← Back to Library

Lies, damned lies, and statistics: Meta's case-in-chief draws sharp words

Brendan Benedict delivers a rare, granular look inside the courtroom theater of the Federal Trade Commission's antitrust battle against Meta, revealing how the defense is attempting to dismantle the very definition of the market it allegedly monopolizes. The most striking element isn't just the legal maneuvering, but the aggressive use of behavioral economics to argue that users are constantly fleeing Facebook and Instagram for rivals like TikTok and YouTube, effectively rendering the government's narrow market definition obsolete.

The Battle Over Market Definition

The case hinges on whether the "personal social networking services" market is a distinct playground where Meta holds a monopoly, or a sprawling arena where competition is fierce. Benedict notes that Meta's defense began by calling witnesses from Snap to establish that the competitive landscape is far wider than the FTC claims. The defense presented a securities filing from Snap that listed YouTube, Apple, and TikTok alongside Facebook as competitors, a strategic move to broaden the scope of the market.

Lies, damned lies, and statistics: Meta's case-in-chief draws sharp words

Brendan Benedict writes, "Much of the rest of the direct and cross were under seal," highlighting how much of the critical evidence remains hidden from public view, yet the available testimony suggests a coordinated effort to prove that attention is fluid, not captive. The argument is that if users can easily switch apps, the monopoly power the government alleges simply doesn't exist. This framing is effective because it shifts the burden from proving Meta's dominance to proving the ease of user exit, a much harder hurdle for the FTC to clear.

"Much of the rest of the direct and cross were under seal."

Critics might note that hiding so much testimony under seal prevents the public and independent analysts from fully scrutinizing the data, potentially obscuring flaws in the defense's narrative. However, the public portions of the trial already reveal a sophisticated strategy to reframe the competition.

The Economist's Gambit

The core of Meta's defense rests on the testimony of John List, a prominent economist from the University of Chicago and chief economist at Walmart. List did not mince words when addressing the FTC's expert, Scott Hemphill. According to Benedict, List characterized Hemphill's report as jumping around and lacking serious economic analysis, specifically criticizing the inclusion of the app MeWe in the proposed market.

Brendan Benedict quotes List's sharp assessment: "It starts as a hypothesis and it ends as a hypothesis." This is a devastating critique of the opposing expert's methodology, suggesting that the FTC's case is built on circular logic rather than empirical evidence. List then deployed a series of field experiments to test his theory of the "attention economy," arguing that the opportunity cost of time drives users to switch platforms constantly.

In one experiment, List paid participants to spend less time on Facebook and Instagram, observing where their time went. The results showed significant diversion to YouTube and TikTok, apps the FTC had largely excluded from its market definition. Benedict explains that List's point was clear: "more users leave Instagram and Facebook for TikTok and YouTube than for Snapchat, and only Snapchat is within the FTC's proposed PSN market, suggesting that TikTok and YouTube should be in that market, too."

This evidence is compelling because it relies on real-world behavior rather than theoretical models. However, a counterargument worth considering is the "cellophane fallacy"—the idea that measuring substitution from a monopoly price (or in this case, a zero-price product) can be misleading. List dismissed this, arguing that because the price is zero, the fallacy doesn't apply, but the FTC objected that this specific argument wasn't disclosed in his expert report, raising questions about the admissibility of his conclusions.

The Two-Sided Market Defense

Perhaps the most sophisticated part of List's testimony involved the concept of "two-sided markets," where the platform serves both users and advertisers. List argued that the FTC's focus on user experience ignored the advertiser's perspective, creating a "seesaw" effect where increasing ad load might hurt users but benefit advertisers by lowering costs.

Brendan Benedict writes, "Standard economic theory 'provides an ambiguous prediction' as to where the seesaw lands, so it's 'an empirical question' for study." This admission is crucial; it acknowledges that the economic reality is complex and that there is no simple answer to whether Meta is harming the market. List's simulation of a breakup suggested that ad loads would actually increase if Instagram and Facebook were separated, directly challenging the FTC's theory that Meta uses its monopoly power to suppress ad quality or quantity.

"Standard economic theory 'provides an ambiguous prediction' as to where the seesaw lands, so it's 'an empirical question' for study."

This argument is a double-edged sword. While it effectively complicates the FTC's narrative, it also relies on the assumption that the benefits to advertisers outweigh the harms to users, a value judgment that the court may be reluctant to make.

Bottom Line

Brendan Benedict's coverage reveals a defense strategy that is aggressively redefining the terms of the debate, using field experiments to argue that the market for social networking is far broader than the government claims. The strongest part of this argument is the empirical evidence of user diversion to non-Facebook platforms, which directly undermines the monopoly claim. However, the biggest vulnerability remains the court's skepticism regarding the admissibility of List's novel experiments and the potential for the "cellophane fallacy" to invalidate the substitution data. Readers should watch closely to see if Chief Judge Boasberg accepts these experimental findings as sufficient to dismantle the FTC's market definition.

Sources

Lies, damned lies, and statistics: Meta's case-in-chief draws sharp words

by Brendan Benedict · · Read full article

Big Tech on Trial is back. Your author has been tied up in a variety of other legal work for clients, so please forgive my belated write-up from Meta’s abbreviated case-in-chief, which took place at the end of Week 5 and beginning of Week 6 of trial. In the next post, we’ll cover Meta’s main expert economist, Dennis Carlton, and the FTC’s rebuttal from its expert, Scott Hemphill. But in this post, we take a quick tour through Meta’s defense, which featured fact witnesses from Snap, WhatsApp, Walmart, and Instagram, plus two expert economists, John List and Catherine Tucker.

Meta’s Case Begins with Snap.

We left off near the end of Week 5, as the FTC had just rested its case. Meta then called its first witness, Saral Jain of Snap. The entire direct was under seal. When court was open again, the FTC was already into cross. There was a rather lengthy colloquy on the admission status of a Snap earnings call, which the FTC said was pre-admitted. Kellogg Hansen’s Ana Nikolic Paul kept objecting that it was hearsay because it contained statements of Snap founder Evan Spiegel, rather than Jain. But as Chief Judge Boasberg pointed out, since the document was in evidence, that was an objection to foundation, not hearsay, so the court instructed the FTC to lay foundation. The objections continued:

MS. PAUL: Objection, Your Honor. He can't speak for Snapchat.

THE COURT: Overruled. He can.

The questions and the objections didn’t go anywhere anyway since the witness answered that he didn’t know the answers to many of the questions being asked, given his role on the infrastructure side of things.

Jain explained how Snap used Google Cloud and then Amazon Web Services in lieu of its own servers, which freed up the company from having to hire as many engineers. This was further down the line from a point we heard earlier in trial, that Instagram didn’t need Facebook’s servers to succeed because it could have just continued using cloud vendors.

On re-direct, Paul pointed out, and Jain agreed, that Snap used to run on the Google App Engine, which could not handle Snap’s growth projections. Snap also eventually concluded that a single cloud provider wasn’t good enough, so de-risked by having multiple providers, which took some time. Much of the re-direct was under seal, too.

Meta’s next witness was also from Snap, David Levenson, ...