Two New Supreme Court Decisions on the Presumption Against Extraterritoriality
July 6, 2023
The end of the Supreme Court’s term brought two decisions on the presumption against extraterritoriality, a significant and contested interpretive canon for federal statutes. Yegiazaryan v. Smagin ruled 6-3 that a civil RICO suit based on an alleged scheme to fraudulently conceal assets belonging to a U.S. judgment debtor had sufficient domestic content to fit within a statute that does not apply extraterritorially. A week later, the Court decided in Abitron Austria GmbH v. Hetronic International, Inc., that the Lanham Act’s prohibition of trademark infringement did not apply to all overseas sales of pirated goods. Bill Dodge has provided excellent accounts of both decisions, as have Linda Silberman and Rochelle Dreyfuss, and I do not quarrel with them. Both opinions leave open many questions about the purpose and application of the presumption, as Bill points out. In this post I identify the most important unresolved issues and suggest an approach to addressing them that seems most consistent with the premises of the doctrine.
A Short History of the Presumption Against Extraterritoriality
Bill Dodge’s discussion of Abitron carefully reviews the evolution of the doctrine that led up to these decisions. It would be useful, however, to regard the 1991 decision EEOC v. Aramco and the 2010 decision Morrison v. National Australian Bank as establishing the modern doctrine. This doctrine, unlike its antecedents, reflects a set of arguments that had informed a number of cases in the 1980s, 1990s, and 2000s involving the presumption against implying statutory private rights of action. What joins these two presumptions is a recognition (or belief, if you don’t like the presumptions) that extravagant use of certain interpretive doctrines can create federal question doctrine out of thin air, trenching on the constitutional authority of Congress to determine (and limit) the scope of the subject matter jurisdiction of federal courts.
The origins of the presumption against implied statutory private rights of action lie in Justice Powell’s dissenting opinion in a 1979 case, Cannon v. University of Chicago. A majority of the Court held that Title IX provides victims of discrimination in education a private right of action, notwithstanding the statute’s silence on the point and its express creation of an alternative remedial mechanism. Sensing unease on the part of his colleagues, Powell chose the case to advance a well-developed criticism of the practice. (In the interest of full disclosure, I repeat here what I have said several times in print: “I worked as Powell’s clerk at the time that he wrote this opinion, which may explain my attraction to his arguments.”)
Powell’s critique made the argument that extending federal court subject matter jurisdiction in this manner was, if not unconstitutional, at least anti-constitutional: it employed the fully legitimate practice of statutory interpretation to reach an illegitimate result, namely locating the authority to define the prescriptive powers of the federal judiciary in the courts, rather than in Congress. Beginning in 1981 and extending well into this century, a series of cases elevated Powell’s articulation of a presumption against such implication to established doctrine, even if the Court in 2005 declined the opportunity to overrule Cannon itself.
As I argued in a 2015 law review article, the modern presumption against extraterritoriality bears a close resemblance to Powell’s presumption against implied causes of action. Both serve as canons of statutory interpretation that induce Congress to make significant choices about the scope and remedial effect of federal statutes. Both target open-ended judicial discretion used to fill in the content of federal regulatory programs and, even more importantly, to make the decision whether to assign this discretion to judges. Both renounce the multi-factored tests that lower courts had used to address the scope-and-remedy questions and replace these efforts with what aspire to be discrete and definite rules. Both presumptions seek to provide greater clarity to those who live under particular legal regimes as well as to lower courts.
Morrison may have been the last Supreme Court decision to realize anything like the clarity that the presumption is supposed to provide. The case involved the Securities Exchange Act of 1934, a statute that, among other things, proscribes the use of fraud in the purchase or sale of securities. Morrison explained that the statute did not apply outside U.S. territory, and what it applied to was the purchase or sale of securities, not fraudulent behavior. In that case some, perhaps much of the alleged fraud took place in the United States, but the sales at issue all occurred elsewhere. The Court used the term “focus” to sort out what it was that the statute sought to regulate.
Relying on a straightforward reading of the statutory language, the Court ruled that domestic fraud was irrelevant unless tied to domestic transactions. Moreover, the transactions had to have a specific location: they had to take place on the domestic market, and not just have the potential to affect that market. It was the attempt to determine when foreign transactions might have an actionable effect on domestic markets that had tied up the lower courts, and Morrison explicitly put an end to that endeavor.
Morrison took two important steps. First, it distinguished between the evil that a law had the purpose of combatting and the specific market where the law sought to root out that evil. Second, it defined the relevant market discretely and territorially, rather than bringing difficult causal premises into play. Thus the Securities Exchange Act, as interpreted by Morrison, did not make all fraud actionable, but only purchases or sales in U.S. capital markets that suffered from fraud. Open questions remained, but the main issues were clear enough. Going forward, people engaging in transactions as well as the lower courts had a pretty good idea of where they stood with respect to the statute.
The 2023 Cases
I will leap over the many extraterritoriality cases decided between 2010 and 2023. Difficulties in getting five votes for a consistent account of what the presumption was meant to do and how it worked led to considerable confusion. A 2020 article by Bill Dodge provides a magisterial account of the doctrine, but not, I regret to say, the last word. The two new decisions show how much remains unresolved and how much the doctrine has strayed from the goal of providing clarity and limiting judicial discretion. The best that one can say about these recent decisions is that they do not close down paths that would bring us closer to that goal, even if they lean in another direction.
Consider first Smagin. Smagin, a Russian resident, sought to enforce an arbitral award against Yegiazaryan, a California resident. Smagin claimed that Yegiazaryan conspired with others to hide assets that Smagin could have attached and then used to satisfy the arbitral award, which through domestication had become a California judicial judgment. Unable to reverse the diversion of Yegiazaryan’s assets to an unreachable offshore financial account, Smagin brought a civil RICO claim against Yegiazaryan and, perhaps more significantly, the U.S. financial institutions that allegedly conspired with him to conceal and transfer Yegiazaryan’s assets.
In RJR Nabisco, Inc. v. European Community, the Court had determined that civil RICO, unlike its criminal counterpart, covers only domestic injuries. Embracing RJR, Yegiazaryan relied on one of the rules of the first Restatement of the Conflict of Laws, the high-water mark of formalism in the U.S. approach to conflicts, to locate Smagin’s injury offshore. The label that best described Smagin’s claim, he argued, was a tort based on fraud that, according to the supposedly applicable conflicts rule, had its location at Smagin’s residence in Russia, taking the claim outside the territorial scope of civil RICO.
Justice Sotomayor’s opinion for the Smagin majority, like the lower courts it affirmed, rejected this argument. It noted that conflicts rules might have some bearing on determining the focus of a statute, here specifically the location of the injury for which the plaintiff sought compensation. It denied, however, that the 1934 Restatement, rather than the 1971 Restatement (Second), did a better job of capturing the relevant law. Instead of looking at either, the Court trotted out a context-specific, multi-factored balancing test to find where the injury occurred. The Court noted that most of the fraudulent activity occurred in the United States. Moreover, the rights to enforce the California judgment, to obtain post-judgment discovery, to seize local assets, and to auxiliary relief all were creatures of California law. These local rights outweighed the fact that Smagin’s experience of the loss resided with him in Russia. Without relying expressly on the Restatement (Second), the opinion captured the spirit of its anything-goes, maximize-judicial-discretion approach to conflicts questions.
Justice Alito, joined in relevant part by Justices Thomas and Gorsuch, complained that Sotomayor’s opinion left the relevant audiences unsatisfied. Justice Alito had authored RJR and disliked what the majority was doing with that precedent. The majority’s opinion offered “virtually no guidance to lower courts” and risked sowing confusion about the Court’s precedents. Alito hinted that he preferred a blanket rule that would assign a single determinant location to all intangible property as a class, but he did not finish off the argument and, I argue below, would have found that outcome difficult to swallow if he had thought about it more carefully. In the end, Alito and his colleagues did not identify an alternative rule or reject the outcome, but rather expressed a preference for dismissing the case so as to create no precedent while leaving the lower court’s resolution intact.
A fair criticism of Sotomayor’s opinion is that it ignores the distinction between the evil combatted – here fraud in commercial transactions – and the place of the relevant transactions. I will return to this point below to show how the Court could have reached the same outcome without returning to the unsettling, open-ended method that the presumption had originally been forged to discourage.
In Abitron, handed down a week later, Alito wrote for the majority, but failed to reach any outcome-determinative resolution. He articulated some limiting principles that dictated the resolution of the largest portion of the case, but he failed to indicate how those principles applied to other aspects of the lower court’s decision. Some of his reticence might have reflected the unwillingness of four Justices (Roberts, Sotomayor, Kagan, and Barrett) to buy into his interpretation of the statute and the inclination of Justice Jackson to apply it differently. Rather than clarify his differences (if any) with Jackson, however, Alito left a significant issue open.
The Lanham Act creates liability for commerce in goods that present a risk of consumer confusion (what I will call pirated goods). Abitron originally distributed under license Hetronic’s trademarked products, but concluded that it had acquired ownership of the trademark. It began to assemble the product in Europe, using components from third parties. It sold that product mostly in Europe, but some of the goods sold in Europe were resold in the United States, and a small percentage went directly from Abitron to U.S. customers.
Lurking in the background was the 1952 case Steele v. Bulova Watch Co. Steele, a U.S. national, bought watch components from U.S. sources, assembled them (apparently rather badly) in Mexico, and sold them in Mexico to U.S. consumers right across the border from Texas. These transactions were legal in Mexico because Bulova had failed to register its mark there and, under Mexican law, anyone could register and then employ an unregistered mark. The Supreme Court, in an opinion by Justice Clark, argued that there existed enough U.S. contacts to justify application of the Lanham Act. Clark’s opinion, however, provided no insight as to reasoning or implications, and in particular as to which contacts mattered to justify U.S. jurisdiction. Even with the proliferation of U.S. connections, Justices Reed and Douglas dissented, invoking the presumption against extraterritoriality. The lower courts, faced with this opacity, mostly used multi-factor, open-ended tests that wound up all over the place. Their range of outcomes, verging on legal chaos, manifest exactly how little guidance Steele had provided.
The Court apparently took Abitron to clean up this mess. Alito’s opinion focused on the relevant statutory language, which prescribes the use “in commerce” of goods that is “likely to cause confusion.” He rejected the argument that the reference to “commerce” encompasses foreign transactions, even though the Act defined that word as encompassing all commerce that Congress lawfully may regulate. Following in the steps of earlier decisions, he held, and all members of the Court agreed, that more specific language was required to overcome the presumption against extraterritoriality. The issue thus became what domestic activity was necessary for the Act to apply. Following Morrison, all Justices agreed that this presented a “focus” question, which separated activity that matters for the statute’s application from that that does not.
Alito held that use “in commerce” meant that a transaction, to be actionable, had to occur in the United States, in addition to creating a risk of consumer confusion within the United States. Sotomayor’s concurrence, by contrast, argued that the statute focused only on confusion, and that foreign sales of goods destined for the United States could cause confusion. Justice Jackson argued that goods tainted by the confusion risk purchased abroad and resold in the United States probably fit within the statute, although she understood Alito’s opinion as not answering that question. She suggested, but did not conclude, that Abitron might be held liable for resold goods.
Justice Sotomayor and her three colleagues would have taken the same path as Smagin and made the place of the conduct producing the evil, rather than the place of the transaction, determinative. Adopting the view of the Solicitor General’s amicus brief, they argued that the Lanham Act applies whenever a pirated good creates confusion in the U.S. market, even if the pirate engages in no U.S. transaction. The argument seems to take a stream-of-commerce approach: once a seller launches a fake good into international commerce, it bears legal risk under the Lanham Act whenever the product winds up being used in U.S. commerce, no matter how attenuated the connection between the initial foreign sale and the later domestic use.
Sotomayor’s approach leaves unaddressed whether an offshore seller bears complete responsibility for any U.S. reselling, or whether some kind of culpability is required. The Solicitor General’s brief hinted that the standard could be whether resales were reasonably foreseeable, a common enough approach but one replete with uncertainty. Other alternatives might be strict liability, on the theory that goods that bear a risk of confusion are like inherently dangerous ultrahazardous products, or some kind of scienter, requiring the initial seller to either know of or be willfully ignorant as to the likelihood of resale. A hypothetical raised by Justice Jackson in oral argument and embraced by her opinion, regarding consumer goods sold legally in Europe to U.S. tourists and later put up for sale in the United States, indicates strict liability.
Bill Dodge faults Justice Alito for not respecting the precedential force of Steele. A more charitable view might be that the Court generously refused to kick the dead horse of Justice Clark, not one of the more admired opinion writers of his era. Before an opinion can have precedential value, it must say something intelligible that might affect the outcome of a later case. In her separate opinion, Justice Sotomayor defended her approach as “consistent” with Steele. But by refusing to reveal the basis of its decision, Steele made it possible for almost any future outcome to achieve consistency. Alito’s opinion, which mostly slighted Steele rather than explaining fully what a poor instance of legal craftsmanship it was, showed kindness more than disrespect.
Focus and Formalism
Going forward, Smagin and Abitron give transactors and lower courts little guidance as to how to apply the focus test with respect to statutes that do not apply extraterritorially, which is to say most statutes. Smagin looks toward earlier practice, much used in the lower courts, of identifying as many relevant factors as possible and then balancing them in some opaque manner. Abitron points toward a more formal, rule-oriented approach, but fails to pin down how to construct relevant rules given the statutory sources. Both hint, although they do not demand, that going forward courts should fix the focus of a federal statute by concentrating on the evil the statute means to combat, and not the transactions that the statute purports to cover. Morrison, of course, pointed in exactly the opposite direction.
Antiformalists might argue that multi-factored balancing represents the essence of common-law method and saves legal regimes from the opportunist manipulation of empty forms. The critics of the first Restatement believed that its categorical rules ultimately failed to live up to their goals because the boundaries of the categories were too unstable and too easy to exploit by sophisticated transactors and litigators to the cost of the naïve. Rather than pigeon-holing facts into a particular category so as to spit out a particular location on which territoriality will turn, antiformalists want judges to look at everything so as to reach the right result in the particular case.
The Supreme Court’s adoption and evolution of the presumption against extraterritoriality, however, rested on a different point of view. The Justices who bought into this doctrine believed that it was possible to distinguish between predictable and open-ended rules and that the former promote both transactional liberty and the accountability of lower courts to the highest judicial authority. So Morrison argued. One can debate whether these are the right goals for courts to pursue, but not whether the Court has generally approved of them in its statutory-interpretation federal-question-jurisdiction cases.
From this perspective, the relevant issue becomes how might Smagin and Abitron be interpreted in light of the premises of formalism that a less distracted Court might embrace. Is it possible to clarify those decisions to help the lower courts, as well as people seeking to live under the law, to determine where they stand? Is the Court on a path to gut Morrison and bring back multi-factor balancing, or are these late-in-the-term decisions instances of defensible outcomes shrouded in confused and unhelpful reasoning?
Let’s start with Smagin. The litigants did no one any favors by presenting the Court with two bad alternatives. Yegiazaryan’s counsel argued that all a court had to do was tailor the boundaries of “intangible property,” with the relevant rule (location of the owner) following automatically. But once one goes beyond tangibility – realty and personalty existing in material form – the question becomes what is property, as opposed to a contractual interest or an interest under tort law. This inquiry becomes extremely murky if not completely unworkable. Nor, under the facts in Smagin, is such a quest necessary. Smagin’s counsel also provided little help, as they insisted on the multitude of factors pointing to a U.S. interest in the claim, rather than working with Morrison to pin down a clear and predictable location for the claim. They understandably were more interested in prevailing than in clarifying the law.
The test announced by RJR is the place of the injury, not of the acts producing injury. A useful and nearly universal rule in conflicts law locates a debt in the jurisdiction that creates it. Yegiazaryan, Smagin’s debtor, resided in California and, under the New York Convention and the Federal Arbitration Act, faced an obligation to honor an outstanding arbitral award. Upon domestication of the arbitral award, he faced a further obligation to honor that California-based judgment. The injury – the attack on Smagin’s legal interest – thus occurred in the United States because the interest was in the United States, no matter where the conduct seeking to produce the injury occurred. Why complicate the matter, as Sotomayor’s unnecessary multi-factor approach requires?
Bolstering the argument for this approach is the observation that over the last thirty years, the Court generally has backed away from using multi-factor balancing tests to determine the limits of U.S. prescriptive jurisdiction. Perhaps the best example of this trend is the 2004 decision F. Hoffman-La Roche Ltd. v. Empagran S.A., an antitrust case that called the this approach “too complex to be workable,” but the trend is general and clear. Faced with the considerations that led the Court to this course, it seems unlikely that Smagin’s attempt to revive the old way of doing things will flourish.
Consider next how to interpret Abitron, especially in light of Justice Jackson’s concurrence. The Court was unanimous on two questions based on territoriality: The Lanham Act applies to sale of pirated goods directly into the U.S. market, and does not apply to offshore sales where the goods do not find their way to the United States. The Court could not come to agreement as to how the Act applies to foreign sales followed by U.S. resales. Four Justices embraced an answer that elevated the place of the ultimate harm over the place of the defendant’s behavior, and Justice Jackson indicated an openness to that outcome.
Alito’s opinion interprets use in commerce as focusing on the location of transactions in pirated goods. In the case of a contract for the sale of goods, the place where the buyer takes possession of the goods usually serves as the locational default. When a foreign person sells the goods to a foreign person for delivery outside the United States, the goods have not been used in domestic commerce, even if the goods would cause confusion if later introduced into the U.S. market.
Under this formalist approach, the purchaser who resells the goods in the United States, and not the person who supplies that purchaser overseas, would come under the Lanham Act. Neither Justice Jackson’s nor Justice Sotomayor’s opinion mentions this possibility, even though Abitron’s brief pointed to this prospect. The Solicitor General’s brief argued that a reseller might lack the necessary scienter to incur damage liability, but this would not prevent an injunction barring future sales. Of course, a rights holder would prefer a legal rule that concentrates liability on a single, easy-to-find defendant. But, as Morrison indicates, the purposes of the presumption against extraterritoriality are not satisfied whenever one carves out an exception solely to minimize indirect domestic effects of a prescribed wrong. As Morrison recognized, such an exception swallows the rule and leaves in place the uncertainty that plagued the lower courts’ approach.
A legal theory that none of the parties raised was the possibility of accessory liability for the original seller in cases where a reseller violates the Lanham Act by importing pirated goods for sale. Accessory rules such as aiding and abetting usually are more stringent, and therefore more definite, than what the Solicitor General proposed, namely that liability should attach whenever U.S. resales are foreseeable. Instead, accessory liability might depend on the state of the original seller’s knowledge regarding likely resales. This approach makes the point of sale – the use in domestic commerce – the relevant test, but still allows an extension of liability to opportunists who abuse the formal dimension of the rule.
Not only does this result rest on a nice formal understanding of the focus of the Lanham Act, but it fits within the broader structure of U.S. trademark law. The purpose of the Act is to protect from confusion consumers purchasing on the U.S. market. Any person who purposely introduces pirated goods into that market, either directly or by using intermediaries to deliberately reach that market, should come under the Lanham Act. Accessory liability is not an extravagant extension of the law, but rather a conventional feature of most rules of liability.
The government’s approach, adopted by four Justices and not clearly rejected by Justice Jackson, would instead treat pirated goods as a kind of ultrahazardous product, to be suppressed by any means possible. The open-ended issue of reasonable foreseeability, the boundaries of which no reasonable person can predict, would substitute for a more definite rule such as purpose, knowledge or willful disregard. As in Smagin, it would reject Morrison’s key insight, that it is domestic markets, and not evil loose in the world, that most regulatory schemes should be presumed to address.
The end of the Term is when the big decisions drop, “big” defined as those that become political footballs or otherwise influence the culture. In the rush to end on time while wrestling with the big cases, the Justices may divert the time and attention necessary to work out useful accounts of the more technical ones, even if these are the opinions more likely to become lawyers’ bread and butter. This happens every year and is understandable but unfortunate. The presumption against extraterritoriality does a lot of work in an interconnected world where much vital activity is neither wholly foreign nor domestic. Smagin and Abitron seem to be good decisions reaching desirable outcomes. Together, however, the opinions leave lawyers knowing less about an important area of the law than they did before they were handed down.