Lauren Feiner

This was true even as the app built out its discovery engine, which powers algorithmic recommendations of content from users who are connected with one another, Alison testifies. The FTC is using Alison’s testimony to establish that in the past few years and through today, a sizable number of users still come to Facebook to connect with their friends, and the company recognizes this even as it expands into other use cases. Still, Alison says, Facebook now thinks about facilitating connections as inclusive of people users don’t know in real life, including content creators.

Not everyone who joins Facebook these days does so to find their friends on the service, Alison says. He testifies there’s a “growing” number of people who join the app with no friends, and don’t see any friend content on the app. Meta has argued that connecting with friends is an increasingly smaller portion of what users come to its apps for. The FTC says that as an absolute number, there’s still a sizable number of users who do want to connect with friends, and Facebook and Instagram are virtually the only games in town.

Alison downplays how much users care about connecting with their Facebook friends versus other potential connections, like content creators. “I have friends I haven’t seen in 30 years or I met once at a party that I don’t care about,” he says.

It’s not enough to look at what users say they want from Facebook, Alison says, you also have to look at what their actions tell you about what they want. Just because users say in surveys they want to see more friend content, doesn’t mean that’s how it plays out. “When people actually got more friend content on Facebook they visited Facebook less,” he testifies.

Alison is expected to be one of the FTC’s last major witnesses in its case-in-chief. The FTC’s lead attorney Dan Matheson is driving home the point that Meta knows that people want to see posts from their friends when they come to Facebook, by looking at a 2021 internal presentation where it found in a survey that “3 of the 4 top user pain points were related to friend content.”

During redirect, Hemphill testifies that some of the points Hansen made about his analysis weren’t relevant to his findings of Meta’s alleged anticompetitive behavior. He counter’s Hansen’s charge that he thinks he knows better than the business people at Meta. “They’ve got their expertise and I’ve got mine,” he says.

Hemphill clarifies that he did disclose to Meta during his deposition over a year ago that he and former Biden official Tim Wu had a “conversation” with the FTC in 2019 about investigating Meta, but the company never asked him for a copy of the presentation. He says he wasn’t compensated for it, and wasn’t retained by the FTC for another two and a half years — during which he also didn’t lobby the regulator to bring a case. He says that public reporting of Meta’s conduct “seemed like the kind of thing that we’d been worried about in our academic work,” but that he hadn’t formed a view at the time that its acquisitions of Instagram and WhatsApp were anticompetitive.

Back to friendlier-questioning from the FTC, Hemphill says that while ad quality might be improving, Meta is behaving like a monopolist in what it chooses to do with those quality gains. In a more competitive environment, Hemphill says, Meta would pass on those gains to consumers by letting them enjoy the same amount of ads, but at a higher quality. Instead, he says, Meta dials up the knob to show them more ads.

This is how Hansen closes what’s been a tense cross-examination of Hemphill. Meta has used the exam to try to discredit the FTC’s key economic expert, and Hansen reiterates that Hemphill had “preconceived” notions about the case, calling back to the 2019 “roadshow” he participated in to convince regulators to investigate Meta. Hemphill has been on the stand since Monday afternoon, and the FTC attorney Krisha Cerilli is back for redirect questioning.

This is what Hansen asks the FTC expert after showing a chart that shows younger cohorts spend a higher percentage of time on the platform engaging with friends than older cohorts, yet they get served fewer ads — seemingly contrary to Hemphill’s claim that Meta taxes users who enjoy friend sharing more. Hemphill says the point is “apples and oranges” and that the older cohort makes up 60 to 70 percent of the user base getting higher ad loads. That prompts Hansen to ask whether he thinks Meta discriminates against such a large portion of its users, and Hemphill says it’s “a standard understanding of price discrimination.”