Monday, April 18, 2022

Demarcation between Section 1 (concerted action) and Section 2 (unilateral conduct): what Apple, Epic Games, and their amici tell the Ninth Circuit

No matter how hard I try, I can't think of a more fundamental question of U.S. antitrust law than where to draw the line between Section 1 (concerted action, cartels) and Section 2 (unilateral conduct, abuse of market power) of the Sherman Act. In the Ninth Circuit appeal of the Epic Games v. Apple judgment from the Northern District of California, that one is a hotly contested issue. Unless the appeals court elects to sidestep that part, the related decision will affect countless cases down the road. That transcendental relevance warrants taking a closer look, though I wouldn't view it as a game changer in the App Store antitrust case if Epic prevailed on that one: Judge Yvonne Gonzales Rogers ("YGR") held that Section 2 applied, but that Epic would have lost under Section 1 anyway.

To continue with the Section 1/2 part, you can click here. If you're interested in why market definition is a much clearer case for reversal and remand than the choice between Section 1 and 2, and why that part may lead the appeals court to afford Judge YGR only limited deference, please read on.

Ninth Circuit may see how terribly wrong the district judge got the single-brand market part

A single-brand market definition--as advanced by Epic--would make a meteoric impact. It's hard to see how Apple could defend itself then, and even if it could in this particular case, various other App Store antitrust cases would become an uphill battle for Apple. Even the tying part eclipses the Section 1/2 dichotomy. Tying is subject to Section 1 anyway, so the key question there is unrelated to the one of broader Section 1 applicability. In order for Epic to prevail on tying, the appeals court would in a first step have to agree with the DOJ that even something that is not sold or licensed separately can constitute a market.

There is light and there is shadow in the district court's ruling. While Judge YGR's analysis of Section 1/2 applicability reflects a complete understanding of the issues involved (and even those who disagree with the result can't claim she didn't reasonably come down on Apple's side), her disagreement with the foremarket part of Epic's single-brand market definition is triply wrong. She couldn't have done a poorer job than getting the law, the economics, and the technology wrong in the same context:

  • wrong on the law because Kodak is all about a competitive (multi-vendor) foremarket, with only the aftermarket being a single-brand market;

  • wrong on the economics because the question is whether a company's conduct in the aftermarket is disciplined by blowback in the foremarket, not whether that company has market power in the foremarket; and

  • wrong on the technology because Apple doesn't have multiple "operating systems" in connection with the App Store (just iOS for the iPhone).

None of the above three points is a matter of opinion. Those are verifiable facts.

It would have been counterproductive or at least risky for Epic aggressively to expose Judge YGR's triple failure in that context. Circuit judges don't like it when an appellant tries to discredit a judge, no matter how incompetent or inexperienced in a certain respect or field a district judge may have been. The appeals court's task is to identify where the lower court must be overruled, not to judge someone's competence or analytical capacity. That's why Epic contented itself with just focusing on the legal question of whether there can be a market for something that isn't sold separately (by the way, the DOJ supports Epic not only but also on that one, and I'm still puzzled by Apple's explicit admission that iOS competes with Android). The appeals court will need to reverse Judge YGR on that part or Epic's single-brand market as well as its tying-related market definition will fail anyway. But once that part is reversed, and assuming the appeals court remands on that basis, Epic can still--if necessary--help Judge YGR get the law, the economics, and the technology right next time. The first time around, Epic's lawyers may have vastly overestimated her ability to read and understand Kodak as well as Epic's pleadings without a lot more hand-holding from counsel.

While Epic had reasons not to expose Judge YGR's triple error on the foremarket part of the single-brand market theory, I still hope that the Ninth Circuit judges and their clerks will figure it out. Once one understand just how bad that part is (0 out of 5 stars; see my posts on a legally and factually nonsensical sentence and a pointless word in the market definition the court settled for), one must be rather skeptical of each and every part of that decision. An appeals court that doesn't take a judgment by the court below too seriously is more likely to just reverse across the board (and to resolve as much of a case as possible prior to remand, at least by way of crystal clear instructions).

Judge YGR was torn between Section 1 and 2

Part of the reason I described Judge YGR's Epic Games v. Apple opinion as an "extremely honest" effort to get things right is that the part about whether Epic could base its entire case (not just the tying part) on Section 1 reflected how she was--rightly--struggling with that tricky question.

After denying that Apple's dealings with developers (the Developer Program License Agreement (DPLA)) constitute concerted action, she acknowledges the existence of "potential conflicts with the goals of antitrust law given this narrow view." And she is spot-on that the case law on tying suggests Section 1 even applies when a seller exploits its control over a product and "[t]he buyer plays no role beyond purchasing the goods under conditions set by the seller."

She then goes on to note that exclusive dealing claims also come down to "the buyer passively accept[ing] conditions set by the vendor" (like app developers just signing the DPLA as a contract of adhesion as opposed to an individually negotiated agreement). And she recalls that the Amex case "involved an anti-steering provision as a vertical restraint imposed by American Express on merchants [which] accepted the provision as a condition of dealing with American Express without further involvement."

The difference between her thoughtful, balanced, and holistic take on Section 1/2 and her triple error concerning the foremarket is like day and night (which makes the flawed parts even more regrettable).

In this post I'll also discuss tying, exclusive dealing, Amex, and a lot more that needs to be considered in the Section 1/2 context. The district court's judgment provided a perfect starting point.

The two camps in this case

Epic got tremendous support from the DOJ, 35 state AGs, academics including Herbert Hovenkamp, Microsoft, and NGOs. By comparison, Apple basically just has its own astoturfers, Google, and Roblox. As for Apple's astroturfing, it's recently been just absurd.

Epic's broad application of Section 1 is supported by the DOJ and the state AGs. As Apple rightly noted in its responsive brief, those governmental entities will find it easier to enforce antitrust law on that basis. But that doesn't necessarily make their arguments wrong.

Microsoft's brief focuses on market dynamics and tying. The professors don't advocate broad Section 1 applicability, though they do note that Amex "involved Section 1 of the Sherman Act," a suggestion also made by Epic and others (and which I'll address further below).

Two antitrust law professors--Amitai Aviram and Douglass Ross--support Apple against Epic's Section 1 argument with an amicus brief that I don't agree with in each and every respect, but which definitely makes some very good points (this post continues below the document):

22-03-31 Two Profs ISO Appl... by Florian Mueller

The "Section 1 or 2" question is not going to be put to a vote among amici curiae, but it is interesting to see who sides with whom on what basis.

Copperweld v. Independence Tube: "independent centers of decisionmaking"--a term that still lacks clarity

It's been almost four decades since the Supreme Court explained in Copperweld that the law "treat[s] concernted behavior more strictly than unilateral behavior" because concerted activity (Section 1) "deprives the marketplace of the independent centers of decisionmaking that competition assumes and demands." (emphasis added)

Unfortunately, the Copperweld fact pattern was on one end of the spectrum. The case was about "coordinated activity of a parent and its wholly owned subsidiary." It's a no-brainer to deny independent decision-making in that kind of scenario. Nevertheless, three justices dissented from Copperweld because they didn't want a new per se rule like that (they thought the conduct at issue in that case was so outrageous that Section 1 should apply).

Very much unlike the Copperweld dissenters, the problem I have is the lack of clarity in terms of fact patterns that absolutely positively do not amount to independent centers of decision-making. It seems that the only clarity we have is that a parent company and its wholly-owned subsidiaries don't meet that definition. That's like if the only answer to the question "what's definitely an even number?" was that 2 is certainly even, but we were left in the dark with respect to all other numbers.

The two camps' doomsday scenarios

As in virtually any delineation debate, either camp here suggests that the other side's logic would render the distinction meaningless.

Apple and its academic amici rightly reject Epic's focus on the statute's reference to "every contract." Sure, "every" means "any" means something maximally broad. But the Supreme Court has explained, such as in its 2010 American Needle v. NFL decision, that a literal interpretation doesn't make sense as it "could be understood to cover every conceivable agreement, whether it be a group of competing firms fixing prices or a single firm’s chief executive telling her subordinate how to price their company’s product." In the end, everything would fall under Section 1, because--as Apple's academic amici explain--commercial activity always involves contracts (even if just go shopping at the supermarket, you conclude a purchase contract, even though not in writing).

My own reading of the statute is closer to that of Apple's camp: I see the term "contract" as being the first element of a list, and attach some importance to that context as well as to the fact that it's not an absolute "every" but qualified by "in restraint of trade or commerce."

It's a bit harder--though not impossible--to argue that an extremely narrow application of Copperweld would theoretically let Section 2 absorb Section 1. In practice, there is hardly a scenario in which the decision-makers have equal negotiating power. Given the complexity of negotiations, it's actually inconceivable that two or more parties could ever do a deal--such as forming a cartel--without someone being more powerful and/or someone else being needier. If any imbalance--no matter how small--meant that the decision-makers aren't "independent," antitrust plaintiffs would be left with only Section 2, a standard they very often couldn't meet.

Where I clearly disagree with Epic's camp is that it would be the end of the world if Section 1 didn't apply to contracts unilateraly imposed by a dominant player on other parties. There would still be Section 2, though some actors would then get away with their behavior who might otherwise be liable.

I have a far greater conceptual problem with an application of Section 1 that may make parties liable for "concerted action" even though they just do what someone with enormous market power or other leverage forces them to do.

Product-definition issues: tying, bundling

Apple and some of its amici argue that Epic v. Apple is ultimately all about product definition. Some even say that the DPLA is "unilateral conduct" because it's a corollary to Apple's product-definition decision to offer a product with a closed ecosystem, a walled garden--regardless of whether it's formally not a unilateral contract because app developers must sign it, too.

This is a slippery slope for Apple. As Judge Gonzalez Rogers noted in her judgment, tying and bundling do fall under Section 1. Well, tying and bundling are very clearly product-definition decisions...

Unless the Supreme Court (be it in this case or another) or Congress (which is, however, rather inclined to strengthen antitrust enforcement) makes a sweeping decision, we'll have to live with this problem that Section 1 is partly applied to unilateral conduct. I don't like that fact at all. The question is now whether Epic can or cannot be denied Section 1 applicability (beyond its tying claim) given that the case law is what it is.

Exclusive dealing arrangements may or may not be due to a vendor wielding so much market power that the buyer has no choice.

Did Amex answer the question?

The fact pattern in Epic v. Apple has a lot in common with tying and bundling cases, but even more so with the two-sided market in Amex (which doesn't mean that Apple should win only because American Express did).

Put another way, if it looks like a duck, walks like a duck, and quacks like a duck, isn't it a duck? If it's somewhere in the triangle between tying, bundling, and the Amex-style two-sided market, isn't it simply a clear Section 1 case?

The DOJ suggests so:

"The Supreme Court nonetheless has repeatedly decided Section 1 challenges to vertical restraints embodied in express contracts without pausing to consider whether the counterparty embraced those restraints. E.g., Ohio v. Am. Express Co., 138 S. Ct. 2274, 2282 (2018) (Amex) ('Amex’s business model sometimes causes friction with merchants')"

On this one, my own views are closest to the conservative and cautious approach by the two antitrust law professors supporting Apple on Section 1:

"In particular, Ohio v. Am. Express Co has little precedential value for examining whether the licenses underpinning Apple’s closed system reflect concerted action with developers. 138 S. Ct. 2274 (2018). Importantly, the courts provided only cursory analysis of whether American Express’s commercial arrangement with merchants constituted concerted action. The district court concluded summarily that concerted action existed because the provisions were 'contained in American Express’s card acceptance agreements with its merchants.' United States v. Am. Express Co., 88 F. Supp. 3d 143, 167 (E.D.N.Y. 2015). The Department of Justice further observed in its brief that American Express’s antisteering provisions were “only rarely subject to negotiation.” Brief for the United States as Respondent Supporting Petitioners, Ohio v. Am. Express Co., No. 16-1454, 2017 WL 6205804, at 6–7 (Dec. 7, 2017). In short, the record of whether the arrangement with merchants reflected concerted action was not fully developed." (emphases added)

Let's look at it this way: at least Amex didn't elaborate on Section 1 applicability to the effect that the doomsday scenario of Section 1 absorbing Section 2 would have become a reality.

Where do we go from here? Either side has strong arguments and either side makes some points that don't convince me. On balance, I come down on Apple's side with respect to this particular question, but Epic may very well prevail on this. What I as an app developer fear is that in the end the Section 1/2 question will get all the attention though there's little for us app developers to gain from it (while the single-brand market definition would be heaven-sent). It would be a nightmare if Epic lost on market definition (again) and if the Supreme Court then didn't help, with the justices instead focusing on a question for review that would be delicious food for thought but doesn't have much potential to help us developers, such as the following hypothetical one:

"Does an intellectual property license and third-party product distribution agreement that is unilaterally imposed by a platform operator as a logical consequence of that company's product-definition decisions constitute concerted action by independent centers of decision-making under Section 1 of the Sherman Act?"

One of the Section 1 cases cited by Apple may become a boomerang: Relevent Sports v. United States Soccer Federation and FIFA

In paragraph 491 of its proposed conclusions of law, filed on April 7, 2021, Apple pointed to U.S. court decisions that denied the existence of an agreement where there was no "meeting of the minds" but a unilateral command. One of those citations was the following:

"Relevent Sports, LLC v. U.S. Soccer Fed’n, Inc., No. 19-CV-8359, 2020 WL 4194962, at *7 (S.D.N.Y. July 20, 2020) (U.S. Soccer Federation’s compliance with its obligation to follow FIFA policies against sanctioning certain soccer matches was unilateral conduct outside the scope of Section 1)"

Because of my long-standing interest in (and past work related to) the application of antitrust law to sports, especially to soccer, I've listened to the recording of the Second Circuit hearing in that case, which took place one year (to the day!) after Apple's filing.

I can see why Apple thought in April 2021 that this citation was going to help. The Southern District of New York is in some ways the East Coast equivalent of the Northern District of California. But I can also see why--even without the benefit of 20/20 hindsight--it wasn't a good idea to bring up that case: the July 2020 decision Apple refers to was an order dismissing Relevent's original complaint without prejudice. It was just an invitation to refile by adding FIFA--the rulemaker (i.e., Apple's equivalent) in that context--to the case. Relevent brought a new complaint naming FIFA as a co-defendant. In July 2021, between the Fortnite trial in Oakland and Judge YGR's ruling, the New York judge (J. Caproni) also dismissed the amended complaint. Relevent appealed, and the Second Circuit is now highly likely to reverse her. One of the circuit judges noted that the defendants had a problem with the merits. To me it's clear that USSF and FIFA have nothing to offer in their defense but some technicalities that they're resorting to. Smokescreens are not enough, especially when you face a hostile court that even drew mafia and drug cartel analogies (disclaiming any intent to offend USSF and FIFA, though it's not really complimentary either).

Winston & Strawn's Jeffrey Kessler is doing an incredible job for Relevent, a sports promotion company from New York that wanted to organize a Spanish professional soccer league game in the United States (as well as a match between two Ecuadorian teams) but the teams had to withdraw under pressure from soccer bodies complying with FIFA's rule.

It's worth noting that both President Trump's Antitrust AAG Makam Delrahim (by way of a stern warning in a letter to FIFA and USSF) and President Biden's DOJ (by way of an amicus brief) supported Relevent because they believe it's in the interest of U.S. consumers and athletes to combat horizontal geographic market division, with literally zero amicus brief support for USSF and FIFA. In policy terms, I think there is a risk of sports-related U.S. antitrust law not paying enough deference to the specifics of professional sports. U.S. case law has recognized that some horizontal cooperation is allowed when it's needed to make a joint product, though that is only part of the picture: sports is the only industry in which competitive balance is key (everywhere else, even someone who currently needs to cooperate with others would rather acquire them all or drive them all out of business). Also, there are historic reasons for which some territorial restrictions are acceptable in sports and not to be equated with horizontal geographic market division in other industries.

Relevent v. U.S. Soccer is distinguishable from Epic Games v. Apple in even more ways. Still, it is possible now that the Second Circuit will say something about Section 1 applicability in its decision--possibly even before, otherwise not long after the Epic v. Apple hearing--that may encourage the Ninth Circuit to side with Epic on this question. For example, the district judge in Relevent based her (second) dismissal on requiring the plaintiff to show that the rule in question had not just been promulgated, but that there had been a prior agreement to vote in its favor. The Supreme Court has been clear, however, that a formalistic approach is wrong in this kind of context, which benefits Apple as it disagrees with Epic's focus on the DPLA being a contract, but which could also lead to a holding or dictum by the Second Circuit that supports Epic's argument that developers don't have to embrace Apple's restrictions in order for there to be a Section 1 violation. A request for judicial notice is likely going to happen. We might even see two requests should either party attempt to spin the Relevent opinion in its favor. Apple made its bed by citing to Relevent at a premature stage and must now lie in it...

Share with other professionals via LinkedIn: