title:: The Antitrust Case Against Facebook url:: author: [[@Dina Srinivasan]] output: tags:: topic: # Notes <!-- Starts with the assertion that privacy was paramount in 2004 when Facebook started, not backed up be a reference --> ## Questions posed - Why do competitors for placement of ad inventory like publishers have to work *with* Facebook and let them monitor their users? - Does the free market represent the consumer welfare standard? --- - Academic case for monopoly has not yet been made p. 43 - Misleading conduct is of particular concern in markets with direct network effects. fn 26, p. 45 - In markets where consumers do not pay a price, antitrust needs to look at quality. p 46. - Case for monopoly can be made by: p. 54 - Indirect evidence: Share of releveant market, structural factors that show that a company has control over the market. - Direct evidence: Ability to increase price or decrease quality beyond what was possible when competition existed/monopoly did not exist. ## Part 1 - Privacy, a crucial form of competition ## Part 2 Facebook's actions - A - B - C p 69 - 1 : Commercial Surveillance : In 2014, enabled tracking across third parties under the garb of better targeting/more relevant ads; change of course from earlier representations that stated that data from third party sites would only be used for user safety and protection. The reversal meant that privacy and profits now bore an inverse relationship. Facebook as much as a publication's customers, and also their cross-publication interests enabling better ad packaging, thus a competitive advantage. Unlike 2007, since the code that enabled these changes was already embedded across millions of pages, opt-out was not really an option for publications, or one without high switching costs. - 2 Tracking was coupled with identification, meaning users could not reset ad-profiling attempts by clearning cookies or resetting ad ids on devices. In a competitive ad market, such horizontal surveillance would not have been possible. - 3 : Facebook actively circumvented opt out attempts : Digital Ad Alliance Do Not Track was buggy (did not work), limited, inconveneient (needed to be done for each device separately) and only worked if third party cookies were enabled; Ignored brower do not track requests; worked around Ad blockers even as their adoption was increasing indicating consumer preferences to not be tracked. - More surveillance meant more attribution of users to actions and thus greater revenues. ### Beacon - Program that allowed users to share activity on 3rd party websites on Facebook. - Chamath answered an NYT interview saying facebook does not get information if a user declines to share activity. But code analysis revealed that this was not true. - It also showed that facebook linked anonymous cookies to real world people ## Part 3 (indirect) Evidence - Fact-pattern: After competion exit, Facebook amassed users, reversed course of privacy and expanded surveillance. Demonstrates inelasiticity of demand for Facebook's product. Facebook surveillance is equivalent to extraction of monopoly rents at the expense of consumer welfare. Has 80% of social media market (not cited) p 81. - Market share-market definition. p 82. #### Define Relevant Market. - Significant substitution occurs to other products if one firm significantly increases prices or lowers quality. - Since Facebook acted after the exit of competing social networks fromthe market, that can bolser the position that relevant market should be limited to social markets. p 82. - Social media enabled a new form of instant one-many communication (phones enabled one-one voice) p 83. - Makes the case for width (pentration) and depth (usage per day). p 84. - Facebook complained that German Cartel office report did not include Twitter, Snapchat and Youtube. YouTube bills itself as a TV subscription replacement. p 86 - Contends that Facebook cannot succesfully argue being a two-sided market (advertising) citing a Supreme Court (US) judgement where it ruled that one type of 2-sided platform will fail the test (of needing to show abuse of dominance in both markets separately) - newspapers. p 87. In a two-sided market more of one side creates surplus for the other side, but with newspapers, more advertisers doesn't mean more surplus of readers/subscribers, it can mean the opposite. Weak indirect networks (across the market-sides) means the 2 sides can be treated distinctly. p 87 - Market definition analysis #### Market Definition - Market share should be defined based on Facebook's share of time spent on social networks because social networks themselves use this as a measure of their dominance and since no fee is collected, time is a metric that can account for all competitors p 87 - Facebook and Instagram together account for 83% of consumer time spent on social networks in the US. This was based on multiplying number of users per platform by minutes spent per platform to arrive at a total minutes based market size. - Entry Barriers #### Entry Barriers - Closed network leading to strong direct network effects and high switching costs. # Part 4 Was its behaviour illegal? Did it acquire its monopoly by engaging in anticompetitive conduct. - Paper relies on the deception standard (since it resulted in a signicant change in market power) while pointing out that scholars debate when it is actionable under antitrust laws. - It contends yes, through false statements, inducements, disingenous excuses, lies of omission and thus deception of users. - In strong direct network effects markets, deception is of additional concern since these are winner take all as more surplus accumulates to firm that grows more than others. Competition is an early-stage feature. - User referendum process adopted but users not notified of an opportunity to vote about its abolishment p 94. - 2011 FTC case it settled for overriding user preferences to share information meant only 'for friends' with third parties - Inflation of ad metrics, not permitting audits (until 2015). Before the release of this first audit, it made disclosures of bugs that allowed it to inflate ad metrics. In Nov and Dec 16 disclosed additional misreporting - all skewed to err on the side of Facebook's benefit. - Contrasts to OTT (video and audio) where paid and ad free versions are driving growth <!-- Does not mean that they don't surveill, no? --> to digital publishing where many are asking for donations/contributions to sustain themselves. # Further References