In the
United States Court of Appeals
For the Seventh Circuit
No. 10-3675
B OIMAH F LOMO , et al.,
Plaintiffs-Appellants,
v.
F IRESTONE N ATURAL R UBBER C O ., LLC,
Defendant-Appellee.
Appeal from the United States District Court
for the Southern District of Indiana, Indianapolis Division.
No. 1:06-cv-00627-JMS-TAB—Jane E. Magnus-Stinson, Judge.
A RGUED JUNE 2, 2011—D ECIDED JULY 11, 2011
Before B AUER, P OSNER, and M ANION, Circuit Judges.
P OSNER, 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
2 No. 10-3675
the judgment in favor of Firestone Natural Rubber Com-
pany.
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 corpora-
tion 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”? “Interna-
tional 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 (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 determina-
tion 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. Further-
No. 10-3675 3
more, 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 interna-
tional 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: 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 (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 interna-
tional 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.
The concept of customary international law is disquieting
in two respects. First, there is a problem of notice: a
4 No. 10-3675
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 promul-
gated by the obligee’s local community. Both problems
are conspicuous in the Alien Tort Statute, which con-
tains 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 (quoting The Paquete Habana,
supra, 175 U.S. at 700), that are “specific, universal, and
obligatory,” 542 U.S. at 732 (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. But like so many statements
of legal doctrine, this one is suggestive rather than
precise; taken literally it could be easily refuted. No norms
No. 10-3675 5
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 “geno-
cide” and “torture.” See, e.g., Ryan Park, “Proving Geno-
cidal 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 Objec-
tive Definition of Torture,” 67 Wash. & Lee L. Rev. 77, 82-84
(2010); Sanford Levinson, “In Quest of a ‘Common Con-
science’: Reflections on the Current Debate about Tor-
ture,” 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 under-
stood as the statement of a mood—and the mood is one
of caution.
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 heart-
beat (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
6 No. 10-3675
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 (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 corpora-
tions 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 ad-
dressed 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
No. 10-3675 7
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 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
8 No. 10-3675
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, con-
sistently with international law, be punished by an in-
ternational tribunal—for the first time in history.
We have to consider why corporations have rarely
been prosecuted criminally or civilly for violating custom-
ary 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 inflic-
tion 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 en-
gaged that a tradition would develop of punishing vio-
lations 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
No. 10-3675 9
shareholders for management to violate the law, in-
cluding the criminal law, including norms of custom-
ary 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 con-
doned 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 (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 busi-
ness activity that society wants to deter inflicts monetiz-
able 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
10 No. 10-3675
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 re-
sources to raise the probability of punishment above
what might be a very low level 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 Crimi-
nal Liability: Exploring the Uniquely American Doctrine
through Comparative Criminal Procedure,” 118 Yale L.J.
126, 129-30 (2008). It is neither surprising nor significant
No. 10-3675 11
that corporate liability hasn’t figured in prosecutions of
war criminals and other violators of customary interna-
tional law. That doesn’t mean that corporations are ex-
empt 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 vio-
lating 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 lia-
ble. 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: Con-
fusion 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 suc-
cessful, even claims against individuals; only the
United States, as far as we know, has a statute that pro-
vides a civil remedy for violations of customary inter-
national law.
We keep harping on criminal liability for violations
of customary international law in order to underscore
12 No. 10-3675
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 sanc-
tion 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
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 liabil-
ity—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 cus-
tomary international law. Kadic v. Karadži, 70 F.3d 232,
246 (2d Cir. 1995). International law imposes substan-
tive obligations and the individual nations decide how
to enforce them. Banco Nacional de Cuba v. Sabbatino, 376
U.S. 398, 422-23 (1964); Kiobel v. Royal Dutch Petroleum
Co., supra, 621 F.3d at 172-74 and n. 30 (concurring opin-
ion); 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,”
No. 10-3675 13
in International Law 411 (Malcolm D. Evans ed., 3d
ed. 2010). Justice Breyer has opined that “universal crimi-
nal jurisdiction necessarily contemplates a significant
degree of civil tort recovery as well.” Sosa v. Alvarez-
Machain, supra, 542 U.S. at 763 (concurring opinion).
This point is supported by treaties that explicitly au-
thorize national variation in methods of enforcement,
allowing civil and administrative remedies as alterna-
tives to criminal liability if the imposition of such
liability would be inconsistent with domestic law. Organi-
zation 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, proportion-
ate and dissuasive non-criminal sanctions, including
monetary sanctions, for bribery of foreign public offi-
cials”); 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. Sup-
pose 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
14 No. 10-3675
the plantation can be prosecuted as an aider and abettor
of a criminal violation of customary international
law. That remedy would stretch customary interna-
tional 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 corpora-
tions (and other employers) incentives to police their
employees that are needed because the employees them-
selves 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 em-
ployed 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 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
No. 10-3675 15
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 (1844); The Marianna Flora, 24 U.S. (11 Wheat.) 1, 40-
41 (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
16 No. 10-3675
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 ques-
tion. They supply pretty much the entire ground on
which the plaintiffs pitch their argument that the treat-
ment of children on the Liberian plantation has violated
customary international law. Two of these conven-
tions—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 Inter-
national 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 inter-
national law. (It would be as if U.S. states could forbid
the enforcement of federal law within their borders.)
No. 10-3675 17
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 viola-
tions 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 imple-
menting a particular treaty or convention (maybe be-
cause 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 educa-
tionally, 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
18 No. 10-3675
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 circum-
stances 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 Organi-
zation’s constitution, “apart from bringing the Recom-
No. 10-3675 19
mendation before the . . . competent authority or authori-
ties, 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 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 condi-
tions of children below the age of 13 that significantly
reduce longevity or create a high risk (or actuality) of
significant permanent physical or psychological impair-
ment would be deemed to violate customary interna-
tional 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 Prostitu-
20 No. 10-3675
tion 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 em-
ployees, 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 Libe-
rian 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
No. 10-3675 21
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 fami-
lies. 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 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 inter-
national 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
22 No. 10-3675
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 fungi-
cides and other chemicals to the trees. But not only do
we not know how many children work on the planta-
tion; 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 help-
ers. 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 in-
ternational 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
No. 10-3675 23
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 be-
tween 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 inter-
national 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. The first is that plaintiffs must exhaust their
legal remedies in the nation in which the alleged viola-
tion of customary international law occurred. The im-
plications of the argument border on the ridiculous;
imagine having been required to file suit in a court in
Nazi Germany complaining about genocide, before
24 No. 10-3675
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 main-
tained. 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.
A FFIRMED.
7-11-11