Wednesday December 9th, 2020: the US Federal Trade Commission (“FTC”) and 48 US States (led by New York State Attorney General Letitia James) filed antitrust lawsuits against Facebook, arguing the tech giant has unlawfully maintained its monopoly power in the market for “Personal Social Networking” in breach of Section 2 of the Sherman Act (see here for the text of the FTC lawsuit and here for the State lawsuit).
TL;DR: The lawsuits are broadly similar in scope and claims – even though the State lawsuit is at times a bit more elaborate (it’s 20 pages longer than the FTC one after all). The gist is that Facebook has maintained its monopoly power in the personal social networking market (more on which below) through a buy-or-bury strategy, that is through acquisitions of actual or potential competitors (Instagram and WhatsApp) and through deploying an open-first, closed-later strategy with respect to its Applications Programming Interfaces (“APIs”), on which many third parties have grown to rely. Facebook’s conduct harms consumers (e.g., in the form of less choice or innovation) but also advertisers (e.g., in the form of higher advertising prices).
Overall impression: these are strong lawsuits, and are based on ample evidence from internal Facebook emails.
Both lawsuits track the origins of Facebook as social network that eclipsed incumbent MySpace by, among others, presenting itself as a more premium and private social network where the user was the “king” (it may seem a bit ironic now, but that’s how Facebook marketed itself in its early days). The State lawsuit details how originally Facebook was forced to improve its products when faced with competitive threats (e.g., the Google+ social network). Yet, once Facebook tipped the market and Google no longer posed a threat, it started to erode privacy protections in the hunt for more data and advertising dollars to the detriment of consumers (this story is not very different from the one in Dina Srinivasan’s 2018 paper trailing Facebook’s monopolistic journey into pervasive user surveillance). This has resulted in a worse user experience and less choice/innovation/lower quality. Let us now have a closer look at the lawsuits.
On the relevant market
According to the lawsuits, Facebook wields monopoly power in the US market for “personal social networking“, comprising online services that enable and are used to maintain personal relationships and share experiences with friends, family, and other personal connections in a shared social space. Three key elements are said to distinguish personal social networking services from other online services:
- Personal social networking services are built on a social graph that maps connections between users and their connections;
- Such services enable users to interact with their connections and share their personal experiences in a shared social space, among others in a one-to-many “broadcast” format;
- Such services enable users to find and connect with other users and expand their connections (note that the last two elements build on the first element, namely the social graph).
Personal social networking services are to be distinguished from (as they are not reasonably interchangeably with): specialized social networking services (e.g., LinkedIn), online video or audio consumption-focused services (think of YouTube or Hulu) and mobile messaging services.
On monopoly power
Facebook is alleged to wield monopoly power in the aforementioned market at least since 2011. The FTC lawsuit cites a market share in excess of 60% in said market. Both suits also place great emphasis on barriers to entry, namely (direct) network effects (acknowledged by Facebook executives themselves) and high switching costs due to “ratchet effects” (the more a user invests in building her graph, the stickier FB becomes and the harder for her to switch to an alternative).
On unlawful conduct
Facebook’s allegedly unlawful conduct comprises two limbs, (1) the acquisition of nascent /potential rivals to build a protective “moat” around its kingdom; and (2) imposing anti-competitive conditions on access to APIs (the State lawsuit refers to Facebook as deploying a dual “buy-or-bury” strategy). As pretty much everyone expected, the Instagram and WhatsApp acquisitions feature prominently in both lawsuits. Another notable acquisition is that of VPN provider Onavo, which Facebook has been using to track emerging services with spectacular growth/nascent threats (“Early Birds”) and thus inform its strategic market decisions (such as whether to acquire firms or not).
The lawsuits explain how consumer shift from desktop computers to mobile devices around 2010 (as a result of the widespread adoption of smartphones) exposed Facebook Blue’s (that’s how Facebook described its social network) vulnerability to social networking services that were optimized for mobile device use, chief among which was Instagram, which focused on sharing photos. The lawsuits cite some juicy internal emails where top Facebook executives felt very anxious about Instagram’s stellar growth and especially the prospect of it falling into “the wrong hands” (e.g., Apple or Twitter). Plenty of emails show Zuckerberg perceived mobile app companies like Instagram or Path as a clear threat, “building networks that are competitive with out own, “which “if they grow to a large scale they could be very disruptive to us.” But as Zuckerberg pointed out to a subsequent email, “[o]ne thing about startups is you can often acquire them” (emphasis added). And that’s what Facebook did; in 2012 it acquired Instagram (which at the time had zero revenue and just 16 employees) for $ 1 billion. Zuckerberg acknowledged one of the main reasons for doing so was to “neutralize” a potential competitor.
According to the lawsuits, ever since Instagram was acquired, Facebook has taken steps to reduce its impact on Facebook Blue, confirming it is a significant threat to Facebook’s monopoly, but Blue nevertheless continues to lose ground to Instagram. The acquisition of Instagram thus served to eliminate a potential competitor, while at the same time building a moat around Facebook, by making it harder for rivals to challenge its position. As Zuckerberg explained, if Facebook incorporated the social mechanics of Instagram, any new rival product would not gain any traction, as Facebook would already have their mechanics deployed at scale.
In response, Facebook will likely argue that the FTC/States have to prove that absent the merger, Instagram would become a credible competitor to Facebook. For an explanation of why this is not the case, see this post by Tim Wu and Scott Hemphill.
A trip to the past (1): Back in 2012, the then-Office of Fair Trading in the UK approved Instagram’s acquisition, holding among others that Instagram was not well placed to compete against Facebook in the supply of social networking services. Yet as the authors of the Lear Report indicate, the evidence at the time did not seem to unambiguously support this conclusion. On the other hand, at least the UK got to review the transaction, even if at the time it acted a bit naively. In the EU the transaction escaped scrutiny altogether as it did not meet the EUMR turnover thresholds, even though the transaction value was the non-negligible sum of $ 1 billion. The Instagram deal has ever since been cited in support of the view that the merger thresholds in the EUMR should be revised.
The story is a bit similar for WhatsApp. Once more, the lawsuits cite plenty of internal emails where Facebook’s leadership expressed discomfort over mobile messaging apps with huge growth – and WhatsApp was the strongest. The concern was that WhatsApp could be used a springboard to launch a successful social network at scale. A Facebook executive perceived mobile messaging services as “the biggest threat to out product […] bigger than G+, and we’re all terrified.” Facebook tried to respond to the threat by launching Facebook Messenger in 2011, but Zuckerberg himself pointed out in a 2012 email that WhatsApp was simply a superior product. As in the case for Instagram, Facebook executives were anxious about the prospect of WhatsApp being acquired by another tech giant such as Google.
We all know what happened next. Facebook acquired WhatsApp in 2014 for $ 19 billion (!), a valuation alleged to reflect the seriousness of the threat WhatsApp posed to Facebook’s monopoly. Internal emails show how Facebook executives rejoiced once WhatsApp was acquired – a “land grab” as one said. According to the lawsuits, by acquiring WhatsApp Facebook eliminated a significant competitive threat, depriving users (and advertisers) from the benefits of competition from an independent WhatsApp, “which would have the ability and incentive to enter the U.S. personal social networking market.” The State lawsuit also reminds us that post merger Facebook reneged on its pre-acquisition promises and combined user data cross services by linking WhatsApp user phone numbers with accounts on Facebook Blue, thus degrading user privacy. For this practice, Facebook was fined 110 million EUR in 2017 for supplying incorrect or misleading information to the European Commission in the context of the Facebook/WhatsApp merger proceedings.
A trip to the past (2): In another world, back in 2014 the European Commission approved Facebook’s acquisition of WhatsApp without even launching an in-depth review of the transaction. In its 35-page decision, the Commission held among others that WhatsApp did not plan to become a competitor to Facebook and even under a broader market definition that would include WhatsApp and Facebook in the same relevant market, these were not close competitors.
While it’s of course easy to criticize the Commission with the benefit of hindsight (I bet that if the case team knew what Facebook’s executives thought of WhatsApp internally the outcome could have been very different), this transaction has nevertheless made us all wiser, and reminds us of the information asymmetry between companies and regulators. In any event, I do not see any reason why such transactions should not be revisited (and the lawsuits do precisely that). This is not weakness; it’s enforcing the law.
Anti-competitive policy with respect to APIs
The second limb to Facebook’s strategy was imposing anti-competitive conditions on access to its APIs – or as the State lawsuit puts it, deploying an open-first, closed-later strategy. Facebook launched Facebook Platform in 2007, encouraging third-party developers to write apps that interoperate with Facebook Blue. In 2010, Facebook provided third-party apps access to APIs such as the Find Friends APIs, which allowed users of third-party apps to invite their Facebook friends to use the app. In 2010 Facebook launched the Open Graph API, allowing third-party apps and websites to add social plug-ins to their websites and apps (e.g, Like or Share button). This looked like a win-win for apps/websites and Facebook: third-party developers had a new distribution channel, while Facebook expanded its data foothold by tracking the off-site activities of its users. The problem seems to be that for many apps access to Facebook APIs became critically important for their growth; losing access to such APIs would place them to survival mode.
Facebook used this power to adopt conditional dealing policies that limited how third-party apps could use Facebook Platform. Between 2011 and 2018, Facebook made access to its Platform available only on the condition that third-party apps neither competed with Facebook nor promoted competitors. Among others, developers were prohibited from exporting user data to competing social networks. Any apps found in breach of such policies were punished by losing access to such APIs. This is said to have altered the incentives of app developers, discouraging them from trying to compete against Facebook.
The State lawsuit details how at the same time Facebook began to selectively enforce its policies to cut off API access to companies Facebook worried might one day threaten its monopoly, including Twitter’s Vine, Path (a feed-based sharing app) and Circle (a local discovery app). In all cases pulling the plug resulted in reduced growth and/or eventual demise of the affected app.
I am pretty certain some will criticize this claim in the lawsuit, asserting that Facebook has every right not to help rivals, that there is no “duty to deal” in this case, that otherwise rivals would free-ride on Facebook and so on. I am no expert on US law but I should note that this is not a black-or-white issue. Perhaps there would be a good case that if Facebook had never granted access to its APIs then it would have been very difficult to compel it to grant rivals such access in order to help them compete with it. But the issue here is different. Facebook itself provided access to its APIs – not out of good will but because it helped it grow and become the “platform” for all social apps. Facebook itself opened its platform, only to then selectively close it when it detected threats. And some of its measures seem to have been taken with the sole aim of raising switching costs and making multi-homing harder (e.g., prohibiting developers from exporting data to rivals; this would certainly not be viewed positively by all those who advocate for data mobility interventions to improve competition).
Harm to competition
According to the lawsuits, Facebook’s conduct has suppressed competition in the market for US personal social networking services. In turn, this has deprived both users and advertisers of the benefits of additional competition:
- In the case of users, such lost benefits include additional innovation, quality improvements and/or increased choice (e.g., enabling users to select a social network that suits their preferences, including on privacy).
- In the case of marketers, such lost benefits include lower advertising prices, additional innovation, choice, and/or quality improvements (e.g., more transparency; Facebook’s reporting is notorious for its lack of transparency)
The lawsuits are pretty bold when it comes to prayer for relief. They request among others that:
- Facebook should divest Instagram and/or WhatsApp;
- Facebook should be subject to a prior notice and prior approval obligation for future M&As;
- Facebook should be permanently enjoined from imposing anti-competitive conditions on access to APIs and data.
Facebook’s worries have just started
The much-expected FTC and State lawsuits against Facebook are probably just the beginning. In the US, Facebook has been already hit with private class actions. In Europe, Commissioner Vestager said Facebook is embroiled in a “huge” data probe, while at the same time its integration of Marketplace in Facebook Blue is the subject of a separate investigation (oh and there is also this 2019 Bundeskartellamt decision barring Facebook from making access to its services conditional on users accepting that it can combine their on-platform activity with off-platform data).
The cherry on the cake is the upcoming EU legislative package on digital gatekeepers and illegal/harmful content, which could massively impact Facebook’s operations. The European Commission is expected to unveil its proposals for the legislative package tomorrow, 15 December, and we will cover them extensively in the Blog!