Harvard Law Review Harvard Law Review Harvard Law Review


The New Presumption Against Extraterritoriality

The full text of this Article may be found by clicking on the PDF link to the left.

Canons of statutory interpretation are sometimes said to promote continuity and stability in the law. Yet it is widely acknowledged that canons themselves often change. The presumption against extraterritoriality is a prime example. It evolved from a rule based on international law, to a canon of comity, to a tool for finding legislative intent. The presumption then fell into disuse for nearly forty years until it was reborn in EEOC v. Arabian American Oil Co. (Aramco) and substantially revised in Morrison v. National Australia Bank Ltd.

This Article makes three contributions. First, it describes the evolution of the presumption against extraterritoriality over two centuries, providing a detailed account of change in an important canon of interpretation. Second, the Article describes the new, post-2010 presumption, arguing — contrary to the conventional wisdom — that the current version of the presumption is superior to previous ones. Third, the Article addresses the problem of changing canons. It argues changing canons constitute a form of dynamic statutory interpretation, which imposes certain responsibilities: to justify the changed canon in normative terms, to explain the need for change, and to mitigate the transition costs.


As the Supreme Court has increasingly relied on canons of statutory interpretation over the past three decades, these canons have received a great deal of scholarly attention.1×1. For important early articles, see William N. Eskridge, Jr. & Philip P. Frickey, The Supreme Court, 1993 Term — Foreword: Law as Equilibrium, 108 Harv. L. Rev. 26, 65–71 (1994) (discussing canons as an interpretive regime); and Cass R. Sunstein, Interpreting Statutes in the Regulatory State, 103 Harv. L. Rev. 405, 451–89 (1989) (discussing the role of interpretive principles and articulating a number of such principles). For recent book-length treatments, see William N. Eskridge Jr., Interpreting Law: A Primer on How to Read Statutes and the Constitution (2016); and Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts (2012). Canons form an important part of what has recently been dubbed the “law of interpretation.”2×2. See William Baude & Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079, 1107–12 (2017) (discussing canons of interpretation). Professor David Shapiro famously defended interpretive canons on the ground that they promote continuity and stability in the law.3×3. See David L. Shapiro, Continuity and Change in Statutory Interpretation, 67 N.Y.U. L. Rev. 921, 925 (1992) (“[T]he dominant theme running through most interpretive guides that actually influence outcomes is that close questions of construction should be resolved in favor of continuity and against change.”); see also Eskridge & Frickey, supra note 1, at 66 (writing that canons constitute an “interpretive regime” that renders statutory interpretation “more predictable, regular, and coherent”); Amanda L. Tyler, Continuity, Coherence, and the Canons, 99 Nw. U. L. Rev. 1389, 1428 (2005) (“[M]any of the canons play a valuable role within a greater interpretive framework that protects the stability of statutory law by elevating the values of continuity, coherence, and predictability . . . .”). Yet it is widely acknowledged that canons of statutory interpretation themselves change.4×4. See Baude & Sachs, supra note 2, at 1136 (“Interpretive rules can change over time.”); Aaron-Andrew P. Bruhl, Communicating the Canons: How Lower Courts React When the Supreme Court Changes the Rules of Statutory Interpretation, 100 Minn. L. Rev. 481, 494 (2015) (“The interpretive regime of the Supreme Court has not been static over time.”); Philip P. Frickey, Interpretive-Regime Change, 38 Loy. L.A. L. Rev. 1971, 1989–90 (2005) (observing that “the particulars of even longstanding canons drift over time” and that “the Court occasionally creates new canons”); Abbe R. Gluck, Intersystemic Statutory Interpretation: Methodology as “Law” and the Erie Doctrine, 120 Yale L.J. 1898, 1988 (2011) (noting that “the canons of interpretation . . . have not been frozen in time” and that the “Supreme Court continues . . . to generate new interpretive rules”); Nina A. Mendelson, Change, Creation, and Unpredictability in Statutory Interpretation: Interpretive Canon Use in the Roberts Court’s First Decade, 117 Mich. L. Rev. 71, 78 (2018) (“[R]ather than stability, it is change that characterizes the Roberts Court’s current collection of interpretive canons.”); Brian G. Slocum, Overlooked Temporal Issues in Statutory Interpretation, 81 Temp. L. Rev. 635, 639 (2008) (“[C]ourts consider the creation and modification of the rules of statutory interpretation to be subject to judicial prerogative and frequently change the rules.”); Adrian Vermeule, The Cycles of Statutory Interpretation, 68 U. Chi. L. Rev. 149, 149 (2001) (“[T]he Court has changed its practice, and sometimes the formally stated rules, with remarkable frequency.”). The retroactive application of changed canons to statutes enacted before the changes may result in interpretations that are different from the ones the enacting Congresses would have expected. This problem has received little attention.5×5. See Slocum, supra note 4, at 640 (“[T]he temporal problems raised by the retroactive application of new or modified interpretive rules are greatly underappreciated and undertheorized in statutory interpretation scholarship.”). Among the few articles that have discussed the stability issues at length are Baude & Sachs, supra note 2, at 1132–40 (discussing what happens when interpretive rules change and who has the power to change them); Frickey, supra note 4, at 1981–86 (discussing the transition costs of interpretive-regime change); and Slocum, supra note 4, at 646–70 (considering when changes in interpretive rules should be applied retrospectively and prospectively). A few other articles have explored related questions, such as why canons change, see Vermeule, supra note 4, or how lower courts respond to those changes, see Bruhl, supra note 4. This Article focuses on the retroactivity question.

The presumption against extraterritoriality is a prime example of a canon that has changed substantially over time.6×6. This Article deals only with the federal presumption against extraterritoriality that courts apply to federal statutes. For a critical discussion of state presumptions against extraterritoriality, see William S. Dodge, Presumptions Against Extraterritoriality in State Law, 53 U.C. Davis L. Rev. 1389 (2020). The presumption began in the nineteenth century as an application of the Charming Betsy canon, requiring that statutes be construed to avoid violations of international law.7×7. See Murray v. Schooner Charming Betsy, 6 U.S. (2 Cranch) 64, 118 (1804) (“[A]n act of Congress ought never to be construed to violate the law of nations if any other possible construction remains . . . .”). When international law evolved to permit greater extraterritorial regulation, the Supreme Court kept the presumption but articulated new rationales — first, international comity8×8. See Am. Banana Co. v. United Fruit Co., 213 U.S. 347, 356 (1909) (stating that to apply the law of a place other than the place of the act “would be an interference with the authority of another sovereign, contrary to the comity of nations”). and then Congress’s primary concern with domestic conditions.9×9. See Foley Bros. v. Filardo, 336 U.S. 281, 285 (1949) (stating that the presumption is “based on the assumption that Congress is primarily concerned with domestic conditions”). The American Banana version of the presumption that the Court applied during the first half of the twentieth century turned entirely on the location of the conduct.10×10. Am. Banana, 213 U.S. at 356 (“[T]he character of an act as lawful or unlawful must be determined wholly by the law of the country where the act is done.”). When this approach would have led to results that seemed inconsistent with Congress’s intent, the Court distinguished or ignored the presumption.11×11. See infra p. 1592. After 1949, the presumption fell into disuse for four decades.12×12. See infra section I.C, pp. 1595–97. It was reborn in the 1991 case EEOC v. Arabian American Oil Co.13×13. 499 U.S. 244 (1991). (Aramco) and was applied regularly, if somewhat inconsistently, thereafter. The Aramco version of the presumption purported to be a clear statement rule,14×14. Id. at 258 (referring to Congress’s “need to make a clear statement that a statute applies overseas”). and, like American Banana’s version of the presumption, it turned entirely on the location of the conduct.15×15. See infra p. 1602.

The Supreme Court’s 2010 decision in Morrison v. National Australia Bank Ltd.16×16. 561 U.S. 247 (2010). articulated a new presumption against extraterritoriality. First, the Court said explicitly that the presumption was not a “clear statement rule” and that “context can be consulted” to determine whether the presumption has been rebutted.17×17. Id. at 265. Second, Morrison abandoned the presumption’s traditional dependence on the location of the conduct. Whether the application of a statute should be considered domestic or extraterritorial would now turn on whether the object of the statute’s “focus” was found in the United States.18×18. Id. at 266. In Morrison, the Court held that the focus of section 10(b) of the Securities Exchange Act was the transaction not the fraud. Id. at 266–67. Because the transaction in that case occurred outside the United States, applying section 10(b) was prohibited as extraterritorial, despite the fact that the fraudulent conduct occurred in the United States. Id. at 273. In RJR Nabisco, Inc. v. European Community,19×19. 136 S. Ct. 2090 (2016). the Court formalized Morrison’s approach, adopting “a two-step framework for analyzing extraterritoriality issues” that looks first for a clear indication of geographic scope and, in the absence of one, applies Morrison’s “focus” test.20×20. Id. at 2101. This new presumption against extraterritoriality has also been restated in the Restatement (Fourth) of Foreign Relations Law.21×21. Restatement (Fourth) of the Foreign Relations Law of the United States § 404 (Am. Law Inst. 2018). I served as one of the co-reporters for Part IV of the Restatement (Fourth). The views expressed in this Article are my own and should not be attributed to the American Law Institute.

Scholars have been critical of the new presumption against extraterritoriality. It has been called a “runaway canon”22×22. Maggie Gardner, RJR Nabisco and the Runaway Canon, 102 Va. L. Rev. Online 134 (2016). and a “Frankenstein’s monster.”23×23. Anthony J. Colangelo, The Frankenstein’s Monster of Extraterritoriality Law, 110 AJIL Unbound 51 (2016). But the Supreme Court shows no inclination to abandon the presumption despite repeated calls to do so.24×24. See Zachary D. Clopton, Replacing the Presumption Against Extraterritoriality, 94 B.U. L. Rev. 1, 20 (2014) (arguing that the presumption “should be abandoned”); see also Julie Rose O’Sullivan, The Extraterritorial Application of Federal Criminal Statutes: Analytical Roadmap, Normative Conclusions, and a Plea to Congress for Direction, 106 Geo. L.J. 1021, 1080–94 (2018) (arguing that the presumption should be abandoned in civil but not criminal cases); cf. Gary B. Born, A Reappraisal of the Extraterritorial Reach of U.S. Law, 24 Law & Pol’y Int’l Bus. 1, 1 (1992) (arguing for abandonment of the earlier Aramco version). The Court’s articulation of a two-step framework for applying the presumption in RJR Nabisco was unanimous, even though the Court split 4–3 on how that framework should be applied to the private right of action in the Racketeer Influenced and Corrupt Organizations Act25×25. 18 U.S.C. §§ 1961–1968 (2018). (RICO).26×26. All seven participating Justices joined the part of the Court’s opinion articulating the two-step framework, see RJR Nabisco, Inc. v. European Community, 136 S. Ct. 2090, 2096, 2101 (2016), though three of them disagreed with how the Court applied that framework to RICO’s private right of action, see id. at 2111–16 (Ginsburg, J., concurring in part, dissenting in part, and dissenting from the judgment). Justice Scalia died before the case was decided. See Adam Liptak, Antonin Scalia, Justice on the Supreme Court, Dies at 79, N.Y. Times (Feb. 13, 2016), https://nyti.ms/1XqvGem [https://perma.cc/7A83-3KAZ]. Justice Sotomayor was recused. See RJR Nabisco, 136 S. Ct. at 2111. In WesternGeco LLC v. ION Geophysical Corp.,27×27. 138 S. Ct. 2129 (2018). the Court applied the new presumption again, with the addition of two Justices who had not participated in RJR Nabisco and without a word of dissent from the new two-step framework.28×28. Id. at 2136–38. Justice Sotomayor joined the majority opinion. Id. at 2134. And while Justice Gorsuch wrote a dissent joined by Justice Breyer, he agreed with the Court’s application of the presumption. See id. at 2139 (Gorsuch, J., dissenting). At present, there does not appear to be a single member of the Court who wants to abandon the presumption against extraterritoriality.29×29. Justice Kavanaugh invoked Morrison as a circuit judge. See Miller v. Clinton, 687 F.3d 1332, 1360 n.8 (D.C. Cir. 2012) (Kavanaugh, J., dissenting). Elsewhere, he has suggested that presumptions like the one against extraterritoriality should be converted to “plain statement rule[s].” Brett M. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2156 (2016) (book review). His suggestion that the Supreme Court is already moving in that direction, see id. at 2156 & n.189, is mistaken however. See Morrison v. Nat’l Austl. Bank Ltd., 561 U.S. 247, 265 (2010) (denying that the presumption is a “clear statement rule”); see also infra notes 193–96 and accompanying text.

I argue that academic criticisms of the new presumption are misguided. The Morrison/RJR Nabisco version of the presumption is significantly more flexible than its Aramco and American Banana predecessors, and thus decidedly better. In combination with other principles of statutory interpretation and appropriate deference to administrative agencies, the new presumption against extraterritoriality provides a useful tool for courts to determine the geographic scope of federal statutory provisions.

But the problem of changing canons remains. In Morrison, the Supreme Court justified the presumption on the ground that it “preserv[es] a stable background against which Congress can legislate with predictable effects.”30×30. Morrison, 561 U.S. at 261. Nowhere did the Court acknowledge that its focus approach represented a significant departure from the Aramco version of the presumption that it had applied since 1991, to say nothing of the American Banana version that the Court was applying (inconsistently) in 1934, when section 10(b) of the Securities Exchange Act31×31. 15 U.S.C. §§ 77a–77aa (2018). was passed. In RJR Nabisco, plaintiffs argued that because Congress modeled RICO’s private right of action on the Clayton Act,32×32. Id. §§ 15–27, 52–53. RICO’s provision should be given the same geographic scope that the Supreme Court had already given the Clayton Act when RICO was passed.33×33. RJR Nabisco, 136 S. Ct. at 2109. But the Court rejected that argument, noting that it had subsequently “honed [its] extraterritoriality jurisprudence”34×34. Id. at 2110. and instead applying its “current extraterritoriality doctrine.”35×35. Id. at 2111. RJR Nabisco seemed to assert the Court’s authority to change the presumption against extraterritoriality, and to apply it retroactively, without regard to the expectations of the enacting Congress.

Rhetorically, the Supreme Court is committed to some combination of textualism and purposivism.36×36. See infra notes 442–54 and accompanying text. Changing canons, on the other hand, constitute a form of dynamic statutory interpretation in which courts apply statutes in ways that might not have been anticipated by Congress.37×37. See William N. Eskridge, Jr., Dynamic Statutory Interpretation 123–28 (1994) (discussing the role of courts in dynamic statutory interpretation). Currently, the retroactive application of changed canons operates as a “backdoor” form of interpretation that the Supreme Court generally fails to acknowledge or justify.38×38. Cf. William N. Eskridge, Jr. & Philip P. Frickey, Quasi-constitutional Law: Clear Statement Rules as Constitutional Lawmaking, 45 Vand. L. Rev. 593, 598 (1992) (describing the creation of clear statement rules in the 1980s as “backdoor constitutional lawmaking”); Anita S. Krishnakumar, Backdoor Purposivism, 69 Duke L.J. (forthcoming 2020) (manuscript at 12) (on file with the Harvard Law School Library) (describing use of canons by textualist Justices as “backdoor purposivism”). This is likely because of the theoretical problems that textualism and purposivism have with applying changed canons retroactively to existing legislation.39×39. See infra section III.B, pp. 1640–44. Some scholars have proposed applying changed canons of interpretation only prospectively, but doing so seems inconsistent with the judicial role.

Even if applying changed canons retroactively is an inevitable form of dynamism, the Supreme Court should be obligated to justify the changed canon in normative terms, to explain the need for change, and to mitigate the transition costs of moving from one interpretive regime to another. These obligations will apply differently to different canons, depending largely on the content of the new canon and the extent of reliance on the old one. Applying this framework to the new presumption against extraterritoriality, this Article concludes that its retroactive application to existing federal statutes is appropriate.

The Article makes three contributions to the literature. Part I describes the evolution of the presumption against extraterritoriality over two centuries, providing a detailed account of change in an important canon of statutory interpretation. Professor Adrian Vermeule has noted that “there are very few longitudinal studies tracing the history of particular canons.”40×40. Vermeule, supra note 4, at 182 n.72. But see Mendelson, supra note 4, at 110–23 (discussing the evolution of canons during the first decade of the Roberts Court). This Article helps fill that gap.41×41. Professor John Knox has covered some of the same ground, for example, distinguishing among the original, international law–based version of the presumption, the American Banana version, and the Aramco version. See John H. Knox, A Presumption Against Extrajurisdictionality, 104 Am. J. Int’l L. 351, 361–78 (2010). Because Knox’s excellent article was published in 2010, it had no chance to address the new presumption against extraterritoriality inaugurated in Morrison.

Part II describes the new, post-2010 presumption.42×42. Much of what was written prior to Morrison about how the presumption should be understood and applied is no longer accurate. See, e.g., William S. Dodge, Understanding the Presumption Against Extraterritoriality, 16 Berkeley J. Int’l L. 85 (1998). This Part also situates the new presumption in a broader interpretive regime for determining questions of geographic scope, a regime that also includes a principle of reasonableness in interpretation and principles of deference to administrative agencies. Finally, Part II offers an evaluation of the new regime, arguing — contrary to the conventional wisdom — that the latest version of the presumption is a decided improvement over previous ones.

Part III moves beyond the presumption against extraterritoriality to suggest a framework that is applicable to changed canons generally. It argues that changing canons constitute an inevitable form of methodological dynamism43×43. I am indebted to Professor Anita Krishnakumar for the phrase. on a Supreme Court that is rhetorically committed to textualism and purposivism. Instead of ignoring this tension, as the Court has done, this Part proposes ways of living with methodological dynamism. If the Court feels the need to change a canon of interpretation, it should explain why it is doing so using the same factors that it uses to decide when to overrule a precedent.44×44. To be clear, I suggest that the Court should consider the same factors, not that the principle of stare decisis should apply generally to canons of interpretation as others have argued. See, e.g., Sydney Foster, Should Courts Give Stare Decisis Effect to Statutory Interpretation Methodology?, 96 Geo. L.J. 1863 (2008). But see Evan J. Criddle & Glen Staszewski, Against Methodological Stare Decisis, 102 Geo. L.J. 1573 (2014). Part III also makes specific proposals for mitigating the transition costs of moving to a new interpretive regime by adhering to prior interpretations of specific statutes under old canons, by honoring Congress’s expectations when it borrows language from statutes that have been construed under old canons, and in appropriate cases by treating old canons as part of the context in applying new ones.

This Article concludes with a few words of caution. It notes that whether the new presumption against extraterritoriality fulfills its potential to produce sensible tests for the geographic scope of federal statutes ultimately depends on what the Supreme Court does with it.


* Martin Luther King, Jr. Professor of Law and John D. Ayer Chair in Business Law, University of California, Davis, School of Law. Earlier drafts were presented at the Colloquium on International Law and Politics at UC Berkeley School of Law, at the ASIL International Law in Domestic Courts Workshop at the University of Pennsylvania School of Law, and at UC Hastings College of the Law. My thanks to Kevin Benish, Ashutosh Bhagwat, Pamela Bookman, Curtis Bradley, Hannah Buxbaum, Zachary Clopton, Anthony Colangelo, John Coyle, Jennifer Daskal, Kristina Daugirdas, Scott Dodson, William Eskridge, Maggie Gardner, Franklin Gevurtz, Abbe Gluck, Andrew Kent, John Knox, Harold Hongju Koh, Anita Krishnakumar, Thomas Lee, Julian Mortenson, Austen Parrish, David Pozen, Richard Re, Shayak Sarkar, Reuel Schiller, Darien Shanske, Aaron Simowitz, Brian Slocum, David Sloss, Paul Stephan, Symeon Symeonides, Aaron Tang, Carlos Vázquez, and David Zaring for their comments, suggestions, and insights.