The Alien Tort Statute's Translation Problem
The Alien Tort Statute (ATS) was enacted by the first Congress as part of the Judiciary Act of 1789 and, with the exception of references to particular federal courts that changed over time, has not been amended since. In current form, it provides: "The district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States."
In the last decades of the twentieth century, human rights lawyers discovered that they could use the ATS as a means for non-citizens to sue in U.S. courts for human rights violations. The first major case to vindicate this strategy was the decision by the U.S. Court of Appeals for the Second Circuit in Filartiga v. Pena-Irala, which involved a Paraguayan citizen suing another Paraguayan citizen for the torture and murder of the plaintiff's son in Paraguay, where the defendant was Inspector General of Police in Asuncion.
The circumstances of the Filartiga case gave rise to anxieties among conservatives, who worried that U.S. courts would be flooded with claims having no genuine connection to the United States. That fear was almost certainly overblown initially because most foreign human rights violators are not in the U.S. and thus not subject to personal jurisdiction in U.S. courts for actions committed in foreign countries. However, over time, the fear became somewhat more realistic, as non-citizen plaintiffs took to suing corporations that, they alleged, were complicit in human rights violations on foreign soil but, because of their activities in the U.S., were also subject to personal jurisdiction here.
In response, the Supreme Court narrowed the scope of prospective liability under the ATS in a series of cases. In Sosa v. Alvarez-Machain (2004), the Court allowed that claims brought within ATS jurisdiction included those that the framers would have recognized per Blackstone--for piracy, violations of safe conducts, and violations of the rights of ambassadors--and perhaps other customary international norms since recognized as "specific, universal, and obligatory," but held that the claim at issue in the case did not rise to that level. In Kiobel v. Royal Dutch Petroleum (2013), the Court held that the presumption against extraterritoriality applies to the ATS, effectively rendering it unusable in the U.S. for human rights violations occurring elsewhere. (One might wonder how this could be, given that piracy, which is one of the three Blackstonian violations, occurs outside the U.S. territory; the key for the Court was that piracy does not occur within the territory of a foreign sovereign.) In Jesner v. Arab Bank (2018), the Court ruled that foreign corporations can't be sued under the ATS. In Nestlé USA v. Doe (2021), the Court disallowed an ATS suit against a U.S. corporation that made some decisions in the U.S. affecting human rights violations in West Africa because it thought the U.S.-based actions were too tangential to the actions at issue on foreign soil.
Speaking for only a plurality in Nestlé, Justice Thomas said that the door left open in Sosa for violations of customary international law beyond the three Blackstonian originals should be shut. On Tuesday of this week, the Court heard argument in a case that presents the opportunity for the full Court to convert Justice Thomas's plurality view into a majority one. In Cisco Systems, Inc. v. Doe, plaintiffs/respondents are Chinese nationals who contend that Cisco designed software to assist the Chinese government in identifying and then torturing Falun Gong members. (Cisco's CEO and VP of its China subsidiary are also defendants/petititioners.) Coming to the Court at the motion-to-dismiss phase of litigation, the case avoids the specific problem identified in Nestlé because the complaint alleges more than mere passive direction or acquiescence in a foreign human rights violation. It alleges that what Cisco did in the U.S. was itself a violation of international law.
The petitioners and the U.S. as amicus push back in two ways. First, they argue that there is no civil liability for aiding and abetting under either the ATS or the Torture Victims Protection Act (TVPA, which is codified as a note to the ATS). The TVPA authorizes civil liability for an "individual who . . . subjects an individual to torture . . . or . . . extrajudicial killing." Citing Central Bank of Denver, N. A. v. First Interstate Bank of Denver, the petitioners and the U.S. invoke a presumption against aiding-and-abetting civil liability. The back and forth with respect to aiding and abetting is complicated and interesting, but I want to set it aside to focus on the petitioners' other main argument--for closing the door Sosa left open and allowing ATS liability only for the original Blackstonian categories of international law violations.
To my mind, with the exception of some questions asked by Justice Barrett, the oral argument did not really get to the heart of this issue. The problem, as Justice Barrett apparently sees it and as I see it, is one of translating a statute that was adopted in an era with profoundly different background assumptions about how law works.
Writing for the Court in Sosa, Justice Souter said that the ATS is a jurisdictional statute but doesn't create a cause of action. That might have made sense in 1789, when the federal trial courts did not have jurisdiction over most federal-law claims, but reading the ATS as purely jurisdictional today renders it redundant with the general federal question statute, 28 U.S.C. § 1331. If a federal cause of action has some source other than the ATS (about which more in a moment), then a plaintiff doesn't need the ATS as the basis for jurisdiction. That other source would provide the cause of action, while § 1331 would provide jurisdiction.
So . . . what is the source of the cause of action for the Blackstonian three? In 1789, the answer would have been the general law, which would have been understood to incorporate international law norms. But since the Supreme Court's 1938 decision in Erie Railroad Co. v. Tompkins, federal courts do not recognize general law as a source of causes of action. Following on criticisms like those of Oliver Wendell Holmes, Jr., who ridiculed the notion of "a brooding omnipresence in the sky," causes of action (and law more generally) were conceptualized by Justice Brandeis in his majority opinion in Erie as the product of some sovereign, whether state or federal, and whether speaking through the legislature or the courts, but if the latter, made rather than discovered.
When the federal courts abandoned the general law, matters that had previously been seen as governed thereby migrated elsewhere. Mostly that meant state law, as in Erie itself. But as Judge Henry Friendly argued forcefully in an important 1964 article in the NYU Law Review, while Erie held that there is no general federal common law of the sort that state courts produce governing torts, contracts, property, and the like, there remains a substantial body of specific federal common law.
For a long time, just about all judges, lawyers, and scholars assumed that incorporation of international law norms into causes of action fell within that body of federal common law. In a 1997 article in the Harvard Law Review, Professors Curtis Bradley and Jack Goldsmith challenged that view, arguing that, to the extent that customary international law is incorporated into domestic law, it is via state law, not federal law. Although that view has attracted some followers, it remains a dissident position. The prevailing view is that today the Blackstonian three are actionable because federal common law incorporates them. (Another approach that has more recently gained some traction is that general law persists. Professor Stephen Sachs is the most prominent proponent of reviving the general law and even overturning Erie. This too remains a dissident position, at least for now.)
Justice Souter's Sosa opinion extrapolated from Erie a lesson that, once courts recognize that they make rather than discover causes of action, they should do so only cautiously. For this cautious approach, he also noted that already by 2004 the Court had become much more hesitant than in prior years to infer private rights of action from federal statutes that did not expressly provide for any. The same cautious attitude, he suggested, should apply when it comes to finding causes of action for violations of customary international law.
The position favored by Justice Thomas in his Nestlé plurality and advanced by the petitioners and the U.S. in Cisco would take that caution to an extreme, disallowing any causes of action beyond the original Blackstonian three. Yet that position is in tension with itself.
Judicial reluctance or refusal to infer causes of action from statutory silence is ostensibly rooted in textualism and separation of powers. But note how difficult it is to harmonize the Blackstone-three-only position with the actual text of the ATS.
If one reads the ATS as purely jurisdictional, then it is difficult to see why it gives rise even to the three causes of action Blackstone recognized. Justice Souter's opinion in Sosa would give effect to those causes of action to keep faith with the intentions of the Congress that enacted the ATS in 1789, lest it otherwise become a dead letter. But doing so keeps faith only with their subjective intentions based on their now-discarded assumption that general law provides the cause of action. And textualists are not supposed to be much concerned with subjective legislative intentions that do not find their way into statutory text.
Meanwhile, however, if one concedes that the Blackstone three survive, one searches in vain for a basis in the ATS text to forbid any further expansion. The ATS refers to torts "committed in violation of the law of nations," not to torts committed in violation of whatever content the law of nations happened to have in 1789.
Perhaps I oughtn't to have pointed to these inconsistencies, however, because I worry that if this Court comes to think that textualism demands either allowing actions brought pursuant to the ATS to cover more than the Blackstonian three or nothing, it will choose nothing. But I point to them nonetheless for two reasons. First, I doubt my own influence over the Court's conservatives. And second, limiting the ATS to the Blackstonian three is effectively limiting it to nothing. Since Filartiga called attention to the ATS in the modern era, it appears that there have been no ATS lawsuits alleging piracy, violation of the rights of ambassadors, or violations of safe conducts.
Judge Friendly (yes, the same one I mentioned above), wrote in a 1975 Second Circuit case that the ATS "is a kind of legal Lohengrin; although it has been with us since the first Judiciary Act, § 9, 1 Stat. 73, 77 (1789), no one seems to know whence it came." Perhaps not, but if the trend of recent years holds in Cisco, everyone will know whither it goes: nowhere.
-- Michael C. Dorf