Boimah FLOMO, et al., Plaintiffs-Appellants,
v.
FIRESTONE NATURAL RUBBER CO., LLC, Defendant-Appellee.

*1014 Terrence P. Collingsworth, Attorney, Conrad & Scherer, LLP, Washington, DC, Paul Hoffman (argued), Attorney, Schonburn Desimone Seplow Harris Hoffman & Harrison, LLP, Venice, CA, Christian Levesque (argued), Attorney, Conrad & Scherer, LLP, Washington, DC, for Plaintiffs-Appellants.

*1015 Brian J. Murray (argued), Attorney, Jones Day, Chicago, IL, for Defendant-Appellee.

William J. Aceves, California Western School of Law, San Diego, CA, for Amicus Curiae, International Law Scholars.

Tyler R. Giannini, International Human Rights Clinic, Harvard Law School, Cambridge, MA, for Amicus Curiae, Professors of Legal History.

Jennifer M. Green, Attorney, Human Rights Litigation and International Advocacy Clinic, University of Minnesota Law School, Minneapolis, MN, for Amicus Curiae, Nuremberg Scholars.

Lisa S. Blatt, Attorney, Arnold & Porter, Washington, DC, for Amicus Curiae, Chamber of Commerce of the United States of America.

Richard A. Samp, Attorney, Washington Legal Foundation, Washington, DC, for Amicus Curiae, Washington Legal Foundation.

Before BAUER, POSNER, and MANION, Circuit Judges.

POSNER, Circuit Judge.

This suit under the Alien Tort Statute, 28 U.S.C. § 1350, pits 23 Liberian children against the Firestone Natural Rubber Company, which operates a 118,000-acre rubber plantation in Liberia through a subsidiary; various Firestone affiliates and officers were also joined as defendants. The district court granted summary judgment in favor of all the defendants, but the plaintiffs have appealed only from the judgment in favor of Firestone Natural Rubber Company.

The plaintiffs charge Firestone with utilizing hazardous child labor on the plantation in violation of customary international law. The Alien Tort Statute confers on the federal courts jurisdiction over "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." The principal issues presented by the appeal are whether a corporation or any other entity that is not a natural person (the defendant is a limited liability company rather than a conventional business corporation) can be liable under the Alien Tort Statute, and, if so, whether the evidence presented by the plaintiffs created a triable issue of whether the defendant has violated "customary international law."

And what is "customary international law"? "International law is part of our law, and ... where there is no treaty, and no controlling executive or legislative act or judicial decision, resort must be had to the customs and usages of civilized nations." The Paquete Habana, 175 U.S. 677, 700, 20 S. Ct. 290, 44 L. Ed. 320 (1900); see also Sampson v. Federal Republic of Germany, 250 F.3d 1145, 1149-50 (7th Cir. 2001); Restatement (Third) of Foreign Relations Law § 102(2) (1987); Curtis A. Bradley & Mitu Gulati, "Withdrawing from International Custom," 120 Yale L.J. 202, 208-15 (2010). "The determination of what offenses violate customary international law ... is no simple task. Customary international law is discerned from myriad decisions made in numerous and varied international and domestic arenas. Furthermore, the relevant evidence of customary international law is widely dispersed and generally unfamiliar to lawyers and judges. These difficulties are compounded by the fact that customary international law—as the term itself implies—is created by the general customs and practices of nations and therefore does not stem from any single, definitive, readily-identifiable source. All of these characteristics give the body of customary international law a `soft, indeterminate character.'" Flores v. Southern Peru Copper Corp., 414 F.3d 233, 247-48 (2d Cir.2003), quoting Louis Henkin, International Law: *1016 Politics and Values 29 (1995). Customary international law thus resembles common law in its original sense as law arising from custom rather than law that is formally promulgated. See 1 William Blackstone, Commentaries on the Laws of England 67-70 (1765).

The Alien Tort Statute was enacted in 1789, when the principal violations of customary international law were piracy, mistreatment of ambassadors, and violation of safe conducts. Sosa v. Alvarez-Machain, 542 U.S. 692, 715, 124 S. Ct. 2739, 159 L. Ed. 2d 718 (2004); 4 Blackstone, supra, at 68 (1769). But in using the broad term "law of nations" Congress allowed the coverage of the statute to change with changes in customary international law. As cautiously stated by the Supreme Court, "the door is still ajar [for further independent judicial recognition of actionable international norms] subject to vigilant doorkeeping, and thus open to a narrow class of international norms today." Sosa v. Alvarez-Machain, supra, 542 U.S. at 729, 124 S. Ct. 2739.

The concept of customary international law is disquieting in two respects. First, there is a problem of notice: a custom cannot be identified with the same confidence as a provision in a legally authoritative text, such as a statute or a treaty. (Modern common law doesn't present that problem; it is a body of judge-created doctrine, not of amorphous custom.) Second, there is a problem of legitimacy—and for democratic countries it is a problem of democratic legitimacy. Customary international legal duties are imposed by the international community (ideally, though rarely—given the diversity of the world's 194 nations—by consensus), rather than by laws promulgated by the obligee's local community. Both problems are conspicuous in the Alien Tort Statute, which contains no clarifying language, although since it's just a statute, Congress could curtail its scope; the statute therefore is not a blanket delegation of lawmaking to the democratically unaccountable international community of custom creators.

The two problems we've just noted are serious enough to have persuaded the Supreme Court in Sosa to limit the statute's scope to "the customs and usages of civilized nations," 542 U.S. at 734, 124 S. Ct. 2739 (quoting The Paquete Habana, supra, 175 U.S. at 700, 20 S. Ct. 290), that are "specific, universal, and obligatory," 542 U.S. at 732, 124 S. Ct. 2739 (quoting In re Estate of Marcos Human Rights Litigation, 25 F.3d 1467, 1475 (9th Cir.1994)), and "accepted by the civilized world and defined with a specificity comparable to the features of the 18th-century paradigms" (that is, violation of safe conducts, infringement of the rights of ambassadors, and piracy). 542 U.S. at 725, 124 S. Ct. 2739. But like so many statements of legal doctrine, this one is suggestive rather than precise; taken literally it could easily be refuted. No norms are truly "universal"; "universal" is inconsistent with "accepted by the civilized world"; "obligatory" is the conclusion not the premise; and some of the most widely accepted international norms are vague, such as "genocide" and "torture." See, e.g., Ryan Park, "Proving Genocidal Intent: International Precedent and ECCC Case 002," 63 Rutgers L.Rev. 129, 133-38 (2010); Michael W. Lewis, "A Dark Descent into Reality: Making the Case for an Objective Definition of Torture," 67 Wash. & Lee L.Rev. 77, 82-84 (2010); Sanford Levinson, "In Quest of a `Common Conscience': Reflections on the Current Debate about Torture," 1 J. Nat'l Security Law & Policy 231, 252 (2005). The Court's effort at definition illustrates rather than solves the problems of notice and legitimacy and is best understood as the statement of a mood—and the mood is one of caution.

*1017 Firestone draws on that mood for its arguments against liability. Its first argument is that conduct by a corporation or any other entity that doesn't have a heartbeat (we'll use "corporation" to cover all such entities) can never be a violation of customary international law, no matter how heinous the conduct. So, according to Firestone, a pirate can be sued under the Alien Tort Statute but not a pirate corporation (Pirates of the Indian Ocean, Inc., with its headquarters and principal place of business in Somalia; cf. U.N. Security Council, "Report of the Monitoring Group on Somalia Pursuant to Security Council Resolution 1853 (2008)" 99 (Feb. 26, 2010).) Firestone argues that because corporations, unlike individuals, have never been prosecuted for criminal violations of customary international law, there cannot be a norm, let alone a "universal" one, forbidding them to commit crimes against humanity and other acts that the civilized world abhors.

The issue of corporate liability under the Alien Tort Statute seems to have been left open in an enigmatic footnote in Sosa, 542 U.S. at 732 n. 20, 124 S. Ct. 2739 (but since it's a Supreme Court footnote, the parties haggle over its meaning, albeit to no avail). All but one of the cases at our level hold or assume (mainly the latter) that corporations can be liable. Romero v. Drummond Co., 552 F.3d 1303, 1315 (11th Cir.2008); Herero People's Reparations Corp. v. Deutsche Bank, A.G., 370 F.3d 1192, 1193, 1195 (D.C.Cir.2004); Wiwa v. Royal Dutch Petroleum Co., 226 F.3d 88, 91-92 (2d Cir.2000); Beanal v. Freeport-McMoran, Inc., 197 F.3d 161, 163 (5th Cir.1999); see also Abdullahi v. Pfizer, Inc., 562 F.3d 163, 174 (2d Cir.2009); Sarei v. Rio Tinto, PLC, 550 F.3d 822, 831 (9th Cir.2008) (en banc). (Our court hasn't addressed the issue.) The outlier is the split decision in Kiobel v. Royal Dutch Petroleum Co., 621 F.3d 111 (2d Cir.2010), which indeed held that because corporations have never been prosecuted, whether criminally or civilly, for violating customary international law, there can't be said to be a principle of customary international law that binds a corporation.

The factual premise of the majority opinion in the Kiobel case is incorrect. At the end of the Second World War the allied powers dissolved German corporations that had assisted the Nazi war effort, along with Nazi government and party organizations—and did so on the authority of customary international law. E.g., Control Council Law No. 2, "Providing for the Termination and Liquidation of the Nazi Organizations," Oct. 10, 1945, reprinted in 1 Enactments and Approved Papers of the Control Council and Coordinating Committee 131 (1945); Control Council Law No. 9, "Providing for the Seizure of Property Owned by I.G. Farbenindustrie and the Control Thereof," Nov. 30, 1945, reprinted in 1 id. 225, www.loc.gov/rr/frd/Military_Law/enactments-home.html (visited June 24, 2011). The second of these Control Orders found that I.G. Farben (the German chemical cartel) had "knowingly and prominently engaged in building up and maintaining the German war potential," and it ordered the seizure of all its assets and that some of them be made "available for reparations." Id.

And suppose no corporation had ever been punished for violating customary international law. There is always a first time for litigation to enforce a norm; there has to be. There were no multinational prosecutions for aggression and crimes against humanity before the Nuremberg Tribunal was created. "Prosecutorial responses to international crimes have occurred at both the national and international levels, with varying degrees of success. The first international tribunal was the Nuremberg IMT [International Military Tribunal] which sat between 1945 *1018 and 1946 to prosecute high-ranking Nazis." Robert Cryer, "International Criminal Law," in International Law 752, 770-71 (Malcolm D. Evans ed., 3d ed.2010); see also Jonathan A. Bush, "Nuremberg: The Modern Law of War and Its Limitations," 93 Colum. L.Rev. 2022, 2023 (1993). Doubts about the Tribunal's legitimacy focused on whether there were established international norms against such conduct, see, e.g., Cryer, supra; Jonathan A. Bush, "`The Supreme ... Crime' and Its Origins: The Lost Legislative History of the Crime of Aggressive War," 102 Colum. L.Rev. 2324, 2329-30 (2002), not on whether, if there were, violators could, consistently with international law, be punished by an international tribunal—for the first time in history.

We have to consider why corporations have rarely been prosecuted criminally or civilly for violating customary international law; maybe there's a compelling reason. But it seems not; it seems rather that the paucity of cases reflects a desire to keep liability, whether personal or institutional, for such violations within tight bounds by confining it to abhorrent conduct—the kind of conduct that invites criminal sanctions. It would have seemed tepid to charge the Nazi war criminals with battery, wrongful death, false imprisonment, intentional infliction of emotional distress, fraud, conversion, trespass, medical malpractice, or other torts. And it was natural in light of the perceived effect of the Nuremberg trials on German and international opinion concerning the type of practices in which Hitler's government had engaged that a tradition would develop of punishing violations of customary international law by means of national or international criminal proceedings; it was a way of underscoring the gravity of violating customary international law.

But this has nothing to do with the issue of corporate liability. Sometimes it's in the interest of a corporation's shareholders for management to violate the law, including the criminal law, including norms of customary international law the violation of which is deemed criminal. Criminal punishment of corporations that commit crimes is not anomalous merely because a corporation cannot be imprisoned or executed. It can be fined; and so if a crime at least ostensibly in the corporation's financial interest is committed or condoned at the managerial or board of directors level of the corporation, the corporation itself is criminally liable. New York Central & Hudson River R.R. v. United States, 212 U.S. 481, 492-94, 29 S. Ct. 304, 53 L. Ed. 613 (1909); John C. Coffee, Jr., "`No Soul to Damn: No Body to Kick': An Unscandalized Inquiry into the Problem of Corporate Punishment," 79 Mich. L.Rev. 386, 447-48 (1981). The burden of a fine on the corporation will be borne by the shareholders, who correspond to the employers of tortfeasing employees, and indirectly by the managers.

Civil liability of corporations, even when it allows the award of punitive as well as compensatory damages, is not a perfect substitute for fines because not all business activity that society wants to deter inflicts monetizable harms. A corporation might engage in fraud yet the victims be unable to prove causation. Suppose the corporation had misrepresented the efficacy of a cancer drug, but the buyers were not harmed (beyond the price of the drug, which let's assume was modest) because no substitute treatment would have been effective either. One might still want to fine the corporation, in order to increase the expected cost of fraud to it. The example illustrates that two of the fundamental techniques of criminal law are applicable to an entity that cannot be punished other than by a fine—the use of public resources to raise the probability of punishment above what might be a very low level *1019 because of efforts taken to conceal criminal responsibility, and the punishment of preparatory activity in order to reduce the net expected gain from crime.

Corporate criminal liability is criticized, see, e.g., Albert W. Alschuler, "Two Ways of Thinking about the Punishment of Corporations," 46 Am.Crim. L.Rev. 1359 (2009); Daniel R. Fischel & Alan O. Sykes, "Corporate Crime," 25 J. Legal Stud. 319 (1996), but one of the principal criticisms is that it is superfluous given civil liability, id. at 330-31, and that would be a poor reason for denying both criminal and civil liability for abhorrent conduct by a corporation. Similarly, while it is true that criminal punishment of corporations is a peripheral method of social control, adopted by few countries outside the Anglo-American sphere, it would move quickly from periphery to center if corporate civil liability were unavailable; and even though civil liability is available, the resistance (outside the Anglo-American sphere) to corporate criminal liability is eroding. See V.S. Khanna, "Corporate Criminal Liability: What Purpose Does It Serve?," 109 Harv. L.Rev. 1477, 1488-91 (1996); Edward B. Diskant, Note, "Comparative Corporate Criminal Liability: Exploring the Uniquely American Doctrine through Comparative Criminal Procedure," 118 Yale L.J. 126, 129-30 (2008). It is neither surprising nor significant that corporate liability hasn't figured in prosecutions of war criminals and other violators of customary international law. That doesn't mean that corporations are exempt from that law.

The Alien Tort Statute, moreover, is civil, and corporate tort liability is common around the world. See, e.g., Paula Giliker, Vicarious Liability in Tort: A Comparative Perspective 46-50 (2010). If a corporation complicit in Nazi war crimes could be punished criminally for violating customary international law, as we believe it could be, then a fortiori if the board of directors of a corporation directs the corporation's managers to commit war crimes, engage in piracy, abuse ambassadors, or use slave labor, the corporation can be civilly liable. Kiobel v. Royal Dutch Petroleum Co., supra, 621 F.3d at 170 (concurring opinion); see Doug Cassel, "Corporate Aiding and Abetting of Human Rights Violations: Confusion in the Courts," 6 Nw. U.J. Int'l Human Rights 304, 322-23 (2008). The board members would be liable as well, but they might not have the resources to compensate the victims of the corporation's violation of international customary law, let alone pay punitive damages as well.

If a plaintiff had to show that civil liability for such violations was itself a norm of international law, no claims under the Alien Tort Statute could ever be successful, even claims against individuals; only the United States, as far as we know, has a statute that provides a civil remedy for violations of customary international law.

We keep harping on criminal liability for violations of customary international law in order to underscore the distinction between a principle of that law, which is a matter of substance, and the means of enforcing it, which is a matter of procedure or remedy. Suppose it's the case that the only actionable violations of customary international law—which is to say violations that all countries are deemed to have a legal obligation to take appropriate action against—are acts so maleficent that criminal punishment would be an appropriate sanction for the actors. It would not follow that civil sanctions would be improper. If a corporation has used slave labor at the direction of its board of directors, then whether the board members should be prosecuted as criminal violators of customary international law—or also or instead be forced to pay damages, compensatory *1020 and perhaps punitive as well, to the slave laborers—or, again also or instead, whether the corporation should be prosecuted criminally and/or subjected to tort liability—all these would be remedial questions for the tribunal, in this case our federal judiciary, to answer in light of its experience with particular remedies and its immersion in the nation's legal culture, rather than questions the answers to which could be found in customary international law. Kadic v. Karadzić, 70 F.3d 232, 246 (2d Cir.1995). International law imposes substantive obligations and the individual nations decide how to enforce them. Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 422-23, 84 S. Ct. 923, 11 L. Ed. 2d 804 (1964); Kiobel v. Royal Dutch Petroleum Co., supra, 621 F.3d at 172-74 and n. 30 (concurring opinion); Tel-Oren v. Libyan Arab Republic, 726 F.2d 774, 777-78 (D.C.Cir.1984) (concurring opinion); Eileen Denza, "The Relationship between International and National Law," in International Law 411 (Malcolm D. Evans ed., 3d ed.2010). Justice Breyer has opined that "universal criminal jurisdiction necessarily contemplates a significant degree of civil tort recovery as well." Sosa v. Alvarez-Machain, supra, 542 U.S. at 763, 124 S. Ct. 2739 (concurring opinion).

This point is supported by treaties that explicitly authorize national variation in methods of enforcing customary international law—allowing civil and administrative remedies as alternatives to criminal liability if the imposition of such liability would be inconsistent with domestic law. Organization for Economic Cooperation and Development Convention on Combating Bribery of Foreign Public Officials in International Business Transactions, arts. 2-3, Nov. 21, 1997 ("in the event that, under the legal system of a Party [to the convention], criminal responsibility is not applicable to legal persons, that Party shall ensure that legal persons shall be subject to effective, proportionate and dissuasive non-criminal sanctions, including monetary sanctions, for bribery of foreign public officials"); United Nations International Convention for the Suppression of the Financing of Terrorism, art. 5, Dec. 9, 1999; United Nations Convention Against Transnational Organized Crime, art. 10, Nov. 15, 2000. The Alien Tort Statute is a further illustration.

We grant that rights and remedies can't be divorced so neatly as we may seem to be suggesting. Suppose the treatment of children at Firestone's Liberian plantation does violate customary international law (our next question), and suppose Spain decreed that anyone who buys tires made from the rubber produced at the plantation can be prosecuted as an aider and abettor of a criminal violation of customary international law. That remedy would stretch customary international law too far. The case of corporate liability is less extreme. But in the United States the liability of a corporation for torts committed by its employees in the course of their employment is strict, on the theory that strict liability for employees' torts gives corporations (and other employers) incentives to police their employees that are needed because the employees themselves will usually be judgment proof and hence not responsive to tort sanctions. The theory attenuates when the employees include local residents of Third World countries, such as the Liberian rubber farmers employed on Firestone's plantation. American corporations that have branches in backward or disordered countries may be incapable of preventing abuses of workers in those countries.

But the concern we've just expressed is an objection not to corporate liability for violations of customary international law but to the scope of that liability; and the plaintiffs concede that corporate liability for such violations is limited to cases in *1021 which the violations are directed, encouraged, or condoned at the corporate defendant's decisionmaking level. That is analogous to the liability of municipalities under the Monell doctrine, where as we noted recently "a person who wants to impose liability on a municipality for a constitutional tort must show that the tort was committed (that is, authorized or directed) at the policymaking level of government—by the city council, for example, rather than by the police officer who made an illegal arrest." Vodak v. City of Chicago, 639 F.3d 738, 747 (7th Cir.2011). We needn't decide how far corporate vicarious liability for violations of customary international law extends; it's enough that we see no objection to corporate civil liability as circumscribed as the plaintiffs concede.

And if precedent for imposing liability for a violation of customary international law by an entity that does not breathe is wanted, we point to in rem judgments against pirate ships. E.g., The Malek Adhel, 43 U.S. (2 How.) 210, 233-34, 11 L. Ed. 239 (1844); The Marianna Flora, 24 U.S. (11 Wheat.) 1, 40-41, 6 L. Ed. 405 (1825). Of course the burden of confiscation of a pirate ship falls ultimately on the ship's owners, but similarly the burden of a fine imposed on a corporation falls ultimately on the shareholders.

One of the amicus curiae briefs argues, seemingly not tongue in cheek, that corporations shouldn't be liable under the Alien Tort Statute because that would be bad for business. That may seem both irrelevant and obvious; it is irrelevant, but not obvious. Businesses in countries that have and enforce laws against child labor are hurt by competition from businesses that employ child labor in countries in which employing children is condoned.

Having satisfied ourselves that corporate liability is possible under the Alien Tort Statute, we turn to the question whether the treatment of child labor at the Firestone plantation alleged by the plaintiffs during a period of undetermined length preceding the filing of this lawsuit violated customary international law (itself a two-part question: what is the applicable customary international law and did the working conditions at the plantation violate it?), and whether Firestone is liable under the narrow standard for corporate liability proffered by the plaintiffs. We don't understand the plaintiffs to be arguing that Firestone's violation (if it was a violation) of customary international law was of criminal gravity. But neither do we understand Firestone to be arguing that only violations that grave are actionable under the Alien Tort Statute, a question we left open earlier in this opinion.

Three international conventions bear on the first question. They supply pretty much the entire ground on which the plaintiffs pitch their argument that the treatment of children on the Liberian plantation has violated customary international law. Two of these conventions—the United Nations Convention on the Rights of the Child and the International Labour Organization Minimum Age Convention—have not been ratified by the United States, though the third has been—the International Labour Organization Worst Forms of Child Labour Convention. It happens to be the one most helpful to the plaintiffs. And anyway conventions that not all nations ratify can still be evidence of customary international law. Abdullahi v. Pfizer, Inc., supra, 562 F.3d at 176-77; Khulumani v. Barclay Nat'l Bank Ltd., 504 F.3d 254, 283-84 (2d Cir.2007) (concurring opinion); see also Kiobel v. Royal Dutch Petroleum Co., supra, 621 F.3d at 137-38. Otherwise every nation (or at least every "civilized" nation) would have veto power over customary international law. (It would be as if U.S. states could *1022 forbid the enforcement of federal law within their borders.) Moreover, a nation's legislature might refuse to ratify a convention for reasons unrelated to the convention's core principle.

The United States has enacted legislation making violations of customary international law actionable in U.S. courts: it is the Alien Tort Statute. And so the fact that Congress may not have enacted legislation implementing a particular treaty or convention (maybe because the treaty or convention hadn't been ratified) does not make a principle of customary international law evidenced by the treaty or convention unenforceable in U.S. courts.

Article 32(1) of the United Nations Convention on the Rights of the Child (1989) provides that a child has a right not to perform "any work that is likely to be hazardous or to interfere with the child's education, or to be harmful to the child's health or physical, mental, spiritual, moral or social development." That's much too vague and encompassing to create an international legal norm. Millions of middle-class American children are working part-time after school at jobs that confer no intellectual or characterological benefits merely to obtain pin money for buying video games also barren of intellectual or other benefits, the jobs and the games actually functioning to diminish the children educationally, mentally, physically, and spiritually. Shall their parents, and their employers, be hauled before an international tribunal to answer charges of child abuse?

ILO Convention 138: Minimum Age Convention (1973) provides that children should not be allowed to do other than "light work" unless they're at least 14 years old. But the concept of light work is vague, and it must vary a great deal across nations because of variance in social and economic conditions.

More promising for the plaintiffs is the International Labour Organization's Convention 182: The Worst Forms of Child Labour (June 17, 1999), which as we said is the one the United States has ratified. It provides, so far as bears on this case, that the worst forms of child labor include "work which, by its nature or the circumstances in which it is carried out, is likely to harm the health, safety or morals of children." Id., art. 3(d). This is still pretty vague, in part because no threshold of actionable harm is specified, in part because of the inherent vagueness of the words "safety" and "morals." And it is weakened by the further statement that "the types of work referred to under Article 3(d) shall be determined by national laws or regulations or by the competent authority." Art. 4(1). That sounds like forswearing the creation of an international legal norm.

The Convention's Recommendation 190 adds some stiffening detail; it explains that Article 3(d) encompasses "work in an unhealthy environment which may, for example, expose children to hazardous substances, agents or processes, or to temperatures, noise levels, or vibrations damaging to their health," and "work under particularly difficult conditions such as work for long hours." But a "Recommendation" creates no enforceable obligations; according to the International Labour Organization's constitution, "apart from bringing the Recommendation before the . . . competent authority or authorities, no further obligation shall rest upon the Members, except that they shall report to the Director-General of the International Labour Office, at appropriate intervals as requested by the Governing Body, the position of the law and practice in their country in regard to the matters dealt with in the Recommendation, showing the extent to which effect has been given, or is proposed to be given, to the *1023 provisions of the Recommendation and such modifications of these provisions as it has been found or may be found necessary to make in adopting or applying them." ILO Constitution, art. 19(6)(d).

Given the diversity of economic conditions in the world, it's impossible to distill a crisp rule from the three conventions. We would like to think that working conditions of children below the age of 13 that significantly reduce longevity or create a high risk (or actuality) of significant permanent physical or psychological impairment would be deemed to violate customary international law, but we cannot be certain even of that. The plaintiffs have furnished no "concrete evidence of the customs and practices of States" to show that states feel themselves under a legal obligation to impose liability on employers of child labor in our hypothetical case. Flores v. Southern Peru Copper Corp., 414 F.3d 233, 250-52 (2d Cir.2003). Such evidence is readily available for the other types of child labor listed in ILO Convention 182, such as sexual exploitation of children and forced child labor. See, e.g., 22 U.S.C. §§ 7102(8), 7104(i); United Nations Optional Protocol to the Convention on the Rights of the Child on the Sale of Children, Child Prostitution and Child Pornography, arts. 3-4, May 25, 2000; Siliadin v. France, 73316/01, Council of Europe: European Court of Human Rights, July 26, 2005; Hadijatou Mani Koraou v. Republic of Niger, Judgment No. ECW/CCJ/JUD/06/08, ¶¶ 76-82, Economic Community of West African States Community Court of Justice (2008). But not for the child labor in our example; and anyway the working conditions at the Firestone plantation, while bad, are not that bad—more precisely, the plaintiffs haven't presented evidence that would create a triable issue of whether they're that bad.

Although Firestone doesn't employ children, at least directly, it sets high daily production quotas for its employees, who are poor Liberian agricultural workers. It is difficult for an employee to make his daily quota without help, and there is evidence that if he fails to make it he loses his job. These jobs are well paid by Liberian standards—in 2007 the average annual income of tappers (rubber farmers) on the Firestone plantation was $1559, though Liberia's per capita GDP was only $218 (but this figure probably excludes a fair amount of in-kind and unreported income)—so the employees have a strong incentive to fulfill their daily quota. They can assure fulfillment by hiring other poor Liberians to help them; and because Firestone's Liberian employees are paid well by local standards, they can hire helpers cheaply. But alternatively they can dragoon their wives or children into helping them, at no monetary cost; and this happens, though how frequently we don't know.

We can't tell from the record whether Firestone has adopted effective measures for keeping children from working on the plantation. The plantation covers 186 square miles, which is roughly the size of Chicago, and thousands of people live there—approximately 6500 employees of Firestone plus the members of their families. We don't know how many supervisors Firestone has deployed on the plantation, and hence whether there are enough of them to prevent employees from using their children to help them. We don't know the supervisors' routines, or how motivated they are to put a stop to any child labor they observe. Firestone claims that it now has a policy of firing employees who use their children as helpers, but it didn't have such a policy prior to 2005. The suit was filed that year (initially in California, but it was transferred to the district court in Indiana the following year, which is why the docket number in the district court has 06 in it), and though it is unclear when *1024 Firestone's alleged violation of international law began—because it is unclear when the principle of customary international law invoked (or imagined) by the plaintiffs came into existence—it certainly began before 2005. And there is evidence that some of the supervisors had observed child labor during the period (whatever exactly it is) of alleged liability and done nothing to stop it. There is also evidence that the company's decisionmakers were aware of, and may even have condoned, some child labor on the plantation.

But does this add up to a violation of customary international law? "Agriculture is the sector with the most child labourers. It is also the sector with the most potential for decent work for rural children and young adolescents who have reached the legal minimum age of employment." International Labour Organization, "Children in Hazardous Work: What We Know, What We Need to Do" 21 (2011). But there is agricultural work and agricultural work. Harvesting rubber is hard, and to a degree hazardous, work. It involves cutting the bark of the rubber trees with machetes to expose the latex inside the tree trunk, draining the latex into buckets, carrying the buckets—which are heavy when they are full of latex—long distances, and applying fungicides and other chemicals to the trees. But not only do we not know how many children work on the plantation; we do not know, except for the 23 plaintiff children, whose ages range from six to sixteen and whose claims may differ from those of many other child workers on the plantation (which is why the district court refused to certify the suit as a class action—and the plaintiffs have not challenged that ruling), how much work the average child does, how hard that work is, and how many children work as hard as the plaintiff children attest to having worked.

Remember too that Firestone doesn't employ children; the argument rather is that by imposing tough quotas it induces its employees to enlist their children as helpers. The plaintiffs' basic objection seems therefore to be to the quotas. This implies that courts must in a case such as this determine on an employer-by-employer basis what level of production quotas violates customary international law by encouraging oppressive child labor.

We also—and this is the biggest objection to this lawsuit—don't know the situation of Liberian children who don't live on the Firestone plantation. Conceivably, because the fathers of the children on the plantation are well paid by Liberian standards, even the children who help their fathers with the work are, on balance, better off than the average Liberian child, and would be worse off if their fathers, unable to fill their daily quotas, lost their jobs or had to pay adult helpers, thus reducing the family's income. There is a tradeoff between family income and child labor; children are helped by the former and hurt by the latter; we don't know the net effect on their welfare of working on the plantation. Pranab Bardhan, "Some Up, Some Down," in Can We Put an End to Sweatshops? 49, 50-51 (Joshua Cohen & Joel Rogers eds.2001); Frederick B. Jonassen, "A Baby-Step to Global Labor Reform: Corporate Codes of Conduct and the Child," 17 Minn. J. Int'l L. 7, 25-26 (2008).

In short, we have not been given an adequate basis for inferring a violation of customary international law, bearing in mind the Supreme Court's insistence on caution in recognizing new norms of customary international law in litigation under the Alien Tort Statute.

So the suit must fail, but for completeness we note two arguments by the defendant against liability that we reject. *1025 The first is that plaintiffs must exhaust their legal remedies in the nation in which the alleged violation of customary international law occurred. The implications of the argument border on the ridiculous; imagine having been required to file suit in a court in Nazi Germany complaining about genocide, before being able to sue under the Alien Tort Statute. What is true is that a U.S. court might, as a matter of international comity, stay an Alien Tort suit that had been filed in the U.S. court, in order to give the courts of the nation in which the violation had occurred a chance to remedy it, provided that the nation seemed willing and able to do that. Sarei v. Rio Tinto, PLC, supra, 550 F.3d at 831-32. Liberia is not able.

And second, the defendant argues that the statute has no extraterritorial application, except to violations of customary international law that are committed on the high seas. Courts have been applying the statute extraterritorially (and not just to violations at sea) since the beginning; no court to our knowledge has ever held that it doesn't apply extraterritorially; and Sosa was a case of nonmaritime extraterritorial conduct yet no Justice suggested that therefore it couldn't be maintained. Deny extraterritorial application, and the statute would be superfluous, given the ample tort and criminal remedies against, for example, the use of child labor (let alone its worst forms) in this country.

To sum up, although we disagree with the district court's ruling that corporations cannot be held liable for violating the Alien Tort Statute and we reject many of the defendant's arguments, we agree with the judgment.

AFFIRMED.