We've now started with Epic's main expert witness Dr. David Evans. Some background in today's MT: https://www.politico.com/newsletters/morning-tech/2021/05/10/white-house-backs-global-anti-online-extremism-effort-795182

TBH one of the most interesting things about this week will be the Evans v Schmalensee showdown. For the past 20ish years, they've mostly co-written work. But in this case, they are on opposite sides.

Evans will actually be on the stand twice, next on rebuttal to push back on Schmalensee's "digital game transaction" market but he's already trying to undermine it citing Amex.

In American Express (the 2018 Supreme Court case), the court found a two-sided market for transactions. One side was merchants of all types, Evans says, everything from hardware stores to restaurants. The other was customers who wanted to pay with credit cards.

Evans argues that Amex offered credit card services, not a specific type of transaction depending on the merchant.

Smartphones are different from game consoles, Evans argues because game consoles aren't mobile (ie you can't carry them around) and they don't have cellular connectivity, only WiFi or plug-in internet capabilities.

Evans said he looked at the top 50 games on iOS both by downloads and revenue. The majority of those games aren't available on game consoles.

The average consumer spends 1 hour a day on a personal computer and 3 hours a day on a smartphone, Evans says (that's just personal use, he says, and doesn't include work)

Evans describes a "general purpose operating system" as something that different types of apps can be built. He gives Windows, Mac, iOS and Android as types of general purpose OS.

A special purpose operating system is one that only one type of app can generally be built. He gives game consoles as an example.

As Evans starts to talk about selling game consoles below cost, Judge YGR pops in. "You’ve seen the records from Microsoft and Sony and Nintendo that prove those facts," she asks

“I don’t know this is based on your understanding," she says, "or whether I’m actually going to be able to verify this theory I’ve heard multiple times.”

Evans says he has seen a document from Microsoft that discusses the profit margin for xBox and that its sold below cost, and that document includes Microsoft's projections about the profit margin for Sony and Nintendo.

Evans describes the idea of a "foremarket" and "aftermarket" using a printer as an example. A consumer buys a printer (foremarket) and will then have to continue buy paper and toner (aftermarket) to continue to use it.

In this case, Evans argues that the foremarket is the smartphone operating system -- either Apple iOS or Google's Android. The aftermarket is the apps that will work on those operating system, both on the particular device and future ones bought with that OS.

Game consoles aren't "substitutes" for smartphones in the antitrust sense. Evans says that almost everyone who owns a game console also has a smartphone. They wouldn't if a game console was an adequate substitute for a smartphone

Same for computers. They aren't adequate substitutes for smartphones because they aren't as portable and don't have the cellular connection.

Apple has "substantial market power" in smartphone operating systems, Evans says.

The smartphone operating system market is a duopoly controlled by Apple's iOS and Google's Android, Evans says. Since 2013-2014, iOS and Android have about 100 percent share of the market. "There are really only two operating systems in play: iOS and Android"

On a global basis (excluding China), Android has about 60 percent share and iOS has 40 percent. Within the U.S., it's closer to 50 percent, Evans says.

Switching costs are high for consumers. In addition to buying a new phone and rebuying/downloading all the apps, some data won't transfer over. For example, iMessage texts won't transfer, photos in iCloud. Related devices, like Apple Watch, also won't work on an Android device

While consumers tend to use only one system, developers have to use both in order to reach all users. "Developers need to be where customers are. They have to create iOS apps to reach iOS users and have to have an Android app to reach Android users,” Evans said.

"To the extent the users don’t switch much, that means the developers pretty much have to keep" creating apps for both systems, he said.

Judge YGR says she has a few questions before the lunch break. When in Evans' view, did the market consolidate into the duopoly?

Evans says in his view it was really 2010. At that point in time, developers stopping working on apps for other OS and started focusing very much on iOS and Android. "They knew that Android and iPhone were going to take over so they shifted their development efforts"

Not all consumers made the switch by then. There were still people using Blackberry because they liked the keyboard or some of the other smartphones on the market "but by 2010 there was a dramatic movement of people over to these new devices."

And I'm just going to leave this right here for the 40-minute lunch break.... https://www.politico.com/news/2021/03/16/google-files-mobile-search-market-475576

Back from lunch. Evans resumes his testimony.

In-app purchases are separate from app distribution, Evans says. (This is actually not controversial. Apple agrees that its possible to have alternatives to in-app purchases since it does for physical goods. It's point is that it has never allowed it for digital goods)

As an example, he says you can sign up for Netflix on your computer. But if you want to watch on your phone, you have to download the app. "That’s the service provided by app distribution."

It's very hard to assess lifecycle costs of joining iOS or Android, Evans says (this is something Apple's witnesses acknowledged. They don't try to track how much an iOS user spends over the life of a device)

This is an important part for the antitrust analysis. Epic's case is modeled after one brought against Kodak over photo copiers. The Supreme Court found that antitrust law covered Kodak's conduct because users were "locked in" and didn't have alternatives for repair parts

Evans is trying to show that Apple customers are really locked in once they pick the Apple ecosystem and don't have all the information they really need upfront to make a determination about the overall costs

Evans looked at anonymized Fortnite user data. What he found is that most Fortnite users just use one platform, either a smartphone, a console or PC and don't switch between them.

Overall, 82.7 percent of Fortnite players played the game only on a single platform, Evans said.

Judge asks if he can say how many iOS users who only play on their iPhone. About 60% of iOS users only play on a smartphone. About 10-15% primarily play on game console/PC and use smartphone sometimes. The remaining 25-30% usually play on iPhone but sometimes on console/PC

Evans did an analysis looking at 10 weeks before and 10 weeks after Fortnite was removed from the App Store. To compare iOS users, he made a control group of "twins" for each iOS user, a person who had similar play habits but always played console/PC.

Evans estimated that if Apple hadn't removed Fortnite, iOS users would have played 56.3 minutes (that's how long the control group played during the period).

Some iOS only users moved to console/PC once Fortnite wasn't available. He found a 16.7% replacement rate. "That shows a very low level of substitution. More than 80 percent of the time" players didn't find a substitute.

Some percentage of those iOS users who might have gotten a game console or PC anyway. Evans said he looked at the 42 weeks prior to Aug. 13, 2020 to figure out how often iOS users were buying a game console/PC after awhile and moving to that

After accounting for users who were likely to have gotten a console anyway, Evans says the replacement rate was only 3.1% -- meaning very few iOS users actually switched platforms once Fortnite was removed

Evans is now talking about a SSNIP test, which stands for a Small but Significant and Non-transitory Increase in Price

YGR: "Did you do any analysis about whether users moved to a comparable game?"

Evans said he didn't have information on what account holders did outside of Epic. He suggests that some iOS users probably did switch to a different game. "How does that affect" your analysis, YGR asks.

"If we learned iOS users who couldn’t play Fortnite app, decided to play Roblox or a casual game on iOS, that’s consistent with my market definition. They are substituting for another iOS game," Evans said.

"So some other gaming company could have benefited," she asks. "That is possible," Evans says. iOS-only users one might have substituted to a different game. "That wouldn’t affect any of my conclusions."

Evans tried to measure whether developer substitution was possible, meaning whether if Apple irrespective of the litigation decided to increase commissions, whether a developer like Epic could just eschew iOS and move to console/PC.

"Not nearly enough revenue moved over to game consoles and PCs," Evans said for Epic to walk away from Apple. (Evans didn't offer any numbers for this. I suspect it's sealed. Epic's lawyers said some of the testimony would involve numbers that are sealed)

Moving on to market power, Evans says “my opinion is Apple has monopoly power in the iOS app distribution market.”

One of the reasons he has concluded that is because of Apple's "high and persistent profit margins." Evans said he calculated Apple's profit margins based on data the company provided in the lawsuit. They are showing the numbers to the judge but not anyone else (sadly)

Though YGR just noted that “all sealing orders are subject to change if I decide, in writing my order, I need to to make the reasoning transparent.”

The numbers come from profit and loss statements that Apple created to present to its board. There was also a set of P&L statements that came from Tim Cook's papers. Evans says the statements relate only to the App Store, not other aspects of Apple's business

This contradicts Apple, which has repeatedly said that it doesn't break out profitability for the App Store as a stand-alone entity.

“In a competitive environment, I would have expected over time the margins would decline," Evans said.
"Do you see a decline in profit margin on this slide?" Epic lawyer Gary Bornstein asked.
"No," Evans said.

Evans compared Apple's profit margins to other online retailers including Alibaba, eBay, Mercadolibra, Rakuten and Etsy. “Apple’s profit margin was vastly higher than the benchmark group of companies, even the most successful of these companies. It’s a multiple” of the group

Evans says he determined that “Apple’s restrictions harm competition” both for app developers and consumers and have led to higher prices, poorer distribution and a slower pace of innovation.

Evans refers to an email from Apple's Schiller in which he suggests lowering the commission rate to 20-25% or capping the App Store's profit at $1 billion.

Evans said he used the P&L statements he got from Apple for the profit and costs of the App Store in 2019. From separate documents, he found the R&D spent by Apple for iTunes in 2019.

The App Store is within iTunes in the Apple organization and acounts for 2/3 of the revenue, he said. Evans said he compared used that ratio to estimate the R&D for the App Store.

Without giving the number, Evans say the App Store's R&D to sales ratio "is very low" and is lower than the median R&D to sales ratio of the benchmark companies.

Another thing I never really expected from this trial: an economist trying to explain to a judge how Tinder works.

We've wrapped for the day. Evans will resume his testimony Tuesday morning.