Sunday, January 2, 2022

Opening brief due this month in Epic's appeal of Apple's antitrust win: not an insane long shot, but definitely an uphill battle after all that went wrong

In a few weeks' time, Epic Games will file its opening brief with the United States Court of Appeals for the Ninth Circuit, trying to turn around a litigation in which Apple undoubtedly won the first round. I expect to focus very much on patent litigation in the coming weeks (there may already be some interesting filings targeting Apple tomorrow), but I would like to share some thoughts ahead of Epic's brief, given the importance of that case to app developers like me.

I've read the 185-page decision three times. The first time--right after it came down in September--I was thoroughly disappointed and suspected (but this view has changed!) the judge of simply hedging her bets for an appeal. By finding for Apple on almost all counts (especially on 100% of those that really mattered), while handing Epic a consolation prize under state competition law, she's highly unlikely to be reversed across the board, as the appellate outcome will most likely be binary. Finding for Apple makes a reversal of the key parts much less likely than if she had sided with Epic, but even if it happened, it wouldn't be as bad as if she had acquitted Apple 100%: she would just come across as having erred on the side of caution. Conversely, if the most important parts are affirmed and the appeals court determines that the state UCL part cannot fare better than the Sherman Act claims, it's only a narrow (contrary to disinformation by some people) injunction that goes down the tubes.

There was something she said early on--at the preliminary injunction stage--that suggested to me she was going to optimize the decision for the inevitable appeal. She wanted the parties to agree to a jury trial (which Apple was seeking on its counterclaims) rather than a bench trial, and she said that the appeals court doesn't really afford much deference to factual findings by the lower courts. That was a clear allusion to FTC v. Qualcomm--and while Judge Lucy Koh decided in that case what she thought was right based on the evidence (by now she's on the appeals court, which I'm very happy about), Judge Yvonne Gonzalez Rogers ("YGR") really would have preferred to let a jury decide, as appeals courts don't overturn jury verdicts as easily as judicial findings.

Just like the mixed outcome is now particularly appeal-proof, it will also serve to discourage other parties from insisting on a bench trial when Judge YGR prefers a jury trial. This, again, is an effect of what happened, but I don't think it's a factor that influenced her decision-making any more than a remark about lawyers making a lot more money than judges (which was interesting as a former colleague of hers--former New York District Judge Katherine Forrest--became a Cravath partner and represents Epic against Apple). The same applies to her misgivings regarding Epic's impulse purchasing-centric business model and the staged and orchestrated crisis surrounding Fortnote Mobile.

After reading the decision three times and reflecting on it, I felt that Epic has no one to blame but itself. It is an extremely honest decision by a judge who--as far as I can tell--really wanted to reach the appropriate conclusions and would have been prepared to find for Epic if she had seen a solid basis for that.

Judge YGR had not only Epic but also Apple all figured out by the end of the trial. Her examination of Tim Cook removed any residual doubts: developers are at Apple's mercy, and whatever concessions Apple made to developers were driven by litigation and regulation, not competition. So the outcome is nonsensical in terms of there being no reasonable doubt that Apple has monopoly power over developers. I mean, we all know that Apple treats developers as serfs, squeezes them like lemons, and tyrannizes them. Just a few days ago, a former Tumblr developer described Apple's app review process as perplexing, random, discordant, and even "asinine." But it is not at all nonsensical when considering that a finding of an antitrust violation depends on a chain that is as strong as its weakest link. It's a step-by-step analysis, and Epic failed at certain hurdles in Judge YGR's more than reasonable opinion.

If one wants to take a cynical perspective on why she arrived at that decision, one could in all fairness also be cynical about Epic and suspect that a very substantial part of its motivation may simply have been to make Apple look bad, with a view to bringing about regulatory action here and there, and especially legislative change. If Epic lost this litigation, but Apple's death grip, or "vice-like grip" (I believe they meant "vise-like grip") that the UK's competition regulator identified), came to an end, so what? As someone told Epic CEO Tim Sweeney on Twitter, "lose the battle, win the war."

Toward the end of the trial, Judge YGR said she takes with a grain of salt the anecdotal evidence both parties present, and mostly cares about the aggregates: about hard facts and statistics. Some of the evidence Epic adduced may have looked more anecdotal than it actually was. Apple's intentions in some regards do matter in an antitrust context, and maybe the judge could or even should have given them more weight. But in general I have to say that some of what Epic tried in court would have been far more likely to influence a jury than a judge--if a judge (in the gatekeeper role) had even allowed to present some of that evidence as opposed to throwing it out for being more confusing than having probative value.

Epic got the most reputable U.S. law firm involved, and how they managed the litigation under those time constraints was superhuman. Maybe a less conservative firm would have done things that Cravath wouldn't consider. Maybe some other firm would have dissuaded Epic from its overly dramatic "Project Liberty." And Epic's jurisdictional choices didn't really work out. Germany actually has the most favorable statutory law of all major markets (with special gatekeeper rules)--but Epic focused only on English-speaking countries. At the end of the day, Epic made the key decisions, and it was Epic's job to provide its counsel with the facts that it takes to win a case like that.

Not only was the case on an unbelievably tight schedule (because of the already-pending class actions) but one may reasonably feel that the requirements Judge YGR outlined in her ruling couldn't possibly have been met without another three or more trial weeks (and Epic actually did ask for more time). In particular, it appears that Epic would have needed much more of an opportunity to address issues surrounding Apple's security pretext and the shortcomings of the app review process. In the latter context, Epic would have had a benefit from talking to some of the "little guys" like Kosta Eleftheriou, who might have been a first-rate trial witness.

Before the appellate proceedings have been begun in earnest, Epic has already "burned" its Coalition for App Fairness. Future amicus briefs will have to be filed by individual developers, and the CAF can orchestrate such campaigns, but it's lost its credibility as a "friend of the court."

If one compares what Epic has achieved against Apple so far to what its key CAF friends Spotify and Match Group (Tinder) have done, Epic has made far more noise than all others combined, but Spotify and Match have been hugely more effective because they achieved breakthroughs on which they and others can expand. Spotify's case is special because it involves direct competition with Apple Music, but a Statement of Objections (SO) by the European Commission is a tremendous accomplishment. Too bad that Spotify's top lawyer is now working for a company that is on pretty friendly terms with Apple. Match Group's Dutch success may appear small at first sight: a tiny market, and just about dating apps. The decision is based not only on Dutch but also EU law, however, and what applies to dating apps can subsequently benefit other complainants. Arguably, what Match has achieved has the greatest potential for expansion, and even though Match has deep pockets, it pretty much used "guerrilla warfare" against Apple. So Epic may ask itself, after presumably spending many tens of millions of dollars on legal fees, losing lots of iOS revenues (and counting), and annoying many of its customers, why it holds next to nothing in its hands against Apple.

The name of the game now is for Epic to focus. There are lots of things in Judge YGR's ruling that one can disagree with. I identified a couple of things that I didn't consider correct, such as that Apple's conversion of IAP prices charged in the U.S. is actually not like they'd charge the equivalent in foreign currencies--there's often a major discrepancy. And there are plenty of typos in that decision, which makes me comfortable knowing that my blog posts are rarely typo-free. But Epic needs things that can really change the outcome. That can turn a defeat into a victory. And that's where it becomes very, very, very complicated.

At this stage, they already have to look even beyond the Ninth Circuit and think about their strategy in the Supreme Court. They can't relitigate everything now. They need one or two game-changers. If they don't have a clear-cut focus, they can still fill 100 pages or more with rhetoric, even including advocatic hyperbole, but it won't be likely to give them a better outcome.

I looked at the decision from the perspective of what might be a promising appellate strategy. To be honest, I don't think there is a "promising" one per se: Apple is more likely than not to defend what it has, and Epic will probably lose whatever little it accomplished, as a California UCL ruling that is at odds with a finding of no federal antitrust violation is disfavored by local Chavez case law. But relatively speaking, there is one area in which Epic may still be able to turn things around, and that's the market definition.

Tim Cook's examination by the judge showed very clearly (as does other evidence and testimony) that Apple does have unfettered monopoly power over iOS developers. That is, by the way, also a problem because Epic may find it hard to get too many other developers to support its appeal through amicus briefs, as they are all afraid of the most powerful tyrant the industry has ever seen. So the question is whether Epic can convince the higher courts that the obviously correct bottom line can be reached on the basis of a legally solid derivation. And that's where it all begins with market definition.

Epic's proposed single-brand market definition involves an operating system foremarket and an app distribution aftermarket. The least convincing part of Judge YGR's decision is where she adopted Apple's nonsensical distinction between devices and operating systems. There is a one-to-one relationship (as programmers like me would say) between iPhones and iOS installations. I could easily see how Epic would convince both the Ninth Circuit and the Supreme Court that its foremarket made sense while the related part of the district court's judgment did not.

With respect to the aftermarket, the judge was unconvinced of what Epic's experts said, though Apple's expert testimony also had shortcomings. I will read with great interest how Epic will try to convince the appeals court that the judge either made a legal error in what she concluded from the testimony or that her fact-finding was so wrong that it should be reversed. It would take a post longer than this one just to speculate on that, and that's why it's better to wait and see what Epic will do.

A single-brand market definition would be a game changer. However, Apple would argue that some of the other findings and conclusions in the judgment would make Epic lose anyway. So Epic will argue that even under the court's digital mobile game transactions market definition, it should win--but I doubt that Epic can successfully appeal the ruling unless the market definition is narrowed, just like some patent appeals are pretty pointless without a modified claim construction.

Under a single-brand market definition, Apple's market share would by definition amount to 100%. Possibly, the appeals court might find that such a major change warrants a remand, a complete do-over on that basis. I believe Apple's strongest argument against monopoly power even under a single-brand market definition would be the following passage:

"[A]s defined by the Court, in terms of distribution of mobile apps, multiple avenues do exist to distribute the content to the consumer. Distribution can occur through web apps, by web access, and through other games stores. This doctrine does not require distribution in the manner preferred by the competitor, here native apps. The availability of these other avenues of distribution, even if they are not the preferred or ideal methods, is dispositive of Epic Games’ claim. The doctrine does not demand an ideal or preferred standard."

This is in the specific context of Epic's essential-facility claim under Section 2 of the Sherman Act. But unless those other avenues of distribution are clearly found unacceptable (which they are, and Epic's CEO explained the issues with web apps way better on Twitter than his company did in its proposed findings of fact and conclusions of law in the Apple case), the web avenue would still defeat Epic's case even under a single-brand market definition.

The single-brand market definition would also be needed, at least in my view, to keep Epic's case against Google alive. However, the judges will all be aware--as was Judge YGR--of the potential effect on other platforms. All walled gardens could be sued, though the facts are different and the outcome wouldn't necessarily always be the same. And a court might also just feel that Epic couldn't afford to go against all walled gardens at the same time, which still doesn't mean that Apple's conduct is lawful.

As I just mentioned Google, I'd like to share one last thought at this stage. The reason for which Google defended itself against Oracle's copyright case--even though it was IP, not antitrust, and Google was the defendant, while Epic is a plaintiff--had a lot do with judicial sympathy for app developers. In the end, the Supreme Court saw that Google had virtually the whole software developer community behind it. While all developers access APIs, most of us never clone an API the way Google did with its incorporation of Java into Android, but some didn't--and some didn't want to--understand that difference. So in the end the majority of the justices made a fair use finding in Google's favor that even exceeded what Google considered a realistic way to win (Google's argument was really about deference to the jury). And let's not forget the Pepper case, though it is a consumer class action.

The biggest difference is, however, that hardly any software developer had to fear any retribution from siding with Google against Oracle. It will be a far--FAR--more difficult decision for developers of all sizes to go against Apple in order to support Epic, though I want to be clear that I for my part feel that Apple accepts my way of speaking out on the issues. Still, I don't blame anyone for being afraid, or even very afraid. If Epic wants to benefit from the kind of dynamics that enabled Google to win (on a basis that was totally inconsistent with the entire body of U.S. fair use case law), it will have to come up with a better plan for garnering support from developers of all sizes than getting an amicus brief from the CAF.

Here, again, judges have the opportunity to help developers. The key to the kingdom is a single-brand market definition.

Share with other professionals via LinkedIn: