Politics Archive

The FTC’s economic expert Scott Hemphill is back on the stand testifying about how Meta’s alleged monopoly power allows it to offer different prices to different users and across its services. For example, Meta didn’t show ads on Reels for a while, but it did show ads in other parts of the Instagram app. Hemphill compares this to what sounds like the antitrust case against Whole Foods — saying that a supermarket specializing in premium, natural foods might have high margins on its organic mangos, but low margins in areas on which it competes with many other stores, like pantry staples. He says this is similar to Meta lowering prices on the product where it competes with other apps (short-form videos) but not on the one where it dominates the market (friends and family sharing on Feed and Stories).

The judge is using Hemphill’s testimony to work through some remaining questions he has in how to think about reaching a verdict in this case. It’s too early to say what this means for his thinking, but his questions indicate that some of Meta’s arguments are at least spinning his wheels. He asks, for example, if users’ feeds are increasingly made up of posts from accounts users aren’t connected to (i.e. algorithmically-recommended influencer posts), then can time spent on the Instagram feed or stories really be used as a proxy for users’ wanting to see more posts from their friends?

Just because an alternative service can fulfill some of what a monopolist offers doesn’t make it an adequate substitute. Hemphill gives the example of Google’s search services. Sure, users can do a shopping search on Amazon or a travel search on Expedia, but these are “slivers of what it does,” Hemphill says. Boasberg’s own colleague in this courthouse found Google to be a monopolist for general search services despite this.

Boasberg posits a hypothetical: would it matter if only 10 percent of users’ time on Meta’s apps were spent on friend and family sharing — the use case the FTC says Meta monopolizes? This seems to be a key lingering question for the judge. “Does it have to encompass their core use? So, the majority of what users go to the platform for?” he adds. Hemphill acknowledges the slippery slope, but says that even at a hypothetical 10 percent of its business, that would still be “large and important and subject to monopolization, and it would still be something that we care about.”

Boasberg asks Hemphill to explain why the way users choose to spend their time when one app is down wouldn’t be relevant to show which apps are substitutes. “In fact, why isn’t that the best indicator about what is a substitute?” he adds. This is exactly the argument Meta’s made since the first day of trial: that users spending more time on Instagram after TikTok’s brief pre-ban shutdown shows users see it as a viable alternative.

Hemphill uses this to show that friends and family posts, found mostly in the Feed and Stories, are still core to the app. Over a few days in April 2022, less than 1 percent of Facebook users and about 5 percent of Instagram users who watched Reels did not also spend any time in the Feed or Stories features. “The point here is that yes, users are in addition spending time on Reels, but that Feed and Stories remain fundamental to at least part of the experience of the vast majority of users.”

Judge Boasberg asks Hemphill to “square” Meta’s claim that user interest in friends and family posts are declining with Hemphill’s claim that Meta’s apps are degrading in quality because they’re showing users less of this content. Hemphill says there’s evidence users really do want to see posts from their friends and they show “frustration that they don’t get it.” So if Boasberg finds users do want this kind of content and Meta isn’t serving it, that can show it’s degraded the quality of its apps.

New York University professor Scott Hemphill takes the stand, fleshing out the economic argument behind the government’s market definition, and how Meta has — in his expert opinion — harmed competition and consumers with its monopoly power.

In a 2022 document, Meta found that Instagram was experiencing “downward pressure” from TikTok “on interest content consumption.” But, it added, “that same pressure is less evident on consumption of friend content, which begs the question: Are we putting enough focus/energy into defending/evolving key jobs we continue to be hired for?”

The FTC makes the point that because Instagram and Facebook have already added most of the roughly 250 million potential eligible users in the US, its user growth rate looks slower, even though it’s still adding users overall. But Schultz says this is exactly why Meta competes with apps like YouTube and TikTok for users to spend more time on their own apps — doubling down on Meta’s argument in the case. “Time is the thing people are all competing for because people are multi-homing, they’re using every app.”

Schultz says he “quibble[s]” with the idea that Instagram had good engagement pre-acquisition. “I didn’t think it had the highest engagement when I got to see the numbers when it was inside,” he says.

The FTC’s theory that Meta’s power is only constrained by these two apps is “ridiculous,” Schultz testifies, concluding his examination by Meta’s attorney.

Creating a social graph of users’ friends, likes, and interests isn’t as meaningful for social apps as it used to be, Schultz says. That’s thanks to advances in AI, which has made it easier to predict and serve content to users they might not have even known they’d like. The FTC says Meta’s network effects and extensive mapping of users’ connections makes it difficult for upstarts to dislodge it.