In the Microsoft antitrust case, Bill Gates’ emails were perhaps the government’s most compelling evidence. Now, as regulators pursue antitrust cases against Facebook and Google, they are looking for similarly incriminating paper trails. While Facebook’s internal emails provide strong support for the FTC’s case, Google’s internal documents are more tight-lipped.


After two decades without a major Section 2 enforcement action, in the past two months the Federal Trade Commission (FTC) and a group of 48 state and territory attorneys general have brought separate suits against Facebook, while the Antitrust Division of the US Department of Justice and two groups of state attorneys general have sued Google. These suits were much anticipated, and were preceded by congressional hearings and years of scholarly discourse on the threats these platforms pose.

Naturally, there is a great deal of public interest in how these cases might turn out and whether one or both of these companies might be broken up. It is likely these cases will drag on for years, and it would be a fool’s errand to try to handicap their outcomes at this early stage. Nonetheless, we can already evaluate one measure of the strength of the governments’ cases—the documentary evidence cited in the FTC and DOJ complaints. 

Antitrust analysis is grounded in economics. Both parties will engage expert economists, who will produce reports on market definition, market power, and consumer harm. These reports, and the testimony of the expert witnesses, will play a significant role in these cases. But documents are important too, perhaps even more so.

For an antitrust plaintiff, the most powerful pieces of evidence often can be the defendant’s own documents. Customer and competitor testimony can be helpful to understand the competitive impact of a merger or specific conduct, though both customers and rivals may have biases that color their views of the case. Third-party testimony—from industry analysts, for example—also can be persuasive. But when the defendant’s own words support the plaintiff’s theory of the case, that can be outcome shifting. In the Microsoft case, the government effectively used Bill Gates’ emails and other internal Microsoft documents against the company. Doug Melamed, the Antitrust Division’s Acting Assistant Attorney General at the time the case was litigated, stated in a recent interview that the government’s case “was built on these emails.” 

“When the defendant’s own words support the plaintiff’s theory of the case, that can be outcome shifting.”

By this measure, the FTC’s Facebook complaint bodes well for its case. It is replete with powerful Facebook documents that directly support the FTC’s theory of the case. The agency alleges that Facebook engaged in a combination of anticompetitive acquisitions (Instagram and WhatsApp) and exclusionary conduct involving access conditions on its application programming interfaces (APIs). Its case with regard to Facebook’s acquisition strategy is neatly summarized by a line from a 2008 email from Mark Zuckerberg: “it is better to buy than compete.” This statement has the same pithy (and damaging) quality as the line attributed to a Microsoft executive that the company’s strategy was to “cut off Netscape’s air supply.” 

The documentary evidence laid out in the FTC’s complaint shows that senior Facebook executives, including Zuckerberg, were increasingly concerned about Instagram’s growth and the threat it posed to Facebook. In 2012, for example, Zuckerberg wrote that “it’s quite possible that our initial thesis was wrong and [Instagram’s] was right—that what people want is more to take the best photos than to put them on FB. If so, [Facebook’s photo application] Snap might be a good first step but we’d be very behind in both functionality and brand on how one of the core cases of Facebook will evolve in the mobile world, which is really scary and why we might want to consider paying a lot of money for this.”

Zuckerberg also made the case for why the purchase of what was merely a nascent competitor was potentially anticompetitive: “The businesses are nascent but the networks are established, the brands are already meaningful and if they grow to a large scale they could be very disruptive to us,” particularly because Zuckerberg believed Facebook was “vulnerable in mobile.” 

The documents regarding the WhatsApp purchase are equally troubling. The FTC’s theory is that while WhatsApp is a mobile messaging service, not a social network, Facebook’s leadership was concerned that a sufficiently large mobile messaging app could successfully enter the social network market at scale. Facebook’s documents support this theory. The company’s Director of Product Management emailed colleagues that mobile messaging “is the biggest threat to our product that I’ve ever seen in my 5 years here at Facebook; it’s bigger than G+ and we’re all terrified.” After the acquisition was completed, a Facebook manager wrote that the $19 billion price tag was “worth it” and that the deal “prevents probably the only company which could have grown into the next FB purely on mobile.” 

The FTC’s conduct case involves Facebook Platform, a set of services that allow third-party developers to create apps and tools that interoperate with Facebook. Facebook Platform grants third-party apps access to Facebook user data and allows them to use valuable features such as the Find Friends API and the Facebook “Like” button. The FTC alleges that Facebook Platform and its “Open Graph” API have been extremely valuable for Facebook because they give the company access to troves of customer data and help increase its user base. At the same time, the FTC contends that Facebook Platform became a critical resource for third-party apps. The FTC’s theory is that Facebook used access to its valuable APIs as a cudgel to harm competition by cutting off access to any app that directly competed with Facebook (by “replicating” Facebook’s “core functionality”) or that linked to or promoted “any other competing social platform.” Facebook allegedly targeted several competitors with this strategy, including Google+ and Twitter’s Vine offering.

Again, the complaint supports this theory with internal Facebook documents, including an email from a Facebook engineer complaining about Platform restrictions: “so we are literally going to group apps into buckets based on how scared we are of them and give them different APIs?” Another engineer characterized this strategy as “sort of unethical.”

Can helpful documents save a weak case on the merits? Probably not. It is very difficult for a plaintiff to prevail on a unilateral refusal to deal claim of the sort the FTC has alleged regarding Facebook Platform, and the documentary evidence may not be enough to overcome the defendant-friendly case law. But where the case is close, the type of documents the FTC has collected can make all the difference.

The power of a defendant’s documents to affect the outcome of an antitrust case is underscored by an unusual paragraph in the DOJ’s complaint against Google. Paragraph 11 of the complaint states that Google “learn[ed] one thing from Microsoft—to choose its words carefully to avoid antitrust scrutiny.” The complaint quotes Google’s Chief Economist as warning that “We should be careful about what we say in both public and private. ‘Cutting off the air supply’ and similar phrases should be avoided.” This paragraph prepares the court for what is missing in the DOJ’s case—any damaging Google documents—and it speaks to the expectation that plaintiffs will be able to produce such documents in most cases. The New York Times has reported on Google’s policies instructing employees not to mention antitrust buzzwords in emails and other documents, as well as Google employees’ standard practice of copying the legal team on emails to lay the basis for claims of attorney-client privilege.

“We should be careful about what we say in both public and private. ‘Cutting off the air supply’ and similar phrases should be avoided.”

Google’s Chief Economist

The Times has also reported that DOJ staff attorneys felt rushed to get a Google complaint out quickly and perhaps before it was fully ready. A rushed process could help explain why the complaint is thin on damaging Google documents. The DOJ agreed to a trial date in 2023, giving the agency time to find any such documents, if they exist.

The apparent success Google has had in all but eliminating smoking-gun documents presents a challenge to enforcers. Sophisticated and disciplined companies with a vigilant in-house legal team may be able to reduce or eradicate altogether the kinds of documentary evidence enforcers have successfully relied on to win antitrust cases over the past thirty years. As a general matter, it is probably more likely that mergers and acquisitions produce more “hot” documents than suspect conduct does. It is natural for proponents of particular acquisitions to claim that they will help the acquirer “crush” the competition or “dominate” the market. That may be one reason why the FTC—which is challenging not only Facebook’s conduct but also its acquisitions—was able to find so many powerful documents and the DOJ—which is challenging only Google’s conduct—was not.

Many of the Facebook documents are a decade or more old, so it is also possible that its leaders were not as careful as they would be now. Indeed, in the 2000s even Google (along with several other Silicon Valley companies) was implicated by its own documents in a scheme to suppress wages by agreeing with its rivals not to cold call each other’s engineers. These documents were sufficiently damaging that the defendants settled cases brought by the DOJ and a private class.

The Facebook and Google cases present a host of fascinating antitrust issues, including how markets should be defined in these spaces and what constitutes consumer harm in non-price markets. But the power of defendant’s documents also bears watching. Can Facebook overcome the damage its leaders’ own words will do to its defense? And can the DOJ prevail in the absence of any powerful Google documents? The answers to these questions might tell us quite a bit about the future of antitrust litigation.