In a 5-4 split decision, the U.S. Supreme Court appears to have reworked a longstanding precedent that has been a foundation of antitrust litigation for more than 40 years—the “direct SCOTUS Blows Down Apple’s House Made of Illinois Brickpurchaser” rule of Illinois Brick, which generally forecloses “downstream” purchasers from suing for alleged violations of the Sherman Act. Apple Inc. v. Pepper addressed a “market” that almost certainly is familiar to the reader: Apple’s iPhone “App Store.” The Supreme Court allowed a putative class action of iPhone application consumers to proceed against Apple, even though the “sting” of Apple’s allegedly monopolistic 30% commission on each app sale first was felt by app developers, not the consumers.

Apple invoked the so-called Illinois Brick defense, which is derived from a 1977 Supreme Court case by the same name. There, the defendant (the Illinois Brick Company) manufactured concrete blocks and sold those blocks to masonry contractors; those masonry contractors sold masonry structures to general contractors; those general contractors sold their services for construction projects to the State of Illinois. The state sued Illinois Brick, alleging that Illinois Brick had engaged in a conspiracy to fix the price of concrete blocks, and that the state consequently paid more for the concrete blocks than it otherwise would have. According to the state, that alleged “overcharge” was “passed on” through the distribution chain from Illinois Brick (the manufacturer), to the masonry contractors, to the general contractors, and finally to the state (as the ultimate consumer).

In Illinois Brick, the Supreme Court ruled that because the state was an “indirect purchaser,” which did not purchase concrete blocks directly from Illinois Brick, the state was not a proper antitrust plaintiff. According to the Supreme Court, it was not clear that all or even some of the alleged overcharge was “passed on” to the state, and allowing the state to bring suit would result in unnecessarily complicated damages calculations about the extent to which the overcharge was “passed on” through the distribution chain, or even duplicative damages. The proper plaintiff, the Supreme Court surmised, would be an entity that had purchased concrete blocks directly from Illinois Brick; any such “direct purchaser” necessarily would have paid the alleged overcharge.

For more than 40 years, Illinois Brick has operated as a bright-line, prudential “standing” rule in antitrust class actions, barring indirect-purchaser plaintiffs from bringing antitrust claims because any alleged overcharge may not have been “passed on” through the distribution chain. Under Illinois Brick, only direct purchasers could sue. But the Supreme Court’s new decision in Apple arguably has disrupted the long-settled understanding of what it means to be a “direct purchaser.”

In Apple, four iPhone owners alleged that Apple unlawfully monopolized the market for iPhone apps and took an anticompetitive 30% commission on each app sale. Consequently, according to the plaintiffs, app consumers have paid higher prices for apps in the App Store. Apple argued that Illinois Brick barred the app consumers’ antitrust claims, even though—in the App Store—consumers pay Apple directly in return for Apple’s granting them access to download the desired apps. According to Apple, even though it transacted directly with the app consumers, it was merely serving as “agent” for the app developers, who set the price of each app, and who, according to Apple, should be considered the real “sellers” of the apps. Plus, the alleged anticompetitive overcharge—Apple’s 30% commission—was levied directly on the developers (who received only 70% of each app sale), not the consumers. It was not clear that any of the 30% commission was passed on to consumers in the form of higher app prices.

The district court agreed with Apple, dismissing the case, but the Ninth Circuit reversed. A divided Supreme Court affirmed the Ninth Circuit, disagreeing with Apple’s characterization of the transaction. Writing for the majority, Justice Kavanaugh (joined by Justices Ginsburg, Breyer, Sotomayor, and Kagan) focused literally on the concept of a “direct purchaser,” not the “pass on” of any alleged overcharge. The Supreme Court’s reasoning was simple: Because app consumers transact directly with Apple, and there is no “intermediary” between them and Apple, the plaintiff-consumers in the case are “direct purchasers,” so Illinois Brick does not bar the antitrust claim. Based on the fact that Apple transacted directly with the consumers, the Supreme Court rejected Apple’s attempt to characterize itself as “agent” for the developers. The Supreme Court also was unmoved by the fact that app developers, not consumers, feel the brunt of the alleged overcharge, and completely dismissed concerns about the difficulty of allocating damages between consumers and developers, and the possibility of duplicative recoveries. Instead, the majority reasoned that, because there was no distribution chain, the alleged overcharge—i.e., the 30% commission—“has not been passed on by anyone to anyone.”

Justice Gorsuch (joined by Chief Justice Roberts and Justices Alito and Thomas) wrote a blistering dissent. He accused the majority of reducing Illinois Brick to a “blind formalism” by focusing exclusively on the fact that app consumers purchase “directly” from Apple, while ignoring the economic relationship among the parties and the fact that app developers, not the consumers, “first felt the sting of the alleged overcharge.” As Justice Gorsuch explained, by allowing app consumers to sue, the majority had opened the door to “unduly complicated” damages computations and “conflicting” claims about allocating the alleged overcharge between consumers and developers—precisely what the Illinois Brick rule was intended to prevent.

The majority’s decision to allow consumers to sue even though they did not “first feel the sting of the alleged overcharge”—and may never have felt that sting—arguably does amount to a formalistic reworking of the Illinois Brick doctrine. But questions remain. For example, the majority, at least implicitly, conceptualized the App Store as a two-sided market, where developers pay for the ability to distribute apps and where consumers buy apps. The majority seemed to bless the possibility that both consumers and developers could recover from Apple. Yet the majority (and the dissent) failed even to mention the Supreme Court’s decision last year in Ohio v. American Express Co., where the Supreme Court rejected an antitrust challenge to a different two-sided market: Amex’s credit card platform, which connects cardholder-buyers to merchant-sellers. In Amex, the Supreme Court held that plaintiffs challenging anticompetitive conduct in a two-sided market must establish that the overall effect of the challenged conduct on both sides of the market is anticompetitive, and affirmed judgment for the defendant credit card companies because the plaintiffs’ analysis of anticompetitive effects focused only on the merchant side of the relevant market. Perhaps the Supreme Court’s failure to mention Amex in the Apple opinion was simply a product of the majority’s express refusal to “assess the merits of the plaintiffs’ antitrust claims” or “any other defenses Apple might have.” Or perhaps it indicates a desire by the Apple majority, which included all four Justices who dissented in Amex, to limit Amex’s holding to the specific two-sided market at issue in that case.

More broadly, what will be Apple’s long-term impact on antitrust jurisprudence? At a minimum, we would expect consumer plaintiffs to attempt to use Apple to bring class actions against other  internet platform companies—e.g., Amazon, Alibaba, eBay—as well as any other retailer who operates on a commission basis but allows a given manufacturer or distributor to set the sale price.  Because of the Supreme Court’s formalistic focus on the identity of the “direct purchaser,” we would not be surprised to see Apple and other tech companies restructure their distribution models in response to this decision.

Finally, does Apple mean the Illinois Brick doctrine is on solid ground for the foreseeable future?  Many commentators, and several amici (including 31 state attorneys general), had argued that the Supreme Court in Apple should reverse Illinois Brick. None of the parties argued for that change in the law, and the Supreme Court did not take up that issue. But in a sense, the Supreme Court did just the opposite, reaffirming Illinois Brick as a bright-line rule that permits only “direct purchasers” to bring antitrust claims. Nevertheless, Justice Kavanaugh’s majority opinion notably downplayed the core rationale of the Illinois Brick doctrine—to avoid the complications, expense, and uncertainty associated with calculating and allocating overcharge damages among participants at different levels of a distribution chain. That could signal the majority’s willingness to revisit, and possibly even to overrule, Illinois Brick down the road.

“Two Wrongs Don’t Make a Right, But a Few More Can Make a Unicorn”Class actions typically involve a proposed class of plaintiffs seeking recovery from the same defendant on similar grounds. But that is not the only animal in the class action corral. Rule 23 makes this clear in its very first sentence: One or more members of a class may sue or be sued as representative parties on behalf of all members….” But this particular class action animal is more often imagined than seen. Indeed, as the Fourth Circuit recently observed, “Defendant class actions are so rare they have been compared to ‘unicorns.’”

The reason defendant class actions are so rare is that they present a host of practical challenges, due process challenges, and adequacy challenges. How can you force an involuntary defendant to look out for the interests of a defendant class? Can’t the defendant class representative defeat certification simply by filing a statement with the court that it has absolutely no intention of looking out for the interests of anyone but itself, and is fully willing to sacrifice the interests of the class in order to obtain a better deal for itself? Doesn’t a corporation’s duty to its own shareholders require just that? Wouldn’t any absent member of the defendant class that was in its right mind opt out anyway? Besides, who pays the counsel for a defendant class, given that the class is being asked to involuntarily pay money and not seeking to recover money? And how does Rule 23 allow the assertion of personal jurisdiction over out-of-state members of a defendant class? Because of these challenges and others like them, a recent law review article by Francis Shen found that only 177 defendant class actions had ever even been proposed as of early 2011, and most of those had to do with declaratory relief as to the meaning of a state or federal statute.

But recently, believe it or not, there has been an actual confirmed sighting of this mythical class action unicorn in the form of a certified (and affirmed) defendant class action. The discovery was made by the Fourth Circuit and happened in a case called Bell v. Brocket. It turns out that the unicorn was actually created by mistake. Or rather, by an accumulation of mistakes.

The case devolved from the SEC’s enforcement action against on an online auction site allegedly being run as an effective Ponzi scheme. The SEC froze the assets of the allegedly offending LLC, whereupon Bell was appointed receiver for the LLC. He proceeded to file a defendant class action against all those who had received at least $1,000 more from the scheme than they put in, for the purposes of returning the difference to those who had a net loss from investing in the scheme. Some of the “net winners” were named as defendants and proposed representatives of the defendant class. Some defendants objected to serving as class reps on cost grounds, arguing that they could not afford to fairly represent the class. But they did not argue that the proposed no-opt-out class violated their due process rights, and they did not argue that some members of the class were not subject to personal jurisdiction in the forum.

The trial court certified the class as a no-opt-out class under 23(b)(1), typically applicable in damage actions only when the target of multiple damage claims has funds too limited to pay all claims. But here there were many defendants who were the target of one receiver’s damage claims. (The phrase square peg in a round hole seems to fit here.) Compounding the problem, the court directed notice to the class, but did not appoint anyone counsel for the defendant class. Remarkably, no absent class member objected to this failure at the time.

As the case proceeded, counsel for the individual named defendants made it clear that they and their counsel were not interested in representing the defendant class unless somebody agreed to pay them to do so. With no such volunteer appearing, counsel for most defendants formally withdrew. One, Mr. Edmundson, stayed on as counsel for his individual clients. Meanwhile, while the defendant class went unrepresented, not only had notice been sent without the input of anyone acting as counsel for the defendant class, but the court had appointed one of the named defendant’s experts as expert for the defendant class.

Well after the expert’s work was underway, the court finally appointed Mr. Edmundson as class counsel, despite his long and understandable reluctance to assume the risk of leaping upon an already running unicorn, but with a catch—Mr. Edmundson would only be class counsel for the liability phase of the case. The expert previously appointed by the court then promptly filed a report conceding that the operation was a Ponzi scheme, and thereby effectively all but conceding the liability of the defendant class. Class counsel then declined to even depose the plaintiff’s expert. The trial court then granted summary judgment against the defendant class and proceeded to notify class members of an unusual and distinctly homemade plan for determining damages, without objection by Mr. Edmundson, whose appointment as class counsel did not extend to the damages phase in any event. The plan called for the receiver to send his damage calculations to each class member for them to either accept or object to.

On the eve of entry of final judgment against class member defendants whose liability was thus determined, sanity attempted to intervene in the form of objector Edward Rourke. He moved to decertify, finally pointing out some of the problems that should have been obvious all along:  “(1) Edmundson had conflicting interests as class counsel; (2) Edmundson failed to obtain an independent expert evaluation of the business and conceded the existence of a Ponzi scheme; and (3) since Edmundson only represented the class through the liability phase of the action, there was no adequate representation of class members during the damages phase and decertification was warranted.” Unphased, the trial court entered final judgment.

Sanity persisted in beating its head against the wall. Several more absent class members sought to intervene and object post judgment, arguing that: “(1) Edmundson failed to vigorously represent the interests of the class and unnecessarily conceded the existence of a Ponzi scheme; (2) following the district court’s entry of summary judgment, Edmundson failed to represent the interests of the class; and (3) Edmundson had potential conflicts of interest throughout his time as class counsel.” Bell opposed, arguing that these objections came too late.  In reply, the objectors finally pointed out that the damages procedure violated the right to due process and a jury trial. But even these objectors said nothing about the failure of the district court to name class counsel at the time it certified the class; the failure of the district court to apply the Rule 23(g) factors in deciding whom to appoint as class counsel; or any challenge to commonality, typicality, jurisdiction, or the absence of opt-out rights.

The trial court rejected all of these objections as untimely second guessing of strategic choices made by class counsel. The late-arriving objectors then appealed.

A clearly pained Fourth Circuit held its nose and affirmed. (Apparently unicorns smell bad, or at least this one did.) The Fourth Circuit flatly declared that “Defendant class actions, like plaintiff class actions, must comply with Rule 23.” It then essentially conceded that this one didn’t. It noted that prompt appointment of class counsel was mandatory upon certification and did not happen here. It noted that consideration of the Rule 23g factors in choosing class counsel was likewise mandatory, not optional. And it agreed there was prejudice to the class from these errors by the trial court, since the key events upon which the liability and damage findings were based occurred in large part while the class went unrepresented.

But then the Fourth Circuit confronted the “horn” of the dilemma: Do we enforce Rule 23 or our own rule that issues not timely raised below cannot be considered on appeal? With a clearly sour stomach, the Fourth Circuit surveyed the situation, noted that many net winners had already individually accepted and paid their calculated liability to net losers, and concluded that “the toothpaste cannot be put back in the tube.” Therefore, while painstakingly pointing out the errors of the court and counsel that forced it to do so and reminding future courts and litigants not to ignore the requirements of Rule 23 in considering defendant classes, the Fourth Circuit reluctantly affirmed, thereby confirming the existence of at least one defendant class unicorn.

The moral of the story? However difficult it may be to conjure up a unicorn intentionally, it’s really not that hard to create one through lack of attention — which is why the Fourth Circuit’s decision is itself suspect. Especially as to adequacy and due process issues, the Supreme Court has long said that absent class members are not supposed to be bound by the failure of their representatives or the certifying court to ensure adequacy of representation and fundamental due process (see Phillips Petroleum Co. v. Shutts, 472 U.S. 797 (1985)). So while there may indeed have been a sighting, this unicorn remains on the endangered species list.

“Sorry, Wrong Number”: Northern District of California Denies Certification in TCPA Class ActionAs we have noted before, whether a claimant under the Telephone Consumer Protection Act (TCPA) gave “prior express consent” to receiving communications from the defendant is frequently a critical issue (and often the only issue standing between the defendant and massive class action liability). A recent decision from the Northern District of California in a “wrong number” class action, Revitch v. Citibank, N.A., is illustrative of the consent battleground.

The TCPA prohibits, inter alia, the making of unsolicited calls to cellular telephones using an automatic dialing system. A call is not unsolicited, however, where the call recipient gave the caller “prior express consent.” In Revitch, the plaintiff sued Citibank (an issuer and servicer of credit cards) after Citibank called him multiple times on his cell phone using an autodialer, despite the plaintiff not being a Citibank customer. Revitch sought certification of a class consisting of all persons in the U.S. who were called by Citibank using an autodialer where such persons were not listed in Citibank’s records as the intended recipients of the calls.

Recognizing that Citibank had prior express consent to call its customers about their accounts, Revitch pinned his certification theory on data from Citibank’s call system, indicating where call operators flagged particular phone numbers as “bad numbers” or “not valid” – thus indicating (according to Revitch) true “wrong numbers” as to which consent could not have been given or obtained. Revitch in his certification evidence also attempted to deal with the fact that a phone number can be flagged as “wrong” in Citibank’s system even when it is in fact the customer’s correct number (for example, an evasive customer might claim “wrong number” when being called about his delinquent account). To control for such false positives, Revitch’s expert used data from a “reverse lookup” service that, according to the expert, allowed for the identification of actual “wrong number” recipients.

The court nonetheless denied class certification, holding that individual issues of consent would predominate in any trial.  First it noted evidence from Citibank that many of the “wrong numbers” proposed by Ravitch’s expert were in fact associated with Citibank customers that the bank was trying to reach (such as family members who also had Citibank accounts). Citibank also proffered expert evidence of its own, claiming that many of the “wrong number” call recipients were called in connection with a different account than the one in which a “wrong number” notation had been made in the call log system. A further problem with Revitch’s theory, according to the court, was that Citibank’s “wrong number” data existed only from November 2017 forward; proof of wrong number calls prior to that date would have to be taken from individual, account-level records. Relying on Citibank’s evidence, the court concluded that adjudicating class members’ claims would “devolve into tedious resolution of individualized issues based on individualized evidence.”

From the defense perspective, there are a couple of takeaways from Revitch:

First, the opinion notes a split in district court decisions on the issue of whether “wrong number” notations in a defendant’s call records will suffice to establish predominance and commonality.  Some courts have said yes and certified classes; others (such as Revitch) have found the issue of consent to be too individualized for certification. To date no court of appeals has addressed the issue.

Second, Revitch illustrates the importance of the defendant’s evidence on the consent point (including, potentially, expert testimony), particularly when the plaintiff’s theory is premised on notations in the defendant’s call records. Several of the decisions certifying classes rejected as overly speculative the defendants’ arguments that “wrong number” notations didn’t necessarily mean an actual wrong number had been called.  Revitch shows, unsurprisingly, that such arguments have much more force if they are proven to be true. Failing to develop the proof needed on this issue at the time certification is decided makes it easy for a court to simply conclude, “I can always revisit this certified class if the defendant’s theoretical objection is later shown to be correct.” In a world that can involve huge class-wide statutory damages, having to file a motion to decertify is not the place any defendant wants to find itself.