In 1984, in Tel-Oren v. Libyan Arab Republic, the United States Court of Appeals for the District of Columbia Circuit affirmed the district court’s dismissal of an international human rights case. In the first sentence of his concurring opinion, Judge Edwards wrote, “This case deals with an area of the law that cries out for clarification by the Supreme Court.”1 Twenty years later, the Supreme Court decided Sosa v. Alvarez-Machain.2 What, if anything, has the Court clarified?
Judge Edwards referred to tort suits brought by aliens for violations of the law of nations under the Alien Tort Statute (“ATS”), 28 U.S.C. § 1350. The ATS originated as Section 9 of the Judiciary Act of 1789. The modern text is little changed from the original. It reads:
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.3
There are two jurisdictionally interesting things about Section 9. First, Section 9 was designed to provide easy access to federal court, authorizing jurisdiction in district courts staffed by resident district judges, with no amount-in-controversy requirement. Second, although Section 9 premised jurisdiction on the presence of an alien as a party, it limited the exercise of that jurisdiction to only two specified sources of substantive rights: violations of United States treaties and torts in violation of the law of nations. There was no difficulty with subject matter jurisdiction over suits based on treaty violations. Such suits clearly came within the “arising under” jurisdiction of Section 2 of Article III, and I will not be concerned in this essay with those suits.
Instead, my concern is tort suits for violation of the law of nations. There is no indication that the adopters of Section 9 thought at the time that subject matter jurisdiction was questionable in such suits. But by the early nineteenth century it had become clear that the general law, including the law of nations, was not federal law in either the jurisdiction-conferring or supremacy-clause sense.4 It had also become clear that there was no party-based jurisdiction in suits between aliens. This meant that subject matter jurisdiction in suits between aliens for torts in violation of the law of nations almost certainly did not exist.5 Perhaps in part because of doubts about subject matter jurisdiction, Section 9 essentially disappeared for almost 200 years.
In 1980, Section 9 reappeared as in Filartiga v. Pena-Irala, decided by the Second Circuit.6 The Paraguayan survivors of a Paraguayan national who had been tortured to death in Paraguay served process on the torturer in New York and brought suit in federal district court under the ATS, alleging a tortious violation of the law of nations. Relying in part on Justice Gray’s famous statement in The Paquete Habana in 1900 that “[i]nternational law is part of our law,” Judge Irving Kaufman held for a unanimous panel that the basis for subject matter jurisdiction was “the law of nations, which has always been part of the federal common law.”7 On the merits, the court held that “an act of torture committed by a state official against one held in detention violates established norms of the international law of human rights, and hence the law of nations.”8
Filartiga’s holding that customary international law is federal common law accomplished two things. First, it solved the problem of subject matter jurisdiction. If customary international law is federal common law, a suit to enforce a right under that law is a suit “arising under” federal law. Second, it instructed American courts that established norms of international human rights under customary international law are binding on all American courts as federal common law. That is, it held that customary international law is federal law in both the jurisdiction-conferring and supremacy-clause senses.
Filartiga was the beginning of a consistent line of cases in which the lower federal courts held that established norms of international human rights based on customary international law are part of the “law of nations” under the ATS and are part of the federal common law.9 But Filartiga was based on an ahistorical premise. When Justice Gray wrote in The Paquete Habana that “international law is part of our law,” he did not mean that international law was federal law in the jurisdiction-conferring and supremacy-clause sense. Rather, he meant what such statements had meant ever since it had been settled that the general law was not federal common law. Justice Gray meant that international law was “part of our law” only in the sense that it was applied by American courts.
In 1997, Professors Bradley and Goldsmith challenged what had become the conventional wisdom under Filartiga, contending that under the concept of general law that prevailed at the time of the framing, international law was not federal law in either the jurisdiction-conferring or supremacy-clause sense.10 I generally agree with their historical thesis. An early dispute over the character of the general law in non criminal cases had been quickly resolved in favor of the view, favored by St. George Tucker, that such law was not federal law.11 As soon as this was settled, there was no possibility that international law could have been considered federal common law. It was simply general law.
When the Supreme Court granted certiorari in Sosa v. Alvarez-Machain,12 it was widely anticipated that the Court would answer some (perhaps most) of the vexing questions about the nature of the customary international law of human rights and, as a result, about the extent of the federal courts’ subject matter jurisdiction. The facts of Sosa are as follows. A Mexican agent of the United States Drug Enforcement Administration (“DEA”) was interrogated and tortured, and then murdered, in Mexico. DEA officials believed that Alvarez-Machain (“Alvarez”), a Mexican doctor, had acted to prolong the DEA agent’s life in order to extend the interrogation and torture. At the behest of the DEA, a group of Mexican citizens including Sosa abducted Alvarez from his house in Mexico, held him overnight in a motel in Mexico, and transported him to the United States where he was arrested by federal officers. Alvarez was then tried in federal court for torture and murder. Alvarez was acquitted. He then returned to Mexico and brought a civil suit against the United States under the Federal Tort Claims Act, and against Sosa under the ATS. I am today concerned only with Alvarez’s suit against Sosa under the ATS.
Alvarez’s ATS suit against Sosa was jurisdictionally identical to the suit in Filartiga. Following Filartiga, the Ninth Circuit in Sosa first stated what had become black-letter law: well established norms of customary international law are federal common law, enforceable in federal court under the ATS. Going beyond Filartiga but following its own precedent, the Ninth Circuit next stated that the ATS authorizes federal courts to create federal common law by “creat[ing] a cause of action for an alleged violation of the law of nations.”13 On the merits, the Ninth Circuit held that there is “a clear and universally recognized norm prohibiting arbitrary arrest and detention,” and that Alvarez had therefore stated a claim under federal common law.14
The Supreme Court granted certiorari “to clarify the scope of . . . the ATS.”15 Justice Souter wrote the majority opinion. Agreeing with the government, he concluded for the Court that the adopters of Section 9 of the Judiciary Act intended that the ATS be a bare grant of jurisdiction, without an accompanying direction to the federal courts to create law. But, disagreeing with the government, he concluded that at least part of the customary international law of human rights is federal common law. He wrote that the “jurisdictional grant is best read as having been enacted on the understanding that the common law would provide a cause of action for the modest number of international law violations with a potential for personal liability at the time.”16 According to Justice Souter, there were three such violations under customary international law when the ATS was adopted—“violation of safe conducts, infringement of the rights of ambassadors, and piracy.”17 Justice Souter did not limit modern courts to enforcing only those three rights under customary international law, but he required that they be defined with comparable specificity: “we think courts should require any claim based on the present-day law of nations to rest on a norm of international character accepted by the civilized world and defined with a specificity comparable to the features of the 18th-century paradigms we have recognized.”18
Justice Souter counseled caution in finding federal common law based on customary international law. First, he wrote, the “prevailing conception of the common law has changed since 1789,” and “a judge deciding in reliance on an international norm will find a substantial element of discretionary judgment in the decision.”19 Second, after the Court’s decision in Erie Railroad. v. Tompkins,20 federal courts have sharply reduced their law-making role: “[A]lthough we have even assumed competence to make judicial rules of decision of particular importance to foreign relations, such as the act of state doctrine, the general practice has been to look for legislative guidance before exercising innovative authority over substantive law.”21 Third, the creation of private causes of action generally should be left to the legislative branch. Fourth, federal courts should be particularly cautious in finding federal common law causes of action based on customary international law because of the “potential implications for the foreign relations of the United States of recognizing such causes.”22 Finally, the federal courts “have no congressional mandate to seek out and define new and debatable violations of the law of nations.”23
On the merits of Sosa’s claim, Justice Souter concluded that no norm of customary international law, defined with sufficient clarity to qualify as federal common law, had been violated: “[A] single illegal detention of less than a day, followed by the transfer of custody to lawful authorities and a prompt arraignment, violates no norm of customary international law so well defined as to support the creation of a federal remedy.”24
Justice Breyer concurred in the majority opinion, and wrote an intriguing separate opinion. In his view, judicially enforceable norms of customary international law should depend on the common practice and agreement among the courts of many nations: “Today international law will sometimes . . . reflect not only substantive agreement as to certain universally condemned behavior but also procedural agreement that universal jurisdiction exists to prosecute a subset of that behavior. . . . That subset includes torture, genocide, crimes against humanity, and war crimes.”25
Justice Scalia, joined by Chief Justice Rehnquist and Justice Thomas, objected to the entire enterprise. He agreed with the Court that the ATS was a bare jurisdictional statute, containing no direction to the federal courts to create federal common law. But he argued vigorously, citing Bradley and Goldsmith, that customary international law could never be federal common law enforceable under the ATS: “The Framers would, I am confident, be appalled by the proposition that, for example, the American peoples’ democratic adoption of the death penalty could be judicially nullified because of the disapproving views of foreigners.”26
So, here we are, twenty-four years after the Second Circuit’s decision in Filartiga, and twenty years after Judge Edwards’ plea for clarification in Tel-Oren.27 We now know two things that perhaps we did not know before.
First, we know—because the Supreme Court has told us—that there is a federal common law of international human rights based on customary international law. We do not know very much about the precise content of that law, for the Court has refused to give us any modern examples of such rights. It has told us only that the rights must be defined with comparable clarity to the definitions of safe conducts, rights of ambassadors, and piracy in 1789. Justice Breyer suggests in his separate opinion that the list of such cognizable international human rights might include at least torture, genocide, crimes against humanity, and war crimes. But at this point we can only guess whether a majority of the Court will agree with any or all of the rights on Justice Breyer’s list.
Second, we also know—though not because the Court has told us—that the federal common law of customary international law is federal law in both the jurisdiction-conferring and the supremacy-clause senses. I am somewhat surprised, given the lead-up to Sosa, that the Court did not discuss the subject matter jurisdiction problem that has haunted the ATS almost from the beginning. But despite its lack of discussion, the Court’s decision necessarily implies that the federal common law of customary international law is jurisdiction-conferring. Alvarez, a citizen of Mexico, sued Sosa, another citizen of Mexico, under the ATS. As the Supreme Court said in 1800 in Mossman v. Higginson, there can be no party-based jurisdiction when there are aliens on both sides of the case.28 Thus, the only basis for the federal court to hear an alien versus alien suit under the ATS is the federal nature of the substantive claim.29
The Court’s decision also necessarily implies that the federal common law of customary international law is federal law in the supremacy-clause sense. Supremacy is an inherent characteristic of any federal law, whether constitutional law, treaty law, statute law, or common law.30 However, to say that the federal common law of customary international law is federal law in the supremacy-clause sense is not to say in which circumstances it will apply, or to say what preemptive force it might have. The difficulty of preemption issues is suggested by the number of recent cases in which the Supreme Court has dealt with preemption in a wholly domestic setting.31 There is no reason to think that preemption issues posed by the new federal common law of customary international law of human rights will be any easier. Indeed, I think they are likely to be harder.
I will address three types of cases in which such preemption issues will arise. They hardly exhaust the universe of possibilities, but they give a sense of the nature and range of the questions presented.
First is the paradigm Filartiga case. I will consider behavior constituting cruel, inhumane and degrading treatment (“inhumane treatment”). Assume that the inhumane treatment took place in a foreign country, and that the victim as well as the perpetrator are citizens of that country. Further, assume that we have federal case law (which, of course, at the moment we do not) that tells us clearly that inhumane treatment does not violate a norm of customary international law established with sufficient clarity to satisfy the criteria of Sosa. That is, assume that under Sosa inhumane treatment does not violate the federal common law of international human rights cognizable in federal district court under the ATS.
The victim learns that the perpetrator is in the United States, files suit in state court, and serves process on him in that state. To what degree, if any, is the state court constrained by the federal courts’ conclusion that the behavior of the perpetrator has not violated federal common law? It is at least clear that the state court cannot apply the norm of customary international law against inhumane treatment as a matter of federal law. But what about two other possibilities: (1) Can the state court apply the norm simply as matter of customary international law? (2) Can the state court incorporate the norm of customary international law into state law, and then apply that norm as a matter of state law? Does the federal common law of international human rights—which does not recognize this norm—preempt the application by the state court of customary international law under either (1) or (2)? In other words, does the federal common law of international human rights operate not only as a floor (requiring state courts to enforce a federal common law norm) but also as a ceiling (preventing state courts from enforcing anything that is not a federal common law norm)?
At this early point in our understanding of Sosa’s implications, we cannot really answer these questions. But we can at least say this: At some point, an expansive definition and enthusiastic enforcement of international human rights by a state court—whether as a matter of pure international law, or of international law incorporated into state law—may well be preempted by the Constitution’s dormant implied international relations clause. Of course, there is no written international relations clause, so the clause is only implied. And in the case we are imagining there is no federal legislation based on the implied international relations clause, so the implied clause is only dormant. But we know from such cases as Zschernig v. Miller that state laws can be preempted by this dormant implied clause.32 To the degree that a state court decision is preempted by the federal common law of international human rights, this part of human rights law would be, of course, controlled by federal law. Indeed, it would not be a misuse of language to say that such preemption is therefore also part of federal common law.
Second is a hybrid case involving an American defendant and an alien plaintiff. For example, assume that an American corporation is building a gas pipeline across a country with which the corporation is in a commercial partnership. Assume that the country, with the knowledge of its American corporate business partner, conscripts its own citizens under inhumane conditions to help build the pipeline. Further, assume, as above, that there is federal court case law telling us that the inhumane treatment in this case does not violate a norm of federal common law under the Sosa analysis. Finally, assume that the corporation is incorporated in California and has its principal place of business in California.
Several of the mistreated aliens file suit in state court in California against the American corporation, alleging violation of a norm of customary international law against inhumane treatment. As above, we assume that no norm of federal common law has been violated. The same two possibilities as above remain: (1) Can the state court apply customary international law? (2) Can the state court incorporate customary international law into state law? But the preemption analysis has to be different in this case, for the state court is applying either international law, or state law incorporating international law, to a corporation of that state. Whatever interest the national government may have in the uniform application of international human rights law in courts in the United States, that interest must be counterbalanced to some degree by the interest of the State of California in regulating the behavior of its own corporation.
Third is the entirely domestic case. Here, I take as my example capital cases in which the defendant contends that customary international law forbids the death penalty. The Supreme Court has recently, as a matter of federal constitutional law, limited the application of the death penalty for juveniles and the mentally retarded, so for my purposes those cases are off the table. My concern is with death penalty cases in which defendants contend, based on customary international law, that capital punishment cannot be imposed under any circumstances.
In numerous federal and state cases, defendants have sought to use customary international law as a defense against any imposition of the death penalty. The defense has failed in all of these cases.33 There is nothing in Sosa indicating that a different answer will ever be compelled as a matter of federal common law. Justice Scalia need not be afraid that customary international law, incorporated into federal common law, will invalidate the death penalty in the United States. But the absence of a federal common law prohibition against the imposition of the death penalty almost certainly does not have any preemptive force. If a state court decides that the death penalty should be forbidden in prosecutions brought under state law, such a decision is entirely that state’s business. Whether the state court so decides as a matter of customary international law, as a matter of state law that incorporates customary international law, or entirely as a matter of state law makes no difference. The federal government simply has no interest that would justify telling the state that it must impose the death penalty for state-law crimes.
So far, I have discussed these three types of cases only as they might be litigated in state court. But the federal courts will see such cases, too. Of the greatest interest are the first two types.
The first—the paradigm Filartiga case—cannot come into federal court on its own if there is no federal common law right that satisfies Sosa’s criteria. But such a case is cognizable in federal court based on supplemental jurisdiction under 28 U.S.C. § 1367(a).34 Indeed, in Sosa itself, Alvarez’s claim against Sosa could have come into federal court under Section 1367(a) because of Alvarez’s claim, in the same case, against the United States under the Federal Tort Claims Act. In other Filartiga-type cases, all that would be required would be a non frivolous (though ultimately losing) claim against an alien based on a federal common law right that satisfies the Sosa criteria.35 Once such a non frivolous common law claim is brought, the federal district court would have supplemental jurisdiction over all other human rights claims, whether those claims are based directly on customary international law or on state law that incorporates customary international law, so long as those claims meet Section 1367(a)’s relatedness requirement.
The second type of case can come into federal district court under alienage jurisdiction. The Supreme Court held in Mossman v. Higginson in 1800 that there was no subject matter jurisdiction over an alien-alien suit, but that holding does not extend to an alien-citizen suit.36 Thus, there is subject matter jurisdiction when an alien sues an American corporation for violation of human rights under customary international law even where the right asserted does not satisfy the Sosa criteria.37
In these types of cases, state court decisions on the role of customary international law may affect litigation in federal court. If the state court directly applies customary international human rights law that does not satisfy the Sosa criteria for federal common law, the federal court may choose to follow the state court’s decision as a matter of general law, though I see nothing in Erie that would require it to do so. However, if the federal court does not choose to follow the lead of the state court, I see nothing in Sosa to prevent the federal court from directly applying customary international law entirely on its own, just as the state court has done. On the other hand, if the state court incorporates customary international law into state law, a federal court sitting in that state will have no choice. It will be required by Erie to follow the state court’s interpretation of its own state law, which, in this instance, has incorporated customary international law.38
It is obvious from the foregoing that the Court’s opinion in Sosa has only begun to tell us what we need to know. To paraphrase Judge Edwards in Tel-Oren, “This case deals with an area of law that still cries out for clarification from the Supreme Court.”
I close by quoting Professor Henkin, who has observed that “the nominal continuity in our jurisprudence masks radical development, much of it in our time.”39 It is true, as Professors Bradley and Goldsmith have pointed out, that customary international law in the nineteenth century was general rather than federal law. It is also true that admiralty law during that period was general rather than federal law.40 And it is also true that there was no such thing as federal common law during that period.
But as Professor Henkin has pointed out, our jurisprudential categories have changed radically in the last ninety years even while retaining a nominal continuity. General law, as a category of domestic law, disappeared in 1938 in Erie Railroad Co. v. Tompkins.41 Admiralty law became federal law in 1917 in Southern Pacific Co. v. Jensen.42 The act of state doctrine became federal common law in 1964 in Banco Nacional de Cuba v. Sabbatino.43 And part of the customary international law of human rights became federal common law in Sosa in 2004.
Given these changes, it is hard to argue that customary international law cannot be federal common law today simply on the ground that it was general law in the nineteenth century. It is equally hard to argue that customary international law should cease to be general law simply because domestic general law disappeared in 1938. If we are to be true to nineteenth century jurisprudential categories, customary international law should remain general law, unless and until specifically incorporated into state or federal law. If it remains general law, it is potentially applicable in the courts of the United States, both state and federal, just as it was in The Paquete Habana.44 But, of course, there is no necessary reason that this jurisprudential category should remain constant, any more than the other categories have done.
International human rights, as we understand them today, are a recent creation, and the Court’s decision in Sosa is but a way station in what promises to be a long journey. To some slight degree Sosa has clarified the law of human rights in American courts, but it has left us with more questions than answers. The answers to those questions may be suggested by nineteenth century jurisprudential categories. But those questions can be fully and properly answered only by adapting our jurisprudence to the modern world, just as those who came before us adapted their jurisprudence to what was, for them, their modern world.
Preferred citation: William A. Fletcher, International Human Rights in American Courts, 93 Va. L. Rev. In Brief 1 (2007), http://www.virginialawreview.org/inbrief/2007/03/22/fletcher.pdf.
[Back to this issue]