-invisible-strangers.html; Jane Gordon,
MySpace Draws a Questionable Crowd, N.Y. TIMES (February 26, 2006),
https://www.nytimes.com/2006/02/26/nyregion/nyregionspecial2/myspace-draws-a-questionable-crowd.html; John
Kreiser, MySpace: Your Kids’ Danger, CBS NEWS (Feb, 6, 2006), https://www.cbsnews.com/news/myspace-your-
kids-danger/; Myspace Murderer: An Epilogue, WIRED (Nov. 17, 2006), https://www.wired.com/2006/11/myspace-
murder-an-epilogue/; Tom Rawstorne, How Pedophiles Prey on MySpace Children, DAILYMAIL.COM (July 21,
2006), http://www.dailymail.co.uk/femail/article-397026/How-paedophiles-prey-MySpace-children.html; Noah
Shachtman, Murder on MySpace, WIRED (Dec. 1, 2006), https://www.wired.com/2006/12/murderblog/; Brad Stone,
U.S. States Fault MySpace on Predator Issues, N.Y. TIMES (May 15, 2007),
https://www.nytimes.com/2007/05/15/technology/15iht-15myspace.5712603.html;
Brad Stone, New Scrutiny for Facebook Over Predators, N.Y. TIMES (July 29, 2007),
https://www.nytimes.com/2007/07/30/business/media/30facebook.html.
37
See, e.g., Julie Rawe, How Safe is MySpace? TIME MAGAZINE (June 26, 2006),
http://content.time.com/time/magazine/article/0,9171,1207808,00.html; Susanna Schrobsdorff, Predator’s
Playground, NEWSWEEK (Jan. 26, 2006), http://www.newsweek.com/predators-playground-108471.
38
Catherine Dwyer, Starr Hiltz & Katia Passerini, Americas Conference on Information Systems, Trust and
Privacy Concern within Social Networking Sites: A Comparison of Facebook and MySpace (2007).

Electronic copy available at: https://ssrn.com/abstract=3247362

was open to anyone who wanted to join, Facebook was closed to all but those who could validate

their identity with a university-issued “.edu” email address. After joining, Facebook users

encountered strict privacy settings—for example, the default setting was that only one’s

university classmates or one’s friends could see one’s profile.39 On the other hand, on MySpace

the default setting was more open—anyone could see one’s profile. This is relevant because

studies show that consumers do not change a website’s default settings because it is cumbersome

and time-consuming40—the modern-day equivalent of leaving the VCR clock blinking. A social

network’s default settings are therefore an important initial sales pitch.

Further signaling a commitment to privacy, Facebook almost immediately hired a chief

privacy officer and articulated to users a short, plain-English, privacy policy that put privacy at

the center of the user experience.41 Only some 950 words long, the initial policy took only a few

minutes to read. Indeed, just as the liberal return policy is designed to engender trust among

prospective customers in e-commerce, so was Facebook’s privacy policy designed to engender

trust with early adopters of social media. The opening sentence proclaimed: “Because we want to

demonstrate our commitment to our users' privacy, we will disclose our information and privacy

39
Initially, Facebook users could not make their profile visible to everyone even if they wanted to. See Boyd &
Ellison, supra note 4; DAVID KIRKPATRICK, THE FACEBOOK EFFECT: THE INSIDE STORY OF THE COMPANY THAT IS
CONNECTING THE WORLD (2011); KATHERINE LOSSE, THE BOY KINGS: A JOURNEY INTO THE HEART OF THE SOCIAL
NETWORK XV (2012).
40
Wendy E. Mackay, Triggers and Barriers to Customizing Software, CONF. ON HUMAN FACTORS IN
COMPUTING SYS.,153–160 (1991). See also Catherine Dwyer, Digital Relationships in the" Myspace" Generation:
Results From a Qualitative Study, 40TH ANNUAL HAW. INT’L CONF. ON SYS. SCIS., 19 (2007) (after interviewing
undergraduates that use social media and instant messenger applications, authors pointed out that most interviewees
did not customize the privacy settings of their chosen platforms.); Ralph Gross & Alessandro Acquisti, Proceeding
of the Ass’n for Computing Machinery Workshop, Information Revelation and Privacy in Online Social Networks
71–80 (2005) (authors studied the behavior of more than 4,000 students on Facebook and found that only 1.2% of
users studied changed their privacy settings).
41
Chris Kelly, who later ran for governor of California, was hired as Facebook’s chief privacy officer in
September of 2005.

Electronic copy available at: https://ssrn.com/abstract=3247362

practices ....”42 Therein, Facebook made important representations regarding what it would and

would not do. One important representation—whose evolution is the focus of this Paper —was

the promise to not usurp privacy by using tracking technology.43 Facebook promised: “We do

not and will not use cookies to collect private information from any user.”44

Without understanding tracking technology, it is impossible to understand how Facebook

would eventually leverage its power to degrade user privacy and build a business model affecting

data and advertising inventory across millions of independent businesses. The most common

tracking methodology to surveil users leveraged the use of “cookies”, small text files that

websites can install on users’ computers.45 At a simple level, a cookie enables a website to

identify a browser using a unique identifier, in order to remember a previous action on the part of

the user. Cookies can be used to remember historical actions—such as if a user had put items

into a shopping cart, if a user had already logged-in, or what a user normally enters in a form

field. On an individual level, cookies can add to a user’s welfare by allowing a website to

improve the quality of the site’s user experience.

42
Facebook Privacy Policy, FACEBOOK.COM (Dec. 30, 2004), http://www.thefacebook.com/policy.php
[https://web.archive.org/web/20050107221705/http://www.thefacebook.com/policy.php] [hereinafter “FACEBOOK
PRIVACY POLICY (2004)”].
43
See id. (“We use session ID cookies to confirm that users are logged in. These cookies terminate once the
users close the browser. We do not and will not use cookies to collect private information from any user.”); see also
Facebook Privacy Policy, FACEBOOK.COM (effective February 27, 2006), http://www.facebook.com:80/policy.php
[http://web.archive.org/web/20060222121022/http://www.facebook.com:80/policy.php] (“We use session ID
cookies to confirm that users are logged in. These cookies terminate once the user closes the browser. By default, we
use a persistent cookie that stores your login ID (but not your password) to make it easier for you to login when you
come back to Facebook.”); Facebook Privacy Policy, FACEBOOK.COM (effective Sept. 12, 2007),
http://www.facebook.com/policy.php
[http://web.archive.org/web/20070912083143/http://www.facebook.com/policy.php] [hereinafter FACEBOOK
PRIVACY POLICY (2007)] (“When you enter Facebook, we collect your browser type and IP address. …… We use
session ID cookies to confirm that users are logged in. These cookies terminate once the user closes the browser. By
default, we use a persistent cookie that stores your login ID (but not your password) to make it easier for you to
login when you come back to Facebook.”).
44
FACEBOOK PRIVACY POLICY (2004), supra note 42.
45
HTTP State Management Mechanism, RFC-EDITOR.ORG (April 2011). https://www.rfc-
editor.org/info/rfc6265.

Electronic copy available at: https://ssrn.com/abstract=3247362

Cookies, however, can also be used to determine what a user was researching, reading, or

buying on a site. When a user types a website URL into a browser (such as nytimes.com), the

user’s computer initiates an HTTP request with the website’s server. The website’s server then

sends back an HTTP response that allows the user to see the webpage. During HTTP responses,

servers can also send back and implant cookies on a user’s device. The cookies thumb-print a

user’s device with a unique identifier, or a “cookie ID” (for example, 123456789).46 These

cookies can then be “read” by the issuing company during a user’s subsequent HTTP requests.

When read, the user’s cookies provide the user’s cookie ID, and other elements of the HTTP

request can provide specific information regarding the user’s browser session, including: the

specific URL the user is visiting (for example, goodmenproject.com/marriage-2/Coming-Out-

To-Your-Wife/), the time the user is visiting said URL, and the IP address of the user (which

discloses the user’s geographic location). In this way, a company can use cookies to know (and

therefore remember) the fact that user 123456789 was reading the article “Coming Out to Your

Wife” on a Sunday morning at 5:30 a.m. from Wichita, KS.

Widespread commercial surveillance of consumer behavior across the internet, however,

was inhibited by two restraining facts. First, a company could only read its own cookies. Second,

a company could read cookies only when a user initiates an HTTP request to the company’s

server. For example, if The New York Times wrote a cookie onto a user’s device, the Times

could not read its cookies when the user was on wsj.com. In other words, The Times could know

what users were doing on nytimes.com but not on other properties. A company could circumvent

these built-in privacy protections by installing a piece of their own code on other websites, that

would invisibly generate an HTTP request on behalf of the user to the company’s server. Of

46
Cookies were initially invented with randomized ID numbers instead of permanent ones attributed to a user’s
browser precisely to avoid ''cookies [being] used as a general tracking mechanism.'' See John Schwartz, Giving Web
a Memory Cost Its Users Privacy, N.Y. TIMES (Sept. 4, 2001),
https://www.nytimes.com/2001/09/04/business/giving-web-a-memory-cost-its-users-privacy.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

course, no publisher would want another publisher to be tracking the behavior of its own

customers. In order to develop complete and accurate profiles of users, there would need to be

cooperation among thousands of sites which would otherwise be competitive. At one point,

competitors and competition kept horizontal cookie collusion in check.

When Facebook first entered the market, and for the next ten years, Facebook promised

to not surveil users for commercial purposes. This promise was not de minimis. As an electronic

communications service provider, Facebook knew people’s real identities and therefore could

correlate anonymous cookie IDs with real world identities. You weren’t just 123456789 reading

the article titled “Coming Out to Your Wife,” you were Jacob Greenberg of Wichita, KS reading

“Coming Out to Your Wife.” Additionally, numerous studies showed that consumers were

concerned about privacy and objected to behavioral (or psychographic) targeted digital

advertising.47 Aside from broader objections to surveillance, which is known to quell research

and speech, surveillance had immediate practical implications within the household. If a

company could know that Ms. Mitchell was shopping for that Star Wars Lego set that was all the

rage that season, it could incessantly display on the family computer ads for the same toy, which

could potentially ruin for young Sam what Santa Claus was bringing for Christmas.

Although this Paper focuses on the evolution of the promise not to surveil users outside

of Facebook, for which there is a rich historical record of Facebook’s conduct both pre and post

market dominance, it is worth noting that Facebook made other significant privacy

47
Online privacy and the ability to browse, read, shop, and think online without being watched has been a
consumer concern since the turn of this century. See Chris Jay Hoofnagle, Jennifer Urban & Su Li, Privacy and
Advertising Mail, BERKELEY CENT. FOR L. AND TECH. (2012), http://ssrn.com/abstract=2183417; M. J. Metzger & S/
Docter, Public Opinion and Policy Initiatives For Online Privacy Protection. 47 J. OF BROADCASTING & ELEC.
MEDIA 3 (2003); Schwartz, supra note 46 (2001 survey by Public Opinion Strategies, a Republican polling
organization, showed 67% of Americans said “online privacy” was a “big concern”); U.S. Internet Users Ready to
Limit Online Tracking for Ads, News.Gallup.Com (Dec. 21, 2010), https://news.gallup.com/poll/145337/internet-
users-ready-limit-online-tracking-ads.aspx (2010 USA Today/Gallup poll showed Americans were aware of cookies
and behavioral advertising, 67% said marketers should not be able to conduct behaviorally targeted advertising, and
61% said free internet services do not justify the practice).

Electronic copy available at: https://ssrn.com/abstract=3247362

representations upon entering the market.48 For example, Facebook initially gave users the ability

to opt-out of having their information shared with third-parties, including advertisers or

marketers, or to prohibit Facebook from collecting additional information about themselves from

third-parties.49 Facebook also initially promised that users could modify or remove information

Facebook had about them at any time. These promises combined signaled to users that, unlike

other social networks, Facebook offered its members “very granular and powerful control on the

privacy … of their personal information.”50 The combination of Facebook’s closed network

approach and strict default privacy settings made people believe that Facebook, and not

MySpace, was the trustworthy choice.

Facebook’s short privacy policy, default privacy settings, and outward signaling as

privacy-centric, were strategic decisions that played an important role in attracting users to the

platform. David Kirkpatrick, who chronicles Facebook’s history in his book The Facebook

Effect, summarizes, “Privacy … has been a major concern of Facebook’s users from the

beginning.”51 For example, for Katherine Losse, one of Facebook’s first employees, Facebook’s

unique stance on privacy was a critical part of her decision to join a social media platform. She

explained, “The privacy protections of the restricted network … made it feel, surprisingly,

okay.”52 Accordingly, early consumer studies revealed that the broader population possessed

similar feelings for privacy concerns.53 The majority of Americans polled in one 2004 national

48
For a general overview of Facebook’s progressive deterioration of user privacy, see Jennifer Shore & Jill
Steinman, Did You Really Agree to That? The Evolution of Facebook’s Privacy Policy, TECH. SCI. (2015),
https://techscience.org/a/2015081102.
49
FACEBOOK PRIVACY POLICY (2004), supra note 42.
50
Alessandro Acquisti & Ralph Gross, Imagined Communities: Awareness, Information Sharing, and Privacy
on the Facebook, INT’L WORKSHOP ON PRIVACY ENHANCING TECHS. 2 (2006) [hereinafter “Imagined
Communities”].
51
See Kirkpatrick, supra note 39 at 13.
52
See Losse, supra note 39 at XV.
53
David Stark & C. Hodge, Consumer Behaviors and Attitudes About Privacy: A Tns/Truste Study, TNS/TRUST
13 (2004), http://www.truste.org/pdf/Q4_2004_Consumer_Privacy_Study.pdf.

Electronic copy available at: https://ssrn.com/abstract=3247362

consumer study said privacy was a “really important issue that [they] care about often.”54

Another survey focused on early Facebook users’ attitudes towards privacy and concluded that

Facebook users cared about privacy policy more than terrorism.55 Others in academia studied and

compared MySpace users’ satisfaction with MySpace’s privacy settings with Facebook users’

satisfaction with Facebook’s privacy settings and concluded that users generally preferred

Facebook’s settings over MySpace’s settings.56

Indeed, increased levels of privacy are broadly understood to be the mechanism that

induces people to trust and then communicate over a particular communications channel. For

example, social exchange theory, with a foot simultaneously in economics, psychology and

sociology, studies human communications and explains that people communicate and form

relationships through a simple cost-benefit analysis57: the benefit is the opportunity to socialize,

and the cost is the risk associated with trusting someone. Privacy increases trust and reduces risk,

thereby causing humans to communicate more. In the real word, the lack of a permanent record

ensures some level of privacy.58 Online, a social network that wants users to choose its platform

over others could generate superior trust through superior promises of privacy.

Privacy continues to play an important role in the current, though less competitive, social

media market. SnapChat is the only social network post-2010 to have gained significant adoption

with users in the United States.59 Like Facebook once did, SnapChat competes on quality, and

54 Id.
55
See Acquisti & Gross, Imagined Communities, supra note 50 at 6.
56
See Acquisti & Gross, Imagined Communities, supra note 50 at 16; Dwyer, Hiltz & Passerini, supra note 38
at 339.
57
See Donna L. Hoffman, Thomas P. Novak & Marcos Peralta, Building Consumer Trust Online, 42 COMMS.
OF THE ACM 80–85 (1999); see also Miriam J. Metzger, Privacy, Trust, and Disclosure: Exploring Barriers to
Electronic Commerce, 9 J. OF COMPUTER-MEDIATED COMM. 942 (2004); MICHAEL E. ROLOFF, INTERPERSONAL
COMMUNICATION: THE SOCIAL EXCHANGE APPROACH (1981).
58
Consider Lawrence Lessig’s prescient essay on internet threats to privacy: Lawrence Lessig, The
Architecture of Privacy, 1 VAND. ENT. L. AND PRACTICE 56-65 (1999).
59 See STATISTA.COM, infra note 237.

Electronic copy available at: https://ssrn.com/abstract=3247362

importantly, on privacy levels. For example, one of the central draws of SnapChat is the fact that

SnapChat communications appear for a fixed period of time, then disappear, being also deleted

from SnapChat’s servers.60 The impermanency of the record ensures a high degree of privacy.

Additionally, when Facebook users discovered in early 2017 that Facebook was collecting and

sharing user data in ways previously not known or understood, users instigated a

#deleteFacebook movement, and millions flocked to startup competitor Vero Social that

promised significantly higher levels of privacy.61

The fact that communications markets turn on privacy is history repeating itself. Wireless

telegraphy, before eventually being used to facilitate what we know today as the modern radio,

initially stalled as a technology because of its inability to facilitate private communications.62

Moreover, this nation’s historic antitrust case against and eventual break-up of AT&T in the

telephone network market was instigated in part by an equipment competitor’s desire to offer

telephone equipment that offered a higher level of privacy in communications.63 Until 1956,

customers of AT&T’s telephone service were required to use AT&T telephone equipment to

make and receive calls. Equipment manufacturer Hush-A-Phone marketed a cup-like device that

attached to AT&T equipment to reduce the risk of conversations being overheard. Consumers

purchased over 125,000 Hush-A-Phones to make for “privacy in communications” and in 1948,

Hush-A-Phone filed a complaint with the Federal Communications Commission requesting that

60
When Does Snapchat delete Snaps and Chats?, SNAPCHAT.COM, https://support.snapchat.com/en-US/a/when-
are-snaps-chats-deleted, (last visited Jan. 21, 2019).
61
See Ben Gilbert, #DeleteFacebook Is Trending: Here’s How to Delete Your Facebook Account,
BUSINESSINSIDER.COM (March 18, 2018), https://www.businessinsider.com/how-to-delete-facebook-2018-3; see
also Ann-Marie Alcantara, Could This Social Media Ad-Free App Go Mainstream Amid Facebook’s Privacy Woes?,
ADWEEK.COM (April 12, 2018), https://www.adweek.com/digital/this-ad-free-social-media-app-thinks-it-can-
capitalize-on-facebooks-privacy-woes/.
62
BHU SRINIVASAN, AMERICANA:THE 400-YEAR HISTORY OF AMERICAN CAPITALISM 293 (2017).
63
JEAN-JACQUES LAFFONT & JEAN TIROLE, COMPETITION IN TELECOMMUNICATIONS 18 (1999).

Electronic copy available at: https://ssrn.com/abstract=3247362

the FCC compel AT&T to officially permit customers’ use of Hush-A-Phone equipment. 64

Hush-A-Phone eventually prevailed and this opened the door to competition in the market for

telephone equipment.65

In the early 2000s, Facebook entered the social media market, disintermediated

MySpace’s momentum, and started to consolidate the social network market by offering

consumers an alternative service of superior quality. Other qualitative features, like Facebook’s

user interface design, real-name policy, and college launch strategy, also played a role in making

Facebook’s platform attractive to users.66 Regardless, the fact remains that superior privacy

levels played a central role in this narrative. Propelled then by direct network effects, whereby

each additional user that chose Facebook made the Facebook network more attractive to the next

incremental user, Facebook started to consolidate the social network market at a rapid rate,

taking market share from MySpace, Friendster, and competitive offerings from Google and

AOL. At the end of 2004, MySpace had 5 million users and Facebook had 1 million users.67 In

2006, when some 250 million people around the world used a social network, 100 million used

MySpace, 12 million used Facebook, and the rest were dispersed across numerous competitors.68

Then, in 2007, user growth at MySpace started to decelerate, while growth at Facebook

accelerated. By mid 2007, Facebook had overtaken MySpace as the most visited social media

64
For the court of appeals decision, see Hush-A-Phone v. United States, 238 F.2d 266 (D.C. Cir. 1956).
65 Id.
66 For example, Sean Parker, Facebook’s first president, has explained that Facebook’s strategy in first

dominating the social media market, and general product development, were important reasons that users chose
Facebook over MySpace. Alexia Tsotsis, Sean Parker on Why MySpace Lost to Facebook, TECHCRUNCH.COM (June
28, 2011), https://techcrunch.com/2011/06/28/sean-parker-on-why-myspace-lost-to-facebook/. Others have partly
attributed Facebook’s superior product to superior design. See, for e.g., Jenna McWilliams, How Facebook Beats
MySpace, THE GUARDIAN (June 23, 2009),
https://www.theguardian.com/commentisfree/cifamerica/2009/jun/23/facebook-myspace-social-networks.
67 Number of Active Users at Facebook Over the Years, YAHOO! FINANCE (Oct. 23, 2012),

https://finance.yahoo.com/news/number-active-users-facebook-over-years-214600186--finance.html.
68 Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

network in the U.S.69 One year later, MySpace stopped growing, while Facebook was registering

half-a-million new users per day.70 Today, of course, MySpace, Orkut, Bebo, and other

competitors, are relics of the early history of social media. Facebook dominates the life of the

average American—99% of adults that use any form of social media use Facebook, and the

average American now spends over an hour per day on Facebook applications.71

II. THE PARADOX OF SURVEILLANCE REFLECTS MONOPOLY POWER

Monopoly power refers specifically to the power to control a market—that is, the power

in a market to raise price above or reduce quality below competitive levels.72 From a legal

standpoint, the question of whether a company has monopoly power in a market is answered

through direct or indirect evidence. Indirect evidence focuses on a company’s percent share of a

relevant market, amongst other structural factors that indicate a company’s hold on a market

(e.g., entry barriers).73 Direct evidence, on the other hand, demonstrates a company’s acquired

ability to increase price above or decrease quality below levels unsustainable in a previous

competitive environment.74 Under the direct evidence approach, the strongest type of evidence is

evidence of price or quality levels pre-power, an event or point in time which results in power,

69 Erick Schonfeld, Facebook Blows Past MySpace in Global Visitors for May, TECHCRUNCH.COM (June 20,

2008), https://techcrunch.com/2008/06/20/facebook-blows-past-myspace-in-global-visitors-for-may/.
70 Id.
71 See Social Media Fact Sheet, PEW RES. CTR. (Feb. 5, 2018), http://www.pewinternet.org/fact-sheet/social-

media/ (69% of U.S. adults use social media; 68% of U.S. adults use Facebook). Average Time Spent per Day with
Facebook, Instagram, and Snapchat, EMARKETER.COM (June 1, 2018), https://www.emarketer.com/Chart/Average-
Time-Spent-per-Day-with-Facebook-Instagram-Snapchat-by-US-Adult-Users-of-Each-Platform-2015-2020-
minutes/219352.
72
See generally United States v. E. I. du Pont de Nemours & Co., 351 U.S. 379, at 391 (1956) [hereinafter “du
Pont”].
73
United States v. Microsoft Corp., 253 F.3d 34, 51 (D.C. Cir. 2001) (“...monopoly power may be inferred
from a firm's possession of a dominant share of a relevant market that is protected by entry barriers.”); United States
v. Grinnell Corp., 384 U. S. 563, 571 (1966).
74
See du Pont, 351 U.S. at 391 (explaining that “monopoly power is the power to control prices”); Standard Oil
Co. of New Jersey v. United States, 221 U.S. 1 (1911).

Electronic copy available at: https://ssrn.com/abstract=3247362

and evidence of price increases or quality degradations that would have previously been

unsustainable.75 This fact-based, retrospective before-and-after analysis—though somewhat rare

in antitrust cases—is particularly relevant in Facebook’s case.

Part II now tells the story of how it came to be that Facebook extracts from consumers an

exchange founded on surveillance. Contrary to what many believe, digital surveillance is not

simply the inevitable byproduct of how the internet works. Rather, promises of privacy were the

deciding factors that tipped the early market in Facebook’s favor, away from MySpace.

Facebook’s conduct, from 2004-2012, provides the benchmark of quality—at least with respect

to commercial surveillance—that the restraining forces of competition demanded. By 2014,

competitors had exited the market, Google’s competitive offering Orkut shut down,76 and

Facebook’s monopoly was complete due to the exit of competition combined with the protection

of the barrier to entry that results from a product with over a billion users on a closed

communications network. Subsequently, in 2014, Facebook leveraged its market power in a

consolidated market to successfully degrade privacy to levels unsustainable in the earlier

competitive market when market participants were subject to consumer privacy demands.

Facebook rapidly unraveled its promise not to conduct commercial surveillance by using the

technical framework it built over the years by perpetuating the belief that it would not leverage

such a framework for a commercial purpose.

A. PRE-POWER: Failure of Beacon and Early Misrepresentations

75
See generally Daniel A. Crane, Market Power Without Market Definition, 90 NOTRE DAME L. REV. 31
(2014).
76
See Ellen Huet, Google Finally Shuts Down Orkut, Its First Social Network, FORBES MAGAZINE (June 30,
2014), https://www.forbes.com/sites/ellenhuet/2014/06/30/google-kills-orkut/#2874fded634b.

Electronic copy available at: https://ssrn.com/abstract=3247362

By 2007, people were sharing more baby pictures and personal family and friend photos

on Facebook than they were on any other social media site—MySpace was edgy, Facebook was

wholesome.77 Then on November 6, 2007, with growing momentum in the market, Facebook

reneged on the promise not to surveil users outside of Facebook through the release of an

advertising product called “Beacon.”78 Beacon was a direct product license to third-parties that

openly allowed Facebook to monitor and record user activity off of Facebook, and reflects

Facebook’s first attempt to track users on the sites of independent businesses.79 But Beacon was

immediately controversial. Its ultimate failure is evidence of what competition demanded of

Facebook from a privacy perspective.

Launched in conjunction with a handful of third-parties, including Blockbuster and The

New York Times, Facebook provided Beacon participants a piece of Facebook code to install on

their own sites.80 When a user triggered an action on a participating site (say, rented a movie or

read an article), the website then presented the user with a pop-up box requesting permission to

share the user’s activity on Facebook.81 Many people today would remember being once jarred

by the pop-up requesting permission to share one's reading or browsing activity with Facebook.

The user could decline permission by selecting the “No, Thanks” option.82 If the user did not

77
Facebook’s Mark Zuckerberg: Hacker. Dropout. CEO., FASTCOMPANY.COM (May 1, 2007),
https://www.fastcompany.com/59441/facebooks-mark-zuckerberg-hacker-dropout-ceo.
78 Leading Websites Offer Facebook Beacon for Social Distribution, NEWSROOM.FB.COM (NOV. 6, 2007),

https://newsroom.fb.com/news/2007/11/leading-websites-offer-facebook-beacon-for-social-distribution/; see also

Facebook Announces Facebook Ads, Beacon in New York, ADWEEK.COM (Nov. 6, 2007),
https://www.adweek.com/digital/facebook-announces-facebook-ads-beacon-in-new-york/.
79 NEWSROOM.FB.COM, supra note 78.
80 Id.
81 Michael Arrington, Ok Here’s At Least Part of What Facebook is Announcing on Tuesday: Project Beacon,

TECHCRUNCH.COM (Nov. 2, 2007), https://techcrunch.com/2007/11/02/ok-heres-at-least-part-of-what-facebook-is-

announcing-on-tuesday/.
82
Users could have opted-in (via Facebook settings) to always include activity on third-party sites on their
News Feed, opted-out to never include it, or opted-in to include it only if the user did not opt-out when given the
pop-up confirmation (default). For a short history of Beacon’s changing user-interface against a backdrop of
consumer uproar, see Louise Story, The Evolution of Facebook’s Beacon, N.Y. TIMES (Nov. 29, 2007),
https://bits.blogs.nytimes.com/2007/11/29/the-evolution-of-facebooks-beacon/.

Electronic copy available at: https://ssrn.com/abstract=3247362

click the “No, Thanks” button, Facebook received information about the user’s activity (title of

movie rented or article read).83 Then, Facebook would publish the activity on the user’s

Facebook page, calling the publication a “social advertisement.”84 For Beacon participants, like

The New York Times or Conde Nast, the social sharing of user activity was a type of free

marketing.

Screenshot of the Beacon pop-up displayed to users on third-party sites.85

Beacon was a transparent attempt to get users to consent to sharing information about

their activity on third-party sites with Facebook. However, the mere presence of Facebook’s

code on these sites made it technically possible for Facebook to track users’ activities on these

sites, even if the user did not grant Facebook permission to do so. This is because, when a user

visited a Beacon site (e.g., blockbuster.com), regardless of whether the user consented or not,

Facebook code initiated an HTTP request on behalf of the user to Facebook’s servers. Through

this newly opened connection, Facebook could write cookies on user computers during HTTP

responses, or read cookies during HTTP requests. The requests and cookies could reveal the

specific page a user was on—effectively allowing Facebook to accomplish surveillance on the

users that had clicked “No, Thanks.”

83
According to Facebook at the time, Facebook tracked and recorded the activity only of logged-in users who
did not click the “No, Thanks” button.
84
See Arrington, supra note 81.
85 Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

At the time, Facebook claimed that Beacon only tracked and monitored the activities of

consenting users, and that Facebook code was not used to conduct less overt surveillance through

cookies. For example, in a follow-up to emerging, strong, user push-back on the new Beacon

product,86 The New York Times interviewed Facebook’s vice president of marketing and

operations, Chamath Palihapitiya.87 The reporter asked, “If I buy tickets on Fandango, and

decline to publish the purchase to my friends on Facebook, does Facebook still receive the

information about my purchase?” Palihapitiya answered, “Absolutely not. One of the things we

are still trying to do is dispel a lot of misinformation that is being propagated unnecessarily.”88

Facebook represented that it did not receive information about users that declined to share

information about their activity.

Only hours after Palihaptiya’s comments in the Times, Stefan Berteau, a senior research

engineer at California’s Threat Research Group, examined the actual contents of Facebook’s

HTTP requests and responses that were normally invisible to users, and revealed that

Palihapitiya’s representations were not true.89 While Facebook claimed that Beacon trackers did

not transmit user information to Facebook if users clicked “No, Thanks”, the contents of the

cookie files did just that.90 Berteau further revealed that Beacon trackers were also used to log

86 See Nick O'Neill, MoveOn.org to Challenge Facebook Beacon, ADWEEK.COM (Nov. 20, 2007),
www.adweek.com/digital/moveonorg-to-challenge-facebook-beacon/.
87
Brad Stone, Facebook Executive Discusses Beacon Brouhaha, N.Y. TIMES (Nov. 29, 2007),
https://bits.blogs.nytimes.com/2007/11/29/facebook-responds-to-beacon-brouhaha/.
88 Id.
89
Stefan Berteau, Facebook’s Misrepresentation of Beacon’s Threat to Privacy: Tracking Users Who Opt Out
or Are Not Logged In, CA SECURITY ADVISOR RES. BLOG, (Dec. 1, 2007),
https://www.techmeme.com/071201/p3#a071201p3.
90
Id. (Berteau showed, for example, that if a user saved a recipe on Epicurious to a favorites folder, and
explicitly declined to have this information publish to his Facebook profile, the Beacon program wouldn’t publish it,
but would nonetheless share the information about what the user was doing on the Epicurious website with
Facebook). See also Juan Carlos Perez, Facebook’s Beacon More Intrusive than Previously Thought,
PCWORLD.COM (Nov. 30, 2007), https://www.pcworld.com/article/140182/article.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

the activity of users who logged-out of Facebook or did not have a Facebook account.91

Furthermore, if a logged-out user ever clicked the “Remember Me” checkbox when logging-in,

Facebook actively associated cookies with a user’s Facebook ID number.92 In other words,

Facebook linked normally anonymized cookie data back to the identities of real people—a

connection Facebook could make because it operated the social network.

Indeed, Facebook’s Privacy Policy at the time did not obtain user consent for this

practice. As displayed on November 29, 2007, Facebook claimed that it only used cookies to

“confirm that users are logged in” and that “[t]hese cookies terminate once the user closes the

browser.”93 Within a few days, Facebook confirmed Berteau’s findings, contradicting its own

earlier representations about user privacy.94 For the first time, Facebook had been publicly

caught using cookies for broader surveillance despite the promise it did not and would not do so.

Almost immediately, Facebook faced user protest, petitions from advocate groups, and

class action lawsuits. A MoveOn.org petition garnered 50,000 signatures within days.95 Class-

action lawsuits on behalf of users were filed in Texas and California.96 The page of another

petition called “Facebook, stop invading my privacy,” stated: “A lot of us love Facebook - it’s

helping to revolutionize the way we connect with each other. But they need to take privacy

91
See Berteau, supra note 89; see also Juan Carlos Perez, Beacon’s User Tracking Extends Beyond Facebook,
CA Says, COMPUTERWORLD.COM (Dec. 3, 2007), https://www.computerworld.com/article/2537951/data-
privacy/beacon-s-user-tracking-extends-beyond-facebook--ca-says.html.
92
See Berteau, supra note 89; Perez, supra note 90.
93
FACEBOOK PRIVACY POLICY (2007), supra note 43.
94
See Perez, supra note 90.
95
See O'Neill, supra note 86;
Ellen Nakashima, Feeling Betrayed, Facebook Users Force Site to Honor Their Privacy, WASH. POST (Nov.
30, 2007), http://www.washingtonpost.com/wp-
dyn/content/article/2007/11/29/AR2007112902503.html?noredirect=on.
96
See Harris v. Blockbuster Inc., 622 F. Supp. 2d 396 (N.D. Tex. 2009); Lane v. Facebook Inc., 696 F.3d 811
(9th Cir. 2012).

Electronic copy available at: https://ssrn.com/abstract=3247362

seriously.” 97 Facebook was founded upon the qualitative promise of no surveillance outside of

Facebook and users did not want this to change. Consumer resistance is early proof of

consumers’ preference for no surveillance.

Rejection of Facebook surveillance on third-party sites was part of a wider rejection of all

third-party cookie tracking as consumers tried to stop this type of commercial surveillance that

the advertising industry coordinated to impose on consumers.98 A 2005 industry study showed

that 67% of Americans wanted to protect their privacy and prevent tracking.99 Some advertising

companies paid businesses for the ability to install their own code on the business’ websites to

track the business’ customers.100 One such company was Seevast. Seevast’s code triggered an

open connection between a business’ website visitors and Seevast’s servers, which Seevast used

to install cookies on, and retrieve cookies from, the business’ users.

Unlike companies like Seevast, Facebook knew users’ real identities, and nearly fifty

million people had a Facebook account.101 Facebook’s surveillance, unlike Seevast’s, could be

tied not only to random cookie variables, but to peoples’ real names. In other words, Facebook

could leverage the ability to identify people through use of the century’s new communications

technology, to conduct a particularly invasive, and permanent, form of surveillance.

97
See Eric Auchard, Facebook Alters Notifications after Privacy Furor, REUTERS.COM (Nov. 29, 2007),
https://www.reuters.com/article/us-facebook-privacy-idUSN2925736120071130.
98
For example, consumers’ cookie deletion rates rose at a rapid clip starting in 2004 when the advertising
industry started to use cookies for tracking. Subsequently, the digital ad industry association group Interactive
Advertising Bureau (IAB) and Safecount announced an initiative to cut cookie deletion rates. See Mickey Khan,
Rising Cookie Rejection Bites into Metrics, DMNEWS.COM (July 11, 2005), https://www.dmnews.com/customer-
experience/news/13074659/rising-cookie-rejection-bites-into-metrics. See also
Chris Jay Hoofnagle, Jennifer King, Su Li & Joseph Turow, How Different are Young Adults from Older Adults
When It Comes to Information Privacy Attitudes and Policies? (2010),
https://repository.upenn.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=1413&co
ntext=asc_papers (finding that 39% of American Internet users delete all their cookies “often”).
99
See Khan, supra note 97 (citing study from online market researcher InsightExpress).
100
See Saul Hansell, Facebook Retreats on Online Tracking, N.Y. TIMES (Aug. 15, 2006),
https://www.nytimes.com/2006/08/15/technology/15search.html.
101
See Julie Sloane, Facebook Got Its $15 Billion Valuation — Now What? WIRED (Oct. 26, 2007),
https://www.wired.com/2007/10/facebook-future/.

Electronic copy available at: https://ssrn.com/abstract=3247362

Additionally, Facebook did not have to pay companies, it could simply leverage the power to

give away something valuable for free.

In the face of backlash, some Beacon participants pulled out of the Beacon program,

effectively declining to extract consent from their own website visitors to Facebook’s

surveillance mechanism. The growing e-retailer Overstock, one of the initial companies to sign

up for Beacon, pulled out. Their senior vice president of corporate affairs said, “we need to make

sure that the Facebook community is accepting of this new type of advertising.”102

In a competitive market, Facebook likely worried that strong user discontent around

privacy would disrupt its momentum. While MySpace faced stagnating user growth, it still had

double Facebook’s number of users.103 Google’s Orkut had lost steam, but Google and MySpace

recently announced they would join forces on Google’s new OpenSocial initiative.104 Orkut,

Friendster, and Netscape co-founder Marc Andreessen announced support for OpenSocial.105 In

the U.K., a key battleground for the social media market, MySpace and Bebo still had the most

users.106 Additionally, competitors displayed an appreciation of having to compete with

Facebook on matters related to user privacy. One such savvy competitor was MySpace’s parent

company, News Corp. An analyst with Jupiter Research remarked to CNN Money, “News Corp.

102
Caroline McCarthy, Facebook’s Zuckerberg: ‘We Simply Did a Bad Job’ Handling Beacon, CNET.COM
(Dec. 5, 2007), https://www.cnet.com/news/facebooks-zuckerberg-we-simply-did-a-bad-job-handling-beacon/. See
also Louise Story, Coke Is Holding Off on Sipping Facebook's Beacon, N.Y. TIMES (Nov. 30, 2007),
https://bits.blogs.nytimes.com/2007/11/30/coke-is-holding-off-on-sipping-facebooks-
beacon/?mtrref=undefined&gwh=59CBD7EC0914ECD5FCAB88B8BBF2201A&gwt=pay.
103
See Sloane, supra note 101.
104
See Jacqui Cheng, Google Goes After Facebook With New OpenSocial Social Networking API,
ARSTECHNICA.COM (Nov. 1, 2007), https://arstechnica.com/uncategorized/2007/11/google-goes-after-facebook-
with-new-opensocial-social-networking-api/.
105
See Marc Andreessen, Open Social: A New Universe of Social Applications All Over the Web,
BLOG.PMARCA.COM (Oct. 31, 2007), http://blog.pmarca.com/2007/10/open-social-a-n.html,
[http://web.archive.org/web/20071102041108/http://blog.pmarca.com/2007/10/open-social-a-n.html] (Andreesen
wrote on his blog, "We will aggressively support Open Social in every conceivable way.”).
106
See Rory Cellan-Jones, Facebook Dismisses Privacy Fears, BBC NEWS (Sept. 12, 2007),
http://news.bbc.co.uk/2/hi/technology/6990767.stm (citing Nielsen netRatings and comScore numbers).

Electronic copy available at: https://ssrn.com/abstract=3247362

and Fox recognize the importance of allowing people to be alone with their friends, so they do

not feel like they are being looked at by Big Brother. They understand how many competitors

they have nipping at their heels right now, so they are doing everything they can not to alienate

users.”107

With numerous competitors nipping at Facebook’s heels, Facebook retreated almost

immediately after launching Beacon. Zuckerberg apologized and announced that Facebook

would allow users to opt-out of the Beacon program.108 But consumers did not accept

Facebook’s opt-out scheme, which required them to navigate Facebook’s default privacy

settings. Consumer uproar persisted, and, by September of the following year, Facebook revealed

it would shut down Beacon entirely. Zuckerberg would later call Beacon a “mistake.”109

Facebook’s retreat with Beacon is evidence of what competition demanded of Facebook

at the time. Following subsequent calls for regulation, Randall Rothenberg, the President and

CEO of the Internet Advertising Bureau, the industry’s trade association, wrote an op-ed in The

Wall Street Journal, proclaiming that users’ ability to resist Facebook’s privacy changes was

testament to the free market’s ability to regulate itself.110 Rothenberg, of course, was wrong.

107
Paul R. La Monica, Move Over, MySpace Social Networking is Hot - and There's More to the Rapidly
Growing Market than MySpace and Facebook, CNN.COM (Mar. 20, 2007),
http://money.cnn.com/2007/03/19/news/companies/socialnetworks/index.htm.
108
See Erick Schonfeld, Zuckerberg Saves Face, Apologizes For Beacon, TECHCRUNCH.COM (Dec. 5, 2007),
https://techcrunch.com/2007/12/05/zuckerberg-saves-face-apologies-for-beacon/; Vauhini Vara, Facebook Rethinks
Tracking Site Apologizes, Makes it Easier to Retain Privacy, WALL STREET J. (Dec. 6, 2007),
https://www.wsj.com/articles/SB119687856122414681?mod=searchresults&page=3&po=18s;
109
See Mark Zuckerberg, Thoughts on Beacon, BLOG.FACEBOOK.COM (Dec. 5, 2007),
http://blog.facebook.com/blog.php?post=7584397130,
[https://web.archive.org/web/20080107025500/http://blog.facebook.com/blog.php?post=7584397130]; and Mark
Zuckerberg, Our Commitment to the Facebook Community, NEWSROOM.FB.COM (Nov. 29, 2011),
https://newsroom.fb.com/news/2011/11/our-commitment-to-the-facebook-community/. For a history of
Zuckerberg’s apologies, see Geoffrey A. Fowler & Chiqui Esteban, 14 Years of Mark Zuckerberg
Saying Sorry, Not Sorry, WASH. POST (April 9, 2018),
https://www.washingtonpost.com/graphics/2018/business/facebook-zuckerberg-apologies/?utm_term=.dbcff8504c9.
110
Randall Rothenberg, Facebook's Flop, WALL STREET J. (Dec. 14, 2007),
https://www.wsj.com/articles/SB119760316554728877?mod=searchresults&page=3&pos=16 (writing “Internet
consumers have shown themselves willing and able to police the medium on their own. Just ask Facebook:

Electronic copy available at: https://ssrn.com/abstract=3247362

On the heels the Beacon controversy, and competitors’ rising awareness of the

importance of privacy to consumers, Facebook took the unprecedented step of announcing that

future privacy changes would be subject to user approval. 111 Under a newly announced

democratic process, incorporated into Facebook’s governing documents, Facebook bound itself

to allowing users to vote on future changes to important documents that contractually change

user privacy—including the Privacy Policy, and other policies such as the Statement of Rights

and Responsibilities and The Facebook Principles. Facebook’s press release announced that the

voting procedure “offers its users around the world an unprecedented role in determining the

future policies governing the service.”112 In a rarely given public press conference, Zuckerberg

explained that Facebook was doing this because social media users “feel a visceral connection to

their rights. ... We are one of the only services on the web where people are sharing pretty

personal and intimate information ... We’re making it so that we can’t just put in a new terms of

service without everyone’s permission. We think these changes will increase the bonding and

trust users place in the service.” 113

The promise to let users vote on future privacy changes assured users that Facebook

would not leverage its growing power to undermine users’ privacy without users’ meaningful

consent. Facebook was already earning healthy profits on its exchange with users. In 2009,

Consumer regulation proved itself to be a far more effective, efficient, economically productive and unforgiving
mechanism than federal regulation ever will be.”).
111
See Facebook Announces First-Ever User Vote on Terms of Service Changes, ADWEEK.COM (Apr. 6, 2009),
https://www.adweek.com/digital/facebook-announces-first-ever-user-vote-on-terms-of-service-changes/; Facebook
Drafts New Governing Documents, Adopts New User Voting Process on Policy Changes, ADWEEK.COM (Feb. 26,
2009), https://www.adweek.com/digital/facebook-drafts-new-governing-documents-process-for-user-voting-on-
policy-changes/.
112
Facebook Opens Governance of Service and Policy Process to Users, NEWSROOM.FB.COM (Feb. 26, 2009),
https://newsroom.fb.com/news/2009/02/facebook-opens-governance-of-service-and-policy-process-to-users/
113
Facebook’s general counsel Ted Ullyot, and Facebook’s vice president of privacy were also present. See
Rafe Needleman, Live Blog: Facebook Press Conference on Privacy, CNET.COM (Feb. 26, 2009),
https://www.cnet.com/news/live-blog-facebook-press-conference-on-privacy/.

Electronic copy available at: https://ssrn.com/abstract=3247362

Facebook earned $229 million in profits on $777 million in revenues.114 In 2010, it earned $606

million on $1.974 billion.115 Profit margins were 29% and 30% respectively—not quite as high

as today’s 47% but healthy nonetheless. But as Facebook’s market power grew through the

foreclosure of competition and the lock-in of network effects, Facebook would eventually

abolish this newly announced voting procedure and reinstate the scope, scale, and invasiveness

of Beacon’s mission. Today, Facebook surveillance is a mandatory tie-in with a third-party’s

(e.g., The New York Times) use and license of other Facebook business products (Like buttons,

Logins, etc.).

B. PRE-POWER: More Backtracking and Pattern of Conduct

This section traces the subsequent development of Facebook’s “social plugin” products

on the heels of Beacon's retreat. This early history of Facebook plugins is relevant to an antitrust

inquiry for three reasons. First, history reveals that competition continued to restrain Facebook’s

ability to initiate surveillance. Second, Facebook’s surveillance framework today requires the

coordination of millions of independent third-parties. Facebook induced publishers and others to

first coordinate with Facebook upon the representation that Facebook would not leverage their

coordination for commercial surveillance. Third, the record opens the door to consider in Part IV

whether Facebook’s pattern of conduct reflects an anticompetitive acquisition of monopoly

power under Section 2 of the Sherman Act.

114
See FACEBOOK INC., REGISTRATION STATEMENT (Form S-1) (Feb. 1, 2012),
https://www.sec.gov/Archives/edgar/data/1326801/000119312512034517/d287954ds1.htm [hereinafter “FACEBOOK
S-1”].
115 Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

The relevant history of Facebook social plugins centers around the “Like” button—

introduced early in 2010, at Facebook’s annual F8 developer conference.116 The Facebook Like

buttons, and even the Login buttons,117 are products offered to publishers and other independent

businesses to improve site functionality and increase revenue. For publishers, the Like buttons,

offered a turn-key review and distribution mechanism.118 Facebook explained, “[e]ach Like

creates distribution on Facebook, which brings more Facebook users back to the article on your

site.”119 Because online publishers generate incremental revenue for each click on an article,

more user visits meant more money. According to Facebook, installation of social plugins

increased traffic on average by 200%.120

Thousands of publishers (competitors of Facebook for digital ad dollars), and other

websites and apps, hoping for incremental ad revenue, flocked to install the Facebook Like

button. All a third-party had to do was install a single line of HTML Facebook code into their

application.121 Within the first week of availability, more than 50,000 sites added social

116
John D. Sutter, Facebook Makes It Easier for Users to Share Interests Across Web, CNN.COM (April 21,
2010), http://www.cnn.com/2010/TECH/04/21/facebook.changes.f8/index.html. Like buttons are small buttons,
which often display the text ‘Like’ alongside a blue thumbs-up icon. Third-parties could install the buttons on their
websites. A news website, for example, might display Like buttons near articles. Readers could click on the button
to indicate support for an article’s point of view. When clicked, the button counter might increase by one, the user be
given an opportunity to comment, and the article be published to the user’s News Feed. For users, the Like button
was marketed as a communications tool—enabling users to “easily share interesting content with friends.” Facebook
Platform Showcase. FACEBOOK.COM (Dec. 5, 2010),
http://developers.facebook.com/showcase/news?p=wallstreetjournal,
[https://web.archive.org/web/*/http://developers.facebook.com/showcase/news?p=wallstreetjournal]. See also Like
Button for the Web, FACEBOOK.COM (2018), https://developers.facebook.com/docs/plugins/like-button.
117
The Facebook Login plugin (called Registration Plugin at the time) was another plugin launched December
2008 that allowed third-parties to deploy Facebook’s user registration and sign-in, as opposed to their own. More
seamless sign-in meant more sign-ins, which often meant more revenue. See Jessica E. Vascellaro, Facebook Plans
Enhanced Ties to Outside Services, WALL STREET J. (July 24, 2008),
https://www.wsj.com/articles/SB121687251639480377?mod=searchresults&page=1&pos=17.
118
See The Value of a Liker. FACEBOOK.COM (Sept. 29, 2010), https://www.facebook.com/notes/facebook-
media/value-of-a-liker/150630338305797.
119
Id.
120
See Facebook Platform Showcase, supra note 116.
121
See generally Brand Permissions Center Usage Guidelines, FACEBOOK.COM (Dec. 9, 2010),
http://www.facebook.com/brandpermissions/logos.php
[https://web.archive.org/web/20101209212738/http://www.facebook.com/brandpermissions/logos.php]; Social

Electronic copy available at: https://ssrn.com/abstract=3247362

plugins.122 CNN, The New York Times, The Wall Street Journal, Slate, and ABC were among

the many initial adopters.

Facebook social plugins opened a vulnerability between users’ devices and Facebook’s

servers much in the same way that Beacon did a few years earlier. Like Beacon, social plugins

required independent businesses to install Facebook code on their websites, which opened a

backdoor communication between users’ devices and Facebook’s servers. Meaning, the Like

button was not simply a static image that one hard-coded into their website. If The New York

Times had installed Like buttons, and the user visited nytimes.com, while the user initiated an

HTTP request with a Times server, the nytimes.com response would include Facebook code that

would then automatically initiate a separate request from the user to Facebook—for the purpose

of retrieving and displaying a Like button. As was the case with Beacon, if it so wanted,

Facebook could leverage third-party initiated requests to glean users’ data, and to write and/or

read tracking cookies.123

For many years, Facebook perpetuated the belief it would not leverage backdoor access,

the way it had with Beacon, to conduct surveillance for commercial purposes. Consumers had

shown an aversion to the idea of Facebook tracking them while not on Facebook. The steady

stream of public claims that Facebook did not and would not use the network of code for social

plugins to monitor and track consumer behavior prompted consumers to continue to trust and

Plugins, FACEBOOK.COM (Nov. 26, 2010), http://developers.facebook.com/docs/reference/plugins/like

[https://web.archive.org/web/20101126215652/http://developers.facebook.com/docs/reference/plugins/like]; Social
Plugins, FACEBOOK.COM (Dec. 7, 2010), http://developers.facebook.com/plugins
[https://web.archive.org/web/20101207235101/http://developers.facebook.com/plugins].
122
How to Use the New Facebook Social Plugins for Your Business, FACEBOOK.COM (May 4, 2010),
https://www.facebook.com/notes/facebook-for-developers/how-to-use-the-new-facebook-social-plugins-for-your-
business/394310302301/.
123
Cookies set during the retrieval of third-party functionality are called third-party cookies (or TPCs). For a
definitional break-down of third-party cookies, first-party cookies, and session cookies, see Jessica Davies, Know
Your Cookies: A Guide to Internet Ad Trackers, DIGIDAY.COM (Nov. 1, 2017), https://digiday.com/media/know-
cookies-guide-internet-ad-trackers/.

Electronic copy available at: https://ssrn.com/abstract=3247362

therefore choose Facebook over alternatives in the market. But Facebook’s claims were also

important in soliciting the coordination of third-parties to spread the presence of Facebook code

across the internet. Many third-parties, publishers for example, competed with Facebook on the

advertising side of the market. They licensed and installed social plugins as a means to distribute

their own content. Surveillance of their own readers, however, could be used against them to

undercut the value of and pricing power over their own proprietary readers.124 Specifically, if

Facebook could compile a list of people that read the Journal, even those who did not use

Facebook, it could simply sell the ability to retarget “Journal readers” with ads across the

internet for a fraction of the cost that the Journal charged.

When Zuckerberg first announced the Like buttons at the 2010 developer’s conference,

he did not mention that Like buttons could be used to track users.125 At the time, Facebook was

under intense privacy scrutiny, some politicians threatened investigations, and Facebook faced

competition over privacy. MySpace announced it was now offering superior privacy controls in

an attempt to get privacy-concerned users to switch. Facebook was on a pre-IPO mission to get

users to choose Facebook over alternatives in the market, and in May 2010, Zuckerberg

convened a press conference to address privacy concerns.126 The consumer technology

124
Facebook surveillance of users on independent publisher sites and apps benefits Facebook advertising
revenue in two ways. One, Facebook can leverage the data to better target users (or conduct superior attribution for)
on Facebook’s own properties. Two, Facebook can leverage the data to directly target a publisher’s readers and price
undercut a publisher’s ad rates.
125
See generally Declan McCullagh, Facebook ‘Like’ Button Draws Privacy Scrutiny, CNET.COM (June 2,
2010), https://www.cnet.com/news/facebook-like-button-draws-privacy-scrutiny/.
126 On May 26, 2010, Facebook convened a press conference at its Palo Alto headquarter. There, in response to

Julia Boorstin, CNBC: How does this controversy and your new approach to privacy affect your approach to
revenue and your business model?, Zuckerberg said: “So it might be kind of crazy... to people...I don't know. It
might seem weird. I don't actually know exactly what the external perception of this is. But I always read these
articles that are like "OK you guys must be doing this because it's going to make you more money." And honestly
for people inside the company that could not ring less true.” KIRKPATRICK, supra note 39; Jessica E. Vascellaro,
Facebook Grapples with Privacy Issues, WALL STREET J. (May 19, 2010),
https://www.wsj.com/articles/SB10001424052748704912004575252723109845974. U.S. senators Charles Schumer
(D-N.Y.), Michael Bennet (D-Col.), Mark Begich (Alaska) and Al Franken (D-Minn.) wrote an open letter to
Facebook founder and CEO Mark Zuckerberg addressing privacy concerns. See generally Senators’ Letter to
Facebook, POLITICO (Apr. 27, 2010), https://www.politico.com/story/2010/04/senators-letter-to-facebook-036406.

Electronic copy available at: https://ssrn.com/abstract=3247362

publication CNET covered Facebook’s Like button launch and addressed users’ concerns around

Like buttons being used to decrease user privacy.127 One issue, even absent the topic of cookies,

was that the presence of Like buttons on other sites enabled Facebook to receive any internet

users’ URLs as they moved around the internet, and Facebook’s privacy policy did not appear to

obtain users’ consent for this practice.128 Facebook responded by saying that its privacy

statement was “not as clear as it should be, and we’ll fix that.”129 Barry Schnitt, a Facebook

spokesman, then allayed concerns by explaining that Facebook plug-ins work like any other

plug-ins on the internet, Facebook does not use social plug-in data for advertising, and Facebook

may use received data to catch bugs in its software.130

But in November of 2011, Dutch researcher Arnold Roosendaal exposed Facebook’s

hidden activity with Like buttons in the way that Berteau did earlier with Beacon. Shortly after

launching the Like buttons, Roosendaal published a paper showing that Facebook was using the

Like button code now installed on third-party sites to write and read user cookies.131 Roosendaal

showed that each time a Facebook user visited a site with a Like button, Facebook retrieved the

user’s Facebook website login cookies, which contained the user’s unique identifying number,

traceable to his or her real identity.132 Facebook again was leveraging login cookies from the

communications network to conduct detailed surveillance possibly for the advertising side of its

business. In addition to a user’s ID number, Facebook retrieved the specific URL the user was

on, which revealed the title of an article the user was reading or the name of the product a user

127
McCullagh, supra, note 125.
128
Id. (for example, CNET pointed out that Facebook’s FAQ stated, “No data is shared about you when you
see a social plug-in on an external website,” and that Facebook’s privacy policy also did not appear to obtain users’
consent.)
129 Id.
130 Id.
131
Arnold Roosendaal, Facebook Tracks and Traces Everyone: Like This!, (Tilburg L. Sch. Legal Studs. Res.
Paper Ser. No. 03/2011, Nov. 30, 2010), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1717563.
132 Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

was buying. Roosendaal then demonstrated that Facebook used these open connections to write

cookies and surveil the behavior of people that did not even have Facebook accounts.

The Wall Street Journal published the results of its own investigative study confirming

Roosendaal’s findings.133 The study, led by a former Google engineer, examined the presence of

social widgets on the world’s top 1,000 most-visited sites. The Journal concluded that Facebook

buttons had been added to millions of websites, including a third of the top 1,000 most-visited

sites. Facebook knew when a user reads an article about “filing for bankruptcy” on MSNBC.com

or about depression on a small blog, even if the user didn’t click any Like button.

With a finger on the pulse of Americans’ sentiment towards tracking, the Journal

recommended users log-out of Facebook to stop Facebook from tracking them and gave

Facebook an opportunity to comment. Bret Taylor, Facebook’s chief technology officer at the

time, responded definitively to the privacy breach allegations, “We don’t use them for tracking

and they’re not intended for tracking.”134 Taylor clarified that Facebook places cookies on the

computers of people that visit facebook.com to protect users’ Facebook accounts from cyber-

attacks.135 The Journal also gave Facebook an opportunity to answer to Roosendaal’s allegation

that Facebook was tracking people that did not even have a Facebook account. Facebook said

that Roosendaal had found a “bug,” and that it had therefore discontinued this practice.136

But the issue—as important as it was—was not put to rest. Later that year, in September

of 2011, Nik Cubrilovic, an Australian internet security contractor, followed up on Roosendaal’s

133
Amir Efrati, 'Like' Button Follows Web Users, WALL STREET J. (May 18, 2011),
https://www.wsj.com/articles/SB10001424052748704281504576329441432995616 (The WSJ study also found that
Facebook trackers could continue to track internet users even if users had closed their browsers or turned off their
computers.). See also Reed Albergotti, Facebook to Target Ads Based on Web Browsing, WALL STREET J. (June 12,
2014), www.wsj.com/articles/facebook-to-give-advertisers-data-about-users-web-browsing-1402561120.
134 Efrati, supra note 133.
135
Id.
136
Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

and the Journal’s discoveries, and published an article showing that Facebook Like and other

plugins were still tracking user activity outside of Facebook even if users had completely logged

out of Facebook.137 Normally, when a user logs out of a service, the service’s login cookies

terminate. But Facebook’s cookies weren’t terminating, they were persistent and still able to

identify and track people. Mainstream media picked up on Cubrilovic’s reporting in an effort to

hold Facebook accountable. In coverage about Facebook’s “alleged cookie snooping,” the

Huffington Post asked, “Is Facebook tracking which websites users visit even after they’ve

logged out of the service?”138

Facebook again acted quickly to pacify the growing drumbeat of public concern over

Facebook’s intentions with regards to leveraging social plugin code for surveillance. A Facebook

spokesperson, Facebook engineers, and the Facebook privacy policy were put forth to dispel

worry. In one instance, Facebook responded to a CBS News inquiry with a statement that,

“Facebook does not track users across the web. … No information we receive when you see a

social plugin is used to target ads, [and] we delete or anonymize this information within 90 days

....”139 Elsewhere, Facebook engineers responded directly in the comments section of blogs and

articles.140 Facebook engineer Arturo Bejar pleaded with users, “please know that ... when you’re

logged in (or out) we don’t use our cookies to track you on social plugins to target ads or sell

your information … We use your logged in cookies ... for safety and protection.”141 When

137
See Jason O. Gilbert, Facebook Logout Tracking: Privacy Concerns Arise Over Alleged Cookie Snooping,
HUFFINGTON POST (Sept. 26, 2011), https://www.huffingtonpost.com/2011/09/26/facebook-logout-cookies-privacy-
tracking_n_980838.html. See also Emil Protalinski, Facebook Denies Cookie Tracking Allegations, ZDNET.COM
(Sept. 25, 2011), https://www.zdnet.com/article/facebook-denies-cookie-tracking-allegations/.
138
Gilbert, supra note 137.
139
Erik Sherman, Facebook’s New Privacy Bust: Users Log In but They Can’t Log Out, CBS NEWS (Sept. 26,
2011), https://www.cbsnews.com/news/facebooks-new-privacy-bust-users-log-in-but-they-cant-log-out-update/.
140
See Protalinski, supra note 137 (Facebook spokesperson replied to a journalist’s request for comment by
pointing to Facebook engineer’s public blog comments).
141 Id.

Electronic copy available at: https://ssrn.com/abstract=3247362

pressed by news outlet ZDNet for an official statement, a Facebook spokesperson directed the

reporter to Bejar’s statement.142 The New York Times also disseminated Facebook’s

representations to the public.143 This all appeared to corroborate what the Facebook Help Center

represented to consumers at the time, “we do not use [information we see when you visit a

website] to deliver ads.”144

Facebook’s reasoning though—regarding users’ safety and protection—still called into

question why Facebook cookies contained users’ Facebook ID numbers. Normally, a company

might include user ID numbers in cookies if it intended to conduct individual tracking. Thus, in

direct response to Cubrilovic’s allegations, Facebook promised to change how its cookies

worked so that users’ ID numbers were not embedded in cookies after users logged out.145 The

removal of account ID numbers from cookies ensured that Facebook could not conduct de-

anonymized user surveillance in spite of claims of its intentions not to do so.

If Facebook, or any other company, did want to conduct mass commercial surveillance

on users, one precondition would be the pervasive consent and coordination of third-parties (e.g.,

The New York Times and Overstock). One way a company could do this would be to do so

overtly, as Facebook had with Beacon—simply ask users with a pop-up box if they want to share

what they are reading or buying on other sites with Facebook. But this direct, overt, attempt at

getting users to share their information had failed. Another way Facebook could go about

accomplishing the same end would be to get third-parties to install Facebook code for an

142 Id.
143
Riva Richmond, As ‘Like’ Buttons Spread, So Do Facebook’s Tentacles, N.Y. TIMES (Sept. 27, 2011),
https://bits.blogs.nytimes.com/2011/09/27/as-like-buttons-spread-so-do-facebooks-tentacles/.
144
See Protalinski, supra note 137.
145
See Alan Henry, Facebook Is Tracking Your Every Move on the Web; Here's How to Stop It,
LIFEHACKER.COM (Sept. 26, 2011), https://lifehacker.com/5843969/facebook-is-tracking-your-every-move-on-the-
web-heres-how-to-stop-it.

Electronic copy available at: https://ssrn.com/abstract=3247362

independent product, like the Like buttons, then eventually leverage use and dependence of Like

buttons (or other social plugins) to later extract a new use permission.

For the American consumer, this architecture of interests in the market was precarious.

Independent businesses were creating consumer privacy vulnerabilities for their own financial

gain; and, it all rested on Facebook living up to its word. This fact did not wash over the

reporters that sought to hold Facebook accountable to its representations. In December of 2012,

The Wall Street Journal revisited the issue, to point out that Facebook plugins now appeared on

two-thirds of websites surveyed.146 But Facebook again publicly assured that it only uses data

from unchecked Like buttons for security purposes and to fix bugs in its software. Noticeably

absent from Facebook’s public statement was the fact that Facebook also filed a patent

application the year prior, on September 22, 2011, for a “method … for tracking information

about the activities of users of a social networking system while on another domain.”147 Michael

Arrington, veteran tech blogger, caught it and called it “Brutal Dishonesty.”148

If Facebook wanted to eventually usurp privacy by using the back-end code from Like

buttons or other social plugins, it faced a roadblock—the user referendum process for privacy

changes introduced a few years earlier. Thus, while publicly representing that it only used social

plugin data for users’ safety and protection and generally deflecting concern over Facebook

intentions, Facebook simultaneously dismantled the user voting process. In late 2012, with over a

billion users, and an historic initial public offering now under its belt, Facebook addressed this

roadblock expediently. Facebook proposed major changes that could pave the way for Facebook

146
See Julia Angwin, It's Complicated: Facebook's History of Tracking You, PROPUBLICA.ORG (June 17,
2014), https://www.propublica.org/article/its-complicated-facebooks-history-of-tracking-you; Geoffrey A. Fowler,
What You Can Do About Facebook Tracking, WALL STREET J. (Aug. 5, 2014), https://www.wsj.com/articles/what-
you-can-do-about-facebook-tracking-1407263246.
147
U.S. Patent No. 2011/0231240A1 (filed Feb. 8, 2011) (issued Sept. 22, 2011).
148
Michael Arrington, Facebook: Brutal Dishonesty, UNCRUNCHED.COM (Oct. 1, 2011),
https://uncrunched.com/2011/10/01/brutal-dishonesty/.

Electronic copy available at: https://ssrn.com/abstract=3247362

to decrease user privacy. One provision proposed was the abolishment altogether of future

referendums for privacy changes.149 After soliciting user feedback, and receiving push-back,

Facebook submitted the proposed changes to a vote. Eighty-eight percent of users voted against

Facebook’s proposed privacy changes.150

But Facebook shrugged. At this point, users had high switching costs and the fine print of

the governing documents requiring referendums had a kill switch. In order for a user vote to be

binding, 30% of users would have had to vote in the election.151 With over a billion users, and

only some 589,000 votes casted, Facebook discarded the results of the election.152 Facebook then

moved forward with privacy erosions and the abolishment of the referendum process. Many

users protested that Facebook had not informed them of the upcoming election, either by

notifying them upon a Facebook login or by simply sending them an email.153 Other users

claimed to receive an email notifying them of a vote, but not informing them how and where to

vote. It was an election where Facebook didn’t want anyone voting.

C. POST-POWER: Deterioration of the Promise Not to Track

“We give lots of f*** about your privacy, so we wrote this. Read it, so you know what the f***
we’re going to do with the s*** you post ...”

—Comments by Facebook employees in 2011 parodying the genuineness of Facebook’s concern for user privacy;
specifically, the Facebook user privacy policy.154

149
See ANTONIO GARCIA MARTINEZ, CHAOS MONKEYS: OBSCENE FORTUNE AND RANDOM FAILURE IN SILICON
VALLEY (2016).
150
See Dan Farber, The Facebook Vote and a Nation-State in Cyberspace, CNET.COM (Dec. 11, 2012),
https://www.cnet.com/news/the-facebook-vote-and-a-nation-state-in-cyberspace/#.
151 Id.
152 Id.
153
See Donna Tam, The Polls Close at Facebook for the Last Time, CNET.COM (Dec. 10, 2012),
https://www.cnet.com/news/the-polls-close-at-facebook-for-the-last-time/.
154
See Alexia Tsotsis, It’s About Time Someone Translated the Facebook TOS Into Bro Speak,
TECHCRUNCH.COM (Aug. 17, 2011), https://techcrunch.com/2011/08/16/code-curls-girls/; Martinez, supra note 149
at 317.

Electronic copy available at: https://ssrn.com/abstract=3247362

By early 2014, rivals that initially competed with Facebook including MySpace,

Friendster, Mixi, Cyworld, hi5, BlackPlanet, Yahoo’s 360, AOL’s Bebo, and dozens of others,

had exited the market.155 Google—Facebook’s rival and nemesis—announced it would shut down

its competitive social network Orkut. By June of 2014, Orkut had exited the market, effectively

ceding the social media market to Facebook.156 For Facebook, the network effects of over a

billion users on a closed communications protocol further locked-in the market in its favor. With

countless friends and family connected to a given consumer, the cost of foregoing Facebook for

the consumer grew in proportion with Facebook’s growth. For Facebook, these circumstances—

the exit of competition and the lock-in of consumers—greenlit a change in conduct.

Facebook’s monopoly power gave it the ability to further deteriorate privacy and extract

more of the user’s data—which, as Dave Wehner, chief financial officer of Facebook, clarified

on Facebook’s Q2 2018 earnings call, is directly correlated with higher ad revenues.157 Part I of

this Paper detailed Facebook’s initial commitment to privacy, and Sections A and B surveyed

Facebook’s inability to extract a condition of surveillance in a competitive market. Section C

traces Facebook’s ability to reverse course post-power—direct evidence of monopoly power.

First, Facebook initiates user surveillance for commercial ad purposes. Second, Facebook ties

user identification with cookies to conduct more intrusive surveillance. Third, Facebook then

circumvents user attempts to opt-out or block Facebook’s quality degradations. The three quality

deteriorations can be understood as the monopoly rents Facebook is able to command in the

market today. Absent competition, Facebook is able to degrade quality levels below that which

was required in a competitive market, and financially profit from this conduct.

155
See supra text accompanying notes 69-70.
156
See Huet, supra note 76.
157
Facebook Q2 Earnings, supra note 12.

Electronic copy available at: https://ssrn.com/abstract=3247362

Facebook’s quality deterioration led to an interesting phenomenon—decreasing user

satisfaction, despite Facebook’s continued ability to retain and grow its user base. According to

the American Consumer Satisfaction Index (ASCI), social media is amongst the lowest scoring

of all industries surveyed. With an industry average of 72, social media’s ASCI score is lower

than even health insurance and airlines.158 Facebook, with a score of 67, and a trailing average of

66, has an ASCI score lower than almost every American airline—and is also lower than the

average industry benchmarks of 95% of the industries covered by the ASCI study. This paradox

defies the law of demand—which says that given a constant price, a decrease in quality must

necessarily lead to a decrease in buyer consumption.

1. Facebook Initiates Commercial Surveillance.━ In June of 2014, Facebook announced

it would leverage its code presence on third-party applications to track consumers, enabling it to

surveil the specific online behavior of this country’s citizens despite widespread preference to

the contrary.159 Facebook would do precisely what it had spent seven years promising it did not

and would not do, and finally accomplished what the previous competitive market had restrained

it from doing. With a relatively quick software update, Facebook would leverage the code on

third-party sites and apps used to deliver other Facebook products—Like buttons, Login buttons,

conversion tracking pixels,160 retargeting pixels, and the Facebook software development kit—

158
ACSI E-Business Report 2018, THEACSI.ORG (July 24, 2018), http://www.theacsi.org/news-and-
resources/customer-satisfaction-reports/reports-2018/acsi-e-business-report-2018.
159
Making Ads Better and Giving People More Control Over the Ads They See, NEWSROOM.FB.COM (June 12,
2014), https://newsroom.fb.com/news/2014/06/making-ads-better-and-giving-people-more-control-over-the-ads-
they-see/; Albergotti supra note 133. For a technical description of how Facebook tracks people, reference a
technical report prepared for the Belgian Privacy Commission in 2015: GÜNEŞ ACAR, BRENDAN VAN ALSENOY,
FRANK PIESSENS, CLAUDIA DIAZ, & BART PRENEEL, FACEBOOK TRACKING THROUGH SOCIAL PLUG-INS,
https://securehomes.esat.kuleuven.be/~gacar/fb_tracking/fb_pluginsv1.0.pdf.
160
The Facebook conversion tracking pixel is a piece of Facebook code that advertisers affix to their websites
to enable Facebook to report back to the advertisers whether a Facebook ad campaign is yielding traffic and sales; in
other words, it a piece of code that allows the advertiser to measure the return-on-investment of Facebook ad

Electronic copy available at: https://ssrn.com/abstract=3247362

for the additional new purpose of tracking users.161 In a previously competitive market,

Facebook was not able to get away with this qualitative degradation. Now Facebook could

significantly degrade its quality because consumers no longer had alternative social networks to

turn to.

Facebook had repeatedly pacified privacy concerns by representing that any data gleaned

from the presence of Facebook code on third-party sites was not to be used for “commercial

purposes” but rather for users’ own “safety and protection.” But now Facebook changed course

and announced that the data derived from tracking consumers would augment Facebook ad

targeting, attribution, and measurement.162 In other words, this deterioration of privacy would be

directly related to increased revenue and profits. First, Facebook would use data from this

commercial surveillance to enhance its ad targeting algorithms, which meant that Facebook ads

could be more targeted and reach a larger relevant advertising base than those of other ad sellers

in the market—such as The New York Times or Hearst. Second, data from commercial

surveillance would allow Facebook to get paid for more advertising through increased

attribution. A significant percentage of marketers only pay Facebook if ads result in a specific

measurable outcome (i.e., a click on an ad or a sale of a product). Facebook calls these action-

based ads. Attribution refers to the process of identifying the set of user actions that lead to a

campaigns. The announcement that Facebook would use data it retrieves from its conversion tracking pixels also
upset advertisers. The tracking pixel was there to report to the paying advertiser whether the ads were working.
Now, Facebook could use data garnered from one campaign to sell advertising to competitors. For example,
Facebook could use data from the Facebook conversion tracking pixels on Audi’s website (to measure Audi’s ad
campaign), to better sell advertising to Mercedes (by targeting users who previously looked at Audis on Audi’s
website).
161
See Cotton Delo, Facebook to Use Web Browsing History for Ad Targeting, ADAGE.COM (June 12, 2014),
adage.com/article/digital/facebook-web-browsing-history-ad-targeting/293656/.
162
See id. (Facebook is using the passive data, where users go on their PCs and phones, to make its own ads
smarter). See also, Parmy Olson, Facebook Moves to Become the World's Most Powerful Data Broker, FORBES
MAGAZINE (April 30, 2014), https://www.forbes.com/sites/parmyolson/2014/04/30/facebok-moves-to-become-the-
worlds-most-powerful-data-broker/#662d16f42006 (Facebook uses personal data as “leverage with advertisers who
are desperate to better-target their ads”).

Electronic copy available at: https://ssrn.com/abstract=3247362

desired result.163 Increased commercial surveillance allows Facebook to fine-tune attribution

models and claim credit for more actions, which further increases Facebook’s ad revenues.

When Facebook reversed course in 2014, unlike when it tried to do so earlier, Facebook

code was deployed across millions of sites and mobile apps, and the intentions of the code were

altered in one fell swoop. Over the course of the seven years that Facebook represented it would

not use social widgets to track consumers, millions of websites had signed up for and installed

Facebook plugins.164 Not including mobile apps, this included approximately 30% of the top 1

million of the most-visited websites,165 including news websites like The Wall Street Journal,

The Washington Post, and The San Francisco Chronicle. Facebook laid the groundwork for

tracking by requiring third-parties to install Facebook code in order to license Facebook’s other

products. For independent publishers and retailers, Facebook would thereafter tie surveillance of

their own customers with the continued use and license of Facebook’s social network products

for businesses.

Proprietary access to subscribers and the identities of readers and visitors is a highly

guarded asset historically by subscription businesses. It is unlikely that publishers would have

shared this information unless they were under the belief that Facebook was a content

distribution platform and traffic generator, not a surreptitious aggregator of consumer data for

Facebook's own internal, and competitive, advertising sales efforts. Facebook obtained the initial

cooperation of third-party businesses through the inducements of content distribution and the

convenience of single login. Now Facebook would receive the ability to monitor the behavior of

their customers—competitors with Facebook in the digital advertising market—by changing the

163
IAB Attribution Hub, IAB.COM (2018), https://www.iab.com/guidelines/iab-attribution-hub/.
164
See BUILTWITH.COM, supra note 6.
165
See Steven Englehardt & Arvind Narayanan, Online Tracking: A 1-million-site Measurement and Analysis,
WEBTAP.PRINCETON.EDU (2018),
http://randomwalker.info/publications/OpenWPM_1_million_site_tracking_measurement.pdf.

Electronic copy available at: https://ssrn.com/abstract=3247362

fine print of permissions. Facebook increasingly knew as much about The Wall Street Journal’s

readers as the Journal did itself. Furthermore, unlike the Journal, Facebook now knew which

Journal readers were avid ESPN readers, giving it the capability to bundle and sell targeted

audiences, which further commoditized the value of competitors’ inventory. Under the new

regime, when a consumer visited a website with a Facebook plugin, Facebook piggy-backed onto

the requests and responses necessary to simply display the plugins, to now also surveil the users

of competitor ad sellers—rendering the Facebook code a Trojan Horse of sorts.

From the consumer’s perspective, consumers could choose whether to reveal information

on Facebook itself, and they did.166 Now, consumer choice, even the choice not to use Facebook,

no longer mattered. In the earlier competitive market, the cooperation of third-parties, on which

Facebook’s tracking depended, was predicated on how consumers felt about the proposition. As

discussed above, previously, when consumers protested, participating companies stopped

coordinating with Facebook.167 In 2014, unlike in 2007, Facebook did not have only a handful of

third-parties working with it. Facebook had a substantial portion of the horizontal market

coordinating with it for some functionality or another—whether for user-registration or article

sharing. These independent businesses now had their own switching costs. They had built their

businesses over the last seven years to depend on Facebook code, and now, that reliance was

correlated with their own revenue performance.168 Finally, the conscientious objector no longer

had any power to alter the direction of the wider market.

Reflecting its ability to influence market actors in the ecosystem, Facebook then required

all businesses to change their own privacy policies to extract from their own users the consent to

166
See Acquisti & Gross, Imagined Communities, supra note 50 at 13 (e.g., Acquisti & Gross’ study of what
Facebook users do to satisfy their desire for privacy revealed that users “claim to manage their privacy fears by
controlling the information they reveal”).
167
See McCarthy, supra note 102.
168
See generally Roosendaal, supra note 131; FACEBOOK.COM, supra note 118.

Electronic copy available at: https://ssrn.com/abstract=3247362

have Facebook track them for commercial purposes. For convenience, Facebook provides exact

copy-and-paste legal language to use:169

“Third parties, including Facebook, may use cookies, web beacons, and other storage

technologies to collect or receive information from your websites, apps and elsewhere on the

internet and use that information to provide measurement services, target ads …”

This is how consumers went from having a preference of privacy to having nearly every

competitor in the market extract from them identical and uniform consent for Facebook’s

commercial surveillance practices.

2. Facebook Leverages Consumer Identity for Stronger Surveillance.━ Facebook further

deteriorated user privacy by tying the newly announced tracking of consumer behavior across the

wider internet with the real, stable, human identities that Facebook knew because of its position

in the social network market.170 When the restraining forces of competition worked, Facebook

had to remove user IDs from cookies to ensure that Facebook could not conduct de-anonymized

169
Facebook Platform Policy, FACEBOOK.COM (2018), https://developers.facebook.com/policy (“Obtain
adequate consent from people before using any Facebook technology that allows us to collect and process data about
them, including for example, our SDKs and browser pixels. When you use such technology, provide an appropriate
disclosure... That third parties, including Facebook, may use cookies, web beacons, and other storage technologies
to collect or receive information from your websites, apps and elsewhere on the internet and use that information to
provide measurement services, target ads and as described in our Data Policy.”).
170
This decision was reflected in Facebook’s re-launch of the Atlas ad server. See generally see Jack Marshall,
Adblock Lets Users Quash Facebook’s Atlas Tracking, BLOGS.WSJ.COM (Oct. 1, 2014),
https://blogs.wsj.com/cmo/2014/10/01/adblock-lets-users-quash-facebooks-atlas-tracking/?ns=prod/accounts-wsj;
Zach Rodgers, With Atlas Relaunch, Facebook Advances New Cross-Device ID Based on Logged in Users,
ADEXCHANGER.COM (Sept. 28, 2014), https://adexchanger.com/platforms/with-atlas-relaunch-facebook-advances-
new-cross-device-id-based-on-logged-in-users/. Also consider comments from competitive ad sellers in the digital
media market, for example, from YieldBot CEO Jonathan Mendez, in an interview with The Wall Street Journal,
“[Facebook has] a lot of advantages … in terms of their data and relationship with consumers and this allows them
to leverage it.” Why You Should Care About Facebook’s Atlas Ad Relaunch, WALL STREET J. (Oct. 15, 2014),
https://www.wsj.com/video/why-you-should-care-about-facebooks-atlas-ad-relaunch/C11CF1DC-C0CA-4D0D-
8F98-2ACEA86CD0B7.html?mod=searchresults&page=1&pos=3.

Electronic copy available at: https://ssrn.com/abstract=3247362

surveillance.171 Now, with the foreclosure of competition, Facebook would reinstate invasive

identity monitoring.172 This reinstatement is further evidence of Facebook’s monopolistic market

power.173

By augmenting tracking with consumer identification, Facebook also circumvented users’

attempts to limit tracking by deleting cookies (or resetting a mobile device’s advertising

identifier). Clearing cookies expunges cookie variables from a user’s device and breaks the link

between one’s device and the cookie’s memory.174 For example, suppose The New York Times

wanted to surveil a user via cookie ID 123456789. If a user deleted the cookie, the next time the

user visited nytimes.com, the Times’ server would try to identify the user but find no cookie. The

profile the Times had compiled on user 123456789 would thereafter be worth little. The ability

to correlate tracking data to one’s identity now circumvented the ability of consumers to wipe

171
See Henry, supra note 145.
172
For an advertising competitor’s perspective on Facebook’s decision to tie Facebook IDs with tracking
cookies, see Olson, supra note 162 (where Mark DiMassimo, CEO of the ad agency DiMassimo Goldstein, explains
“Facebook [now] knows who you are”).
173
Part IV of this paper examines whether Facebook illegally acquired monopoly power in the social network
market. A separate issue to be examined is whether, by bundling social network user IDs into its own advertising
inventory and the inventory of other market actors like Hearst, Facebook is illegally “leveraging” its monopoly
position in the social network market to inappropriately monopolize the advertising market. Though the Supreme
Court curtailed Section 2 “leveraging” theory in Verizon v. Trinko, some Courts have interpreted Trinko to leave the
door open to some types of leveraging claims. Verizon Commc'ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540
U.S. 398 (2004); Z-Tel Communications, Inc. v. SBC Communications, Inc., 331 F. Supp. 2d 513 (E.D. Tex. 2004).
The Areeda treatise contends the Sherman Act permits “leveraging” claims where (a) a firm uses monopoly power
in market A, (b) to place rivals in market B at a competitive disadvantage (by raising B’s costs or decreasing the
quality of B’s product), and (c) higher prices, reduced output, or reduced quality (normally associated with
monopoly power) results in market B. In the technical sense, the authors of the Areeda treatise argue that proof of
(c) meets the definition of “monopolization” or “attempt to monopolize” within the literal language of §2, that
monopoly market share is not necessary to a finding of “monopolization” or “attempt to monopolize”, and that
therefore, “leveraging” is an unnecessary distinguishable §2 theory. AREEDA & HOVENKAMP, §652, supra note 26.
Furthermore, Facebook’s tracking of consumers on third-party sites and leveraging of user IDs may be challenged as
illegally maintaining and perpetuating the Facebook monopoly. New entrants in the social media market that also
rely on attracting advertisers cannot compete with Facebook’s commercial surveillance. Consider, for example, the
fact that Facebook monetizes U.S. & Canada users at a $17.07 average revenue per user (ARPU), but that SnapChat
can only monetize North American users at a $1.81 ARPU (even though SnapChat users only spend half the amount
of time on SnapChat than they do on Facebook). See Alexei Oreskovic, Look at the Big Gap Between SnapChat’s
Revenue Per User and Facebook’s, BUSINESSINSIDER.COM (May 10, 2017),
https://www.businessinsider.com/snapchat-arpu-versus-facebook-arpu-charts-2017-5.
174
See Chris Jay Hoofnagle, Ashkan Soltani, Nathaniel Good, & Dietrich J. Wambach, Behavioral Advertising:
The Offer You Can't Refuse, 6 HARV. L. & POL. REV. 273, 277 (2012).

Electronic copy available at: https://ssrn.com/abstract=3247362

their slate clean—Facebook, unlike The New York Times, could immediately match observed

behavior with a stable identity. The social network had a real-name policy, and occasionally

required users to prove their names with state-issued identification.175 To Facebook, it was not

user 123456789 that was reading Coming Out to Your Wife, it was simply Jacob Greenberg. If

the user deleted Facebook cookies, the profile Facebook had compiled could still live on under

the user’s real-name profile. Furthermore, Facebook could re-cookie a user the next time the user

visited the Facebook social network, which users did multiple times per day.

With code that used a persistent singular identification mechanism now pervasive across

competing websites and apps, Facebook could connect the dots on consumers as they moved

from site to site, and from computer to mobile phone, to compile rich dossiers on users.

Facebook was the single eye that could see what John Doe was doing on both The New York

Times and on The Wall Street Journal.176 The competitive market once ensured that competitors

on the advertising side of the business did not track and monitor what users were doing across

the horizontal market. The unique ability to conduct horizontal surveillance, for both Facebook

and Google, explains the current duopoly in digital advertising, where nearly all industry growth

goes to only two companies. Both companies, however, only achieve this end by leveraging a

monopoly position in another market—for Facebook, the social network market, and for Google,

the search market.

175
For Facebook’s name policy, see What Names Are Allowed on Facebook?, FACEBOOK.COM,
https://www.facebook.com/help/112146705538576 (stating that “The name on your profile should be the name that
your friends call you in everyday life. This name should also appear on an ID or document from our ID list.”).
Sometimes, Facebook has required users to prove their identity by submitting a copy of government-issued
identification. See, e.g., Hamm Samwich, A Drag Queen’s Open Letter to Facebook, HUFFINGTON POST (Sept. 18,
2014), https://www.huffingtonpost.com/hamm-samwich/nominative-dysphoria-a-dr_b_5839148.html.
176
See generally Rodgers, supra note 170 (explaining that the Facebook ID, uses the “login” as the
“foundation” of tracking, measurement, and ad personalization).

Electronic copy available at: https://ssrn.com/abstract=3247362

Armed with data derived from Facebook’s new surveillance capabilities, Facebook could

bill advertisers for more conversions, and sell advertising based on surveillance data.177 For

example, if Sally Smith read on the family’s shared computer an article about marital problems

on a small hometown news site, Facebook could know and save it to Smith’s dossier. When

Smith wakes up the next morning refreshed from a night’s sleep, the marital drama now tucked

in the past, and logs onto Facebook, she might now be presented with an ad from a divorce

lawyer in her Facebook News Feed. Alternatively, her husband might have woken up, logged

opened ESPN on the same computer, and himself been presented with an ad for a divorce

lawyer. Behind the scenes in both cases, Facebook data derived from Facebook surveillance is

the facilitator.

3. Facebook Circumvents Consumer Attempts to Opt-Out.━ Consumers did not want

Facebook to track their behavior across the Internet,178 so they tried to circumvent Facebook’s

new quality deteriorations. On the one hand, this consumer behavior is further evidence of users’

preference for no surveillance. On the other, Facebook’s behavior is evidence of Facebook using

its market power to forcibly impose on consumers that which they are still—in a consolidated

market—trying to resist. First, Facebook itself did not and does not allow consumers to opt-out of

the new off-site tracking. Second, Facebook chose to ignore consumers’ explicit requests,

177
See Delo, supra note 161.
178
See C. J. Hoofnagle, J. M. Urban, and S. Li, Privacy and Modern Advertising: Most US Internet Users Want
'Do Not Track' to Stop Collection of Data about their Online Activities, AMSTERDAM PRIVACY CONF. (Oct. 2012);
Lymari Morales, U.S. Internet Users Ready to Limit Online Tracking for Ads, NEWS.GALLUP.COM (Dec. 21, 2010),
https://news.gallup.com/poll/145337/internet-users-ready-limit-online-tracking-ads.aspx; Kristen Purcell, Joanna
Brenner, & Lee Rainie, Search Engine Results 2012, PEW RESEARCH CENTER,
http://www.pewinternet.org/2012/03/09/search-engine-use-2012/; TRUSTE AND HARRIS INTERACTIVE, Privacy And
Online Behavioral Advertising, TRUSTE.COM (July 2012), http://truste.com/ad-privacy/TRUSTe-2011-
ConsumerBehavioral-Advertising-Survey-Results.pdf; J. Turow, J. King, C. J. Hoofnagle, A. Bleakley, & M.
Hennessy, Americans Reject Tailored Advertising and Three Activities That Enable It, UNIV. OF PENN. SCHOLARLY
COMMONS (Sept. 2009), https://repository.upenn.edu/cgi/viewcontent.cgi?article=1138&context=asc_papers.

Electronic copy available at: https://ssrn.com/abstract=3247362

enacted via the browsers’ Do No Track option, to not be tracked. Third, when consumers

installed ad blockers to circumvent tracking and targeted advertising, Facebook responded by

circumventing the users’ installed ad blockers.

First, Facebook did not give users the option to opt-out of Facebook’s tracking and

monitoring of their online behavior.179 Instead, Facebook informed users they could stop

Facebook from showing them ads based on this new surveillance data by opting out on the

Digital Advertising Alliance (or DAA) website. The DAA is an industry alliance formed in

response to FTC investigations into the industry’s privacy practices and reflects an industry

effort to police itself. The DAA’s stated mission is to give consumers the choice to opt out of

behaviorally targeted advertising.180 Facebook, Google and others are alliance members. On the

DAA’s website, the opt-out process was, conveniently, painstakingly inconvenient—the user had

to go through a multiple-step process for each Facebook account, browser, and device in each

household. One might recall the inconvenience consumers faced opting out of direct marketing

calls before adoption of the Do Not Call list. For a household of three, opting out required going

through the opt-out process about nine times—just for Facebook.181 If the consumer did go

through the DAA’s opt-out process, the DAA website often informed the consumer that the opt-

out requests “were not completed. This may be the result of a temporary technical issue.”182

Furthermore, the DAA’s opt-out solution only worked if a consumer set her browser security

settings to permit third-party cookies—the very mechanism that allows companies like Facebook

179
See Angwin, supra note 146.
180 See DIGITAL ADVERTISING ALLIANCE, https://digitaladvertisingalliance.org (last visited Jan. 21, 2019).
181
Assuming 3 configurations per person (for example, 2 browsers & 1 mobile device).
182
This has been my experience as a consumer from 2016 to today. My request to opt-out has resulted in an
error message stating that the DAA could not process my request. When I try to opt out of Facebook tracking, the
DAA presents the following message: “Opt-out requests for 1 participating companies were not completed. This
may be the result of a temporary technical issue. Select "Try Again" to request opt outs from those companies again.
Click "Understand Your Choices" for more information.” See DAA Webchoices Browser Check, DIGITAL
ADVERTISING ALLIANCE, http://optout.aboutads.info/?c=2&lang=EN (accessed Sept. 2, 2018).

Electronic copy available at: https://ssrn.com/abstract=3247362

to do what the consumer was now trying to avoid.183 Almost to the point of comedy, the DAA’s

website then informed users that if they cleared their cookies (to rid tracking cookies), doing so

would inadvertently have the effect of allowing Facebook to track them all over again.184 Even if

the consumer succeeding at opting out, he only opted out of being shown targeted advertising,

not of Facebook surveillance.

Second, Facebook also circumvented those users who had explicitly set their browser

privacy settings to Do Not Track. Since the DAA’s solution did not provide users with

meaningful choice, many users instead activated the Do Not Track settings in their web

browsers. The Do Not Track setting in browsers was another industry response to threatened

regulatory action. Back in 2010, the FTC toyed with the idea of creating a singular national “Do

Not Track” list for digital advertising companies, to parallel the Do Not Call list for

telemarketers. Such a Do Not Track list would have given consumers a “kill switch,” which

would turn off tracking across the horizontal market.185 In an effort to ward off regulation,

various industry players promised to give users Do Not Track opt-outs. Microsoft updated the

Internet Explorer browser to provide users with a Do Not Track setting. Other browsers—

Firefox, Safari, Opera, and Chrome—also adopted the Do Not Track protocol.186 With Safari, for

example, a user could go to Preferences Settings, toggle to the Privacy tab, then select the

checkbox “Ask websites not to track me.”187 The Do Not Track protocol though didn’t

183
See id.
184 Id.
185
See Kenneth Corbin, FTC Mulls Browser-Based Block for Online Ads, INTERNETNEWS.COM (July 28, 2010),
http://www.internetnews.com/ec-
news/article.php/3895496/FTC+Mulls+BrowserBased+Block+for+Online+Ads.htm.
186 TRACKING PROTECTION WORKING GROUP, https://www.w3.org/2011/tracking-protection/ (controls and issues the

protocol); Emil Protalinski, Everything You Need to Know About Do Not Track: Microsoft vs Google & Mozilla,
THE NEXT WEB (Nov. 25, 2012) https://thenextweb.com/apps/2012/11/25/everything-you-need-to-know-about-do-
not-track-currently-featuring-microsoft-vs-google-and-mozilla/ (listing the browsers that have adopted the protocol).
187 Safari Privacy Settings (Open Safari application; then open “Preferences” from the menu bar; then enter the

“Privacy” settings; then check the “Ask websites not to track me” box).

Electronic copy available at: https://ssrn.com/abstract=3247362

technically block companies from tracking users. Rather, a user’s browser would simply send a

message notifying companies that he or she does not wish to be tracked.188 It was simply a polite

request, which a company could choose whether or not to heed.

In another demonstration of market power, Facebook would ignore users’ activation of

Do No Track.189 In 2013, Erin Egan, the chief privacy officer of Facebook, explained that

Facebook would bypass consumer Do Not Track settings because Facebook does not track

consumers for advertising purposes, in effect arguing that consumers do not understand what Do

Not Track means. 190 “We don’t use that data for an advertising purpose,” she emphasized. In

2014, after Facebook changed course and began tracking consumers for commercial purposes,

Facebook simply continued to ignore consumers’ Do Not Track signals.

Sensing rising consumer frustration, the private market responded with software that

consumers could use to both stop surveillance and block targeted ads from loading on pages

altogether.191 Ad blockers prevented not only the visual display of advertising, but also third-

party tracking. Generally, this worked by blocking the user’s device from making third-party

initiated HTTP requests with advertising companies. In the Fall of 2014, after Facebook’s new

tracking announcements, online searches for “how to block ads” spiked at an unprecedented

188 Id.
189
See generally Liam Tung, Google, Facebook 'Do Not Track' Requests? FCC Says They Can Keep Ignoring
Them, ZDNET.COM (Nov. 9, 2015), https://www.zdnet.com/article/google-facebook-do-not-track-requests-fcc-says-
they-can-keep-ignoring-them/.
190
Elise Ackerman, Google And Facebook Ignore "Do Not Track" Requests, Claim They Confuse Consumers,
FORBES MAGAZINE (Feb. 27, 2013), https://www.forbes.com/sites/eliseackerman/2013/02/27/big-internet-
companies-struggle-over-proper-response-to-consumers-do-not-track-requests/#a99039822fc.1.
191
Common ad blockers work by communications blocking, where the client’s request to an ad servers or ad
company is prevented from occurring. They can also work by element hiding, where HTML elements are loaded
onto the client’s page but hidden from the user (for example, hide elements with class = “Ad”). Disconnect and
Privacy Badger are two products that include anti-trackers with their ad blockers, but Privacy Badger was the only
company that could block tracking conducted by Facebook Like buttons.

Electronic copy available at: https://ssrn.com/abstract=3247362

rate.192 Shortly after, Pew Research Center reported that 91% of Americans felt they had lost

control over the way their personal data is collected and used.193 A Forrester Research report

showed that 19% of consumers had taken steps to activate the Do Not Track feature in their

browsers—even though Do Not Track had no teeth.194 The marketing materials of AdBlock, the

#1 ad blocking software on the market, touted “Privacy is Paramount.”195 However, publishers

whose livelihoods depended on advertising started to panic at the prospect of mass consumer

adoption of ad blockers. AdBlock, and another anti-tracking software company called

Disconnect, joined hands with privacy advocacy group Electronic Frontier Foundation, in an

attempt to broker an agreement with publishers—if publishers would only respect users’

activation of Do Not Track, they would drop their fences and stop blocking their ads.196

The use of ad blocking software by consumers rose in tandem with the market’s ability to

extract the rent of commercial surveillance. A 2015 joint study between Adobe and PageFair

showed that there were 198 million people actively blocking ads, with a U.S. year-over-year

growth rate of 48%.197 In August of 2015, Apple announced that its new mobile operating

system, IOS 9, would permit developers to introduce apps that enabled content blocking into the

app store. When IOS 9 released in September, the top app downloads were immediately for ad

192
Google Trends, GOOGLE.COM (2013-2014), https://trends.google.com/trends/explore?date=all&q=adblock;
see generally PageFair, The Rise of Adblocking, PAGEFAIR.COM (2014),
https://downloads.pagefair.com/downloads/2016/05/The-Rise-of-Adblocking.pdf.
193
See Mary Madden, Public Perceptions of Privacy and Security in the Post-Snowden Era, PEW RES. CTR.
(Nov. 12, 2014), http://www.pewinternet.org/2014/11/12/public-privacy-perceptions/.
194
Privacy Is Far From Dead: Introducing Contextual Privacy, FORRESTER RESEARCH (Dec. 19, 2013),
https://www.forrester.com/Privacy+Is+Far+From+Dead+Introducing+Contextual+Privacy/-/E-PRE6624.
195
See ADBLOCK, www.adblock.com.
196
Press Release, Electronic Frontier Found., Coalition Announces New ‘Do Not Track’ Standard for Web
Browsing (Aug. 3, 2015),
https://www.eff.org/press/releases/coalition-announces-new-do-not-track-standard-web-browsing
197
PAGEFAIR, The 2015 Ad Blocking Report, PAGEFAIR.COM (Aug. 10, 2015),
https://pagefair.com/blog/2015/ad-blocking-report/.

Electronic copy available at: https://ssrn.com/abstract=3247362

blockers.198 By 2016, reports showed that one in five smartphone users—or 420 million people

worldwide—were blocking ads when browsing on the mobile web.199 By 2017, a report from the

advertising research company eMarketer estimated that one-quarter of U.S. internet users were

blocking ads one way or another.200 In tandem, independent studies conducted by the Pew

Research Center, and the Annenberg Center, continued to show that Americans were

overwhelmingly opposed to being shown ads targeted to them based on information derived

from surveillance.201 Was this the largest boycott in human history?202

Facebook raced to engineer a way to circumvent users’ installation of ad blockers.

Initially, Facebook prevented its public-facing pages from loading on user devices that had ad

blockers installed. If consumers landed on forbes.com and Forbes prevented its page from

loading, consumers could switch to a Forbes competitor to read news. With Facebook,

consumers did not have any alternative product they could switch to. Then, in August of 2016,

Facebook announced it had found a way to circumvent ad blockers entirely.203 Facebook “flipped

a switch on its desktop website that essentially renders all ad blockers … useless.”204 This time,

198
Sarah Perez, A Day After iOS 9’s Launch, Ad Blockers Top the App Store, TECHCRUNCH.COM (Sept. 17,
2015), https://techcrunch.com/2015/09/17/a-day-after-ios-9s-launch-ad-blockers-top-the-app-store/.
199
PAGEFAIR, Mobile Adblocking Report, PAGEFAIR.COM (2016), https://pagefair.com/blog/2016/mobile-
adblocking-report/; and Mark Scott, Rise of Ad Blocking Software Threatens Online Revenue, N.Y. TIMES (May 30,
2016), https://www.nytimes.com/2016/05/31/business/international/smartphone-ad-blocking-software-mobile.html.
200
Facing Up to Ad Blocking: How Publishers, Advertisers and their Digital Media Partners are Responding,
EMARKETER.COM (June 21, 2017), https://www.emarketer.com/Report/Facing-Up-Ad-Blocking-How-Publishers-
Advertisers-Their-Digital-Media-Partners-Responding/2002077.
201
See Mary Madden & Lee Raine, Americans’ Attitudes About Privacy, Security and Surveillance, PEW RES.
CTR. (May 2, 2015), http://www.pewinternet.org/2015/05/20/americans-attitudes-about-privacy-security-and-
surveillance/ (study revealed that 93% of American adults believe that having control over who gets their
information is important; 88% said it is important that they are not watched or eavesdropped on without their
permission ; 84% of respondents wanted control over what online marketers knew about them); Joseph Turow,
Michael Hennessy & Nora Draper, The Tradeoff Fallacy, NEW ANNENBERG SURVEY RESULTS (2015),
https://www.asc.upenn.edu/sites/default/files/TradeoffFallacy_1.pdf.
202 Beyond Ad Blocking – The Biggest Boycott in Human History, DOC SEARLS BLOG: BLOGS.HARVARD.EDU (Sept.

28, 2015), https://blogs.harvard.edu/doc/2015/09/28/beyond-ad-blocking-the-biggest-boycott-in-human-history/.

203
Facebook only had to block ads on desktop, because Facebook had already found way to serve ads in mobile
apps that could not be touched by ad blockers.
204
Mike Isaac, Facebook Blocks Ad Blockers, but It Strives to Make Ads More Relevant, N.Y. TIMES (Aug. 9,
2016), https://www.nytimes.com/2016/08/10/technology/facebook-ad-blockers.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

Rothenberg proclaimed, “Facebook should be applauded for its leadership on preserving a

vibrant exchange with its users.”205 Here too, Facebook’s dominance played a role in its ability to

devise a method to circumvent blockers.206 Before long, Wehner was sharing on earnings calls

the ad revenue growth Facebook was able to sustain by evading ad blockers. For example, on

Facebook’s Q3 2016 earnings call, Wehner pointed out that half of the 18% year-over-year

revenue growth in desktop ads was “largely due to our efforts on reducing the impact of ad

blocking.”207 From Q3 2016 to the end of Q2 2017, Facebook was able to make $709 million

dollars circumventing ad blockers.208

With rapid consumer adoption of ad blockers and Facebook’s unique ability to force onto

consumers the presence of behaviorally targeted ads, Facebook approached publishers with a

proposition. Facebook offered publishers the ability to publish content not on their own websites,

but inside the walls of the impenetrable Facebook, where Facebook could ensure the delivery of

behaviorally targeted advertising that commanded higher rates in ad markets.209 Despite

normally being competitors in the ad market, participating publishers like The New York Times

worked in tandem with Facebook to sell the advertising. If a participating publisher sells the

advertising, it keeps 100% of the ad revenue.210 If Facebook sells the advertising for The New

York Times, Facebook retains a 30% cut. Facebook had the power to force invasive advertising

on consumers through a capability that other publishers like The New York Times did not

205
See id.
206
See Casey Johnston, Why Facebook is Really Blocking the Ad Blockers, THE NEW YORKER (Aug. 12, 2016),
https://www.newyorker.com/business/currency/why-facebook-is-really-blocking-the-ad-blockers.
207
Facebook Inc. Q3 2016 Earnings Conference Call, NASDAQ.COM (Nov. 2, 2016),
https://www.nasdaq.com/aspx/call-transcript.aspx?StoryId=4018524&Title=facebook-fb-q3-2016-results-earnings-
call-transcript.
208
Johnny Ryan, Facebook’s Hackproof Ads Turned its Adblocking Problem in to a $709 Million Revenue
Stream, PAGEFAIR.COM (Nov. 2, 2017), https://pagefair.com/blog/2017/facebook-adblock-audience/.
209
This is Facebook’s Instant Articles program. See Instant Articles (July 2018), https://instantarticles.fb.com/.
210
Facebook does not receive zero consideration in return. Facebook can monitor and measure the behavior of
readers of Instant Articles and monetize this data in various ways.

Electronic copy available at: https://ssrn.com/abstract=3247362

have.211 Facebook could successfully fight against users’ preference for privacy—users had to

submit to the terms of trade imposed upon them by this century’s new communications network.

III. INDIRECT EVIDENCE CONFIRMS FACEBOOK’S MONOPOLY POWER

Facebook entered a competitive social media market and disintermediated competition by

offering superior consumer privacy protections. Subsequently, Facebook tried to undermine

privacy to initiate consumer surveillance for the purpose of delivering more targeted advertising,

but the competitive market would not allow it. Only after other social networks like MySpace

and Orkut exited the market, and Facebook amassed over a billion users, was Facebook able to

reverse course, and initiate consumer surveillance. The fact-pattern demonstrates an inelasticity

of demand for Facebook’s product, and in turn, constitutes a direct showing of monopoly power

under Section 2 of the Sherman Act. Facebook’s ability today to extract surveillance in its

exchange with consumers merely reflects an ability to extract monopoly rents from consumers

that contradicts their own welfare. A review of indirect, or circumstantial, evidence of market

power—to which I now turn—further explains Facebook’s exhibited monopoly power over

product quality. Facebook controls over 80% of the social network market and Facebook’s

control is protected by strong entry barriers which dissuade new-entrants.

Evaluating a market’s structure by estimating a particular firm’s percentage share of a

market can be a useful mechanism for predicting a firm’s market power, but should not be

necessary to prove Facebook’s monopoly power.212 The goal of antitrust law is to protect

211
See Johnston, supra note 206.
212
See generally AREEDA & HOVENKAMP, §652, supra note 26 (articulating that Section 2 of the Sherman Act
prohibits monopoly, which primarily refers to monopoly conduct, not monopoly shares, and going so far as to say,
“Nothing in the language of the Sherman Act limits its conception of monopoly to large market share.”) Note that in
Amex, the Supreme Court did recently state that defining the relevant market is necessary despite direct evidence in
cases alleging improper vertical restraints. Ohio v. Am. Express Co., 138 S. Ct. 2274 (2018).

Electronic copy available at: https://ssrn.com/abstract=3247362

consumers from the harms associated with a lack of competition—namely, increased prices,

decreased quality, lower output, and less innovation. Often, direct proof of a company’s ability to

act like a monopolist is not available. In the absence thereof, circumstantial evidence of the

market’s structure can indicate whether a particular firm has monopoly power. Under this market

share-market definition approach to analyzing power, the exercise becomes one of calculating

Facebook’s share of the relevant market. If one shows that Facebook controls a large percentage

of a market that is protected by entry barriers, then this assertion serves as a proxy for direct

proof of Facebook’s ability to set price or define quality at levels that deviate from the

competitive norm.213 In Facebook’s case, direct evidence of monopoly power exists, and this fact

should obviate the need for isolating Facebook’s share of a relevant market. Nonetheless, some

scholars have argued that framing a defendant’s power in terms of market structure may still be

necessary to satisfy Section 2 statutory requirements.214 In this respect, pleading a dominant

share of a relevant market may serve a purpose outside the scope of proving power.

Under the market share-market definition framework, the first step is to isolate the

relevant market from which firm’s market share is deduced. The relevant market is that in which

“significant substitution in consumption” occurs,215 and should only include other products that

consumers can turn to if one firm increases price or decreases quality.216 In Facebook’s case, the

appropriate contours of the relevant market should only include other social networks that

consumes use interchangeably. The purpose of defining the relevant market is to identify the

swath of other products that can restrain a company’s ability to extract monopoly rents.

213 AREEDA & HOVENKAMP, §652, supra note 26.

214
See generally AREEDA & HOVENKAMP, §531, supra note 26 (also noting that identifying a market’s structure
and defendant’s share may be useful to distinguish between monopoly rents extracted by a single monopolist versus
coordinating oligopolists”).
215
See generally AREEDA & HOVENKAMP, §5.02 supra note 26; see also United States v. Grinnell Corp., 384
U. S. 563, 571 (a market is restricted to reasonably “interchangeable services”).
216
See generally Grinnell Corp., 384 U. S. at 571.

Electronic copy available at: https://ssrn.com/abstract=3247362

Empirically, an earlier competitive social network market appeared to do just this.217

Competition from Murdoch’s MySpace or Google’s Orkut restrained how boldly Facebook (and

others) could deteriorate quality given a constant price in a way that competition from instant

messaging or email services today do not. In other words, Facebook’s ability to deteriorate

quality below competitive levels only after the exit of other social networks lends credence to the

position that the relevant market should be limited to social networks.

Furthermore, the relevant market should be constrained to social networks because social

networking is a new, unique, form of communications, unrivaled in its ability to distribute and

amplify a consumer’s communication. Whereas the telephone in the 20th century unleashed a

watershed of one-to-one voice communications, the social network has opened the floodgates on

one-to-many communications, distributed instantly from one person to others in a person’s

“social graph.”218 From the consumer’s perspective, Facebook is merely a tool for digital

communications.219 The Facebook social media network enables instant one-to-many

communication via text, audio, image, or video, exponentially decreasing the consumer’s

transactional costs that are associated with such widely distributed communications. Initially,

Facebook marketed its social media platform as “a social utility.”220 By simply hitting the

singular “post” button, a mother might share wedding pictures with hundreds, if not thousands,

of close and extended family members, long-lost college friends, and acquaintances worldwide.

217
See generally infra Part II.
218
The term “social graph”, initially popularized by Facebook itself and first referred to at the 2007 Facebook
F8 conference, refers to a person’s web of connections, and is the underlying business asset that a social network
uses to distribute one’s communication to one’s contacts. Facebook Unveils Platform for Developers of Social
Applications, NEWSROOM.FACEBOOK.COM (May 24, 2007), https://newsroom.fb.com/news/2007/05/facebook-
unveils-platform-for-developers-of-social-applications/.
219
On recent advertisements (as seen on nytimes.com medium rectangle ads on June 30, 2018), for example,
Facebook marketed itself as a tool to “connect with friends and family.” Also consider generally the marketing
language on Facebook’s website. Bringing the World Closer Together, FACEBOOK.COM (July 20, 2018),
https://www.facebook.com/pg/facebook/about/?ref=page_internal.
220
Facebook Homepage, FACEBOOK.COM (Nov. 29, 2007), http://www.facebook.com/
[http://web.archive.org/web/20071129130140/http://www.facebook.com/].

Electronic copy available at: https://ssrn.com/abstract=3247362

In this nation’s earlier historic antitrust action against AT&T in the telephone market, the

court defined the relevant market simply as “telecommunications,” because the public interest

was so served since telecommunications had come to play a dominant role in “modern economic,

social, and political life.”221 Social media has come to dominate the American way of life in

much the same way that the telephone did for earlier generations. The fabric of American

politics, the roll-out of new consumer products, and the dissemination of news all unfold on

social media. The sitting U.S. President’s preferred way of communicating directly with the

American people is through Twitter, one of the handful of social networks remaining today.222

Two-thirds of Americans now receive news via social networks, with the majority receiving

news via Facebook.223 In some countries, the sheer amount of time citizens spends on social

networks has become a matter of public health concern. Earlier this year, to address the time

spent on social media, and the potentially addictive nature of such platforms, France passed

legislation banning cell phones entirely from school grounds.224

In the United States, and across the world, Facebook in particular dominates the social

network market. In the U.S. alone, 210 million consumers have a Facebook account,225 with

roughly three-fourths of those consumers using the platform at least once per day.226 To put this

in perspective, it was not until some 100 years after the invention of the telephone that the

221
United States v. American Tel. and Tel. Co., 552 F. Supp. 131, 165 (D.D.C. 1983).
222 Tamara Keith, Commander-In-Tweet: Trump’s Social Media Use and Presidential Media Avoidance,
NPR.ORG (Nov. 18, 2016), https://www.npr.org/2016/11/18/502306687/commander-in-tweet-trumps-social-media-
use-and-presidential-media-avoidance.
223
Elisa Shearer and Jeffrey Gottfried, News Use Across Social Media Platforms 2017, PEW RES, CTR. (Sept. 7,
2017), http://www.journalism.org/2017/09/07/news-use-across-social-media-platforms-2017/.
224
Sam Schechner, France Takes on Cellphone Addiction with Ban in Schools, WALL STREET J. (Aug. 13,
2018), https://www.wsj.com/articles/france-takes-on-cellphone-addiction-with-a-ban-in-schools-1534152600.
225
The Facebook ads interface claims to reach 210 million consumers in the U.S. ages 13 and over. Facebook
Ads, FACEBOOK.COM (July 20, 2018), https://www.facebook.com/business/products/ads.
226
See Social Media Fact Sheet, PEW RES. CTR. (Feb. 5, 2018), http://www.pewinternet.org/fact-sheet/social-
media/.

Electronic copy available at: https://ssrn.com/abstract=3247362

telephone rivaled Facebook’s penetration into the lives of American households.227 But Facebook

does not only cut wide, it also cuts deep. Americans currently spend over 40 minutes per day on

just Facebook, and over an hour per day on Facebook owned-and-operated platforms, like

Instagram.228 U.S. consumers spend approximately 150 million hours per day on Facebook.229

Facebook’s dominance also crosses international borders. Across the world, around one in every

four humans has a Facebook account. This means that of the population that has access to the

internet, nearly one in every two persons that could have a Facebook account does.230

Within the realm of social networks, only platforms that consumers use interchangeably,

and that bear on Facebook’s demand elasticity, should be considered part of the relevant market

for antitrust analysis. Here, particularly problematic for Facebook is the fact that of the U.S.

adult population that uses any form of social media, nearly 99% use Facebook.231 Subsets of

consumers use Facebook in addition to one or more additional social networking platforms. This

fact alone suggests that consumers do not find other social networking platforms to be adequate

substitutes. Perhaps this is why when Senator Lindsey Graham (R-SC) asked Zuckerberg to

name a product consumers could use instead of Facebook, Zuckerberg was unable to.232

The fact that consumers find it indispensable to use Facebook simply reflects the fact that

speech on Facebook is more powerful than other methods of personal speech. As the only social

227 U.S. Census Bureau, Statistical Abstract of the United States: 1999 885 (1999),

www2.census.gov/library/publications/1999/compendia/statab/119ed/tables/sec31.pdf.
228 EMARKETER.COM, supra note 71.
229 Id.
230 BROADBAND COMMISSION FOR SUSTAINABLE DEVELOPMENT, The State of Broadband 2017: Broadband

Catalyzing Sustainable Development Report (Sept. 2017), https://www.itu.int/dms_pub/itu-s/opb/pol/S-POL-

BROADBAND.18-2017-PDF-E.pdf (estimating around 48% of global population had access to the internet at the
end of 2017).
231 See PEW RES. CTR., supra note 226 (69% of U.S. adults use social media; 68% of U.S. adults use Facebook).
232
Unable to name a direct competitor, Zuckerberg explained that Google, Apple, Amazon, Microsoft, Twitter
and other services generally overlap with Facebook in different ways. See Facebook Congressional Hearing Before
the Committees on the Judiciary and Commerce, Science and Transportation, 115th Cong. (April 2018). For
transcript of hearing see Transcript of Zuckerberg’s Senate Hearing, WASH. POST (April 10, 2018),
https://www.washingtonpost.com/news/the-switch/wp/2018/04/10/transcript-of-mark-zuckerbergs-senate-
hearing/?noredirect=on&utm_term=.17ef62e494e8.

Electronic copy available at: https://ssrn.com/abstract=3247362

network with a user-base of over 2 billion, Facebook connects its users to the largest number of

people, and can distribute their speech more broadly than users could otherwise. On a recent

earnings call, Sheryl Sandberg, chief operating officer of Facebook, touted Facebook as the place

to “reach everyone [in] almost every country in the world.”233 One may not always want to speak

loudly, but when one does, or one is selling a product, selling oneself for political office, or

distributing news, Facebook distribution becomes a necessity.

Only Facebook has a user-base of over 2 billion. Competitive social networks with much

smaller user-bases cannot directly compete with the built-in utility of Facebook’s product.234

This explains why competitors instead narrowly focus on carving out a sub-niche in the social

network market—short tweets, disappearing messages, a social network for professionals. For

example, the second largest social network, Instagram, has 1 billion users globally, about 88

million in the U.S., and allows people to communicate almost entirely through visuals.235

Facebook, however, owns Instagram. LinkedIn is in third place with about 500 million users,

with 133 million in the U.S., but focuses specifically on being a “professional network,” enabling

digital communication for “economic opportunity.”236 SnapChat has 191 million daily-active

users worldwide, with approximately 72 million in the U.S., but distinguishes itself with short,

disappearing pictures or videos.237 Consider Twitter—with some 336 million monthly-active

233
Sheryl Sandberg, Facebook Inc. Q4 2017 Earnings Conference Call, NASDAQ.COM (Jan. 31, 2018),
https://www.nasdaq.com/aspx/call-transcript.aspx?StoryId=4141984&Title=facebook-s-fb-ceo-mark-zuckerberg-
on-q4-2017-results-earnings-call-transcript.
234
Since social networks today are closed communications protocols, each platform’s utility is a function of the
number of users found on a particular platform.
235
See Facebook Q1 2018 Quarterly Earnings Report, FACEBOOK INVESTOR RELATIONS (April 25, 2018),
https://investor.snap.com/~/media/Files/S/Snap-IR/reports-and-presentations/1q-18-10q.pdf; Number of Instagram
users in the United States from 2015 to 2021, STATISTA.COM (Aug. 2017),
https://www.statista.com/statistics/293771/number-of-us-instagram-users/.
236
About LinkedIn, LINKEDIN.COM (July 20, 2018), https://about.linkedin.com/.
237
See Snap Inc. Q1 2018 Quarterly Earnings Report, SNAP INVESTOR RELATIONS (May 2, 2018),
https://otp.tools.investis.com/clients/us/snap_inc/SEC/sec-
show.aspx?Type=html&FilingId=12721842&Cik=0001564408 (reporting 191 monthly Daily Active Users (DAUs)
and 81 million DAUs in North America). SnapChat does not provide a breakdown of U.S. only users. But

Electronic copy available at: https://ssrn.com/abstract=3247362

users, with 68 million in the U.S., which constrains user text communication to no more than 280

characters.238 For antitrust purposes, competing social networks—Instagram, LinkedIn,

SnapChat, and Twitter—are not adequate substitutes because their user-bases are much smaller,

and they serve a more narrow purpose.239

With regards to instant messaging services such as WhatsApp and video sharing services

such as YouTube, consumers use these services in inherently different ways than they use social

media platforms. Thus, they too are not substitutes. Sandberg recently explained the difference

between Facebook and instant messaging platforms Messenger and WhatsApp to investors:

social networks such as Facebook are about one-to-many communications, Messenger and

WhatsApp are mainly about one-to-one communications.240 Even so, Facebook also owns

Messenger and WhatsApp. In response to an abuse of power investigation by Germany’s cartel

office, Facebook complained that the office’s report “paints an inaccurate picture” of dominance

because the report did not include Twitter, SnapChat and YouTube.241 But for the consumer,

YouTube is more interchangeable with television or subscription-video on demand (Amazon,

HBO, etc.) than it is with Facebook. In a reflection of this, YouTube markets itself to advertisers

as a substitute for television network spend.242

Facebook’s U.S. reach is about 90% of Facebook’s North America reach (210 million in the U.S. vs. 23 million in
Canada). Assuming that SnapChat’s U.S.-North America ratio is similar, we can estimate that SnapChat’s U.S.
DAU number is approximately 72.3 million. Statistica estimates that SnapChat’s reach in the U.S. is approximately
77 million (not a DAU number). Number of Snapchat Users in the United States from 2015 to 2021 (in millions),
STATISTA.COM (Feb. 2017), https://www.statista.com/statistics/558227/number-of-snapchat-users-usa/.
238
Twitter Q1 2018 Quarterly Report, TWITTER INVESTOR RELATIONS (April 25, 2018),
https://investor.twitterinc.com/results.cfm. For a break-down of U.S. monthly active users; Q2 2017 Letter to
Shareholders, TWITTER.COM (July 27, 2017), http://files.shareholder.com/downloads/AMDA-
2F526X/4882764939x0x951006/4D8EE364-9CC3-4386-A872-ACCD9C5034CF/Q217_Shareholder_Letter.pdf.
239 15 U.S.C. § 2 (2000).
240
See Sandberg, supra note 233.
241
See Sam Schechner, Germany Says Facebook Abuses Market Dominance to Collect Data, WALL STREET J.
(Dec. 20, 2017), https://www.wsj.com/articles/facebook-abuses-its-dominance-to-harvest-your-data-says-german-
antitrust-enforcer-1513680355?.
242
YouTube’s marketing literature for advertisers focuses on this comparison, advertising the fact that “In an
average month, 18+ year-olds in the United States spend more time watching YouTube than any television

Electronic copy available at: https://ssrn.com/abstract=3247362

Facebook will not be successful in arguing that it is a two-sided platform and that the

boundaries of the relevant market definition should account not only for Facebook’s share of the

social network consumer market, but also its share of the digital advertising market. According

to the Supreme Court’s opinion in Amex, defendant American Express, as a credit card company,

operated a two-sided “transaction” platform, so plaintiffs needed to prove defendant’s conduct

was cumulatively anticompetitive in both markets.243 The district court had erred in treating the

credit-card market as two separate markets—one for merchants and one for cardholders—and

then only evaluating the anticompetitive effects to the merchant side of the market.244 The Court

in Amex, however, pigeonholed its holding to two-sided “transaction” platforms, where a

company merely facilitates a simultaneous transaction between two parties. A transaction

platform displays strong bi-lateral indirect network effects between the two distinct markets in

which a defendant operates. With American Express, this appears true: more Amex cardholders

always increases the value proposition for Amex merchants; and, more Amex merchants always

increases the value proposition for Amex cardholders. Because of this, the relevant market has to

include both sides. The Court pointed out a type of two-sided platform that might fail this rule—

companies that deliver content to consumers, but advertising to marketers. Newspapers, it

explained, fail the strong indirect networks effects test: more consumers are always good for

advertisers, but more advertisers are not always good for consumers.245 Two-sided platforms that

network.” Case Study: The Evolution of Digital Video Viewership, NIELSEN.COM (Sept. 25, 2015),
http://www.nielsen.com//content/dam/corporate/us/en/reports-downloads/2015-reports/nielsen-google-case-study-
sept-2015.pdf.
243
Ohio v. Am. Express Co., 138 S. Ct. 2274 at 2280 (2018).
244
Id.
245
See Am. Express Co. at 2281-82 (“A market should be treated as one sided when the impacts of indirect
network effects and relative pricing in that market are minor. … But in the newspaper-advertisement market, the
indirect networks effects operate in only one direction; newspaper readers are largely indifferent to the amount of
advertising that a newspaper contains. … Because of these weak indirect network effects, the market for newspaper
advertising behaves much like a one-sided market and should be analyzed as such.) (citing Lapo Filistrucchi,
Damien Geradin, Eric van Damme, & Pauline Affeldt, Market Definition in Two-Sided Markets: Theory and
Practice 1, 5 (Tilburg L. Sch. Legal Studs. Res. Paper Ser. No. 09/2013), available at

Electronic copy available at: https://ssrn.com/abstract=3247362

serve consumers on one side, but advertisers on the other, exhibit weak indirect network effects

and should therefore be evaluated as one-sided markets for antitrust purposes. The Court’s earlier

precedent in the antitrust case against the Times-Piccayune newspaper in New Orleans supports

this view. In Times-Piccayune, the Court considered a challenge to a newspaper’s advertising

policy and held that the relevant market includes only the newspaper’s market share in

advertising, not its market share in consumer readership.246

Under the second step of the market share—market definition analysis, market share

should be derived from Facebook’s quantitative share of consumer time on social networks.

Share of consumer time on social networks is the relevant measure of market share for two

distinct reasons. First, Facebook does not charge users a fee, some competitors do, and a measure

must be able to account for all competitors in the market.247 More importantly, consumer time is

the barometer which Facebook and other social networks use in the market—on the ground so to

speak—to signal their relative dominance. On this point, the Supreme Court has indicated that

one should calculate market share in the way that market participants think and talk about it. For

example, in Ohio v. American Express Co., the Supreme Court calculated the market share of

credit card companies based on transaction volume, not transaction revenues—to reflect how

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2240850 and Times-Picayune Publishing Co. v.

United States, 345 U. S. 594, 610 (1953)).

246
Times-Picayune Publishing Co., 345 U. S. at 610.
247
While Instagram, SnapChat and Twitter do not charge users a price, competitor Vero Social, which surged
to the #1 downloaded app on the Apple app store early 2017 following news of Facebook privacy breaches, offers
an ad-free, subscription model. See Todd Spangler, Vero Hits No. 1 on Apple’s App Store, But Can Subscription
Social Network Sustain the Hype? VARIETY (Feb. 27, 2018), https://variety.com/2018/digital/news/vero-social-app-
store-ad-free-social-hype-1202711654/. Additionally, Zuckerberg and Sandberg have made comments to U.S.
Congress and NBC news respectively that suggest Facebook itself may introduce an ad-free, subscription version of
Facebook. See generally, Callum Borchers, Would You Pay $18.75 for Ad-Free Facebook? WASH. POST (April 14,
2018), https://www.washingtonpost.com/news/the-fix/wp/2018/04/14/would-you-pay-18-75-for-ad-free-
facebook/?utm_term=.833d67a1fefc.

Electronic copy available at: https://ssrn.com/abstract=3247362

market participants themselves signal their relative dominance.248 Social networks, including

Facebook, Instagram, Twitter, and SnapChat speak of their relative strength in the social

network market by referring specifically to the number of users and the number of minutes spent

by users on their platform.249

However, even if one were to consider Instagram, Twitter, and SnapChat as part of the

relevant market for antitrust analysis, Facebook still dominates market share. Including time

spent on these other platforms, approximately 83% of the consumers’ time goes to Facebook and

Instagram.250 Consumers spend about 66% of time across all on just Facebook, 17.5% of time on

Instagram, 16% on SnapChat, and 0.5% of time on Twitter. Anecdotally, Americans spend 19%

of time spent across all mobile applications on just the Facebook mobile app.251

Facebook’s control of over 80% of consumer time reflects Facebook’s monopolistic

power in the social network market. While there is no bright-line rule regarding market shares,

courts have assumed monopoly power when market shares exceed over 70-plus percent.252 For

example, before the dissolution of AT&T for monopoly power in the interexchange market,

AT&T’s share of interexchange revenue was about 77%.253 Standard Oil, found to have illegally

monopolized the oil market, controlled about 90% of refinery output.254

248
Ohio v. Am. Express Co., 138 S. Ct. 2274 (2018) (American Express was found to only control 26.4% of
the credit card market because it controlled 26.4% of transaction volume).
249
Consider the fact that social network advertising media kits focuses on the number of users and number of
minutes spent on their social networks.
250
I multiplied most-recent U.S. user numbers of each platform (Facebook 210 million, Instagram 88 million,
SnapChat 77 million, Twitter 68 million) by the number of average minutes spent by users per day on each
(Facebook 41 minutes, Instagram 26, SnapChat 27, Twitter 1) to find a total market size of 13,045,000,000 minutes
per month. See EMARKETER.COM, supra note 71.
251
Simon Khalaf, U.S. Consumers Time-Spent on Mobile Crosses 5 Hours a Day, FLURRY ANALYTICS BLOG
(March 2, 2017), http://flurrymobile.tumblr.com/post/157921590345/us-consumers-time-spent-on-mobile-crosses-5.
252
See United States v. Grinnell Corp., 384 U. S. 563, 571 (1966) (market share exceeding 87% is
predominant); Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451, 481 (1992) (market share
exceeding 80%); United States v. E. I. du Pont de Nemours & Co., 351 U.S. 379, at 391 (1956) (market share
exceeding 75%).
253
United States v. Am. Tel. and Tel. Co., 552 F. Supp. 131 (D.D.C. 1983).
254
Standard Oil Co. of New Jersey v. United States, 221 U.S. 1 (1911).

Electronic copy available at: https://ssrn.com/abstract=3247362

Notwithstanding Facebook’s dominant share of the relevant market, market share

analysis also considers whether firms are protected by barriers to entry.255 Without entry barriers,

new entrants have the opportunity to quickly disintermediate even the most dominant firms. The

mere threat of entry can prevent monopolists from being able to profitably extract monopoly

rents from consumers. But, there are entry barriers associated with Facebook’s closed

communications protocol and over 2 billion users. When a communications network is closed, a

user can only communicate with another user of the same network. This creates a powerful

phenomenon known as direct network effects.256 Just as the utility of owning a phone in the late

19th century grew as phones became more accessible, so does the utility of a closed social

network depend on how many other people use it. A new market entrant cannot easily get users

to switch to a platform with less users. Additionally, the lack of perfect substitutes for Facebook

has allowed the company to have proprietary control over one’s social graph. Facebook’s users

are often connected exclusively through Facebook’s network, may lose cell phone numbers for

older contacts, and Facebook becomes the primary method for users to remain in contact with

one another. Because the communications protocol is closed, and Facebook controls one’s social

graph, consumers face high switching costs and competitors face a significant barrier to entering

the market.

255
United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001).
256
A product has direct network effects if one’s use of a product increases the product’s utility for others. For
example, in a market with only two competing social networks, assuming constant price and quality, the network
with the larger user-base has greater value because it allows new users to communicate with more people. Other
businesses with strong network effects include those of Uber, Wikipedia, and Airbnb. In such markets, the company
that achieves early gains begins immediately to benefit from the superiority of one’s product due to network effects.
The concept of network effects was first identified in the context of the long-distance telephony market in the 1970s.
See Jeffrey Rohlfs, A Theory of Interdependent Demand For A Communications Service, THE BELL J. OF ECON. AND
MANAGEMENT SCI, 16–37 (1974). Subsequent papers on this topic include W. BRIAN ARTHUR, ON COMPETING
TECHNOLOGIES AND HISTORICAL SMALL EVENTS: THE DYNAMICS OF CHOICE UNDER INCREASING RETURNS (1983);
Joseph Farrell & Garth Saloner, Standardization, Compatibility, and Innovation, THE RAND J. OF ECON. 70–83
(1985); Michael L Katz & Carl Shapiro, Network Externalities, Competition, and Compatibility, 75 THE AM. ECO.
REV. 424 (1985). During the first dot-com rush, venture capitalists and entrepreneurs in Silicon Valley were keenly
in-tune and in-favor of investing in companies that benefited from direct network effects, or “first-mover
advantage.” See generally Eric Ransdell, Network Effects, FASTCOMPANY.COM (Aug. 31, 1999),
https://www.fastcompany.com/37621/network-effects.

Electronic copy available at: https://ssrn.com/abstract=3247362

IV. FACEBOOK’S PATTERN OF CONDUCT RAISES CONCERNS ABOUT ILLEGAL

MONOPOLIZATION

A remaining question is whether Facebook’s pattern of behavior reaches the level of

anticompetitive conduct with which antitrust law concerns itself. Parts I and II of this Paper

traced the particular history of Facebook’s conduct with respect to the quality of its product. The

qualitative aspects of its product, particularly the level of privacy, is not abstract—Facebook’s

product is free, privacy was a critical form of competition in a functioning market, citizens favor

no surveillance, and extracted surveillance is strongly tied to current Facebook revenues and

profits. Part II argued that Facebook’s broad-scale surveillance of American consumers today

reflects monopoly rents, a conclusion also supported by indirect evidence of Facebook’s

monopoly power considered in Part III. Part IV now contends that Facebook’s conduct, taken

collectively, raises serious issues of anticompetitive practices. Facebook’s conduct engendered

trust in consumers, but the record suggests that Facebook’s commitment to user privacy was

disingenuous. Facebook’s course of misleading conduct resulted in precisely the type of harm

that antitrust law concerns itself with—the exit of rivals and the subsequent extraction of

monopoly rents in contravention to consumer welfare.

A. Heightened Scrutiny in Markets with Direct Network Effects

Though Facebook may be a monopoly, antitrust law, and the Sherman Act specifically,

only condemns monopolies that acquired their power by engaging in anticompetitive conduct.257

257
15 U.S.C. § 2 (2000).

Electronic copy available at: https://ssrn.com/abstract=3247362

The classic definition of anticompetitive conduct is "the willful acquisition or maintenance of

[monopoly] power as distinguished from growth or development as a consequence of a superior

product, business acumen, or historic accident.”258 In other words, anticompetitive conduct falls

outside the bounds of “competition on the merits.”259 It includes behavior that can be described

as “predatory,” “exclusionary,” “unethical,” or “deceptive.”260 Though scholars debate when

deception is actionable under antitrust laws, as opposed to merely under consumer protection

statutes, even the leading antitrust treatise concedes that deception falls into the category of

prohibited, anticompetitive, conduct, when deception has made “a durable contribution to the

defendant’s market power.”261

Courts have a broad mandate to capture and regulate a range of conduct that harms the

competitive process and is likely to or does result in harm to consumers.262 The breadth and

flexibility that courts have reflects the fact that “monopolization [conduct] has tended to be

258
In United States v. Grinnell Corp., the Court articulated what remains the two-part test for a Section 2
violation: "(1) 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." 384 U. S. 563, 571.
259
Microsoft Corp., 253 F.3d (finding Microsoft conduct outside the scope of “competition on the merits” to be
anticompetitive).
260
See Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 605 (1985); Allied Tube & Conduit
Corp. v. Indian Head, Inc., 486 U.S. 492, 500 (1988) (stating that “unethical and deceptive practices can constitute
abuses of administrative or judicial processes that may result in antitrust violations”); Spectrum Sports, Inc. v.
McQuillan, 506 U.S. 447, 458 (1993) (defining anticompetitive conduct generally as “conduct which unfairly tends
to destroy competition itself"). See also AREEDA & HOVENKAMP, supra note 26; ROBERT H. BORK, THE ANTITRUST
PARADOX 138 (1978) (“If a firm has been ‘attempting to exclude rivals on some basis other than efficiency,’ it is fair
to characterize its behavior as predatory.”). For an overview of how federal agencies and courts evaluate a
monopolist’s deception, see Maurice E. Stucke, How Do (and Should Competition Authorities Treat a Dominant
Firm’s Deception, 63 SMU L. REV. 1069 (2010). For a conversation on deception specifically, see Kevin S.
Marshall, Product Disparagement Under the Sherman Act, Its Nurturing and Injurious Effects to Competition, and
the tension Between Jurisprudential Economics and Microeconomics, 46 SANTA CLARA L. REV. 231, 244 (2006)
(explaining that deception can “create entry barriers, lead to capricious market exit, create artificial market
equilibrium, or even lead to oligopolies and monopolies.”).
261
AREEDA & HOVENKAMP, §782, supra note 26 (additionally, scholars point out that in situations where
deception by a monopolist is at issue, the concern should be deception of consumers).
262
See Herbert Hovenkamp, The Monopolization Offense, 61 OHIO ST. L. J. 1035 (2000),
https://kb.osu.edu/bitstream/handle/1811/70413/OSLJ_V61N3_1035.pdf;sequence=1.

Electronic copy available at: https://ssrn.com/abstract=3247362

nonrepetitive and specific to industry.”263 Standard Oil’s strategies in the oil market differed

from American Tobacco’s strategies in the tobacco market, AT&T’s approaches in the telephony

market, or Microsoft’s tactics in the operating system market.

Moreover, anticompetitive conduct in the acquisition of monopoly power is of heightened

concern in markets exhibiting direct network effects. In such markets, meaningful competition

only exists at the early stages. At the beginning competition is fierce—each company vies to

edge-out competitors to “tip” the market in its favor.264 If a company can achieve early adoption

with consumers, it can thereafter benefit not only from its product’s price or quality, but also

simply from its user-base—a feature that strongly influences consumer choice. Because of these

market dynamics, the Court of Appeals for the D.C. Circuit in the monopolization case against

Microsoft explained that competition in networked industries is often "for the field" rather than

"within the field."265 Misleading, deceptive, or otherwise unethical conduct at the early stages

can induce market participants to choose the firm that they think increases their welfare, but the

very act of mistaken choice can lock in the market to their detriment. Carl Shapiro, then deputy

assistant attorney general for economics in the Antitrust Division of the Department of Justice,

addressed the need for heightened concern: “Even more so than in other areas, antitrust policy in

network industries must pay careful attention to firms’ business strategies, the motives behind

263
Id. at 1037.
264
Theoretical economist W. Brian Arthur has published several papers discussing the unique nature of in
markets characterized by network effects. His seminal paper is ON COMPETING TECHNOLOGIES AND HISTORICAL
SMALL EVENTS: THE DYNAMICS OF CHOICE UNDER INCREASING RETURNS, supra note 256. See also MALCOLM
GLADWELL, THE TIPPING POINT: HOW LITTLE THINGS CAN MAKE A BIG DIFFERENCE (2006); William J. Kolasky,
Jr. & William F. Adkinson, Jr., Single Firm Conduct: Who's Big? What's Bad? Presentation Before the A.B.A. Sec.
of Antitrust L. 30 (Apr. 15, 1999) (on file with author) (stating that "if the ultimate market outcome is likely to be a
monopoly of the surviving firm, with the opportunity to earn substantial rents, competition among firms to be the
survivor will be intense.").
265
United States v. Microsoft Corp., 253 F.3d 34, 50 (D.C. Cir. 2001) (citing Harold Demsetz, Why Regulate
Utilities?, 11 J. OF L. & ECON. 55, 57 (1968)).

Electronic copy available at: https://ssrn.com/abstract=3247362

these strategies, and their likely effects….”266 In that vein, even Robert Bork, leader of Chicago-

school antitrust jurisprudence, may have agreed when he advised that one must pay attention to

the “route by which [monopoly] was gained.”267

Such was the focus of the Court’s analysis of Microsoft’s misleading behavior in the

operating system market.268 Microsoft had made false public representations about its own

products which induced developers to develop applications compatible with Microsoft’s

operating system. While Microsoft represented that applications developers wrote would be

cross-compatible with Sun, developers ended up writing applications that were only compatible

with the Windows operating system. Microsoft’s false statements induced developers to choose

Microsoft to their detriment. Since the operating system market exhibits direct network effects,

misleading behavior to induce choice is anticompetitive. This was anticompetitive conduct that

supported the finding that Microsoft had illegally monopolized the operating system market.

B. Pattern of False Statements, Misleading & Deceptive Conduct

A substantive issue is whether Facebook’s pattern of conduct was anticompetitive and

provides a basis for a claim of illegal acquisition of monopoly power. To begin, Facebook

entered the market with superior privacy protections making promises to not violate user privacy

which induced early consumer reliance.269 For several years, in a hotly competitive market,

Facebook did not merely change its privacy representations, it continued to perpetuate them. But

266
Shapiro, supra note 26 (explaining that “the very nature of the ‘positive feedback’ cycle means that
monopolization may be accomplished swiftly. And, once achieved, the network effects that helped create dominance
may make it more difficult for new entrants to dislodge the market leader than in other industries lacking network
characteristics.”).
267
BORK, supra note 260 at 164.
268
Microsoft Corp., 253 F.3d at 76-77.
269
FACEBOOK PRIVACY POLICY (2004), supra note 42.

Electronic copy available at: https://ssrn.com/abstract=3247362

while perpetuating the belief that Facebook did not track consumers using third-party code,

Facebook was caught doing precisely that on multiple occasions.270 The investigative efforts of

multiple independent researchers unveiled that the claims Facebook was making to consumers

via Facebook’s own policies and the public comments of Facebook executives were false.271

Facebook itself even conceded their falsity.272

Facebook then deflected consumer concern over the discovered hidden activity and false

statements with words and actions that implied the sincerity of Facebook’s commitment to user

privacy. For example, when Facebook’s hidden activity and false statements with Beacon were

exposed, Facebook retreated and Zuckerberg called Beacon a “mistake.”273 Later, when

Facebook’s hidden activity and false statements with social plugins were exposed, Facebook

claimed that discovered tracking was due to inadvertent software “bugs” or that the tracking was

innocuous because it was for users’ own “safety and protection.”274 At the time, the chief

technology officer of Facebook assured, “[social plugins are] not intended for tracking,” rather

user data collected was used “to protect the site from cyber-attacks by people who try to break in

to users’ accounts.”275 An evidentiary question is whether this activity was in fact inadvertent

and for users’ safety and protection or not. An investigation into the credibility of Facebook’s

alleged reasoning by the German Data Protection Authorities in Schleswig-Holstein found it

without merit.276 If Facebook’s reasoning was factually inaccurate, as the GDPA investigation

suggested, then Facebook’s behavior was deceptive.

270
See infra Section I.B & I.C.
271
Id.
272
See Perez, supra note 90.
273
See Zuckerberg, supra note 109.
274
See infra Section I.B.
275
See Efrati, supra note 133.
276
KIEL, UNABHÄNGIGES LANDESZENTRUM FÜR DATENSCHUTZ SCHLESWIG-HOLSTEIN (ULD):
DATENSCHUTZRECHTLICHE BEWERTUNG DER REICHWEITENANALYSE DURCH FACEBOOK (2011),
https://www.datenschutzzentrum.de/uploads/facebook/20110930-facebook-verantwortlichkeit.pdf (conclusion based

Electronic copy available at: https://ssrn.com/abstract=3247362

In addition to perpetuating factual inaccuracies, Facebook deceived users by omitting

salient facts. 277 Notably absent from public commentary was the fact that Facebook was thinking

about, or planning on, using social plugins for surveillance. In 2011, Facebook had filed a patent

for engaging in this type of behavior that was of public concern.278 The Federal Trade

Commission’s case for illegal monopolization against Intel centered on Intel’s misleading

behavior.279 For example, the FTC’s complaint alleged that Intel had engaged in deceptive acts

by failing to disclose material information about the effects of its products. When Facebook’s

surveillance conduct with plugins was discovered, Facebook provided affirmative, innocuous,

and possibly false reasoning for its actions while omitting information related to Facebook’s

internal dialogue about using plugins to conduct surveillance. Facebook likely deceived market

participants.

In retrospect, Facebook’s widely-covered public announcement of a user referendum

process for future privacy changes in the competitive market of 2009 may also raise issues of

anticompetitive behavior. Facebook’s poor notice to consumers in 2011—when it proposed and

passed an abolishment of user voting—suggests that Facebook’s commitment to the user

referendum was less than genuine. Consumers ultimately felt cheated that Facebook did not

notify them of an opportunity to vote.280 Indeed, internal comments by employees suggest that

Facebook made public representations about valuing privacy when in fact internally Facebook’s

on finding that Facebook internal technical documentation for consumer protection and prevention did not depend
on the use or reference of user cookies). See also Arnold Roosendaal, Privacy and Identity § Massive Data
Collection By Mistake, in PRIVACY AND IDENTITY MANAGEMENT FOR LIFE 274-282 (Camenisch et al. ed., 2012).
277
Deception can occur when one fails to disclose a fact that wrongfully causes a false belief in another. See
generally Gregory M. Klass, The Law of Deception: A Research Agenda, 89 U. COLORADO. L. REV. 707
(forthcoming 2018), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3156625.
278
See U.S. Patent, supra note 147.
279
Complaint in the Matter of Intel Corp., Case 0610247 (F.T.C. 2009) (No. 9341); Decision and Order in the
Matter of Intel Corp. No. 9341 (F.T.C. 2009).
280
See supra notes 150-153 and accompanying text.

Electronic copy available at: https://ssrn.com/abstract=3247362

culture and practices were not concerned with user privacy. Facebook employees themselves

sometimes parodied this disingenuous public-private disconnect.281

C. Wider Pattern of False Statements and Misleading Conduct

Importantly, Facebook’s wider history may point to a larger pattern of misleading

conduct within the company. In 2011, Facebook settled charges with the Federal Trade

Commission alleging a range of false and misleading material statements made to consumers

related to user privacy—all of which fell outside of the scope of Facebook’s conduct that is the

focus of this Paper.282 For instance, the FTC complaint alleged that while Facebook’s privacy

controls allowed users to restrict their information to “only friends,” Facebook was actually

overriding user choice and sharing users’ information with third-parties. More recently,

Facebook has come under congressional scrutiny for deceiving consumers by knowing but not

disclosing to users that their data was misappropriated by political consulting firm Cambridge

Analytica. When asked by Senator Kamala Harris (D-CA) if Facebook made an explicit decision

to not inform users, Zuckerberg answered “yes,” and called it another “mistake.”283 The wider

record of deceptive conduct with respect to user privacy may also be relevant to a larger case of

unlawful acquisition (and even perpetuation) of monopoly power.

Moreover, buyers of Facebook advertising would be keenly familiar with other

“mistakes” and alleged “bugs” that Facebook claimed inadvertently caused a string of inflations

281
See Tsotsis, supra note 154; see Martinez, supra note 154.
282
Facebook Settles FTC Charges That It Deceived Consumers by Failing to Keep Privacy, FTC.GOV (Nov.
29, 2011), https://www.ftc.gov/news-events/press-releases/2011/11/facebook-settles-ftc-charges-it-deceived-
consumers-failing-keep; Complaint for violation of the Federal Trade Commission Act in the matter of Facebook
Inc., No. 0923184 (F.T.C. Dec. 5, 2011),
https://www.ftc.gov/sites/default/files/documents/cases/2011/11/111129facebookcmpt.pdf.
283
Mark Zuckerberg Testimony: Senators Question Facebook’s Commitment to Privacy, N.Y. TIMES (April 10.
2018), https://www.nytimes.com/2018/04/10/us/politics/mark-zuckerberg-testimony.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

in ad metrics, which financially benefited Facebook to the detriment of buyers. A full review of

Facebook’s conduct (and pricing power) on the advertising side of the market is outside the

scope of this Paper. However, I review here two incidents on the advertising side that suggest a

larger pattern of willful disregard for truth. Aside from lending credence to a pattern and practice

of false statements and misleading conduct, Facebook’s behavior generally reflects pricing

power on the advertising side of the market.

For example, in September of 2017, media buyers became aware that Facebook materials

claimed to reach more people in the U.S. than Census data shows even existed.284 The false

representations are made in Facebook’s self-serve interface that businesses use to evaluate and

purchase Facebook advertising and are the subject of a recently filed class action complaint.285

To give another example, Facebook recently admitted to a succession of false statements

related to ad metrics which induced marketers to purchase Facebook ads. For many years,

Facebook did not allow ad buyers to audit and verify Facebook representations directly

correlated to Facebook ad billing—despite audit rights being standard practice in the industry.286

Facebook advertising is subject to elevated risks of fraud. Unlike print advertising, digital

advertising is shown only briefly to a user and poof disappears forever. In other words, without

audit rights, digital media buyers suffer from the risk of sellers charging them for goods not

delivered at all. Facebook has long been a hold-out in permitting buyers to audit Facebook

284
The initial discrepancy was caught and pointed out in a note to investors by respected industry analyst Brian
Wieser. See supra note 15; see also Allison Schiff, Facebook Data May Be at Odds with Census Data, But
Advertisers Won’t Stop Spending, ADEXCHANGER.COM (Sept. 6, 2017),
https://adexchanger.com/platforms/facebook-data-may-odds-census-data-advertisers-wont-stop-spending/
285
Complaint, Danielle A. Singer v. Facebook Inc., Case 4:18-cv-04978-KAW, (USDC N.D. Cal., Aug. 15,
2018).
286
A refusal to submit to audits is a demonstration of pricing power on the advertising side of the market.
Facebook’s conduct reminds us of Yellow Pages’ behavior in the print advertising market. Yellow Pages was also a
hold-out, not allowing buyers to audit Yellow Pages’ distribution figures. Yellow Pages, of course, was also a
monopoly. ASS’N OF NAT’L ADVERTISERS TELEPHONE DIRECTORY COMMITTEE AND AM. ASS’N OF ADVERTISING
AGENCIES DIRECTORY ADVERTISING COMMITTEE, THE NEED FOR YELLOW PAGES THIRD-PARTY CIRCULATION
AUDITING (2005), available at http://www.ana.net/getfile/67.

Electronic copy available at: https://ssrn.com/abstract=3247362

deliverables. The head of digital marketing at Wendy’s, who is currently the chief marketing

officer of Papa John’s, once explained, “[w]hat frustrates us when we run a campaign [on

Facebook] is that there’s almost no acknowledgement that the campaign even existed in the first

place.”287

After mounting industry pressure and calls of fraud, Facebook capitulated and announced

in the fall of 2015 that it would allow third-parties to audit select metrics.288 With the impending

release of the first audited metrics, Facebook came forward with disclosures of discovered

“bugs” that had caused it to inadvertently inflate ad metrics for multiple years.289 In September

2016, Facebook revealed that it had overestimated customers’ video ad view-times for the last

two years. Facebook sent a letter to Publicis, one of the largest ad agencies in the world with a

market cap of over $12 billion, estimating that it had overstated view times by 60-80%. Publicis

had spent $77 million on Facebook ads the prior year. Publicis, in turn, sent a note out to clients:

“This once again illuminates the absolute need to have ... verification on Facebook’s platform.

Two years of reporting inflated performance numbers is unacceptable.”290

287
Allison Schiff, Facebook: Counting Viewed Impressions is a ‘No Brainer’, ADEXCHANGER.COM (Feb. 18,
2015), https://adexchanger.com/online-advertising/facebook-counting-viewed-impressions-is-a-no-
brainer/#comment-911780. See also Jessica Davies, ‘Facebook doesn’t operate with real-world metrics’: GroupM
talks tough on Facebook, DIGIDAY.COM (Sept. 21, 2018), https://digiday.com/media/facebook-doesnt-operate-real-
world-metrics-group-m-talks-tough-facebook/ (where an executive at WPP, Facebook’s largest single client, states:
“We are still not able to verify delivery of our clients’ advertising via Facebook … They control the delivery of
consumption data back to us. We also have major issues with the quality of the environment our ads are delivering
in, especially when it comes to Facebook. Also, completion rates on Facebook are appallingly low.”).
288
Tim Peterson, Facebook to Sell 100% In-view Ads, Let Brands Fact-check Video Ad Views, ADAGE.COM
(Sept. 17, 2015), http://adage.com/article/digital/facebook-adds-ad-viewability-verification-options/300409/.
289
Lindsay Stein, ANA Calls Facebook Metrics to be Audited and Accredited, ADAGE.COM (Sept. 29, 2016),
http://adage.com/article/cmo-strategy/ana-calls-facebook-metrics-audited-
accredited/306096/?CSAuthResp=1515363470777:0:2255529:0:24:success:5A7B779E9C26F932284F303D3DF67
7EA; Suzanne Vranica & Jack Marshall, Facebook Overestimated Key Video Metric for Two Years, WALL STREET
J. (Sept. 22, 2016), https://www.wsj.com/articles/facebook-overestimated-key-video-metric-for-two-years-
1474586951.
290 Vranica & Marshall, supra note 289.

Electronic copy available at: https://ssrn.com/abstract=3247362

Between November and December of 2016, Facebook disclosed additional measurement

misreporting.291 Facebook had been inflating the number organic visits to brand posts, over-

reporting by 7-8% the average length of time people spent reading Instant Articles, over-stating

video ad completion rates, over-reporting the number of clicks it had reported it sent to advertiser

websites, over-reporting the number of Likes for live videos, and inflating the number of times

people shared links of posts on Facebook. Advertisers used these metrics to ask and inform

imperative questions: is this advertising campaign working? and should I spend more or less

money in the future with Facebook?

The Association of National Advertisers (ANA) subsequently publicly called for an in-

depth general audit, as they once did with the Yellow Pages—which is still pending.292 P&G’s

chief brand officer, Marc Pritchard, has warned that systemic fraud in ad markets may explain

the U.S. economy’s anemic growth.293 Pritchard has publicly pressured Facebook to submit to

transparency, and likened Facebook’s self-policing to letting a “fox guard the hen house.”

Advertising agencies, the middlemen in ad buys, were and continue to be in an awkward

position, having been entrusted to spend their brands’ money wisely. The chief operating officer

of GroupM, the media buying arm of ad agency behemoth WPP, lamented at an industry

291
James Hercher, Facebook Jumps Out of the Frying Pan and Into Fire with More Measurement Errors,
ADEXCHANGER.COM (Nov. 16, 2016), https://adexchanger.com/platforms/facebook-jumps-frying-pan-fire-metric-
errors/; Garett Sloane, Now Facebook Says it Gave Some Publishers Bad Traffic Numbers on Instant Articles,
ADAGE.COM (Dec. 16, 2016), http://adage.com/article/digital/facebook-gave-publishers-bad-traffic-
numbers/307192/.
292
Jack Marshall, ANA Pushes Facebook for Greater Measurement Transparency, WALL STREET J., (Sept. 30,
2016), https://www.wsj.com/articles/ana-pushes-facebook-for-greater-measurement-transparency-1475186796;
Allison Schiff, Facebook Gets Its First MRC Accreditation, But There’s Still More To Go, ADEXCHANGER.COM
(April 5, 2018), https://adexchanger.com/platforms/facebook-gets-its-first-mrc-accreditation-but-theres-still-more-
to-go/.
293
Jack Neff, P&G Tells Digital to Clean Up, Lays Down New Rules for Agencies and Ad Tech to Get Paid,
ADAGE.COM (Jan. 29, 2017), http://adage.com/article/media/p-g-s-pritchard-calls-digital-grow-up-new-
rules/307742/; Marc Pritchard, Chief Brand Officer, Speech given at Interactive Advertising Bureau's Annual
Leadership Meeting (Jan. 29, 2017), available at https://www.youtube.com/watch?v=NEUCOsphoI0.

Electronic copy available at: https://ssrn.com/abstract=3247362

conference that Facebook is one of the only ones that can measure themselves, and it is a result

of “Facebook’s market dominance.”294

D. Antitrust Harm: Monopoly Rents & Allocative Inefficiency

Ultimately, Facebook’s course of conduct misled consumers and resulted precisely in the

type of harm with which antitrust law concern itself.295 Facebook today is a monopoly that has

the power to extract monopoly rents from consumers, as detailed in Section II-C. Facebook’s

collective conduct—specifically, false statements, misleading statements, and omissions—

reviewed in Part I and Part II, contributed to this ultimate destination. The harm is not

speculative, it is complete. This new age’s communications utility extracts the cost of

widespread digital surveillance despite users’ preference to the contrary.

The tendency is to think that Facebook’s free service reflects consumer surplus, yet

nearly every advertising market in the U.S. is in decline as American consumers indicate a

preference for ad-free communications and media. In the world of television and video,

consumers have flocked from ad-supported TV to ad-free, over-the-top competitors. In that

world, Netflix, Amazon, and HBO, all paid, ad-free platforms, are bathing in a watershed

moment of creativity and high consumer satisfaction.296 With radio, terrestrial loses ground to

digital alternatives that offer subscriptions to ad-free music streaming. While digital music

294
Allison Schiff, Facebook And GroupM Tussle on Third-Party Viewability Verification, ADEXCHANGER.COM
(June 3, 2015), https://adexchanger.com/online-advertising/facebook-and-groupm-tussle-on-third-party-viewability-
verification/
295
Consumers have standing to sue for quality reductions. See AREEDA & HOVENKAMP, ¶¶ 345, 502, supra
note 26 (stating that “clearly a consumer has standing to sue a cartel that reduces quality of the product that the
consumer purchased”). Facebook’s 2014 decisions to initiate consumer surveillance are also anticompetitive in the
sense that they make it more difficult for new entrants like SnapChat to compete with Facebook on the advertising
side of the market.
296
The American Customer Satisfaction Index (ACSI), an annual survey of consumer attitudes to services
including pay TV, video streaming, ISPs and fixed and wireless phone.

Electronic copy available at: https://ssrn.com/abstract=3247362

platforms can be ad-supported or ad-free, the paid but ad-free models currently drive the largest

share of growth in the industry.297 In the digital advertising market, most sellers of digital

advertising struggle to obtain incremental, year-over-year revenue growth. In the midst of a

booming economic cycle in 2017, The Guardian, Britain’s nearly 200-year-old daily newspaper,

started to ask readers for donations after digital ad revenues fell.298 BuzzFeed followed suit.299

However, the market is growing. The U.S. internet advertising market grew 21.8% in 2016 and

21% in 2017. Facebook and Google accounted for over 99% and over 90% of that growth in

2016 and 2017, respectively.300 Consumer studies show that consumers today are averse to

advertising.301Against this backdrop, Facebook’s free but ad-supported communications service,

built upon a massive commercial surveillance apparatus, may indeed reflect one great allocative

inefficiency in markets.

CONCLUSION

“Monopoly in trade or in any line of business in this country is odious to our form of government
…. its tendency is repugnant to the instincts of a free people.”

—Chief Justice Sherwood of the Supreme Court of Michigan.302

297
See MORGAN STANLEY, LEADERS CONSOLIDATING: OUR 4TH ANNUAL MUSIC & RADIO SURVEY 8 (Dec. 17,
2017),
https://fa.morganstanley.com/balog/mediahandler/media/111221/Morgan%20Stanley%204th%20Annual%20Music
%20and%20Radio%20Survey.pdf.
298
David Bond, Guardian Relies on Readers’ Support to Stave Off Crisis, FINANCIAL TIMES (May 13, 2017),
https://www.ft.com/content/9044ff9a-358b-11e7-99bd-13beb0903fa3.
299
Benjamin Mullin, BuzzFeed News Asks Readers to Chip in With Donations, WALL STREET J. (Aug. 27,
2018), https://www.wsj.com/articles/buzzfeed-news-asks-readers-to-chip-in-with-donations-1535395575
300
Sarah Sluis, Digital Ad Market Soars To $88 Billion, Facebook And Google Contribute 90% Of Growth,
ADEXCHANGER.COM (May 10, 2018), https://adexchanger.com/online-advertising/digital-ad-market-soars-to-88-
billion-facebook-and-google-contribute-90-of-growth/.
301
MILLWARD BROWN, AdReaction Gen X, Y and Z Executive Summary,
http://www.millwardbrown.com/adreaction/genxyz/global/gen-x-y-and-z/how-media-habits-differ (consumer study
finding that “gen Z” (16-19-year-olds) is “more averse to advertising in general”).
302
Richardson v. Buhl, 77 Mich. 632 (1889).

Electronic copy available at: https://ssrn.com/abstract=3247362

The fact that this century’s new communications utility is free but necessitates

widespread surveillance of consumers is a paradox in a democracy. Facebook watches, monitors,

and remembers what over 2 billion people do and say online. Contrary to what those in the

advertising industry would regulators to think, American consumers value a state of no

surveillance and have attempted to protect this aspect of their privacy since the beginning. The

fact that the free market today offers no real alternative to this exchange is a reflection only of

the failure of competition.

At least for this titan of tech, antitrust law provides a framework for appreciating and

correcting for the foreclosure of consumer choice. Facebook is a monopoly that tipped the early

market with promises of data privacy and then engaged in a long line of misleading conduct,

which foreclosed competition. The historical record tells the story of Facebook’s monopoly

power in the social media market. Facebook tried, but could not, degrade the quality of its

product to impose commercial surveillance on users through Beacon in the competitive market

of 2007. Thereafter, Facebook pivoted to licensing Like buttons, Logins, and other products to

independent businesses, which Facebook could leverage for the same purpose. Yet competition

between 2008 and 2014 continued to restrain Facebook’s ability to initiate tracking for the

purpose of targeted advertising. Facebook had to retreat from alleged accidental tracking, assure

consumers and other market participants that the underlying code for social plugins was not used

for commercial surveillance, and then promise users an ability to vote on future privacy changes.

Only after the exit of competitors, and the barrier to entry that comes with over a billion users on

a closed communications protocol, was Facebook able to reverse course. The history of

Facebook’s market entry and subsequent rise is the story of Facebook’s monopoly power.

Electronic copy available at: https://ssrn.com/abstract=3247362

Facebook’s pervasive and intrusive commercial surveillance of citizens’ digital footprints is

merely this titan’s form of monopoly rents.

Consumers today turn from Facebook to other websites and apps and face an identical

degradation of quality across millions of sites and competitors on the advertising side of the

market. For publishers like The New York Times and others, Facebook extracts commercial

surveillance of their customers through publishers’ licenses of Facebook’s business products

(e.g., Like buttons etc.). Facebook has commoditized these publishers’ own user data, once a

prized proprietary possession, for its own benefit either to sell Facebook advertising or the

advertising of a publisher’s competitors. This market structure has deteriorated the pricing power

of market actors across the horizontal market and resulted in the duopoly of Facebook and

Google—which account for just about the entirety of the growth in the digital advertising market

against a backdrop of publishers such as BuzzFeed or The Guardian soliciting reader donations.

The historical record that elucidates Facebook’s monopoly power raises the question of

whether Facebook’s decade-long course of conduct was anticompetitive—especially in the

winner-take-all market of a closed communications platform. The record is replete with reliance-

inducing future promises, false statements, disingenuous excuses, and convenient omissions

which, collectively, likely deceived users. The adoption of a user referendum process for future

privacy changes coupled with the failure to meaningfully notify users of an opportunity to vote

further raises the specter of a pattern of anticompetitive conduct. Indeed, the wider record of

misleading and deceptive conduct—whether that conduct was the subject of an FTC or

congressional investigation or the complaints of advertisement buyers—may point to a more

systemic problem which harms not only consumer welfare but also presents risk to market

stability. Antitrust scholar Robert Steiner once warned that deception by a dominant firm could

have a domino effect within an industry, leading smaller firms to engage in similar patterns of

Electronic copy available at: https://ssrn.com/abstract=3247362

conduct and inefficiency in the industry.303 Indeed, today, the digital advertising industry is

considered one of the most fraud-stricken in the world—the industry expects to absorb $19

billion of waste due to fraud in 2018.304

To correct for consumer harm and reduction of choice in the market, a remedy must

induce competition and stop horizontal coordination. To induce viable opportunity for new

entrants, consumers must be able to export their social graph,305 and Facebook should migrate

from a closed to an open communications protocol. A user on Facebook should be able to send a

message to, or receive a message from, a user of a competitive social network—in the same way

that users of AT&T can call or text a user of Sprint, Verizon, or T-Mobile. The adoption of an

open application programing interface for user messages, chats, posts, and other communications

could aid this process. The social network LinkedIn permits communications to be distributed

across users’ Twitter feeds for example. Additionally, it is paramount that a remedy put a stop to

coordination amongst competitors. For this, we need to empower consumers with a singular Do

Not Track switch that can counter the collusion in the horizontal market. Consumers must be

able to just say no to commercial surveillance—a broad interconnected apparatus that uniquely

303
Robert L. Steiner, Double Standards in the Regulation of Toy Advertising, 56 CINCINNATI L. REV. 1259,
1264 (1988). See also Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 474 n.21 (1992) (noting that
"in an equipment market with relatively few sellers, competitors may find it more profitable to adopt Kodak's
service and parts policy than to inform the consumers").
304
ASS’N OF NAT’L ADVERTISERS, THE BOT BASELINE: FRAUD IN DIGITAL ADVERTISING 2017 REPORT (2017),
available at http://www.ana.net/content/show/id/botfraud-2017?mod=article_inline; Alexandra Bruell, Ad Fraud
Declines Offer Hope as Marketers Fight Sophisticated Bots, WALL STREET J. (May 24, 2017),
https://www.wsj.com/articles/ad-fraud-declines-offer-hope-as-marketers-fight-sophisticated-
bots-1495645098?mod=article_inline (economic loss due to fraud estimated at $6.5 billion in 2017);
Estimated cost of digital ad fraud worldwide in 2018 and 2022 (in billion U.S. dollars), STATISTA.COM (2018),
https://www.statista.com/statistics/677466/digital-ad-fraud-cost/; Lara O’Reilly, Google Issuing Refunds to
Advertisers Over Fake Traffic, Plans New Safeguard, WALL STREET J. (Aug. 25, 2017),
https://www.wsj.com/articles/google-issuing-refunds-to-advertisers-over-fake-traffic-plans-new-safeguard-
1503675395; Suzanne Vranica, P&G Contends Too Much Digital Ad Spending Is a Waste, WALL STREET J. (March
1, 2018), https://www.wsj.com/articles/p-g-slashed-digital-ad-spending-by-another-100-million-1519915621.
305
Others, including economist Luigi Zingales and Guy Rolnik, have also advised for social graph portability
to increase competition. Luigi Zingales & Guy Rolnik, A Way to Own Your Social-Media Data, N.Y. TIMES (June
30, 2017), https://www.nytimes.com/2017/06/30/opinion/social-data-google-facebook-europe.html.

Electronic copy available at: https://ssrn.com/abstract=3247362

serves the digital duopoly. While politicians and regulators grapple with how to make sense of

current market structures in and consumer frustrations with Big Tech, the principles of antitrust

provide clarity for this era’s dominant communications platform.

Electronic copy available at: https://ssrn.com/abstract=3247362