Khulumani v. Barclay National Bank Ltd.

05-2141 -cv, 05-2326-cv
Khulumani v. Barclay National Bank, Ltd., Ntsebeza v. Daimler Chrysler Corp.

UNITED STATES COURT OF APPEALS
FOR THE SECOND CIRCUIT

August Term, 2005

(Argued: January 24, 2006 Decided: October 12, 2007)

Docket Nos. 05-2141-cv, 05-2326-cv
__________________________________________________________

SAKWE BALINTULO KHULUMANI, as personal representative of Saba Balintulo,
FANEKAYA DABULA, as personal representative of Lungile Dabula, NOKITSIKAYE
VIOLET DAKUSE, as personal representative of Tozi Skweyiya, BERLINA DUDA, as personal
representative of Donald Duda, MARK FRANSCH, as personal representative of Anton Fransch,
SHERIF MZWANDILE GEKISO, as personal representative of Ntombizodwa Annestina
Nyongwana, ELSI GUGA, as personal representative of James Guga, JOYCE HLOPHE, as
personal representative of Jeffrey Hlophe, NOMVULA EUNICE KAMA, as personal
representative of Mncedisi Dlokova, JOYCE LEDWABA, as personal representative of Samuel
Ledwaba, JOHANA LERUTLA, as personal representative of Matthews Lerutla, FRIEDA Z.
LUKHULEI, as personal representative of Tokkie Lukhulei, ELIZABETH MAAKE, as personal
representative of Jackson Maake, ARCHITON MADONDO, as personal representative of
Mandla Madondo, BENJAMIN MAIFADI, TSHEMI MAKEDAMA, as personal representative
of Lugile Makedama, MABEL MAKUPE, as personal representative of Andrew Makupe,
MABEL MALOBOLA, as personal representative of Malobola Mbuso, EVELYN MATISO, as
personal representative of Pitsi Matiso, BETTY MGIDI, as personal representative of Jeffrey
Mgidi, ELIZABETH MKHONWANA, as personal representative of Obed Mkhonwana,
CATHERINE MLANGENI, as personal representative of Bheki Mlangeni, CECIL MLANJENI,
as personal representative of Kele Mlanjeni, SAMUEL MORUDU, as personal representative of
Sannah P. Leslie, TSHIDISO MOTASI, as personal representative of John and Penelope Moloke,
WILLIE NELANI, as personal representative of Mongezi Nelani, CATHERINE NGQULUNGA,
as personal representative of Brian Ngqulunga, CATHERINE PHIRI, as personal representative
of Thomas Phiri, ELIZABETH SEFOLO, as personal representative of Harold Sefolo, MARIA
SIBAYA, as personal representative of Jeffrey Sibaya, PATRICIAL M. SONGO, as personal
representative of Dipulo Songo, MPOLONTSI TYOTE, as personal representative of Boyboy
Tyote, NOMKHANGO SKOLWENI DYANTYI, CLIFFORD ZIXELILE FUDUKILE,
WINDOVOEL GAAJE, CHARLES HLATSHWAYO, MOSES HLONGWANE, LESIBA
KEKANA, SANAKI MAHLATSHI, ROBERT MAKANA, ZAKHARIA FIKILE MAMBA,
ELLIOT SITHEMBISO MARENENE, ALFRED MASEMOLA, MAUREEN THANDI
MAZIBUKO, MICHAEL MBELE, LAETITIA NOMBAMBO MFECANE, as personal
representative of Rubin Mfecane, DENNIS MLANDELI, TEFO MOFOKENG,
MOTLALETSATSI MOLATEDI, AZARIEL MOLEBELELI, SIMON MOLOTSI, LINA
MOREANE, as personal representative of Albert Xaba, THABISO SAMUEL MOTSIE, SONTO
NDLOVU, MANGINDIVA ROBERT RHENENE, THOBILE SIKANI, BUBELE STEFANE,
NOLUTHANDO BILETILE, LESLIE MNCEDISI BOTYA, LEON DUKASCHE, ELSIE
GISHI, DORTHIA GOMO-PEFILE, ZAMIKHAYA BISHOP KHALI, JAMES MAGABANA,
NOSIPHO MANQUBA, NOTATHU EUGENIA MATOMELA, NOMISA THERSIA MAY,
MBONGENI NELSON MBESHU, MZUHLANGENA NAMA, ELIAS NGAMANI, as personal
representative of Elizabeth Nagamani, GESHIA NGOXZA, LUCAS NDUKWAYIBUZWA
NGWENYANA, WELLINTON MTYUKATO NKOSIPHENDULE, VUYANI NONGCAMA,
SINDISWA MIRRIAM NUNU, THULANI NUNU, BONIWE PHALAZA, PATHISWA
PRINGANE, as personal representative of Mthozama Theophilus Pringane, MTHUTUZELI
SIKANI, NOLUTHANDO SILETILE, THEMBEKA VICTORIA SIPHAHO, JOHANNES
TITUS, MPOLONTSI TYOTES, MTHUZIMELE MEFORD YAMILE, NTUNANI WILLIAM
ZENANI, THANDIWE SHEZI, ELIAS B. BONENG, DENNIS VINCENT FREDERICK
BRUTUS, MORALOKI A. KGOBE, REUBEN MPHELA, LULAMILE RALRALA,

Plaintiffs-Appellants,

-v.-

BARCLAY NATIONAL BANK LTD., BRITISH PETROLEUM, PLC, CHEVRONTEXACO
CORPORATION, CHEVRONTEXACO GLOBAL ENERGY, INC., CITIGROUP, INC.,
COMMERZBANK, CREDIT SUISSE GROUP, DAIMLERCHRYSLER AG, DEUTSCHE
BANK AG, DRESDNER BANK AG, EXXONMOBIL CORPORATION, FORD MOTOR
COMPANY, FUJITSU, LTD., GENERAL MOTORS CORPORATIONS, INTERNATIONAL
BUSINESS MACHINES CORP., J.P. MORGAN CHASE, SHELL OIL COMPANY, and
UBS AG,

Defendants-Appellees,

AEG DAIMLER-BENZ INDUSTRIE, FLUOR CORPORATION, RHEINMETALL GROUP
AG, RIO TINTO GROUP, TOTAL-FINA-ELF and DOE CORPORATIONS,

Defendants.
________________________________________________________________

LUNGISILE NTSEBEZA,* HERMINA DIGWAMAJE, ANDILE MFINGWANA, F.J. DLEVU,
LWAZI PUMELELA KUBUKELI, FRANK BROWN, SYLVIA BROWN, NYAMEKA
GONIWE, SIGQIBO MPENDULO, DOROTHY MOLEFI, THEMBA MEQUBELA, LOBISA
IRENE DIGWAMAJE, KAELO DIGWAMAJE, LINDIWE PETUNIA LEINANA,
MATSHIDISO SYLVIA LEINANA, KELEBOGILE PRUDENCE LEINANA, DAVID
MOTSUMI, SARAH NKADIMENG, MOEKETSI THEJANE, MOSHOESHOE THEJANE,

*
We direct the Clerk of Court to amend the official caption to reflect this spelling of Ntsebeza’s
name, which is consistent with the spelling used in his briefs and complaint.

2
PASCALINAH BOOKIE PHOOFOLO, KHOBOTLE PHOOFOLO, GLADYS MOKGORO,
JONGANI HUTCHINGSON, SEFUBA SIDZUMO, GOBUSAMANG LAURENCE
LEBOTSO, EDWARD THAPELO TSHIMAKO, RAHABA MOKGOTHU, JONATHAN

MAKHUDU LEDIGA, ANNA LEBESE, SIPHO STANLEY LEBESE, WILLIAM NBOBENI,
JOHN LUCAS NGOBENI, CLEMENT HLONGWANE and MASEGALE MONNAPULA,

Plaintiffs-Appellants,

SAKWE BALINTULO KHULUMANI, P.J. OLAYI, WELLINGTON BANINZI GAMAGU,
Violations of Pass Laws, unlawful detention 1981-1983, torture subjected to discriminatory labor
practices 1981 and WILLIAM H. DURHAM,

Plaintiffs,

v.

DAIMLER CHRYSLER CORPORATION, NATIONAL WESTMINSTER BANK PLC,
COLGATE PALMOLIVE, BARCLAYS BANK PLC, UBS AG, CITIGROUP INC.,
DEUTSCHE BANK AG, DRESDNER BANK AG, COMMERZBANK AG, FORD MOTOR
COMPANY, HOLCIM, INC., EXXON MOBIL CORPORATION, SHELL OIL COMPANY,
J.P. MORGAN, MINNESOTA MINING AND MANUFACTURING CO. (3M CO.),
GENERAL ELECTRIC COMPANY, BRISTOL-MEYERS SQUIBB CO., E.I. DUPONT DE
NEMOURS, XEROX CORPORATION, IBM, GENERAL MOTORS, HONEYWELL
INTERNATIONAL, INC., BANK OF AMERICA, N.A., THE DOW CHEMICAL COMPANY,
COCA-COLA CO., CREDIT AGRICOLE S.A., HEWLETT-PACKARD COMPANY, EMS-
CHEMIE (NORTH AMERICA) INC., CHEVRON TEXACO CORPORATION, AMERICAN
ISUZU MOTORS, INC. and NESTLE USA, INC.,

Defendants-Appellees,

SULZER AG, SCHINDLER HOLDING AG, ANGLO-AMERICAN CORPORATION,
DEBEERS CORPORATION, NOVARTIS AG, BANQUE INDO SUEZ, CREDIT LYONNAIS,
and Unknown officers and directors of DANU INTERNATIONAL, STANDARD
CHARTERED, P.L.C., CORPORATE DOES, CREDIT SUISSE GROUP, CITIGROUP AG,
SECURITIES INC., as successor to Morgan Guaranty, MANUFACTURERS HANNOVER,
CHEMICAL BANK & CHASE MANHATTAN BANK, UNISYS CORPORATION, SPERRY
CORPORATION, BURROUGHS CORPORATION, ICL, LTD., AMDAHL CORP., Computer
Companies, JOHN DOE CORPORATION, HOLCIN, LTD., HENRY BLODGET, JUSTIN
BALDAUF, KRISTEN CAMPBELL, VIRGINIA SYER GENEREUX, SOFIA GHACHEM,
THOMAS MAZZUCCO, EDWARD McCABE, DEEPAK RAJ, JOHN 1-10 DOE, OERLIKON

3
CONTRAVES AG, OERLIKON BUHRLE AG, CORPORATE DOES 1-100, ROYAL DUTCH
PETROLEUM CO., SHELL TRANSPORT & TRADING COMPANY PLC and SHELL
PETROLEUM, INC., MERRILL LYNCH & CO. INC., KENNETH SEYMOUR,

Defendants.
__________________________________________________________________

BEFORE: KATZMANN and HALL, Circuit Judges, and KORMAN, District Judge.**

Appeal from a judgment of the United States District Court for the Southern District of

New York (Sprizzo, J.), dismissing, inter alia, plaintiffs-appellants’ claims under the Alien Tort

Claims Act (“ATCA”) and the Torture Victim Protection Act (“TVPA”). We affirm the district

court’s dismissal of the TVPA claims. We vacate that portion of the district court’s judgment

dismissing the plaintiffs’ ATCA claims, as well as the district court’s denial of the Digwamaje,

and Ntsebeza motions to amend, and remand for further proceedings consistent with this opinion.

Judge KATZMANN files a separate concurring opinion. Judge HALL files a separate concurring

opinion. Judge KORMAN files a separate opinion, concurring in part II and dissenting from

parts III, IV, and V of the per curiam opinion.

Affirmed in part, vacated in part, and remanded.

MICHAEL D. HAUSFELD
Cohen, Milstein, Hausfeld & Toll, P.L.L.C.
Washington, D.C. for
Khulumani Plaintiffs-Appellants.

PAUL L. HOFFMAN
Schonbrun DeSimone Seplow Harris
& Hoffman LLP
Venice, California for
Ntsebeza & Digwamaje Plaintiffs-Appellants.

**
The Honorable Edward R. Korman, United States District Judge for the Eastern District of
New York, sitting by designation.

4
 FRANCIS P. BARRON
Cravath, Swaine & Moore LLP
New York, New York for
Defendants-Appellees.

ROBERT M. LOEB
United States Department of Justice
Washington, D.C. for
Amicus Curiae United States of America.***

PER CURIAM :

I

The plaintiffs in this action bring claims under the Alien Tort Claims Act, 28 U.S.C. §

1350 (“ATCA”), against approximately fifty corporate defendants and hundreds of “corporate

Does.” The plaintiffs argue that these defendants actively and willingly collaborated with the

government of South Africa in maintaining a repressive, racially based system known as

“apartheid,” which restricted the majority black African population in all areas of life while

providing benefits for the minority white population.

Three groups of plaintiffs filed ten separate actions in multiple federal district courts

asserting these apartheid-related claims. See In re S. African Apartheid Litig., 346 F. Supp. 2d

538, 542 (S.D.N.Y. 2004). One group, the Khulumani Plaintiffs, filed a complaint against

twenty-three domestic and foreign corporations, charging them with various violations of

international law.1 The other two groups, the Ntsebeza and Digwamaje Plaintiffs, brought class

***
A dozen entities participated as amici, but only the United States participated in oral argument
of the appeal.
1
The Khulumani Plaintiffs include the Khulumani Support Group, a South African non-
governmental organization that “works to assist victims of apartheid-era violence and has 32,700
members who are survivors of such violence,” as well as ninety-one individual plaintiffs who are “the

5
action claims on behalf of the “victims of the apartheid related atrocities, human rights’

violations, crimes against humanity and unfair [and] discriminatory forced labor practices.” The

Digwamaje Plaintiffs also brought claims under the Torture Victim Protection Act of 1991, Pub.

L. 102-256, 106 Stat. 73 (1992), codified at 28 U.S.C. § 1350 note (“TVPA”), and the Racketeer

Influenced and Corrupt Organizations Act, 18 U.S.C. §§ 1961 et seq. (“RICO”).

In August 2002, the Ntsebeza Plaintiffs filed a motion with the Judicial Panel on

Multidistrict Litigation (“MDL Panel”) to transfer all of the actions to the Southern District of

New York, and in December 2002, the MDL Panel ordered that transfer for coordinated pre-trial

proceedings. See In re S. African Apartheid Litig., 238 F. Supp. 2d 1379, 1380–81 (J.P.M.L.

2002). In July 2003, thirty-one of the fifty-five defendants in the Ntsebeza and Digwamaje

actions filed a joint motion to dismiss. Following the transfer of the Khulumani complaint to the

Southern District of New York, eighteen of the twenty-three defendants in that action also filed a

joint motion to dismiss.2

Later that month, Penuell Mpapa Maduna, who was then the Minister of Justice and

Constitutional Development for South Africa, submitted an ex parte declaration to the district

court, stating that the South African government regarded these proceedings as interfering “with

personal representatives of victims of extrajudicial killing, or were themselves tortured, sexually
assaulted, indiscriminately shot, or arbitrarily detained by the apartheid regime.”
2
Apparently, not all of the named defendants in the three actions have been served with
complaints, and some defendants have indicated that they plan to contest personal jurisdiction. The
district court’s order granting the motion to dismiss stated that it was “limited to those defendants as to
whom the Court’s personal jurisdiction is not contested.” In re S. African Apartheid Litig., 346 F. Supp.
2d at 543 n.3. In addition, eleven of the defendants who joined the motion to dismiss filed separate
motions to dismiss on the ground that the plaintiffs’ claims were conclusory and failed to meet the
pleading standards of Fed. R. Civ. P. 8(a). The district court did not reach those arguments in deciding
the motion to dismiss.

6
a foreign sovereign’s efforts to address matters in which it has the predominant interest” and

asking that the proceedings be dismissed.3 After receiving the South African declaration, the

district court, sua sponte, solicited the views of the United States Department of State.4 The

State Department responded by submitting a “Statement of Interest” asserting that “continued

adjudication of the above-referenced matters risks potentially serious adverse consequences for

significant interests of the United States.”

Ruling on the defendants’ motions to dismiss, the district court held that the plaintiffs

failed to establish subject matter jurisdiction under the ATCA. The district court ruled further

that the plaintiffs, having asserted diversity as an alternate basis for jurisdiction, could not

establish subject matter jurisdiction on that ground. The district court also held that the plaintiffs

failed to state a claim under the TVPA and failed to establish subject matter jurisdiction under

RICO. See In re S. African Apartheid Litig., 346 F. Supp. 2d at 554–57. The district court

therefore dismissed the plaintiffs’ complaints in their entirety. See id. at 557. In March 2005, the

Ntsebeza and Digwamaje Plaintiffs moved for permission to file an amended consolidated

complaint,5 which the district court denied. See Part IV, infra. Following the district court’s

3
The Ntsebeza and Digwamaje Plaintiffs moved to strike Maduna’s declaration and submissions,
asserting that the declaration contained “legal argument by a non-party” and “conclusory allegations not
supported by any creditable citation to the record,” and recited “disputed evidentiary facts.” They also
argued that international comity could only be raised by a party as an affirmative defense, and even if
comity were available, “it would be inappropriate for the Court to dismiss this case without making fact-
bound determinations that cannot be made on this record . . . .” The docket sheet indicates that the
district court never ruled on this motion.
4
Specifically, the district court inquired whether “adjudication of these cases would have an
adverse impact on the interests of the United States and, if so, the nature and significance of any such
impact.”
5
The plaintiffs sought to provide particularized allegations directed at particular defendants, to
“meet the new Sosa standard,” and to clarify for the district court that their ATCA claims were not based

7
issuance of an amended judgment containing an amended Rule 54(b) certification, the plaintiffs

filed timely notices of appeal.6

II

All members of the panel join to affirm the district court’s dismissal of the Digwamaje

Plaintiffs’ TVPA claims. The Digwamaje Plaintiffs asserted a claim under the TVPA, alleging

that the defendants “aided and abetted the apartheid regime’s subjecting the Plaintiffs to torture

and extrajudicial killing within the meaning of the Torture Victim Protection Act . . . under

actual or apparent authority, or under color of law.” The TVPA provides:

An individual who, under actual or apparent authority, or color of law, of any foreign
nation-
(1) subjects an individual to torture shall, in a civil action, be liable for damages to
that individual; or
(2) subjects an individual to extrajudicial killing shall, in a civil action, be liable for
damages to the individual’s legal representative, or to any person who may be a
claimant in an action for wrongful death.

28 U.S.C. § 1350 note § 2(a). For purposes of the TVPA, an individual “acts under color of law .

. . when he acts together with state officials or with significant state aid.” Kadic v. Karadzic, 70

F.3d 232, 245 (2d Cir. 1995). The Digwamaje Plaintiffs, although twice having amended their

complaint, failed to link any defendants to state aid or the conduct of state officials.

Further, based on the district court’s discussion of the matter of diversity jurisdiction, we

affirm the dismissal of the complaints insofar as they seek to assert jurisdiction under 28 U.S.C.

§ 1332(a)(3).

upon the corporations “merely doing business” in South Africa.
6
The Digwamaje Plaintiffs have not challenged the dismissal of their RICO claim.

8
 III

Two members of this panel join to vacate the district court’s dismissal of the plaintiffs’

ATCA claims because the district court erred in holding that aiding and abetting violations of

customary international law cannot provide a basis for ATCA jurisdiction. We hold that in this

Circuit, a plaintiff may plead a theory of aiding and abetting liability under the ATCA. The

respective rationales of Judges Katzmann and Hall are set forth in separate concurring opinions.

IV

We further vacate the district court’s order denying plaintiffs’ motion for leave to amend.

In denying this motion, the district court relied, in part, on the erroneous premise that subject

matter jurisdiction did not inhere and reasoned that any additional amendments to the pleadings

would be futile. Because the denial of the motion rested, in part, on this erroneous premise, we

vacate that order. See Sista v. CDC Ixis N. Am., Inc., 445 F.3d 161, 177 (2d Cir. 2006) (“The

standard for reviewing the denial of a motion to amend a complaint is abuse of discretion . . . .”

(internal quotation marks omitted)); see also Zervos v. Verizon N.Y., Inc., 252 F.3d 163, 169 (2d

Cir. 2001) (“A district court ‘abuses’ or ‘exceeds’ the discretion accorded to it when . . . its

decision rests on an error of law (such as application of the wrong legal principle) . . . .”).7 Until

7
Although the district court rested its decision to deny the plaintiffs’ motion to replead on
several grounds, it is not clear from the district court’s order that it would have reached the same result in
the absence of its erroneous belief that any amendment would be futile. It seems most respectful of the
district court’s considerable discretion in this area to allow it to determine in the first instance whether to
allow the plaintiffs to replead. In that same vein, we also leave to the district court on remand the first
opportunity to consider any motion that may be filed by the Khulumani plaintiffs seeking permission to
amend their complaint.

9
the district court has an opportunity to rule on the motion to amend, we cannot be sure that the

pleadings in the record before us represent the final version of the plaintiffs’ allegations. We

therefore decline to determine whether the plaintiffs have adequately pled a violation of

international law sufficient to avail themselves of jurisdiction under the ATCA and remand to the

district court to allow it to address the pleadings after amendment as may be permitted has

occurred.

V

Moreover, we decline to affirm the dismissal of plaintiffs’ ATCA claims on the basis of

the prudential concerns raised by the defendants.8 In Sosa v. Alvarez-Machain, 542 U.S. 692,

733 n.21 (2004), the Supreme Court identified two different respects in which courts should

consider prudential concerns in deciding whether to hear claims brought under the ATCA. First,

We do not, however, believe it is necessary to vacate the district court’s order to the extent that it
denied the plaintiffs’ motion with respect to their TVPA claims. Given that the plaintiffs had previously
amended their complaint twice and did not bring the present motion to amend until after the district
court’s order dismissing the case, the district court did not abuse its discretion in denying them leave to
replead. See, e.g., Gurary v. Winehouse, 235 F.3d 792, 801 (2d Cir. 2000) (“A district court has broad
discretion in determining whether to grant leave to amend, and we review such determinations for abuse
of discretion.”).
8
In his dissent, Judge Korman adamantly asserts that our opinion fails to show deference to the
position of the Republic of South Africa, a position that also commands the support of the United States
Department of State. Opinion of Judge Korman at 92–100. He presents an analysis of prudential
considerations, the strength of which we decline to address, that may well suggest a roadmap for future
motion practice before the district court on remand. Id. at 73–92. Implicit throughout is the suggestion
that by correcting the district court’s error and clarifying the applicability of accessorial liability to
violations of international law, we have somehow irrevocably, and to the derogation of South Africa’s
sovereignty, relegated the plaintiffs’ claims to the judicial processes of the United States. With all due
respect, Judge Korman’s position in this regard is overstated and ignores the fact that well before any
steps can be taken to address the merits of plaintiffs’ claims, the district court may allow for a full airing
of prudential concerns and, if it so chooses, engage in an analysis of, and decision regarding, those and
other issues, something that has been heretofore lacking.

10
the Supreme Court held that courts should consider prudential concerns in the context of

determining whether to recognize a cause of action under the ATCA. Specifically, the Court

explained that “the determination whether a norm is sufficiently definite to support a cause of

action should (and, indeed, inevitably must) involve an element of judgment about the practical

consequences of making that cause available to litigants in the federal courts.” Sosa, 542 U.S. at

732-33 (internal footnote omitted). Second, the Supreme Court recognized that, in certain cases,

other prudential principles might operate to “limit[] the availability of relief in the federal courts

for violations of customary international law.” Id. at 733 n.21.9

One such principle specifically identified by the Court was “a policy of case-specific

deference to the political branches.” Id. This policy of “[j]udicial deference to the Executive

Branch on questions of foreign policy has long been established under the prudential

justiciability doctrine known as the ‘political question’ doctrine.” Whiteman v. Dorotheum

GmbH & Co KG, 431 F.3d 57, 69 (2d Cir. 2005). Another prudential doctrine that the

defendants raise in this case is “international comity.” See, e.g., Sosa, 542 U.S. at 761 (Breyer,

J., concurring) (suggesting that courts should consider “whether the exercise of jurisdiction under

the [ATCA] is consistent with those notions of comity that lead each nation to respect the

sovereign rights of other nations by limiting the reach of its laws and their enforcement”). This

doctrine, the application of which “ordinarily lies within the discretion of the district court,” asks

9
We reject the proposition endorsed by Judge Korman that the Supreme Court, in a footnote
written while deciding a different case, would instruct us on how to decide this case, which was not
before it. Opinion of Judge Korman at 73. Instead, we take the Supreme Court’s language in footnote 21
of Sosa at face value, as simply observing that there is a strong argument that the views of the Executive
Branch on the issue of the case’s impact on foreign policy should be given “serious weight.” Sosa, 542
U.S. at 733 n. 21. We view summary dismissal at the behest of a footnote as premature. See Cent. Va.
Cmty. College v. Katz, 546 U.S. 356, 363 (2006).

11
whether “adjudication of [the] case by a United States court would offend amicable working

relationships with [a foreign country].”10 Bigio v. Coca-Cola Co., 448 F.3d 176, 178 (2d Cir.

2006) (internal quotation marks omitted).

In dismissing the plaintiffs’ complaints below, the district court explicitly refrained from

addressing the defendants’ arguments that the ATCA claim presented a non-justiciable political

question.11 In re S. African Apartheid Litig., 346 F. Supp. 2d at 543 n.4 (“Defendants also argue

that . . . the matter is a non-justiciable political question. Given the Court’s finding that

defendants are entitled to relief on other grounds, the Court need not address these remaining

grounds for defendants’ motion.” (internal citation omitted)). Although the district court noted

some collateral consequences that might result from the adjudication of these kinds of claims, the

consequences it noted were primarily those “that would result from allowing courts in this

country to hear civil suits for the aiding and abetting of violations of international norms across

the globe.” Id. at 551. Furthermore, the district court expressly characterized its consideration of

the collateral consequences as fulfilling its “duty to engage in ‘vigilant doorkeeping.’” Id. at 550

(quoting from Sosa, 542 U.S. at 729); see also id. at 553. As Sosa makes clear, this duty is

fulfilled in the decision of a federal court to exercise its judicial discretion to recognize a cause of

action for a violation of customary international law, an issue distinct from whether the

10
The parties agree that Sosa’s reference to “case-specific deference” implicates either the
political question or international comity doctrine.
11
Judge Korman suggests that the content of the Defendant’s Joint Motion to Dismiss (“Joint
Motion”), cited by the district court, supports the view that the issue of deference to other political
branches was addressed below. Opinion of Judge Korman at 93. We fail to see how the content of a
motion, the basis for which Judge Korman concedes the district court “declin[ed] to address,” id., lends
any support to the proposition that the district court reached the issues addressed therein.

12
adjudication of a given suit is barred by political question doctrine. Sosa, 542 U.S. at 729. We

see no reason to read the district court’s citation to this part of Sosa as an indication that the court

was also (and contrary to its explicit disclaimer) engaged in a consideration of the political

question or other prudential doctrines.12 To the extent that limited portions of the district court’s

discussion were addressed to the statements of interest submitted by the governments of the

United States and South Africa, In re S. African Apartheid Litig., 346 F. Supp. 2d at 553–54,

these statements were merely illustrative of a general concern with what the court saw as the

“far-reaching” consequences of the specific norm it was then discussing (i.e., doing business in

South Africa), id. at 553. The citation to Sosa’s footnote 21 indicates only that the district court

considered the views of those governments in assessing “the collateral consequences that would

result from finding a new international law violation,” id., and does not suffice to demonstrate

that the court (again contrary to its stated intentions) adopted sub rosa the defendants’ political

question arguments.

We decline to address these case-specific prudential doctrines now and instead remand to

the district court to allow it to engage in the first instance in the careful “case-by-case” analysis

that questions of this type require. This approach is particularly appropriate here because the

plaintiffs have indicated that, if given the opportunity, they would narrow their claims and clarify

12
It was error for the district court to consider these collateral consequences in the context of
deciding preliminarily whether it had jurisdiction to hear this case under the ATCA. However, even if
we construed the district court’s discussion of the “collateral consequences” as a decision not to
recognize a cause of action for plaintiffs’ claims, in which context consideration of these consequences
would have been appropriate, remand would still be necessary because the district court’s decision not to
recognize a cause of action might still have rested, in part, on its erroneous view that the ATCA does not
allow for claims of aiding and abetting liability. Because we cannot know whether the district court
would have declined to recognize the cause of action plaintiffs bring in the absence of that error, we
believe it is appropriate to remand, so it can have the opportunity to decide this issue in the first instance.

13
the nature of their allegations against the various defendants, changes that may affect how the

district court ultimately decides to resolve these issues. Oral Argument Tr., Jan. 24, 2006 at 7–8,

14–15; cf. Zivotofsky v. Sec’y of State, 444 F.3d 614, 619–20 (D.C. Cir. 2006) (holding that,

where the specific relief sought by the plaintiff had changed, remand was appropriate to allow the

district court to develop a more complete record as to whether the revised claim presented a

nonjusticiable political question).13

On remand, the district court will have an opportunity to consider the guidance provided

by our prior cases regarding the relevant weight of statements of interest submitted by the United

States and other governments. In Whiteman, for example, we addressed “when, and to what

extent, . . . the stated foreign policy interests of the United States [should] be accorded

deference,” and we held that we should be guided in this determination by “our application of the

political question doctrine.” 431 F.3d at 69, 71. In that context, we held that “not every case

‘touching foreign relations’ is nonjusticiable and judges should not reflexively invoke these

doctrines to avoid difficult and somewhat sensitive decisions in the context of human rights. We

believe a preferable approach is to weigh carefully the relevant considerations on a case-by-case

basis.” Id. at 69 (quoting Kadic, 70 F.3d at 249). We have held that “even an assertion of the

political question doctrine by the Executive Branch, entitled to respectful consideration, would

not necessarily preclude adjudication.”14 Kadic, 70 F.3d at 250. Likewise, although the views of

13
While we cannot know how these developments will affect the positions of the United States
and South Africa with respect to this litigation, the district court may wish to solicit anew the views of
these governments, and thus that fact, too, counsels against us reaching these issues at this time.
14
We do not believe the Supreme Court’s statement in Sosa is to the contrary. Sosa, 542 U.S. at
733 n.21 (noting only that “there is a strong argument that federal courts should give serious weight to
the Executive Branch’s view of the case’s impact on foreign policy”). Indeed, to give dispositive weight

14
foreign nations are an important consideration under the doctrine of “international comity,” we

have not held them to be dispositive. See, e.g., Jota v. Texaco Inc., 157 F.3d 153, 159–61 (2d

Cir. 1998). At this stage in the litigation, we express no view as to what level of deference to

their views is appropriate in this particular case. Instead, we remand to the district court so that it

may carefully consider whether any of these doctrines require dismissal.

VI

We therefore AFFIRM the district court’s dismissal of the Digwamaje Plaintiffs’ TVPA

claims. We also AFFIRM the district court’s determination that the plaintiffs have failed to

satisfy the diversity requirements of 28 U.S.C. § 1332(a)(3). We VACATE the district court’s

dismissal of the plaintiffs’ ATCA claims, as well as the district court’s denial of the Digwamaje

and Ntsebeza Plaintiffs’ motions to amend and REMAND for further proceedings consistent with

this opinion.

KATZMANN, Circuit Judge, concurring:

This case calls upon us in principal part to determine whether the district court erred in

concluding that it does not have jurisdiction over the plaintiffs’ claims brought under the Alien

Torts Claim Act (“ATCA”), 28 U.S.C. § 1350. I respectfully believe that the district court erred

to the Executive Branch’s views would likely raise serious separation-of-powers concerns. Cf., e.g., First
Nat’l City Bank v. Banco Nacional de Cuba, 406 U.S. 759, 763 (1972) (Powell, J., concurring) (“I would
be uncomfortable with a doctrine which would require the judiciary to receive the Executive’s
permission before invoking its jurisdiction. Such a notion, in the name of the doctrine of separation of
powers, seems to me to conflict with that very doctrine.”).

15
in its analysis of plaintiffs’ ATCA claims in two fundamental respects. First, it conflated the

jurisdictional and cause of action analyses required by the ATCA. As a result, the district court

mistakenly incorporated a discretionary analysis into the determination of whether it has

jurisdiction under the ATCA. Second, it erroneously held that aiding and abetting liability does

not exist under international law.

I

The ATCA provides that “[t]he 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.” Id. When we last substantively grappled with the meaning of this statute, we

noted that “neither Congress nor the Supreme Court ha[d] definitively resolved the complex and

controversial questions regarding the meaning and scope of the ATCA.” Flores v. S. Peru

Copper Corp., 414 F.3d 233, 247 (2d Cir. 2003). The next year the Supreme Court weighed in

on the modern resurgence of the statute in Sosa v. Alvarez-Machain, 542 U.S. 692 (2004). While

the Court did not definitively resolve all of the complex and controversial questions the ATCA

raises, it did clarify to a significant degree how claims brought under the ATCA should be

analyzed.

Sosa endorsed our Court’s prior approach to the ATCA to the extent that we recognized

that the Act created jurisdiction for a narrow set of violations of international law, see id. at 720

(“Congress intended the [ATCA] to furnish jurisdiction for a relatively modest set of actions

alleging violations of the law of nations.”); see also Filartiga v. Pena-Irala, 630 F.2d 876,

16
887–88 (2d Cir. 1980) (construing the ATCA “as opening the federal courts for adjudication of

. . . well-established, universally recognized norms of international law”), and that a plaintiff’s

claim “must be gauged against the current state of international law,” Sosa, 542 U.S. at 733; see

also Filartiga, 630 F.2d at 881 (“[I]t is clear that courts must interpret international law not as it

was in 1789, but as it has evolved and exists among the nations of the world today.”). The Court

also generally endorsed the cautious approach that we had long applied in determining whether to

hear cases brought under the ATCA. See, e.g., Sosa, 542 U.S. at 728–29 (noting that “great

caution” must be exercised in deciding which “norms of today’s law of nations may . . . be

recognized legitimately by federal courts”); see also Flores, 414 F.3d at 248 (“[I]n determining

what offenses violate customary international law, courts must proceed with extraordinary care

and restraint.”).

Sosa did, however, deviate from our case law in one crucial respect. We had allowed

cases to proceed under the ATCA on the assumption that when plaintiffs alleged violations of

well-established international law, their “causes of action are statutorily authorized.” Kadic v.

Karadzic, 70 F.3d 232, 246 (2d Cir. 1995); see also Flores, 414 F.3d at 245 (discussing the

reception of “Filartiga’s holding that the ATCA creates a private right of action for violations of

United States treaties or customary international law”). The Supreme Court flatly rejected this

notion. See Sosa, 542 U.S. at 713-14 (citing William R. Casto, The Federal Courts’ Protective

Jurisdiction Over Torts Committed in Violation of the Law Of Nations, 18 Conn. L. Rev. 467,

479, 480 (1986)). Sosa construed the statute instead to be of a “strictly jurisdictional nature,” in

the sense that it “address[ed] the power of the courts to entertain cases concerned with a certain

subject.” 542 U.S. at 713, 714.

17
 This holding led the Court to explore “a new question, this one about the interaction

between the [ATCA] at the time of its enactment and the ambient law of the era.” Id. at 714.

The Court rejected the argument that the First Congress passed the ATCA “as a jurisdictional

convenience to be placed on the shelf for use by a future Congress or state legislature that might,

some day, authorize the creation of causes of action.” Id. at 719. Rather, the historical materials

suggested that “the statute was intended to have practical effect the moment it became law.” Id.

at 724. This practical effect would be achieved through the invocation of causes of action

already available at common law. Because some “torts in violation of the law of nations were

understood to be within the common law” of 1789, the First Congress would have “understood

that the district courts would recognize private causes of action for certain torts in violation of the

law of nations.” Id. at 720, 724. No further substantive legislation was required. Id. at 724

(“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.”).

Finding that “no development in the two centuries from the enactment of § 1350 to

[today] has categorically precluded federal courts from recognizing a claim under the law of

nations as an element of common law,” id. at 724–25, the Court held that federal courts retained

the ability to “adapt[] the law of nations to private rights” by recognizing “further international

norms as judicially enforceable today,” id. at 728, 729. Most importantly, these norms were

enforceable not by virtue of this statutory creation or authorization of a private right of action, as

we had previously assumed, but rather through an exercise in “residual common law discretion”

to create causes of action under federal common law to remedy the violation of those norms. Id.

18
at 738; see also id. at 728–31. Thus, Sosa makes clear that all ATCA litigation is in fact based

on federal common law, rather than a statutory cause of action. But a federal court’s power to

create these causes of action is not without limits. Indeed, the Court identified a number of

“good reasons for a restrained conception of the discretion a federal court should exercise in

considering a new cause of action of this kind.”1 Id. at 725, 728. Its consideration of these

factors led the Court to identify a minimum requirement “for accepting a cause of action subject

to jurisdiction under § 1350,” namely, that “federal courts should not recognize private claims

under federal common law for violations of any international law norm with less definite content

and acceptance among civilized nations than the historical paradigms familiar when § 1350 was

enacted.” Id. at 732.

Based on the Supreme Court’s holdings in Sosa that (1) the ATCA is purely jurisdictional

and (2) the common law provides the cause of action for claims brought under that jurisdiction, a

federal court faced with a suit alleging a tort in violation of international law must undertake two

distinct analytical inquiries. One is whether jurisdiction lies under the ATCA. The other is

whether to recognize a common-law cause of action to provide a remedy for the alleged violation

of international law. Requiring this analytical separation in ATCA litigation comports with the

general principle that whether jurisdiction exists and whether a cause of action exists are two

1
This “series of reasons” comprises (1) a change in the “prevailing conception of the common
law” such that there is now “a general understanding that the law is not so much found or discovered as it
is either made or created,” Sosa, 542 U.S. at 725; (2) a “rethinking of the role of the federal courts in
making it,” id. at 726; (3) the recognition that “a decision to create a private right of action is one better
left to legislative judgment in the great majority of cases,” id. at 727; (4) the fact that “attempts by
federal courts to craft remedies for the violation of new norms of international law would raise risks of
adverse foreign policy consequences,” id. at 727–28; and (5) the lack of a “congressional mandate” to
engage in judicial lawmaking in this area, id. at 728.

19
distinct inquiries. See TCG N.Y., Inc. v. City of White Plains, 305 F.3d 67, 74 (2d Cir. 2002)

(noting the Supreme Court’s holding that “‘the question whether a federal statute creates a claim

for relief is not jurisdictional’” (quoting Northwest Airlines, Inc. v. County of Kent, 510 U.S. 355,

365 (1994)); cf. Rasul v. Bush, 542 U.S. 466, 484–85 (2004) (reversing the D.C. Circuit’s

holding that jurisdiction did not lie under the ATCA without engaging in a cause of action

inquiry). Moreover, it reflects the fact that it is Congress, and not the courts, that possesses the

power to define the scope of the courts’ jurisdiction. Cf. Whitmore v. Arkansas, 495 U.S. 149,

155–56, 161 (1990) (noting that “[a] federal court is powerless to create its own jurisdiction” and

that a court may not “employ untethered notions of what might be good public policy to expand

our jurisdiction in an appealing case”). Thus, one might question the extent to which a federal

court should exercise its own judgment about the practical consequences of allowing a suit to go

forward in the context of its jurisdictional inquiry when, according to the plain language of the

ATCA, that inquiry is resolved solely by reference to international law. Cf. Harrison v. PPG

Indus., Inc., 446 U.S. 578, 593 (1980) (noting that courts should look to Congress’s intent, and

not policy considerations, when construing the terms of a jurisdictional statute).2 By contrast, it

is entirely appropriate for courts to consider such concerns in the context of determining whether

to recognize a cause of action under federal common law.

2
Federal courts may, of course, decline to accept jurisdiction over a particular case in
appropriate circumstances and in accordance with the well-established principles that guide such
decisions. I merely suggest that because federal courts “have only the jurisdiction granted to them by
Congress,” Carlyle Towers Condo. Ass’n, Inc. v. FDIC, 170 F.3d 301, 306 (2d Cir. 1999), their exercise
of common-law discretion is better viewed as creating a cause of action than jurisdiction. Cf. Bolin v.
Sears, Roebuck & Co., 231 F.3d 970, 974 (5th Cir. 2000) (holding that Congress’s grant of authority to
the Supreme Court to make rules that affect only the timing of appeals does not constitute an improper
delegation of its power to confer jurisdiction).

20
 Having identified the two steps of the inquiry, I now elaborate on what each inquiry

requires. Aside from noting that “Congress intended the [ATCA] to furnish jurisdiction for a

relatively modest set of actions alleging violations of the law of nations,” Sosa, 542 U.S. at 720,

the Supreme Court did not discuss the requirements for invoking this jurisdictional grant in a

particular case. The ATCA, by its terms, “confers federal subject-matter jurisdiction when the

following three conditions are satisfied: (1) an alien sues (2) for a tort (3) committed in violation

of the law of nations.” Kadic, 70 F.3d at 238. In Flores, we held that the law of nations, or

customary international law,3 “is composed only of those rules that States universally abide by, or

accede to, out of a sense of legal obligation and mutual concern.” 414 F.3d at 248; see also The

Paquete Habana, 175 U.S. 677, 708 (1900) (describing a rule of international law as one reached

by “general consent of the civilized nations of the world” and “founded on considerations of . . .

the mutual convenience of belligerent states”). In determining whether a given offense meets

these requirements, we look to the sources of law identified by the Statute of the International

Court of Justice (“ICJ Statute”) as the proper sources of international law. See Flores, 414 F.3d

at 250–51. These include:

a. international conventions, whether general or particular, establishing rules

expressly recognized by the contesting states;

b. international custom, as evidence of a general practice accepted as law;

c. the general principles of law recognized by civilized nations;

3
In Flores we noted that the law of nations, as used in the ATCA “refers to the body of law
known as customary international law.” 414 F.3d at 247.

21
 [and]

d. . . . judicial decisions and the teachings of the most highly qualified publicists

of the various nations, as subsidiary means for the determination of rules of law.

ICJ Statute, art. 38, June 26, 1945, 59 Stat. 1055, 1060.

If jurisdiction is established, the second inquiry is whether a common-law cause of action

should be created to provide a remedy for the alleged violation of international law. Recognizing

that “there are good reasons for a restrained conception of the discretion a federal court should

exercise in considering a new cause of action of this kind,” the Court in Sosa “require[d] 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,” namely, “violation of safe conducts, infringement

of the rights of ambassadors, and piracy.” 542 U.S. at 724–25; see also id. at 732 (“[W]e are

persuaded that federal courts should not recognize private claims under federal common law for

violations of any international law norm with less definite content and acceptance among

civilized nations than the historical paradigms familiar when § 1350 was enacted.”). In setting

out this standard, the Court cited with approval our decision in Filartiga, noting that “[t]his limit

upon judicial recognition is generally consistent with the reasoning of many of the courts and

judges who faced the issue before it reached this Court.” Id.

The Supreme Court instructed that determining whether to recognize a new cause of

action “should (and, indeed, inevitably must) involve an element of judgment about the practical

consequences of making that cause available to litigants in the federal courts.” Id. at 732–33. I

22
do not read Sosa as requiring that a court individually analyze each of the five reasons it

identifies as “argu[ing] for judicial caution when considering the kinds of individual claims that

might implement the jurisdiction conferred by the . . . statute.” Sosa, 542 U.S. at 725. These

reasons are already captured by the “high bar to new private causes of action” set by the

requirement that a claim be accepted by the civilized world and defined with a sufficient degree

of specificity. See id. at 727–32; see also id. at 725 (noting that “there are good reasons for a

restrained conception of [a federal court’s] discretion” and that “[a]ccordingly . . . courts should

require any claim” to meet its standard of acceptance and specificity); David H. Moore, An

Emerging Uniformity for International Law, 75 Geo. Wash. L. Rev. 1, 40–41 (2006) (“Those

concerns, which arise whenever [customary international law] is incorporated as federal common

law, are mitigated by the specific definition and mutuality requirements.”). I do, however, view

the Court’s instruction that an element of judgment must be involved in the decision to recognize

a cause of action as an invitation to lower courts to consider other prudential concerns consistent

with Sosa’s approach. As Sosa suggests, courts considering the “practical consequences” of

recognizing a new cause of action should assess the consequences that might result from making

the cause of action generally available to all potential plaintiffs.

In sum, a district court analyzing a claim under the ATCA will normally be required to

engage in a two-part analysis. The district court here erred by failing to undertake separately the

two parts of this analysis. By conflating these two questions, the district court inappropriately

injected a discretionary element into the determination of whether it had jurisdiction under the

ATCA. See In re S. African Apartheid Litig., 346 F. Supp. 2d at 551, 553–54. Of greater

significance to the district court’s ultimate disposition of the plaintiffs’ claims, though, was its

23
error in analyzing whether the plaintiffs had alleged a “violation of the law of nations,” as is

required for plaintiffs to establish both jurisdiction and a cause of action. It is to this analysis that

I now turn.

II

Asking whether “aiding and abetting international law violations . . . [is a] violation[] of

the law of nations that [is] ‘accepted by the civilized world and defined with a specificity

comparable to the features of the 18th-century paradigms,’” Id. at 549 (quoting Sosa, 542 U.S. at

725), the district court concluded that it is not, noting that it found “little [in its review of

international law] that would lead [it] to conclude that aiding and abetting international law

violations is itself an international law violation that is universally accepted as a legal

obligation,” id. Although I believe the district court was correct to look to international law, I

disagree with its analysis.

A.

The district court’s conclusion that its jurisdiction under the ATCA should depend on

whether international law specifically recognizes liability for aiding and abetting violations of the

law of nations is consistent with our prior case law. We have repeatedly emphasized that the

scope of the ATCA’s jurisdictional grant should be determined by reference to international law.

See Kadic, 70 F.3d at 238 (requiring courts to engage in a “searching review” of international law

to verify that subject matter jurisdiction lies for a particular action); see also Flores, 414 F.3d at

248 (noting that “courts must proceed with extraordinary care and restraint” in “determining

24
what offenses violate customary international law”); Filartiga, 630 F.2d at 887 (“The paucity of

suits successfully maintained under the section is readily attributable to the statute’s requirement

of alleging a ‘violation of the law of nations’ [] at the jurisdictional threshold.”).

It is also consistent with the Supreme Court’s opinion in Sosa. The Court observed in a

footnote that “whether a norm is sufficiently definite to support a cause of action” raises a

“related consideration [of] whether international law extends the scope of liability for a violation

of a given norm to the perpetrator being sued, if the defendant is a private actor such as a

corporation or individual.” Sosa, 542 U.S. at 732 & n.20. While this footnote specifically

concerns the liability of non-state actors, its general principle is equally applicable to the question

of where to look to determine whether the scope of liability for a violation of international law

should extend to aiders and abettors. Furthermore, in Sosa, the Supreme Court echoed our prior

cases’ emphasis on the narrowness of the ATCA’s jurisdictional grant. See id. at 720 (“Congress

intended the [ATCA] to furnish jurisdiction for a relatively modest set of actions alleging

violations of the law of nations.”); see also id. at 712 (describing the statute as a “limited,

implicit sanction to entertain the handful of international law cum common law claims

understood in 1789”). I believe that we most effectively maintain the appropriate scope of this

jurisdiction by requiring that the specific conduct allegedly committed by the defendants sued

represents a violation of international law.

Most importantly, the district court’s approach is consistent with Sosa’s broader

characterization of the relationship between federal common law and international law. The

ATCA’s jurisdictional grant “enable[s] federal courts to hear claims in a very limited category

defined by the law of nations.” Id. at 712 (emphasis added). Once a court determines that the

25
defendants’ alleged conduct falls within one of “the modest number of international law

violations with a potential for personal liability” on the defendants’ part, it then considers

whether “the common law would provide a cause of action” to enable the plaintiffs to bring their

claim. Id. at 724. The common law thus permits the “independent judicial recognition of

actionable international norms,” but the courts must, as Sosa cautioned, be “vigilant

doorkeep[ers].” Id. at 729. We recognized this important role for domestic law in Kadic when

we explained that the “law of nations generally does not create private causes of action to remedy

its violations, but leaves to each nation the task of defining the remedies that are available for

international law violations.”4 70 F.3d at 246. A federal court is free, in the exercise of its

common-law discretion, to decline to provide a cause of action for a violation of international

law. See Sosa, 542 U.S. at 732–33; Tel-Oren v. Libyan Arab Republic, 726 F.2d 774, 778 (D.C.

Cir. 1984) (Edwards, J., concurring) (explaining that each state may choose whether to impose

civil liability for violations of international law). But to assure itself that it has jurisdiction to

hear a claim under the ATCA, it should first determine whether the alleged tort was in fact

“committed in violation of the law of nations,” 28 U.S.C. § 1350, and whether this law would

recognize the defendants’ responsibility for that violation.

B.

I conclude that the recognition of the individual responsibility of a defendant who aids

and abets a violation of international law is one of those rules “that States universally abide by, or

4
In Kadic we assumed that the ATCA itself created the cause of action, see 70 F.3d at 246, but
the same principle applies to the common law following Sosa.

26
accede to, out of a sense of legal obligation and mutual concern.” Flores, 414 F.3d at 248.

Recognized as part of the customary law which authorized and was applied by the war crimes

trials following the Second World War, it has been frequently invoked in international law

instruments as an accepted mode of liability. During the second half of the twentieth century and

into this century, it has been repeatedly recognized in numerous international treaties, most

notably the Rome Statute of the International Criminal Court, and in the statutes creating the

International Criminal Tribunal for the Former Yugoslavia (“ICTY”) and the International

Criminal Tribunal for Rwanda (“ICTR”).5 Indeed, the United States concedes, and the

defendants do not dispute, that the concept of criminal aiding and abetting liability is “well

5
The district court seems to have dismissed the significance of some of these sources because
they imposed criminal and not civil responsibility. See In re S. African Apartheid Litig., 346 F. Supp. 2d
at 550. This distinction finds no support in our case law, which has consistently relied on criminal law
norms in establishing the content of customary international law for purposes of the ATCA. In Kadic,
for instance, we held that a defendant could be held liable under the ATCA based on international
criminal law norms prohibiting genocide and war crimes. Kadic, 70 F.3d at 241–42. In concluding that
genocide was actionable under the ATCA, the Kadic court relied on a United Nations resolution
declaring that genocide is a “crime under international law,” G.A. Res. 96(I), at 188–89, U.N. Doc. A/64
(Dec. 11, 1946), the London Charter, Agreement for the Prosecution and Punishment of the Major War
Criminals of the European Axis, Aug. 8, 1945, E.A.S. 472, and the Genocide Convention, Convention on
the Prevention and the Punishment of the Crime of Genocide, Dec. 9, 1948, 102 Stat. 3045, 78 U.N.T.S.
277, which confirms that “genocide, whether committed in a time of peace or in time of war, is a crime
under international law which [the contracting parties] undertake to prevent and to punish,” id. at art. I
(emphasis added); see Kadic, 70 F.3d at 241. Kadic also held that jurisdiction existed under the ATCA
for “claims of war crimes” because “[t]he liability of private individuals for committing war crimes has
been recognized since World War I and was confirmed at Nuremberg after World War II,” id. at 243,
notwithstanding that the Nuremberg trials were criminal proceedings and that the proscription of war
crimes is self-evidently criminal in nature.
Our past reliance on criminal law norms seems entirely appropriate given that, as Justice Breyer
observed in Sosa, international law does not maintain the kind of hermetic seal between criminal and
civil law that the district court sought to impose. See Sosa, 542 U.S. at 762–63 (Breyer, J., concurring).
“[T]he criminal courts of many nations combine civil and criminal proceedings, allowing those injured
by criminal conduct to be represented, and to recover damages, in the criminal proceeding itself.” Id.
Moreover, the ICTY has recognized the propriety of civil remedies for violations of international
criminal law in certain circumstances, noting for example that a torture victim might “bring a civil suit
for damage in a foreign court.” Prosecutor v. Furundzija, Case No. IT-95-17/1, Trial Chamber
Judgment, ¶ 155 (Dec. 10, 1998).

27
established” in international law.6 Brief for the United States as Amicus Curaie, at 21.

The London Charter, which established the International Military Tribunal at Nuremberg,

was entered into by the allied powers of World War II, “acting in the interests of all the United

Nations,” to establish a tribunal to punish violations of international law. See Agreement for the

Prosecution and Punishment of the Major War Criminals of the European Axis, pmbl., Aug. 8,

1945, E.A.S. 472 (hereinafter London Charter). We have previously recognized the London

Charter as an authoritative source of customary international law. See Flores, 414 F.3d at 244

n.18; United States v. Yousef, 327 F.3d 56, 105 nn.39–40 (2d Cir. 2003) (citing the London

Charter for definitions of war crimes and crimes against humanity under international law).

Moreover, other courts, international bodies, and scholars have recognized that the principles set

out in the London Charter and applied by the International Military Tribunal are significant not

only because they have garnered broad acceptance, but also because they were viewed as

reflecting and crystallizing preexisting customary international law. See Princz v. Federal

Republic of Germany, 26 F.3d 1166, 1174 (D.C. Cir. 1994) (“The trials for the first time made

explicit and unambiguous what was theretofore, as the Tribunal has declared, implicit in

6
The United States also notes that the law governing the initial military commissions established
to prosecute acts of international terrorism following the September 11 attacks recognized aiding and
abetting as a mode of liability. See Crimes and Elements for Trials by Military Commission, 32 C.F.R. §
11.6(c)(1) (2003). The jurisdiction of the military tribunals was limited to established violations of the
law of war or to “offenses that, consistent with that body of law, are triable by military commission.” 32
C.F.R. § 11.3(a). The law of war has long been recognized as a subset of international law. See Ex parte
Quirin, 317 U.S. 1, 27–28 (1942) (“From the very beginning of its history this Court has recognized and
applied the law of war as including that part of the law of nations which prescribes, for the conduct of
war, the status, rights and duties of enemy nations as well as of enemy individuals.”). While the Supreme
Court eventually found that the military commissions lacked the power to proceed, see Hamdan v.
Rumsfeld, 126 S. Ct. 2749, 2759 (2006), nothing in that decision addressed the propriety of including
aiding and abetting liability as part of the commissions’ jurisdiction.

28
International Law . . . .” (citation, alteration, and internal quotation marks omitted)); Siderman de

Blake v. Republic of Argentina, 965 F.2d 699, 715 (9th Cir. 1992) (“The legitimacy of the

Nuremberg prosecutions rested . . . on the nature of the acts [the defendants] committed: acts that

the laws of all civilized nations define as criminal. The universal and fundamental rights of

human beings identified by Nuremberg . . . are the direct ancestors of the universal and

fundamental norms recognized as jus cogens.” (citations omitted)); Theodor Meron, Reflections

on the Prosecution of War Crimes by International Tribunals, 100 Am. J. Int’l L. 551, 559

(2006) (“[T]he Nuremberg and Tokyo Tribunals drew heavily on the 1929 Geneva Prisoner of

War Convention and the Fourth Hague Convention of 1907 as establishing the substantive law to

be applied—that is, as customary law and general principles of criminal law, and as norms of

both state responsibility and individual criminal liability.” (footnotes omitted)). Indeed, shortly

after the conclusion of the initial war crimes trials, the General Assembly of the United Nations

unanimously approved a resolution affirming “the principles of international law recognized by

the Charter of the Nürnberg Tribunal and the judgment of the Tribunal.” Affirmation of the

Principles of International Law Recognized in the Charter of the Nürnberg Tribunal, G.A. Res.

95 (I), at 188, U.N. Doc. A/236 (Dec. 11, 1946) (hereinafter Nuremberg Principles Resolution I).

The London Charter extended individual responsibility for crimes within its jurisdiction

not only to “[l]eaders, organizers, [and] instigators” but also to “accomplices participating in the

formulation or execution of a common plan or conspiracy to commit” any of the crimes triable by

the Tribunal. London Charter art. 6. While the Charter’s language taken “literally . . . would

seem to imply that the complicity rule did not apply to crimes perpetrated by individual action,”

as opposed to by common plan, in practice the Tribunal “applied general principles of criminal

29
law regarding complicity.” International Law Commission, Principles of International Law

Recognized in the Charter of the Nürnberg Tribunal and in the Judgment of the Tribunal, with

commentaries, G.A.O.R., 5th session, Supp. No. 12, U.N. Doc. A/1316, ¶¶ 126-27 (1950) (“ILC

Principles”). Accordingly, when the International Law Commission (“ILC”) formulated the

“principles recognized in the Charter . . . and in the judgment of the Tribunal” at the direction of

the General Assembly, see Nuremberg Principles Resolution I, it omitted any indication of a

limitation on accomplice liability. Principle VII provides that “[c]omplicity in the commission of

a crime against peace, a war crime, or a crime against humanity . . . is a crime under international

law.” ILC Principles, Principle VII. The ILC’s formulation of the principles is considered to be

an authoritative rendering of the formal holdings of the Nuremberg Tribunal and is consulted as

an authoritative source of customary international law by the ICTY and ICTR. See, e.g.,

Prosecutor v. Akayesu, Case No. ICTR-96-4-T, Trial Chamber Judgment, ¶ 526 (Sept. 2, 1998)

(citing Principle VII to establish that “participation by complicity in the most serious violations

of international humanitarian law was considered a crime as early as Nuremberg”); see also

Prosecutor v. Milosevic, Case No. IT-02-54, Trial Chamber Decision on Preliminary Motions, ¶¶

29–30 (Nov. 8, 2001) (finding that “[t]he customary character of [a] rule [of individual

responsibility] is further supported by its incorporation in a wide number of other instruments,”

including, inter alia, the ILC Principles).

That the London Charter’s use of the term “accomplice” was understood to include those

who aid and abet a crime is further confirmed by the law applied in the war crimes trials held in

the United States zone of occupation following World War II. War criminals in the United

States zone of occupation were tried under Control Council Law No. 10, an act promulgated by

30
the joint allied body that coordinated the governance of post-war Germany. Allied Control

Council Law No. 10 (Dec. 20, 1945) in Trials of War Criminals Before the Nuernberg Military

Tribunals Under Control Council Law No. 10, at XVIII (William S. Hein & Co., Inc. 1997)

(1949). Control Council Law No. 10 was patterned after the London Charter and enacted under

its authority with the declared purpose to “give [it] effect.” Flick v. Johnson, 174 F.2d 983,

985–86 (D.C. Cir. 1949). According to Telford Taylor, the Chief of Counsel for War Crimes and

the Chief Prosecutor for the United States under the London Charter, the “underlying principles”

of the London Charter and Law No. 10 “are identical” and the tribunals operating under each

“worked within the same framework” with respect to “the basic principles of international penal

law.” Telford Taylor, Final Report to the Secretary of the Army on the Nuernberg War Crimes

Trial Under Control Council Law No. 10, at 107 (1949), available at

http://www.loc.gov/rr/frd/Military_Law/NT_final-report.html. Control Council Law No. 10

imposed criminal liability on anyone who was “an accessory to the commission of any such

crime or ordered or abetted the same,” Control Council Law No. 10, art. II, sec. 2 (emphasis

added), and tribunals applying Control Council Law No. 10 are viewed by international bodies as

having recognized aiding and abetting liability. See Prosecutor v. Furundzija, Case No. IT-95-

17/1, Trial Chamber Judgment, ¶¶ 195–225, 236–40 (Dec. 10, 1998) (reviewing the case law).

We have previously acknowledged the contribution that Control Council Law No. 10 and

the tribunals that applied it have made to customary international law. In Flores, to support our

conclusion that “[c]ustomary international law rules proscribing crimes against humanity,

including genocide, and war crimes, have been enforceable against individuals since World War

II,” 414 F.3d at 244 n.18, we pointed specifically to Brigadier General Taylor’s assessment that

31
“the major legal significance of the Law No. 10 judgments lies . . . in those portions of the

judgments dealing with the area of personal responsibility for international law crimes,” Taylor,

supra, at 109, quoted in Flores, 414 F.3d at 244 n.18. The United States Government, as amicus

in this case, similarly acknowledges the role this law has played in establishing the availability of

aiding and abetting liability in modern international criminal tribunals. Brief for the United

States as Amicus Curaie, at 21 n.11.

Having been accepted as one of the core principles of the post–World War II war crimes

trials, the individual criminal responsibility of those who aid and abet violations of international

law was repeatedly reflected in international treaties thereafter. These treaties include major

agreements addressing fundamental human rights concerns such as torture, apartheid, slavery,

and genocide. See United Nations Convention Against Torture and Other Cruel, Inhuman or

Degrading Treatment or Punishment, art. 4, Dec. 10, 1984, 1465 U.N.T.S. 85; International

Convention on the Suppression and Punishment of the Crime of Apartheid, art. III(b), Nov. 30,

1973, 1015 U.N.T.S. 243; Supplementary Convention on the Abolition of Slavery, the Slave

Trade and Institutions and Practices Similar to Slavery, art. 6, Sept. 7, 1956, 18 U.S.T. 3201, 266

U.N.T.S. 3; Convention on the Prevention and Punishment of the Crime of Genocide, art III(e),

Dec. 9, 1948, 78 U.N.T.S. 277, 280 (hereinafter Genocide Convention). Aiding and abetting

liability has also been recognized in treaties addressed to crimes of general international concern

such as bribery of foreign officials in international business transactions and drug trafficking.7

7
The fact that some of these treaties do not directly criminalize aiding and abetting, but instead
require state parties to criminalize it in their domestic law, see, e.g., United Nations Convention Against
Transnational Organized Crime, art. 5, G.A. Res. 55/25, at 5-6, U.N. Doc. A/RES/55/25 (Jan. 8, 2001)
(“Each State Party shall adopt such legislative and other measures as may be necessary to establish as
criminal offences . . . aiding, abetting, facilitating or counselling the commission of serious crime

32
See Convention on Combating Bribery of Foreign Public Officials in International Business

Transactions, art. 1 (2), Dec. 17, 1997, reprinted in 37 I.L.M. 1 (1998); United Nations

Convention on Psychotropic Substances, art. 22 (2)(a)(ii), Feb. 21, 1971, 32 U.S.T. 543, 1019

U.N.T.S. 175. More recently, aiding and abetting has been included in a number of the treaties

concerning organized crime and terrorism, which have become prominent concerns of the

international community. See United Nations Convention Against Transnational Organized

Crime, art. 5 (1)(b), G.A. Res. 55/25, at 5–6, U.N. Doc. A/RES/55/25 (Jan. 8, 2001) ;

International Convention for the Suppression of the Financing of Terrorism, art. 2 (5)(a), opened

for signature Jan. 10, 2000, 39 I.L.M. 270; International Convention for the Suppression of

Terrorist Bombings, art. 2 (3)(a), May 23, 2001, 2149 U.N.T.S. 256; see also Protocol Against

the Smuggling of Migrants by Land, Sea and Air, Supplementing the United Nations Convention

Against Transnational Organized Crime, art.5 (1)(b), G.A. Res. 55/25, at 40, U.N. Doc.

A/RES/55/25 (Jan. 8, 2001). The United Nations Security Council (“Security Council”) also

appears to have recognized the importance of accomplice liability in the international response to

terrorism, directing states in the wake of the September 11 attacks to “[e]nsure that any person

who participates in the financing, planning, preparation or perpetration of terrorists acts or in

supporting terrorists acts is brought to justice.” S.C. Res. 1373, ¶ 2(e), U.N. Doc. S/RES/1373

involving an organized criminal group.”), does not limit their value as relevant sources of customary
international law. Customary international law addresses “those wrongs that are of mutual, and not
merely several, concern to States,” Flores, 414 F.3d at 249 (internal quotation marks and emphases
omitted), and where states join a binding agreement to enforce a mode of liability in their domestic law,
that decision reflects a matter of mutual, not several, concern and is thus properly considered a subject of
customary international law. See id. (“Matters of mutual concern between States are those involving
States’ actions performed towards or with regard to the other . . . . Matters of several concern among
States are matters in which States are separately and independently interested.” (citation and internal
quotation marks omitted)).

33
(Sept. 28, 2001) (emphasis added).

Aiding and abetting liability continues to be recognized and enforced in international

tribunals. The Statutes creating the ICTY and ICTR were adopted by resolutions of the Security

Council. In their respective sections on individual criminal responsibility, both statutes impose

individual liability on any person “who planned, instigated, ordered, committed or otherwise

aided and abetted in the planning, preparation or execution” of a crime. Statute of the

International Tribunal for the Former Yugoslavia, art. 7, S.C. Res. 827, U.N. Doc. S/RES/827

(May 25, 1993) (hereinafter ICTY Statute); Statute of the International Criminal Tribunal for

Rwanda, art. 6, S.C. Res. 955, U.N. Doc. S/RES/955 (Nov. 8, 1994) (hereinafter ICTR Statute).

As with the London Charter, the recognition of aiding and abetting liability in the ICTY

Statute is particularly significant because the “Individual Criminal Responsibility” section of that

statute was intended to codify existing norms of customary international law. In his report to the

Security Council regarding the creation of the ICTY, the Secretary-General explained that “in

assigning to the International Tribunal the task of prosecuting persons responsible for serious

violations of international humanitarian law, the Security Council would not be creating or

purporting to ‘legislate’ that law. Rather, the International Tribunal would have the task of

applying existing international humanitarian law.” Report of the Secretary-General Pursuant to

Paragraph 2 of Security Council Resolution 808, ¶ 29, U.N. Doc. S/25704 (May 3, 1993)

(“Sec’y-General Report”). Indeed, international law principles “require[d]” that the Tribunal’s

jurisdiction be limited to “rules of international humanitarian law which are beyond any doubt

part of customary [international] law.” Id. ¶ 34. Accordingly, the provision of aiding and

abetting liability in the ICTY statute reflects a determination by both the Secretary-General and

34
the Security Council, which approved the Secretary-General’s report when it enacted the statute,

that such liability is firmly established in customary international law. The inclusion of

substantively identical language in the statute creating the ICTR presumably reflects a similar

determination.8

Consistent with its statutory authorization, the ICTY has recognized and applied aiding

and abetting liability for violations of international law. See, e.g., Furundzija, Trial Chamber

Judgment, ¶¶ 249, 275; Prosecutor v. Tadic, Case No. IT-94-1-T, Trial Chamber Opinion and

Judgment, ¶¶ 689–92, 730, 735, 738 (May 7, 1997). Furthermore, it has done so only after

confirming that such liability was part of customary law. As the Tribunal recognized, it was

required to determine “the objective basis for such individual responsibility as a matter of

customary international law . . . since the International Tribunal is only empowered to apply

international humanitarian law that is ‘beyond any doubt customary law.’” Tadic, Trial Chamber

Opinion and Judgment, ¶ 662 (quoting Sec’y General Report ¶ 34). The Tribunal therefore

conducted a probing and thoughtful analysis of international law sources in its early decisions to

confirm that aiding and abetting liability is recognized in customary international law. See

Furundzija, Trial Chamber Judgment, ¶¶ 190–249; Tadic, Trial Chamber Opinion and Judgment,

¶¶ 661–91.

8
The Security Council did “elect[] to take a more expansive approach to the choice of the
applicable law” with respect to punishable violations by providing for liability for violations of
Additional Protocol II and common article 3 of the Geneva Convention despite the fact that Additional
article II had “not yet been universally recognized as part of customary international law,” and violations
of common article 3 had never before been criminalized. Report of Secretary-General Pursuant to
Paragraph 5 of Security Council Resolution 955, ¶ 12, UN doc. S/1995/134 (Feb. 13, 1995). With
respect to individual responsibility, though, the Security Council adhered to customary law by adopting
the language used in the ICTY Statute.

35
 More recently, the Rome Statute of the International Criminal Court (“Rome Statute”),

July 17, 1998, 2187 U.N.T.S. 90, provides that a person shall be criminally responsible and liable

for punishment for a crime within the jurisdiction of the ICC if that person:

(c) For the purpose of facilitating the commission of such a crime, aids, abets
or otherwise assists in its commission or its attempted commission,
including providing the means for its commission; [or]

(d) In any other way contributes to the commission or attempted commission
of such a crime by a group of persons acting with a common purpose.
Such contribution shall be intentional and shall either:
(i) Be made with the aim of furthering the criminal activity or criminal
purpose of the group, where such activity or purpose involves the
commission of a crime within the jurisdiction of the Court; or
(ii) Be made in the knowledge of the intention of the group to commit
the crime[.]

Id. art. 25 (3)(c), (d). The Rome Statute is particularly significant for the present inquiry because,

unlike other sources of international legislation, it articulates the mens rea required for aiding and

abetting liability. The Statute makes clear that, other than assistance rendered to the commission

of a crime by a group of persons acting with a common purpose, a defendant is guilty of aiding

and abetting the commission of a crime only if he does so “[f]or the purpose of facilitating the

commission of such a crime.” Id. art. 25 (3)(c).

In drawing upon the Rome Statute, I recognize that it has yet to be construed by the

International Criminal Court; its precise contours and the extent to which it may differ from

customary international law thus remain somewhat uncertain. Nevertheless, the Statute has been

signed by 139 countries and ratified by 105, including most of the mature democracies of the

36
world.9 It may therefore be taken “by and large . . . as constituting an authoritative expression of

the legal views of a great number of States.” Furundzija, Trial Chamber Judgment, ¶ 227.

Furthermore, the Rome Statute’s mens rea standard is entirely consistent with the

application of accomplice liability under the sources of international law discussed above.10 For

example, in the Ministries Case conducted under Control Council Law No. 10, the tribunal

declined to impose criminal liability on a bank officer who was alleged to have “made a loan,

knowing or having good reason to believe that the borrower w[ould] use the funds in financing

enterprises [conducted] in violation of either national or international law,” but was not proven to

have made the loan with the purpose of facilitating the enterprises’ illegal activities.11 United

States v. von Weizsaecker (The Ministries Case), in 14 Trials of War Criminals Before the

9
The United States has not ratified the Rome Treaty for reasons unrelated to any concern over
the definition of aiding-and-abetting. See Brief for the United States as Amicus Curaie, at 24 n.14; see
also Presbyterian Church of Sudan v. Talisman Energy, Inc., 374 F. Supp. 2d 331, 339–40 (S.D.N.Y.
2005) (“[T]he United States feared ‘unchecked power in the hands of the prosecutor’ that could lead to
‘politicized prosecution.’”). Indeed, the Amicus Brief appears to endorse the elements set out in Article
25(3). Brief for the United States as Amicus Curaie, at 26.
10
The same standard was also adopted by the United Nations Transitional Administration in East
Timor (“UNTAET”). See UNTAET Reg. No. 2000/15 § 14.3(c).
11
There are occasions when this intent could be inferred from such sales. In Direct Sales Co. v.
United States, 319 U.S. 703 (1943), the Supreme Court held that the sale of restricted goods with an
inherent capacity for harm, such as the opiates involved in that case, combined with other factors, may be
sufficient to prove that the seller was engaged in a conspiracy with the buyer. Id. at 711. Nevertheless, it
observed that “not every instance of sale of restricted goods, harmful as are opiates, in which the seller
knows the buyer intends to use them unlawfully, will support a charge of conspiracy.” Id. at 712. For
example, the Court held that a conspiracy charge might not lie against a company engaged in a
“continuous course of sales, made either with strong suspicion of the buyer’s wrongful use or with
knowledge, but without stimulation or active incitement to purchase.” Id. at 712 n.8. On the other hand,
sales of unlimited quantities stimulated “by all the high-pressure methods, legal if not always appropriate
in the sale of free commodities,” id. at 712, would suffice to support such a charge. Another example
that would suffice is the Zyklon B case, Trial of Bruno Tesch and Two Others, in 1 Law Reports of War
Criminals 93 (William S. Hein & Co., Inc. 1997) (1946), where the principal defendant not only supplied
prussic acid to the S.S. but undertook to train its members how it could be used to kill human beings. Id.
at 95.

37
Nuernberg Military Tribunals Under Control Council Law No. 10 308, 622 (William S. Hein &

Co., Inc. 1997) (1949). Meanwhile, those who assist in the commission of a crime with the

purpose of facilitating that crime would be subject to aiding and abetting liability under the

statutes governing the ICTY and ICTR.12 My research has revealed no source of international

law that recognizes liability for aiding and abetting a violation of international law but would not

authorize the imposition of such liability on a party who acts with the purpose of facilitating that

violation (provided, of course, that the actus reus requirement is also satisfied).

With respect to the actus reus component of the aiding and abetting liability, the

international legislation is less helpful in identifying a specific standard. However, in the course

of its analysis of customary international law, the ICTY concluded that “the actus reus of aiding

and abetting in international criminal law requires practical assistance, encouragement, or moral

support which has a substantial effect on the perpetration of the crime.” Furundzija, Trial

Chamber Judgment, ¶ 235 (second emphasis added). My research has uncovered nothing to

indicate that a standard other than “substantial assistance” should apply.

Accordingly, I conclude that a defendant may be held liable under international law for

12
These Tribunals would also extend liability to individuals who merely had “knowledge that
[their] acts assist the commission of the specific crime of the principal.” Prosecutor v. Vasiljevic, Case
No. IT-98-32-A, Appeals Chamber Judgment, ¶ 102 (ii) (Feb. 25, 2004). Any individual who acts with
the purpose to facilitate the commission of a crime would necessarily act with such knowledge. Thus, I
do not view the articulation of a broader definition by the ICTY as detracting from my position that
liability in accordance with the purposefulness standard is well-established and universally recognized
under international law. The critical question is whether there is a discernable core definition that
commands the same level of consensus as the 18th-century crimes identified by the Supreme Court in
Sosa. See Sosa, 542 U.S. at 732. I believe that the standard I adopt is such a definition. Cf. United
States v. Smith, 18 U.S. (5 Wheat.) 153, 161 (1820) (noting that “whatever may be the diversity of
definitions, in other respects, all writers concur, in holding, that robbery, or forcible depredations upon
the sea, animo furandi, is piracy”).

38
aiding and abetting the violation of that law by another when the defendant (1) provides practical

assistance to the principal which has a substantial effect on the perpetration of the crime, and (2)

does so with the purpose of facilitating the commission of that crime. Furthermore, based on this

review of international law’s treatment of aiding and abetting liability over the past sixty years, I

conclude that aiding and abetting liability, so defined, is sufficiently “well-established[] [and]

universally recognized” to be considered customary international law for the purposes of the

ATCA. See Kadic, 70 F.3d at 239 (internal quotation marks omitted). This conclusion comports

with the decisions of several other federal courts that have considered the issue.13 See Almog v.

Arab Bank, PLC, 471 F. Supp. 2d 257, 287 (E.D.N.Y. 2007) (noting the “vast body of law

finding aiding and abetting liability available under the [ATCA]”); Presbyterian Church of

Sudan v. Talisman Energy, Inc., 453 F. Supp. 2d 633, 668 (S.D.N.Y. 2006) (“Aiding and abetting

liability is a specifically defined norm of international character that is properly applied as the

law of nations for purposes of the [ATCA].”); Bowoto v. Chevron Corp., No. C 99-02506 SI,

2006 WL 2455752, *3–4 (N.D. Cal. Aug. 22, 2006); In re “Agent Orange” Prod. Liab. Litig.,

373 F. Supp. 2d 7, 52–54 (E.D.N.Y. 2005).

C.

While I conclude that, at present, only aiding and abetting liability imposed in accordance

with the standard outlined above is sufficiently well-established and universally recognized under

13
In light of this conclusion, I need not address the plaintiffs’ argument that a theory of aiding
and abetting liability could be supplied by domestic federal common law even if such liability did not
exist under international law.

39
international law to trigger jurisdiction under the ATCA, I appreciate that this definition is not

necessarily set in stone. International law, like our domestic law, can change, and the ATCA was

intended to change along with it. See Sosa, 542 U.S. at 728 (noting Congress’s intention in

passing the TVPA that the ATCA “‘remain intact to permit suits based on other norms that

already exist or may ripen in the future into rules of customary international law’” (quoting H.R.

Rep. No. 102-367, pt. 1, p. 3 (1991)); see also id. at 725 (articulating the standard for courts

hearing “any claim based on the present-day law of nations” (emphasis added)). In this regard, I

note that there is some support, principally in tribunal decisions from the ICTY and ICTR, for a

definition of aiding and abetting that would lead to liability where an individual provides

substantial assistance with “the knowledge that the acts performed by the aider and abettor assist

the commission of the specific crime of the principal.”14 Prosecutor v. Vasiljevic, Case No. IT-

98-32-A, Appeals Chamber Judgment, ¶¶ 102 (i)–(ii) (Feb. 25, 2004); see also Furundzija, Trial

Chamber Judgment, ¶¶ 249, 275; Tadic, Trial Chamber Opinion and Judgment, ¶¶ 689–92, 730,

735, 738.

To be sure, I acknowledge limitations on the extent to which we may rely on the

decisions of the ICTY and ICTR for the present purposes.15 Those decisions arise out of

14
The Appeals Chamber, which is shared by both the ICTY and the ICTR, has made clear that
the same law relating to modes of liability applies in both Tribunals. See Prosecutor v. Ntakirutimana,
Case Nos. ICTR-96-10-A, ICTR-96-17-A, Appeals Chamber Judgment, ¶ 468 (Dec. 13, 2004) (“Given
the fact that both the ICTY and the ICTR have mirror articles identifying the modes of liability by which
an individual can incur criminal responsibility, the Appeals Chamber is satisfied that the jurisprudence of
the ICTY should be applied to the interpretation of Article 6(1) of the ICTR Statute.”).
15
I note a possible tension in the tribunals’ definition aiding and abetting under which the
necessary mens rea is knowing assistance, Vasiljevic, Appeals Chamber Judgment, ¶ 102(ii), yet requires
that the act of assistance be “specifically directed to assist . . . the perpetration of a specific crime,” id. ¶
102(i). I express no view on how, if at all, this possible tension might be resolved.

40
completely distinct factual contexts and often involve defendants who might have been convicted

on alternate theories of liability. Moreover, as a scholar and participant in the ICTY noted,

panels of that Tribunal occasionally (and consciously) engaged in discussions “peripheral to the

ratio decidendi of a case” in order to provide “clarification [that] might have some value for the

future development of international criminal law.” Antonio Cassese, The ICTY: A Living and

Vital Reality, 2 J. Int’l Crim. Just. 585, 589–90 (2004). Such clarification through dicta can be

useful and is not necessarily without justification. See id. at 590 (“[T]he absence of an

international law-maker and an international court with compulsory universal jurisdiction entails

that many rules are not clear, particularly when they are of customary origin, and are thus open to

differing interpretations—hence the need for courts gradually to spell out the contents of those

rules, if need be through obiter dicta.”); see also 1 Oppenheim’s International Law: Peace 41

(Robert Jennings & Arthur Watts eds., 9th ed. 1992) (noting that “in view of the difficulties

surrounding the codification of international law,” international tribunals are expected to “fulfil,

inconspicuously but efficiently, a large part of the task of developing international law”). But it

“may also prove simply academic and sometimes also misleading for future courts pronouncing

on the same matters.” Cassese, supra, at 590.

Nevertheless, the opinions of the ICTY and ICTR provide evidence of the state of

customary international law. As a leading treatise explains, “decisions of international tribunals

. . . exercise considerable influence as an impartial and considered statement of the law by jurists

of authority in light of actual problems which arise before them.” 1 Oppenheim’s, supra, at 41;

see also Ian Brownlie, Principles of Public International Law at 19 (noting that “the practical

significance of the label ‘subsidiary means’ in Article 38(1)(d) is not to be exaggerated”). The

41
ICTY has been specifically recognized as having provided “a number of significant findings on

issues of law.” Brownlie, supra, at 23; see also Cassese, supra, at 591–93 (arguing that the ICTY

“clarified the contours of many notions of international criminal law,” including “the notion of

aiding and abetting”); cf. Sosa, 542 U.S. at 734 (“‘[W]here 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; and, as evidence of these, to the works of jurists and commentators, who by

years of labor, research and experience, have made themselves peculiarly well acquainted with

the subjects of which they treat.’” (quoting The Paquete Habana, 175 U.S. at 700) (alteration in

original)). While I am unable to find that liability predicated on the definition of aiding and

abetting offered in the decisions of the ICTY and ICTR is sufficiently well-established and

universally recognized to trigger jurisdiction for a tort suit under the ATCA, particularly in light

of the higher standard articulated in the Rome Statute, I am mindful of the conclusions of these

tribunals.16

16
I note further that other federal courts have consulted decisions of the ICTY and ICTR in
determining whether various rules are part of customary international law. See, e.g., Hamdan, 126 S.Ct.
at 2785 n. 40 (ICTY); Sosa, 542 U.S. at 762 (Breyer, J. concurring) (ICTY); Ford ex rel. Estate of Ford
v. Garcia, 289 F.3d 1283, 1290–93 (11th Cir. 2002) (ICTY and ICTR); Tagaga v. INS, 228 F.3d 1030,
1035 n.10 (9th Cir. 2000) (ICTY); Almog, 471 F. Supp. 2d at 286 (noting that “[s]tandards for imposing
liability where a non-primary actor is alleged to be liable for a violation of the law of nations emerge
from,” in part, the ICTY and ICTR case law); Presbyterian Church of Sudan, 453 F. Supp. 2d at 666–68
(analyzing ICTY cases to determine the elements of aiding and abetting liability under international law);
In re “Agent Orange” Prod. Liab. Litig., 373 F. Supp. 2d at 54 (same); Mehinovic v. Vuckovic, 198 F.
Supp. 2d 1322, 1356 (N.D. Ga. 2002).

42
 D.

1.

Judge Korman argues that it is insufficient to inquire whether aiding and abetting is

generally recognized under customary international law and that the appropriate inquiry is instead

to engage in a “norm-by-norm analysis” to determine whether aiding and abetting liability exists

for the violation of a particular rule. Opinion of Judge Korman at 134. Yet this is not how the

inquiry is undertaken by international tribunals whose jurisdiction is limited by customary

international law.

In Tadic, the Appeals Chamber of the ICTY considered whether a defendant could be

held criminally responsible under international law on a common purpose or joint criminal

enterprise (JCE) theory of liability. Prosecutor v. Tadic, Case No. IT-94-1-A, Appeals Chamber

Judgment, ¶ 185 (July 15, 1999). After concluding that the ICTY Statute permitted such liability,

the tribunal turned to customary law.17 The Appeals Chamber reviewed relevant sources of

customary law including case law from post–World War II war crimes cases, international

treaties, and the domestic law of many countries. It then concluded that “the notion of common

design as a form of accomplice liability is firmly established in customary international law.” Id.

¶ 220. The Appeals Chamber made no effort, though, to distinguish among the different crimes

(or even broad categories of crimes) over which it had jurisdiction or to verify the existence of

17
See Prosecutor v. Milutinovic, Case No. IT-99–37-AR72, Appeals Chamber Decision on
Dragoljub Ojdanic’s Motion Challenging Jurisdiction - Joint Criminal Enterprise, ¶ 17 (May 21, 2003)
(noting that each chamber has a “duty to ascertain that a . . . form of liability charged in the indictment is
both provided for under the Statute and that it existed at the relevant time under customary international
law.”).

43
JCE liability at customary law for each individual crime. Later panels have confirmed that

Tadic’s review of “state practice and opinio juris . . . was sufficient” to establish that “such a

norm existed under customary international law.” Prosecutor v. Milutinovic, Case No. IT-

99–37-AR72, Appeals Chamber Decision on Dragoljub Ojdanic’s Motion Challenging

Jurisdiction - Joint Criminal Enterprise, ¶ 29 (May 21, 2003); see also Prosecutor v. Brdjanin,

Case No. IT-99-36-A, Appeals Chamber Judgment, ¶ 363 (Apr. 3, 2007).

These tribunals took the same approach with respect to the application of aiding and

abetting liability. In Tadic, the ICTY trial chamber viewed international treaties and trials

following the Second World War as “establish[ing] the basis in customary international law for

both individual responsibility and of participation in the various ways provided by Article 7 of

the [ICTY] Statute.” Tadic, Trial Chamber Opinion and Judgment, ¶ 669. In Furundzija, the

trial chamber consulted post–World War II case law, as well as modern authoritative

international instruments, to “establish the content” of aiding and abetting liability under

customary international law. Furundzija, Trial Chamber Judgment, ¶¶ 191, 195–231. In neither

case did the Tribunal engage in the norm-specific inquiry advocated by Judge Korman. Indeed,

the Furundzija Trial Chamber concluded that “[t]he definitions and propositions concerning

aiding and abetting [it] enunciated . . . apply . . . to all crimes,” even though it had not purported

to identify sources relating to every crime over which it had jurisdiction. Id. ¶ 250. Yet these

analyses are recognized as authoritative by other chambers of the Tribunals. See, e.g.,

Prosecutor v. Aleksovski, Case No. IT-95-14/1-A, Appeals Chamber Judgment, ¶ 162 (Mar. 24,

2000) (noting that “[t]he liability of a person charged with aiding and abetting another person in

the commission of a crime was extensively considered by Trial Chamber II in the Furundzija

44
Judgment” and adopting its conclusions); Prosecutor v. Bagilishema, Case No. ICTR-95-1A-T,

Trial Chamber Judgment, ¶ 32 n.22 (June 7, 2001) (citing Tadic for discussion of “the customary

nature of the[] principles” of aiding and abetting liability).

The international tribunals’ approach is consistent with the understanding that aiding and

abetting is a theory of liability for acts committed by a third party. As we have recognized in our

domestic criminal law, “aiding and abetting ‘does not constitute a discrete criminal offense but

only serves as a more particularized way of identifying persons involved’” in the underlying

offense. United States v. Smith, 198 F.3d 377, 383 (2d Cir. 1999) (quoting United States v.

Oates, 560 F.2d 45, 54 (2d Cir. 1977) (internal quotation marks omitted)); see also Hefferman v.

Bass, 467 F.3d 596, 601 (7th Cir. 2006) (explaining that “aiding and abetting is a theory for

holding the person who aids and abets liable for the tort itself”). International law is consistent

with domestic law on this point. See, e.g., Prosecutor v. Kunarac, Case Nos. IT-96-23-T & IT-

96-23/1-T, Trial Chamber Judgment, ¶ 391 (Feb. 22, 2001) (“As opposed to the ‘commission’ of

a crime, aiding and abetting is a form of accessory liability.”); Akayesu, Trial Chamber

Judgment, ¶ 527 (defining an accomplice as “someone who associates himself in an offence

committed by another”); see also Hamdan, 126 S. Ct. at 2785 n.40 (“The International Criminal

Tribunal for the former Yugoslavia (ICTY), drawing on the Nuremberg precedents, has adopted a

‘joint criminal enterprise’ theory of liability, but that is a species of liability for the substantive

offense (akin to aiding and abetting), not a crime on its own.”). Because aiding and abetting is a

generally applicable means of identifying who should be held responsible for a particular act,

rather than a necessary element of the act itself, it is more reasonable to consider whether the

theory is accepted as a general principle of customary international law than to ask whether each

45
substantive norm that proscribes a specific conduct encompasses liability for aiding and abetting.

See, e.g., Rome Statute (including the provision on individual criminal responsibility in “Part III:

General Principles of Criminal Law” and the definition of the substantive offenses in Part II);

ICTR Statute (segregating the article defining individual criminal responsibility from the articles

describing the substantive offenses); ICTY Statute (same); Draft Code of Crimes Against the

Peace and Security of Mankind, Report of the International Law Commission on the work of its

forty-eighth session, U.N. GAOR, 51st Sess., Supp. No. 10, U.N. Doc. A/51/10 (1996)

(including the article on individual responsibility in “Part One: General Provisions” and the

definitions of crimes in “Part Two: Crimes Against the Peace and Security of Mankind”).

2.

Viewing aiding and abetting in this way, as a theory of identifying who was involved in

an offense committed by another rather than as an offense in itself, also helps to explain why a

private actor may be held responsible for aiding and abetting the violation of a norm that requires

state action or action under color of law. It is true, as we held in Kadic, that “certain forms of

conduct” violate the law of nations only when undertaken by state actors or those acting under

color of law. Kadic, 70 F.3d at 239. But imposing liability on private actors who aid and abet

such conduct does not, contrary to Judge Korman’s suggestion, detract in any way from this

requirement. Indeed, the imposition of aiding and abetting liability under international law

requires “a predicate offence committed by someone other than the accomplice,” in this case, a

state actor or someone acting under color of law. Akayesu, Trial Chamber Judgment, ¶ 529.

Recognizing the responsibility of private aiders and abettors merely permits private actors who

substantially assist state actors to violate international law and do so for the purpose of

46
facilitating the unlawful activity to be held accountable for their actions.

It is of no moment that a private actor could be held liable as an aider and abettor of the

violation of a norm requiring state action when that same person could not be held liable as a

principal. In our domestic law, it is “well settled that one may be found guilty of aiding and

abetting another individual in his violation of a statute that the aider and abettor could not be

charged personally with violating.” In re Nofziger, 956 F.2d 287, 290 (D.C. Cir. 1992); see also

United States v. Tannenbaum, 934 F.2d 8, 14 (2d Cir. 1991) (“The fact that the accused does not

possess the legal capacity to commit the substantive offense does not mean that he cannot be

convicted . . . of aiding and abetting the commission of the substantive offense by another. Thus,

the inability to commit the substantive offense is immaterial.” (citations omitted)). Indeed, “[t]he

doctrine is of ancient origin.” Nofziger, 956 F.2d at 291. International law, too, recognizes that

criminality is assessed by reference to the actions of the principal, not the aider and abettor. See

Akayesu, Trial Chamber Judgment, ¶ 528 (“[I]t should be understood that the physical act which

constitutes the act of complicity does not have its own inherent criminality, but rather it borrows

the criminality of the act committed by the principal perpetrator of the criminal enterprise. . . .

The accomplice has not committed an autonomous crime, but has merely facilitated the criminal

enterprise committed by another.”).

The defendants and the government argue that the Supreme Court’s decision in Central

Bank of Denver, N.A. v. First Interstate Bank of Denver, N.A., 511 U.S. 164 (1994), precludes us

from recognizing aiding and abetting liability for claims brought under the ATCA because

“where Congress has not explicitly provided for aider and abettor liability in civil causes of

action, it should not be inferred.” See In re S. African Apartheid Litig., 346 F. Supp. 2d at 550.

47
The Court’s holding in Central Bank was primarily premised on a recognition that § 10(b) does

not prohibit aiding and abetting, and “the private plaintiff may not bring a 10b-5 suit against a

defendant for acts not prohibited by the text of § 10(b).” Central Bank, 511 U.S. at 173. Here,

however, the ATCA provides jurisdiction for the courts to hear torts “committed in violation of

the law of nations.” The Supreme Court’s instruction in Central Bank that “when Congress

enacts a statute under which a person may sue and recover damages from a private defendant for

the defendant’s violation of some statutory norm, there is no general presumption that the

plaintiff may also sue aiders and abettors,” 511 U.S. at 182, is thus inapposite. Under the ATCA,

the relevant norm is provided not by domestic statute but by the law of nations, and that law

extends responsibility for the violations of its norms to aiders and abettors. See William R.

Casto, The New Federal Common Law of Tort Remedies for Violations of International Law, 37

Rutgers L.J. 635, 650 (2006) (noting that “[t]he precise holding in Central Bank . . . does not

translate easily to [ATCA] litigation,” and that “at a fundamental level” Central Bank in fact

suggests that a theory of aiding and abetting liability may be recognized if “a norm of

international law forbids private persons to assist violators”). Thus, I conclude that customary

international law recognizes liability for aiding and abetting violations of international law, and

that the district court erred when it reached the contrary conclusion.

3.

Judge Korman further argues that the defendants cannot be held liable as aiders and

abettors because the sources that establish accessorial liability do not extend that liability to

corporations. Opinion of Judge Korman at 118. This argument was not raised by the defendants

on appeal and therefore the issue was not briefed by the parties. It is perhaps not surprising that

48
neither the defendants nor the United States raised this issue as a bar to liability: We have

repeatedly treated the issue of whether corporations may be held liable under the ATCA as

indistinguishable from the question of whether private individuals may be. See Bigio v. Coca-

Cola Co., 239 F.3d 440, 447 (2d Cir. 2000) (asking “whether Coca-Cola can have violated ‘the

law of nations’ if it acted solely as a non-governmental entity”); see also Flores, 414 F.3d at 244

(making no distinction between private individuals and corporations and noting that “certain

activities are of ‘universal concern’ and therefore constitute violations of customary international

law not only when they are committed by state actors, but also when they are committed by

private individuals” (citing Kadic, 70 F.3d at 239–40)); cf. Sosa, 542 U.S. at 732 n.20

(classifying both corporations and individuals as “private actor[s]”). Regardless, because the

defendants have not objected to the imposition of liability on this basis, we need not reach the

issue at this time.

E.

The Ntsebeza and Digwamaje Plaintiffs argue that the district court also erred in focusing

on whether allegations of accessorial liability would support jurisdiction under the ATCA

because they also brought claims of direct liability. To the extent that the court did confront

these direct liability claims, it appears to have rejected them either on the ground that the

plaintiffs failed to allege state action or on the ground that the treaties upon which the claims

were based were not self-executing and therefore could not satisfy the ATCA’s jurisdictional

prerequisite. Both of these bases appear to rest on misunderstandings of our ATCA

49
jurisprudence.

First, it appears that the district court did not distinguish between norms of international

law that apply only to state actors and those norms that proscribe conduct even when not taken

under the color of law. For example, the district court held that the plaintiffs in this case “do not

allege actions by the defendants that elevate them to the status of state actors in the commission

of torture, genocide, killings, and other serious crimes.” In re S. African Apartheid Litig., 346 F.

Supp. 2d at 548–59. As we squarely held in Kadic, however, “the proscription of genocide [by

customary international law] has applied equally to state and non-state actors,” and “acts of rape,

torture, and summary execution . . . are actionable under the [ATCA], without regard to state

action, to the extent that they were committed in pursuit of genocide or war crimes.” 70 F.3d at

242, 244. It is therefore not relevant whether the plaintiffs sufficiently allege that the defendants

acted under color of law in the commission of genocide as long as they sufficiently allege that the

defendants committed genocide.

Second, I believe that the district court overstated the weight we have placed on the self-

executing status of a treaty in our consideration of its weight as evidence of customary

international law. In Flores, we explained that “a treaty that is self-executing or that has been

executed through an Act of Congress—and therefore gives rise to rights legally enforceable in

our courts—provides greater evidence of the customs and practices of the United States than a

treaty that has not been executed.” 414 F.3d at 257. We did not hold that non-self-executing

treaties are without any evidentiary value with regard to the state of current customary

international law, much less that “no liability based upon any alleged violation of the[] norms

[articulated in such treaties] can form an adequate predicate for jurisdiction under the ATCA.”

50
In re S. African Apartheid Litig., 346 F. Supp. 2d at 552. Indeed, in Kadic we relied on the

Genocide Convention to determine the shape of the international proscription of genocide and

made clear that “the legislative decision not to create a new private remedy does not imply that a

private remedy is not already available under the [ATCA].”18 Kadic, 70 F.3d at 241–42. The

relevant inquiry for establishing jurisdiction under the ATCA, as our case law makes clear, is

whether the conduct alleged by the plaintiffs violates a norm “that States universally abide by, or

accede[] to, out of a sense of legal obligation and mutual concern.” Flores, 414 F.3d at 248.

Whether a treaty that embodies that norm is self-executing is relevant to, but is not determinative

of, that question.

III

For the foregoing reasons, I join in the per curiam opinion.

18
Apparently recognizing that the Genocide Convention may be used to support the existence of
genocide as a violation of the law of nations, Judge Korman nonetheless suggests that the district court
should not recognize a cause of action for genocide because “Congress has quite plainly indicated its
intention that such a cause of action should not be available” through its passage of the Genocide
Convention Implementation Act, 18 U.S.C. § 1091. Opinion of Judge Korman at 116–17; see also 18
U.S.C. § 1092 (providing that nothing in the Act should be “construed as creating any substantive or
procedural right enforceable by law by any party in any proceeding”). It is unnecessary to resolve this
issue as part of the jurisdictional analysis, and I would leave it to the district court to address in the first
instance should it undertake to decide whether to recognize a cause of action.

51
Hall, Circuit Judge, concurring:

As reflected in the per curiam opinion, I agree with Judge Katzmann with respect to the

ultimate disposition of this appeal. The district court erred when it ruled that it lacked

jurisdiction under the ATCA to determine plaintiffs’ claims based on defendants’ accessorial

liability. In ruling that it lacked subject matter jurisdiction under the ATCA, the district court

required that “aiding and abetting international law violations [be] itself an international law

violation that is universally accepted as a legal obligation.” In re S. African Apartheid Litig., 346

F. Supp. 2d 538, 549 (S.D.N.Y. 2004). In other words, the district court assumed that a federal

court must look to international law to divine not only the applicable primary violation of

international law cognizable under the ATCA, but also the standard for aiding and abetting

liability. The district court went on to conclude that aiding and abetting liability did not exist as a

matter of customary international law and thus that federal subject matter jurisdiction did not lie.

This conclusion was error. As Sosa makes clear, a federal court must turn to international law to

divine standards of primary liability under the ATCA. To derive a standard of accessorial

liability, however, a federal court should consult the federal common law.

I

A

We begin with the text. In its entirety, the ATCA provides:

The district courts shall have original jurisdiction of any civil action
by an alien for a tort only, committed in violation of the law of
nations or a treaty of the United States.

52
28 U.S.C. § 1350. The First Congress enacted the statute as part of the Judiciary Act of 1789.

Over the course of the last 200-odd years, the text of the statute has changed only slightly, but

“little is known of the framers’ intentions in adopting it—the legislative history of the Judiciary

Act does not refer to section 1350.” Presbyterian Church of Sudan v. Talisman Energy, Inc., 244

F. Supp. 2d 289, 304 (S.D.N.Y. 2003). Few parties invoked the ATCA until our decision in

Filartiga v. Pena-Irala, 630 F.2d 876 (2d Cir. 1980), which heralded the modern era of ATCA

litigation. But from the very beginning, the ATCA’s straightforward text belied complex and

controversial problems of exegesis and praxis. Struggling with these problems in the wave of

litigation that followed Filartiga, we observed that “neither Congress nor the Supreme Court

ha[d] definitively resolved the complex and controversial questions regarding the meaning and

scope of the ATCA.” Flores v. S. Peru Copper Corp., 414 F.3d 233, 247 (2d Cir. 2003).

One year later, in an attempt to clarify the problems posed by the ATCA, the Supreme

Court issued Sosa v. Alvarez-Machain, 542 U.S. 692 (2004). Principally, Sosa resolved that the

ATCA was not “a jurisdictional convenience to be placed on the shelf for use by a future

Congress . . . that might, someday, authorize the creation of causes of action.” Id. at 719.

Instead, the ATCA recognizes “claim[s] based on the present-day law of nations” so long as they

“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.” Id. at 725.

B

In their complaints, the plaintiffs make claims that rest on such norms “defined with a

53
specificity comparable to the features of the 18th-century paradigms.” The allegations include

genocide, torture, cruel and degrading treatment, systematic racial discrimination, forced

removals, and various other avowed crimes against humanity. In essence, the plaintiffs assert

that apartheid itself is a crime against humanity, 1 on a par with genocide2 and slavery, and that

the depredations of the apartheid system (such as torture, sexual assault, extrajudicial killing, and

forced labor) were perpetrated in the course of genocide and crimes against humanity. In their

brief and at oral argument, the corporations did not contest that such allegations reflected

violations of customary international law.3 Though their complaints are not models of precision,

1
The Charter of the International Military Tribunal (Nuremberg Tribunal) describes crimes
against humanity as:
murder, extermination, enslavement, deportation, and other inhumane acts committed
against any civilian population, before or during the war, or persecutions on political, racial,
or religious grounds in execution of or in connection with any crime within the jurisdiction
of the Tribunal, whether or not in violation of domestic law of the country where
perpetrated.
Charter of the International Military Tribunal, art. 6(c), in Agreement for the Prosecution and
Punishment of the Major War Criminals of the European Axis, Aug. 8, 1945, E.A.S. No. 472.
2
The Genocide Convention defines genocide as:
any of the following acts committed with intent to destroy, in whole or in part, a national,
ethnical, racial or religious group, as such:
(a) Killing members of the group;
(b) Causing serious bodily or mental harm to members of the group;
(c) Deliberately inflicting on the group conditions of life calculated to bring
about its physical destruction in whole or in part.
Convention on the Prevention and Punishment of the Crime of Genocide, art. 2, Jan. 12, 1951, 78
U.N.T.S. 277.
Notably, the Apartheid Convention echoes the definition of genocide, in that it applies to
“murder of members of a racial group or groups,” “the infliction upon the members of a racial group or
groups of serious bodily or mental harm,” and the “[d]eliberate imposition on a racial group or groups of
living conditions calculated to cause its or their physical destruction in whole or in part.” Convention on
Apartheid, art. 2, Nov. 30, 1973, 1015 U.N.T.S. 243.
3
At oral argument, counsel for the corporations added that he did not accept the premise that the
plaintiffs had actually pled those allegations, as the complaints did not allege that particular plaintiffs
suffered particular injuries at the hands of specific South African officials whom the corporations aided
and abetted in a specific way at a specific time and place. As reflected in the per curiam opinion,

54
the plaintiffs, if given the opportunity to replead, likely would allege primary violations of

international law cognizable under the ATCA.

C

In addition to its delineation of the standard by which federal courts derive primary

violations of international law, the Sosa opinion also contains numerous dicta. In his concurring

opinion, Judge Katzmann thoroughly summarizes these dicta. It remains inescapable, however,

that Sosa at best lends Delphian guidance on the question of whether the federal common law or

customary international law represents the proper source from which to derive a standard of

aiding and abetting liability under the ATCA. Lacking the benefit of clear guidance, I presume a

federal court should resort to its traditional source, the federal common law, when deriving the

standard. Because I find that federal common law provides a standard by which to assess aiding

and abetting liability, I do not address the alternative argument that such a standard may be

derived from international law.4

It is a “hornbook principle that international law does not specify the means of its

domestic enforcement.” Brief for the International Law Scholars as Amici Curiae at 5–6; see

also Brief for the United States of America as Amicus Curiae at 5 (“[A]lthough the substantive

plaintiffs may have an opportunity to replead these violations with greater specificity on remand.
4
I note that Judge Katzmann, in his concurring opinion, declines to address whether federal
common law provides a standard by which to determine aiding and abetting liability in this case.
Opinion of Judge Katzmann at 39 n.12. It is thus left to a future panel of this Court to determine whether
international or domestic federal common law is the exclusive source from which to derive the applicable
standard.

55
norm to be applied is drawn from international law or treaty, any cause of action recognized by a

federal court is one devised as a matter of federal common law.”). As Judge Edwards has

explained, “the law of nations never has been perceived to create or define the civil actions to be

made available by each member of the community of nations.” Tel-Oren v. Libyan Arab

Republic, 726 F.2d 774, 778 (D.C. Cir. 1984) (Edwards, J., concurring); see also Kadic v.

Karadzic, 70 F.3d 232, 246 (2d Cir. 1995) (“The law of nations generally does not create private

causes of action to remedy its violations, but leaves to each nation the task of defining the

remedies that are available for international law violations.”); Xuncax v. Gramajo, 886 F. Supp.

162, 180 (D. Mass. 1995) (“While it is demonstrably possible for nations to reach some

consensus on a binding set of principles, it is both unnecessary and implausible to suppose that,

with their multiplicity of legal systems, these diverse nations should also be expected or required

to reach consensus on the types of actions that should be made available in their respective courts

to implement those principles.”); Restatement (Third) of Foreign Relations Law § 703 cmt. c (“In

general, individuals do not have direct international remedies against a state violating their

human rights except where such remedies are provided by international agreement. Whether

they have a remedy under the law of a state depends on that state’s law.” (internal citation

omitted)); Beth Stephens, Sosa v. Alvarez-Machain: “The Door Is Still Ajar” for Human Rights

Litigation in U.S. Courts, 70 Brook. L. Rev. 533, 558 (2004) (“Sosa does not require that every

ancillary rule applied in an AT[CA] case meet the level of international consensus required for

the definition of the underlying violation. As in any case in which the federal courts exercise

discretion to recognize federal common law, the courts will fashion rules to fill gaps, borrowing

from the most analogous body of law.”).

56
 Numerous international law scholars have described this principle. See L. Henkin,

Foreign Affairs and the Constitution 224 (1972); Oppenheim, 1 International Law 44–46 (8th ed.

1955); L. Henkin, R. Pugh, O. Schachter & H. Smit, International Law 116 (1980); 4

Blackstone’s Commentaries 72 (Welsby ed., 1854) (noting that the law of nations recognized

certain universal offenses but that accessorial liability is made available “by statute”). As amicus

International Law Scholars persuasively argue, these “means of domestic enforcement”

encompass at least some theories of accessorial liability, including aiding and abetting. Brief for

International Law Scholars as Amici Curiae at 6. I believe our Court should stand by this

principle. Moreover, when international law and domestic law speak on the same doctrine,

domestic courts should choose the latter. Oppenheim, supra, at 44–46. This, too, is a principle

our Court should respect. Here, customary international law and the federal common law both

include standards of aiding and abetting. In a situation such as this, I opt for the standard

articulated by the federal common law.

Supreme Court precedent commands the same result. As Judge Reinhardt noted when

concurring in Doe I v. Unocal Corp., 395 F.3d 932 (9th Cir. 2002), the ATCA is silent “as to

what body of law applies to ancillary issues that may arise, such as whether a third party may be

held liable in tort” for a violation of international norms. Id. at 965 (Reinhardt, J., concurring).

Typically, federal courts look to the federal common law to fill such an interstice. Id. at 966

(citing United States v. Kimbell Foods, 440 U.S. 715, 727 (1979)); see also Texas Indus., Inc. v.

Radcliff Materials, Inc., 451 U.S. 630, 641 (1981) (stating that courts should apply the federal

common law “in such narrow areas as those concerned with the rights and obligations of the

United States, interstate and international disputes implicating the conflicting rights of States or

57
our relations with foreign nations”). Until the Supreme Court provides us more explicit guidance

regarding accessorial liability than it has to date, I remain convinced that our federal common

law embodies a clearly extant standard of aiding and abetting liability. See Unocal, 395 F.3d at

967. It is to this standard that federal courts should resort and to which I now turn.

The Supreme Court has described Halberstam v. Welch, 705 F.2d 472 (D.C. Cir. 1983) as

“a comprehensive opinion on the subject [of aiding and abetting].” Cent. Bank, N.A. v. First

Interstate Bank, N.A., 511 U.S. 164, 181 (1994). Halberstam relied heavily upon the

Restatement (Second) of Torts to set the parameters of aiding and abetting liability. Section

876(b) of the Restatement provides: “For harm resulting to a third person from the tortious

conduct of another, one is subject to liability if he . . . (b) knows that the other’s conduct

constitutes a breach of duty and gives substantial assistance or encouragement to the other so to

conduct himself.” Based on the Restatement, Halberstam held that aiding and abetting included

three elements:

(1) the party whom the defendant aids must perform a wrongful act that causes an
injury; (2) the defendant must be generally aware of his role as part of an overall
illegal or tortious activity at the time that he provides the assistance; [and] (3) the
defendant must knowingly and substantially assist the principal violation.

Halberstam, 705 F.2d at 477. The Halberstam Court then adopted and applied the factors

enumerated in § 876 to assess whether the defendant’s encouragement or assistance was

sufficiently substantial to support liability. “The Restatement suggests five factors in making this

determination: ‘the nature of the act encouraged, the amount of assistance given by the

defendant, his presence or absence at the time of the tort, his relation to the other tortfeasor and

his state of mind.’” Id. at 478 (quoting Restatement (Second) of Torts § 876 cmt. d) (alteration

58
omitted). In the almost quarter-century since Halberstam was decided, many state courts and

Circuit Courts, including the Second Circuit, have adopted the Restatement’s aiding and abetting

standard. Following the lead of the Halberstam Court, I believe that § 876 provides the proper

standard under which to assess whether a particular defendant may be held liable for aiding and

abetting a primary violation of international law.5 I also agree with the Halberstam Court that “a

person who assists a tortious act may be liable for other reasonably foreseeable acts done in

connection with it.” Id. at 484.

Under a proper application of § 876 to ATCA civil aiding and abetting claims, liability

should be found only where there is evidence that a defendant furthered the violation of a clearly

5
I agree with Judge Katzmann that Central Bank, 511 U.S. at 164, poses no bar to finding aiding
and abetting liability under the ATCA, albeit for a different reason. In Central Bank, the Supreme Court
held that the Securities Acts of 1933 and 1934 did not encompass aiding and abetting liability. 511 U.S.
at 171. Noting that the Acts provided for some forms of “indirect” liability, the Supreme Court reasoned
that “Congress knew how to impose aiding and abetting liability when it chose to do so.” Id. at 176.
This same reasoning cannot apply to the ATCA, whose textual brevity and dearth of legislative history
leave us with inconclusive evidence of Congress’s intent to include or exclude aiding and abetting
liability. It would appear, however, that the Founding Generation nevertheless understood the ATCA
encompassed aiding and abetting liability. For example, in Attorney General Bradford’s 1795 opinion,
Breach of Neutrality, 1 Op. Att’y Gen. 57, 59 (1795), he opined that the ATCA allowed civil suits for
damages for those who had “taken part” in violating international law. Sosa, 542 U.S. at 721. In fact,
Bradford’s opinion specifically covered those American citizens who “voluntarily joined, conducted,
aided and abetted” the French fleet in their attack on a British settlement. Breach of Neutrality, 1 Op.
Att’y Gen. at 58 (emphasis added). Attorney General Bradford furthermore referred to an April 1793
Proclamation issued by George Washington which declared that “all those who should render themselves
liable to punishment under the laws of nations, by committing, aiding or abetting hostilities” against the
merchants of foreign nations at peace with the United States would not receive the protection of the
United States. Id. at 59. Cases from that era, moreover, indicate that secondary liability was recognized
as an established part of the federal common law. See Talbot v. Jansen, 3 U.S. (3 Dall.) 133 (1795)
(holding a defendant liable for aiding the unlawful capture of a neutral ship and ordering restitution);
Henfield’s Case, 11 F. Cas. 1099, 1103 (C.C.D. Pa. 1793) (Chief Justice John Jay) (charging a grand jury
that United States citizens may be held liable under the laws of the United States for “committing, aiding
or abetting hostilities” in violation of the law of nations); see also Congress’s Act of April 30, 1790,
which deemed “an accessary [sic] to . . . piracies” anyone who “knowingly and willingly aid[ed]” piracy.
1 Stat. 114 (1790).

59
established international law norm in one of three ways: (1) by knowingly and substantially

assisting a principal tortfeasor, such as a foreign government or its proxy, to commit an act that

violates a clearly established international law norm; (2) by encouraging, advising, contracting

with, or otherwise soliciting a principal tortfeasor to commit an act while having actual or

constructive knowledge that the principal tortfeasor will violate a clearly established customary

international law norm in the process of completing that act; or (3) by facilitating the commission

of human rights violations by providing the principal tortfeasor with the tools, instrumentalities,

or services to commit those violations with actual or constructive knowledge that those tools,

instrumentalities, or services will be (or only could be) used in connection with that purpose.

All members of this panel understand that corporations must transact business in a less

than perfect world. I do not understand defendants to argue, however, that business imperatives

require a license to assist in violations of international law. Rather, I understand defendants to

express the concern that the recognition of ATCA aiding and abetting liability could expose

corporations to liability for merely doing business in countries with repressive regimes or for

participating in activities that, with twenty-twenty hindsight, can be said to have been indirectly

linked to human rights abuses. I share Judge Katzmann’s understanding, however, that private

parties and corporate actors are subject to liability under the ATCA. I therefore concur in Part

II.D.2 of Judge Katzmann’s opinion to the extent that it recognizes the general legal principle

that aiding and abetting liability obtains even where the culpable actor is incapable of violating

the relevant legal norm as a principal, and I concur in Part II.D.3 of his opinion in full.

Defendants raise important concerns about such liability, but those concerns do not counsel in

favor of the per se rejection of corporate liability, private party liability, and aiding and abetting

60
liability under the ATCA. Instead, they require the narrow and careful extension of such liability

to cases in which a defendant played a knowing and substantial role in the violation of a clearly

recognized international law norm. Furthermore, the collateral consequences predicted by

defendants and the dissent remain relevant considerations when making these judgments, as the

common law nature of the inquiry allows for a limited but meaningful consideration of the

practical consequences of extending ATCA liability in the context of each particular case.

Because I intend aiding and abetting liability to attach only in this limited way, I think it

helpful to provide examples illustrating the three ways in which I believe a defendant may incur

liability for aiding and abetting violations of customary international law. The first type of aiding

and abetting liability is designed to capture the case of a principal tortfeasor who seeks assistance

from a defendant to commit an act that violates international law norms, such as the extrajudicial

killing of an opposition political figure. The second is designed to cover circumstances where

the alleged aider and abettor is accused of having purchased security services with the knowledge

that the security forces would, or were likely to, commit international law violations in fulfilling

their mandate. The allegations raised in the cases of Unocal, 395 F.3d 932, and Wiwa v. Royal

Dutch Petroleum Co., No. 96-Civ-8386, 2002 WL 319887 (S.D.N.Y. Feb. 28, 2002), would be

reached by this prong. In Unocal, the alleged aider and abettor corporation was accused of

having purchased security services from a military government to further develop its oil

operations, with the knowledge that the security forces would likely commit international law

violations in fulfilling this mandate. 395 F.3d at 938–42. In Wiwa, the plaintiffs alleged that the

defendants directed and aided government security forces in violating plaintiffs’ rights by

providing logistical support, transportation, and weapons to government security forces to ensure

61
that the corporation’s business activities could proceed “as usual.” Wiwa, No. 96-Civ-8386,

2002 WL 319887, at *2.

The Zyklon B Case provides a clear example of when liability would attach in the third

circumstance, when a defendant provides “the tools, instrumentalities, or services to commit

[human rights] violations with actual . . . knowledge that those tools, instrumentalities or services

will be (or only could be) used in connection with that purpose.” See Trial of Bruno Tesch and

Two Others (The Zyklon B Case), 1 Law Reports of Trials of War Crim. 93 (1947) (British

Military Ct., Hamburg, Mar. 1-8, 1946). In that case, Bruno Tesch was the sole owner of a firm

that distributed Zyklon B, a highly dangerous poison gas, to Auschwitz and other concentration

camps from 1941 to 1945. Zyklon B previously had been used as a disinfectant in public

buildings. The evidence showed that Tesch himself proposed using the gas to exterminate

human beings, undertook to train the S.S. in this “new method of killing,” and was aware that, by

June 1942, the gas was being used for such a purpose. Id. at 95. The Prosecutor successfully

argued that “knowingly to supply a commodity to a branch of the State which was using that

commodity for the mass extermination of Allied civilian nationals was a war crime, and that the

people who did it were war criminals for putting the means to commit the crime into the hands of

those who actually carried it out.” Id. at 94.

These examples are illustrative rather than exhaustive, and I offer them in an effort to

provide greater substantive content for a doctrinal framework that, like most common law rules,

is “vague” and abstract in its articulation of legal obligations and culpable conduct. Opinion of

Judge Korman at 133 (criticizing Restatement aiding-and-abetting standard as “vague and

inappropriate in the present context”). Standards such as “substantial assistance” and “actual or

62
constructive knowledge” are hardly “newly minted,” however, and judicial decisions interpreting

the meaning of these standards in the context of individual cases will provide guidance to courts

considering accessorial liability under the ATCA. See, e.g., Henry v. Lehman Commercial

Paper, Inc., 471 F.3d 977, 993 (9th Cir. 2006) (applying California law on the “knowledge”

requirement of § 876); Aetna Cas. & Sur. Co. v. Leahey Constr. Co., 219 F.3d 519, 535–37 (6th

Cir. 2000) (applying knowledge and substantial assistance requirements of § 876); In re

Temporomandibular Joint (TMJ) Implants Prods. Liab. Litig., 113 F.3d 1484, 1496 (8th Cir.

1997) (addressing “substantial assistance” requirement in context of products liability litigation);

Rochez Bros., Inc. v. Rhoades, 527 F.2d 880, 886 (3d Cir. 1975) (“It has been held that liability

for aiding and abetting may be found on less than actual knowledge of the illegal activity. How

much or how little knowledge would seem to vary with the facts of each case.” (citation

omitted)). Common law decisionmaking proceeds through the incremental, analogical

application of broadly-stated principles, and it is therefore not amenable to the formulation of

finely detailed rules in the manner of a regulatory code. Contrary to the dissent’s suggestions,

however, the contextual nature and factual sensitivity of common law judicial rulemaking takes

account of the “practical problems” that can result from ill-designed legal rules, and the

flexibility of the common law process allows those problems to be addressed and avoided as they

arise.

In the case at bar, plaintiffs have alleged, albeit in insufficiently specific terms, that the

defendant corporations (a) knowingly and substantially assisted a principal tortfeasor to commit

acts that violate clearly established international law norms, and (b) facilitated the commission of

international law violations by providing the principal tortfeasors with the tools,

63
instrumentalities, or services to commit those violations with actual or constructive knowledge

that those tools, instrumentalities, or services would be (or only could be) used in connection

with that purpose. Such allegations, if proven, clearly satisfy the standard for asserting ATCA

liability under an aiding and abetting theory.

As to the arguments raised by the Ntsebeza and Digwamaje plaintiffs regarding direct

liability claims and the district court’s treatment of them, I concur with and join in Part II.E of

Judge Katzmann’s concurring opinion.

III

This case has confronted our panel with a number of difficult and unsettled questions in a

controversial area of the law. As have other courts, see, e.g., Tel-Oren, 726 F.2d at 775, we have

struggled with these issues in an effort to find some ground on which a majority of our panel

could fully agree. Unfortunately, despite tireless debate, we remain diverse in our perspectives.

Though our already lengthy opinions scarcely demand lengthening, I wish to articulate in a bit

more detail, one objection to the dissent covered in part in footnote 14 of the per curiam opinion.

The dissent contends that the district court lacked subject matter jurisdiction as a result of

certain justiciability doctrines, such as case specific deference, the political question doctrine,

and international comity. Respectfully, this contention is in error. In Baker v. Carr, 369 U.S.

186 (1962), the Supreme Court distinguished between the two. The existence, vel non, of subject

matter jurisdiction is simply a question (with few exceptions not relevant here) of whether a

claim arises under federal law. Id. at 198. If subject matter jurisdiction exists, a court may

64
then—and only then—inquire as to the applicability of any of the various justiciability doctrines:

The District Court was uncertain whether our cases withholding
federal judicial relief rested upon a lack of federal jurisdiction or
upon . . . “[]justiciability.” The distinction between the two grounds
is significant. In the instance of nonjusticiability, consideration of the
cause is not wholly and immediately foreclosed; rather, the Court’s
inquiry necessarily proceeds to the point of deciding whether the duty
asserted can be judicially identified and its breach judicially
determined, and whether protection for the right asserted can be
judicially molded. In the instance of lack of jurisdiction the cause
either does not “arise under” the Federal Constitution, laws or treaties
(or fall within one of the other enumerated categories of Art. III, § 2),
or is not a “case or controversy” within the meaning of that section;
or the cause is not one described by any jurisdictional statute.

Id. (citations omitted). As this quote makes clear, the dissent errs in conflating the anterior

question of the existence of subject matter jurisdiction with the posterior question of

justiciability. Cf. Sinochem Int’l Co. v. Malay. Int’l Shipping Corp., 127 S. Ct. 1184, 1191

(2007) (“[A] federal court generally may not rule on the merits of a case without first

determining that it has jurisdiction over the category of claim in suit . . . .”). The dissent’s error

is not a minor one. By conflating these questions, the dissent provides itself the opportunity to

discuss case specific deference, the political question doctrine, and international comity. This

discussion, in turn, contains a number of missteps, at least two of which merit our attention in

this context. The dissent suggests a district court—or even an appellate court, sitting in

review—must dismiss a case when the executive branch, through a State Department Statement

of Interest or other document, deems a case to be a political “irritant.” This is not so. Mere

executive fiat cannot control the disposition of a case before a federal court. Our principle of

separation of powers not only counsels the judiciary to conduct an independent inquiry—it

65
requires us to do so. Regardless of what else Sosa holds, it did not doubt that ATCA suits are

law suits constitutionally entrusted to the judiciary. Cf. Whiteman v. Dorotheum GmbH & Co.

KG, 431 F.3d 57, 69 (2d Cir. 2005) (“[N]ot every case touching foreign relations is

nonjusticiable and judges should not reflexively invoke these doctrines to avoid difficult and

somewhat sensitive decisions in the context of human rights.” (quoting Kadic, 70 F.3d at 249

(internal quotation marks omitted))). Thus a district court must weigh the Statement of Interest,

as well as other relevant facts, in applying the Baker v. Carr factors and exercising its own

discretion before deciding whether to dismiss a complaint.

IV

For the foregoing reasons, I join in the per curiam opinion.

66
Korman, District Judge, concurring in part and dissenting in part:

Today, the majority allows three class actions on behalf of all persons who lived in South

Africa between 1948 and the present and who suffered damages as a result of apartheid to go forward

in a United States court against American, Canadian, and European corporations that sold goods and

materials or made loans to the Union of South Africa during the apartheid era. It does so over the

vigorous objections of the United States, its allies, and, most notably, the Republic of South Africa,

which is justifiably proud of the ability of its legal system to adjudicate legitimate human rights

claims. In doing so, the majority also ignores a direct signal from the Supreme Court of the United

States regarding the non-viability of these claims. The majority also declines to dismiss the case,

even though the legal foundation of the complaints was expressly rejected by a judgment rendered

in Nuremberg that “[l]oans or sale of commodities to be used in an unlawful enterprise may well be

condemned from a moral standpoint and reflect no credit on the part of the lender or seller in either

case, but the transaction can hardly be said to be a crime.” United States v. von Weizsaecker (“The

Ministries Case”), 14 Trials of War Criminals Before the Nuernberg Military Tribunals Under

Control Council Law No. 10 308, 622 (William S. Hein & Co., Inc. 1997) (1949). Indeed, the

defendant in that case was an officer of the Dresdner Bank, which is a named defendant in this case.

The theory of liability was that the Dresdner Bank provided loans to businesses knowing that the

funds would be used to finance enterprises that employed slave labor. In acquitting the corporate

officer of the charges relating to slave labor, the tribunal at Nuremberg observed that:

The real question is, is it a crime to make a loan, knowing or having
good reason to believe that the borrower will us[e] the funds in

67
 financing enterprises which are employed in using labor in violation
of either national or international law? . . . A bank sells money or
credit in the same manner as the merchandiser of any other
commodity. It does not become a partner in enterprise and the
interest charged is merely the gross profit which the bank realizes
from the transaction, out of which it must deduct its business costs,
and from which it hopes to realize a net profit. Loans or sale of
commodities to be used in an unlawful enterprise may well be
condemned from a moral standpoint and reflect no credit on the part
of the lender or seller in either case, but the transaction can hardly be
said to be a crime. Our duty is to try and punish those guilty of
violating international law, and we are not prepared to state that such
loans constitute a violation of that law . . .

Id. Indeed, the theory of liability rejected by the tribunal is the same as that alleged here against

Dresdner Bank. A00211-12. Specifically, along with other banks, it is accused of providing

“[f]oreign capital in the form of trade loans, large international bonds and credits, direct loans . . .

to South African borrowers, and project financing [that] supported the apartheid regime.” A00194.

Instead of deferring to this reasoned judgment, one member of the majority, Judge Hall,

eschews any reference to sources of customary international law and applies a principle of aiding-

and-abetting in civil cases that the Supreme Court observed “has been at best uncertain in application

. . . with the common-law precedents ‘largely confined to isolated acts of adolescents in rural

society.’” Cent. Bank of Denver v. First Interstate Bank of Denver, 511 U.S. 164, 181 (1994)

(quoting Halberstam v. Welch, 705 F.2d 472, 489 (D.C. Cir. 1983)). The second member of the

majority, Judge Katzmann, concludes that a defendant’s liability for aiding-and-abetting a crime

against humanity must be determined by reference to customary international law, and he then goes

on to articulate the standard for determining such liability. While the standard he enunciates is

entirely consistent with the holding in the Ministries Case, and also reflects an emerging consensus

68
on the appropriate standard for holding a private party liable for aiding-and-abetting, he joins Judge

Hall in voting to vacate the judgment dismissing the case without deciding whether the allegations

in the complaints are sufficient to satisfy that standard, and without addressing the objections of the

United States and the Republic of South Africa to the exercise of jurisdiction over the complaints.

Background

Because the allegations in the complaint go unmentioned in the per curiam and concurring

opinions, it is necessary to provide an overview of the case beyond describing the procedure by

which the case made its way here. Judge Sprizzo’s opinion in the district court contains a detailed

and accurate description of the hundreds of pages of allegations contained in the three voluminous

complaints. See In re South African Apartheid Litig., 346 F. Supp. 2d 538, 543-46 (S.D.N.Y. 2004).

This enables me to provide the following brief background.

In 1948, the National Party came to power in South Africa. From then until the early 1990s,

that ruling party imposed and enforced a set of laws under which the non-white population was the

subject of disenfranchisement, state-sponsored discrimination, and repression. These laws,

establishing what became known as apartheid, included severe curtailments on the liberty of non-

whites with respect to residency, travel, assembly, education, employment, and marriage.

Apartheid’s restrictions were enforced by members of South Africa’s military and police, and the

history of apartheid includes many instances of arbitrary detention, torture, and killings by those state

actors. The complaints recount many of those instances of wrongful conduct by apartheid

government officials, including the Sharpeville Massacre of 1960, the Soweto Massacre of 1976, and

the killing of the Craddock Four in 1985.

69
 The allegations are very different with respect to the defendants. The portions of the

complaints relating to defendants’ alleged conduct focus principally on their trade with South Africa.

Thus, car companies are accused of selling cars, computer companies are accused of selling

computers, banks are accused of lending money, oil companies are accused of selling oil, and

pharmaceutical companies are accused of selling drugs. The theory of the complaints is that in this

way defendants facilitated or “aided-and-abetted” apartheid and its associated human rights

violations. To support that theory, the complaints allege generally that defendants knew of the racist

policies of apartheid; that they nevertheless so engaged in the transactions in and with the Union of

South Africa; and that, had they not done so, the apartheid regime would have collapsed, apartheid

would have ended sooner, and plaintiffs would not have suffered some or all of their injuries. The

causal theory advanced by the Khulumani plaintiffs is even weaker: “Apartheid would not have

occurred in the same way without the participation of defendants.” A00166 (emphasis added).

Typical of the allegations are:

• “Apartheid needed the cooperation, financing and supplies
from the defendant financial institutions and companies
and/or their predecessors or successors to ensure that they had
the technology, equipment, systems, infrastructures and
weaponry to insure that their system could function.” A00304.

• “Without oil, the police and military could not have
functioned and the economy of South Africa would have
come to a standstill.” A00171.

• “More than any other single technological advancement, the
computer fostered the concentration of administrative power
in the hands of Africa’s white elite.” A00427.

70
 • “Many computer uses were crucial for the apartheid regimes’
[sic] keeping track of political activists who were later
targeted for assassinations.” A00430.

• “[A]ny transfer of capital to South Africa had military
implications: loans to the railways and harbors systems
assisted in the mobilization of the armed forces; trade
financing provided the computers and telecommunications
equipment necessary to the efficient functioning of a modern
army; financing for housing project perpetuated the
segregated housing of apartheid.” A00201.

• “The money from defendant German banks directly benefitted
and supported the apartheid reign of terror in South Africa.”
A00440.

• “Limitations on the employment of non-whites in salaried,
administrative jobs puts a premium on automating such tasks.
In this sense, U.S. computer firms helped to solve the skilled
white labor problem.” A00428; see also A00306-07.

• “Defendant vehicle manufacturers knowingly supplied
vehicles, parts, and other equipment to the South African
Police (SAP), South African Defense Force (SADF) and
South African Army. . . . These vehicles were used to patrol
African townships, homelands, and other areas and were used
to suppress dissent.” A00216; see also A00312.

The complaints themselves seek relief on behalf of all persons who lived in South Africa

between 1948 and the present who suffered damages as a result of apartheid. Moreover, they seek

to hold responsible each of the defendants, whom they characterize as aiders-and-abetters, liable for

“all acts comprising the entire system of apartheid – a criminal enterprise.” A00255. Nevertheless,

they fail to link the conduct of a specific defendant to an injury suffered by a particular plaintiff. On

the contrary, the gravamen of the complaints is not that individuals should recover from particular

71
defendants in tort, but that millions of people who “lived under the apartheid system”, A00095,

should recover from all the defendants for having been “subject to . . . the overwhelming injustices

that characterized apartheid.” Id. In sum, these are reparations cases, seeking at least $400 billion

in reparations, rather than torts cases for damages. They fail to allege a cognizable cause of action.

Moreover, the plaintiffs conceded at oral argument that they are not anxious to amend their

complaints to link a particular plaintiff to injuries caused by a particular defendant. Tr. at 64.

Indeed, the attorney for the Khulumani plaintiffs acknowledged that he would not necessarily be

satisfied with such an amendment, but that he would “accept it because it would be a beginning.”

Tr. at 80. Nevertheless, he confessed that “[w]hat’s at stake here is almost by definition of the crime

an inability to make that exact trace or that degree of particularity.” Tr. 80-81. This concession

alone establishes that the allegations in any amended complaints would be insufficient “to raise a

reasonable expectation that discovery will reveal evidence” to support them. Bell Atlantic Corp. v.

Twombly, 127 S. Ct. 1955 (2007). Instead, these complaints would simply provide a vehicle to

coerce a settlement. Nevertheless, the per curiam opinion reverses the order dismissing the

complaints and orders Judge Sprizzo to rule again on plaintiffs’ motions for leave to replead, because

plaintiffs have indicated “that, if given the opportunity, they would narrow the claims and clarify the

allegations against the various defendants.” Per Curiam Op. ante at 13-14.

I dissent from the holding keeping these claims alive for the reasons alluded to earlier and

because (1) the Supreme Court, in Sosa v. Alvarez-Machain, 542 U.S. 692 (2004), has instructed us

that this is the very sort of case in which jurisdiction should not be exercised; (2) the State

Department has filed a persuasive Statement of Interest in this matter urging dismissal because of

the adverse effect the continued prosecution of these cases would have on the interests of the United

72
States and our relations with other countries; and (3) the Republic of South Africa, a democratically

elected government representative of all South Africans, including the victims of apartheid, has

asserted the right to define and finalize issues related to reparations for apartheid-era offenses within

its own legal framework – thus making this lawsuit an insult to the post-apartheid, black-majority

government of a free people. These grounds, among others, reinforce the compelling policy

argument articulated that a decision to hear these cases would “reflect[] the worst sort of ‘judicial

imperialism’ . . . [and] send the message that the United States does not respect the ability of South

African society to administer justice by implying that U.S. courts are better placed to judge the pace

and degree of South Africa’s national reconciliation.” Elliot J. Schrage, Judging Corporate

Accountability in the Global Economy, 42 Colum. J. Transnat’l L. 153, 166 (2003).

Discussion

1. Deference to the United States and the Republic of South Africa

This appeal is unique. Departing from its usual practice, the Supreme Court, in the context

of deciding a different case, has already given us guidance as to how this appeal should be decided.

In Sosa, it held that, in determining whether the Alien Tort Claims Act (the “ATCA”) provides a

basis for the exercise of jurisdiction over an alleged tort in violation of the law of nations, courts

must apply “principle[s] limiting the availability of relief” beyond the requirement that the

international law norm whose violation is alleged be sufficiently defined. 542 U.S. at 733 n.21.

Specifically, because the decision to permit a case to proceed involves “an element of judgment

about the practical consequences of making that cause available to litigants,” id. at 732-33, the

73
Supreme Court suggested that “case-specific deference to the political branches,” id. at 733 n.21,

provided one limitation on the exercise of jurisdiction under the ATCA in certain instances.

Among the reasons for such deference is that the recognition of private rights of action for

violations of international law may give rise to “collateral consequences.” Id. at 727. Specifically,

the Court cautioned that “the potential implications for the foreign relations of the United States of

recognizing such causes should make courts particularly wary of impinging on the discretion of the

Legislative and Executive Branches in managing foreign affairs.” Id. The Court continued: “Since

many attempts by federal courts to craft remedies for the violation of new norms of international law

would raise risks of adverse foreign policy consequences, they should be undertaken, if at all, with

great caution.” Id. at 727-28. Then, focusing on the case before us on this appeal, the Supreme

Court continued:

For example, there are now pending in Federal District Court several
class actions seeking damages from various corporations alleged to
have participated in, or abetted, the regime of apartheid that formerly
controlled South Africa. See In re South African Apartheid
Litigation, 238 F. Supp. 2d 1379 (JPML 2002) (granting a motion to
transfer the cases to the Southern District of New York). The
Government of South Africa has said that these cases interfere with
the policy embodied by its Truth and Reconciliation Commission,
which “deliberately avoided a ‘victors’ justice’ approach to the crimes
of apartheid and chose instead one based on confession and
absolution, informed by the principles of reconciliation,
reconstruction, reparation and goodwill.” Declaration of Penuell
Mpapa Maduna, Minister of Justice and Constitutional Development,
Republic of South Africa, reprinted in App. to Brief for Government
of Commonwealth of Australia et al. as Amici Curiae 7a, ¶ 3.2.1
(emphasis deleted). The United States has agreed. See Letter of
William H. Taft IV, Legal Adviser, Dept. of State, to Shannen W.
Coffin, Deputy Asst. Atty. Gen., Oct. 27, 2003, reprinted in id., at 2a.
In such cases, there is a strong argument that federal courts should
give serious weight to the Executive Branch’s view of the case’s
impact on foreign policy. Cf. Republic of Austria v. Altmann, 541

74
 U.S. 677, 124 S. Ct. 2240, 159 L. Ed. 2d 1 (2004) (discussing the
State Department’s use of statements of interest in cases involving the
Foreign Sovereign Immunities Act of 1976, 28 U.S.C. § 1602 et seq.).

Id. at 733 n.21.

The Executive Branch’s view of the impact of these cases on foreign policy is set forth at

some length in a letter to the district court, dated October 27, 2003, from the Legal Advisor of the

Department of State. The letter advises “that continued adjudication of the above-referenced matters

risks potentially serious adverse consequences for significant interests of the United States.”

A01090. After outlining the objections of the Republic of South Africa to the continued prosecution

of these cases in the United States and noting the “extensive steps” it has taken “to promote

reconciliation and redress for apartheid-era injustices,” the Legal Advisor observed that the Republic

of South Africa “is broadly representative of the victims of the apartheid regime” and “is uniquely

charged with a popular mandate to deal with the legacy of apartheid.” Id. He then concluded his

discussion of this aspect of the Executive Branch’s concern as follows:

Support for the South African government’s efforts in this area is a
cornerstone of U.S. policy towards that country. For that reason, we
are sensitive to the view of the South African government that
adjudication of the cases will interfere with its policy goals, especially
in the areas of reparations and foreign investment, and we can
reasonably anticipate that adjudication of these cases will be an
irritant in U.S.-South African relations. To the extent that
adjudication impedes South Africa’s on-going efforts at reconciliation
and equitable economic growth, this litigation will also be detrimental
to U.S. foreign policy interests in promoting sustained economic
growth in South Africa.

Id.

75
 The State Department also voiced other concerns. For one thing, the Legal Advisor wrote,

“[v]arious other foreign governments, including those of the United Kingdom and Canada, have also

approached us via diplomatic channels to express their profound concern that their banks,

corporations and other entities have been named as defendants.” A01090-91. Because of those

governments’ “strong belief that the issues raised in the litigation are most appropriately handled

through South Africa’s domestic processes,” the State Department “anticipat[ed] possible,

continuing tensions in our relations with these countries over the litigation.” A01091.

Moreover, the Legal Advisor expressed concern over the chilling effect that actions of this

kind may have on future foreign investment in developing countries:

The United States relies, in significant part, on economic ties and
investment to encourage and promote positive change in the domestic
policies of developing countries on issues relevant to U.S. interests,
such as respect for human rights and reduction of poverty. However,
the prospect of costly litigation and potential liability in U.S. courts
for operating in a country whose government implements oppressive
policies will discourage the U.S. (and other foreign) corporations
from investing in many areas of the developing world, where
investment is most needed and can have the most forceful and
positive impact on both economic and political conditions. To the
extent that the apartheid litigation in U.S. courts deters such
investment, it will compromise a valuable foreign policy tool and
adversely affect U.S. economic interests as well as economic
development in poor countries.

Id.

In Whiteman v. Dorotheum GmbH & Co. KG, 431 F.3d 57 (2d Cir. 2005) (Cabranes, J.), we

recognized that “[j]udicial deference to the Executive Branch on questions of foreign policy has long

been established under the prudential justiciability doctrine known as the ‘political question’

doctrine . . . .” Id. at 69. In deciding what deference to accord the position of the Executive Branch,

76
we referenced the foundational case of Baker v. Carr, which identified six independent tests for

recognition of a political question. 369 U.S. 186, 217 (1962). In Whiteman, we directed the

dismissal of the complaint without examining each of the Baker tests because it was clear that the

case met the fourth test, namely that “‘a court’s undertaking independent resolution’ of this claim

is impossible ‘without expressing lack of the respect due’ the Executive Branch.” 431 F.3d at 72

(emphasis deleted) (citations omitted).

We need not examine any of the six independent tests identified in Baker because Sosa

plainly held that the determination of whether to exercise jurisdiction over a cause of action for a

violation of the law of nations is subject to “case-specific deference to the political branches.” 542

U.S. at 733 n.21. More than that, the Supreme Court advised us that in the instant cases “there is

a strong argument that federal courts should give serious weight to the Executive Branch’s view of

the case’s impact on foreign policy.” Id. Indeed, one commentator, who has emphasized the need

for a “searching review” of the Bush administration’s positions in human rights cases, has

acknowledged that the present cases are different:

[These] cases reflect the unique history of South Africa and its
transition from apartheid to democracy. The government that
replaced the apartheid regime was recognized internationally as
representative both of the majority of the nation and, in particular, of
the victims of past human rights abuses. The transition included a
negotiated process by which abuses would be investigated and
perpetrators given the opportunity to testify about their actions in
return for amnesty. When the democratically elected, representative
government of South Africa objected to the impact of the U.S.
litigation on the negotiated transition process, the executive branch
asked the courts to defer to this judgment. All of these factors
provide support for the Court’s suggestion that there is a “strong
argument” that executive branch views “in such cases” are entitled to
“serious weight.”

77
Beth Stephens, Sosa v. Alvarez-Machain, “The Door Is Still Ajar” for Human Rights Litigation in

U.S. Courts, 70 Brook. L. Rev. 533, 562 (2005). (Professor Stephens and a fellow academic co-

authored an amicus curiae brief on behalf of international human rights organizations and bar

associations in support of the plaintiffs-appellants. While it urges reversal of the judgment of the

district court, her amicus brief does not address the effect of footnote 21 in Sosa.)

In any event, the force of the advice of the State Department is enhanced by the fact that,

under the doctrine of international comity, dismissal would be justified solely out of deference to the

Republic of South Africa. Comity, a doctrine more easily invoked than defined, may be viewed as

either “the degree of deference that a domestic forum must pay to the act of a foreign government

not otherwise binding on the forum,” Laker Airways Ltd. v. Sabena, Belgian World Airlines, 731

F.2d 909, 937 (D.C. Cir. 1984), or as “a discretionary act of deference by a national court to decline

to exercise jurisdiction in a case properly adjudicated in a foreign state, the so-called comity among

courts.” In re Maxwell Commc’n Corp., 93 F.3d 1036, 1047 (2d Cir. 1996). A court’s exercise of

discretion to dismiss a suit on comity grounds is independent of its treatment of the views of the

Executive Branch. See Ungaro-Benages v. Dresdner Bank AG, 379 F.3d 1227 (11th Cir. 2004)

(dismissing a claim on the grounds of international comity after declining to defer to the Executive

Branch).

The decision whether to dismiss on this basis depends on the degree to which the interests

of a foreign sovereign are “legitimately affronted by the conduct of litigation in a United States

forum,” Jota v. Texaco, Inc., 157 F.3d 153, 160 (2d Cir. 1998), steps the foreign sovereign may have

taken to address the issues in the litigation, and the extent of our own interest in the underlying

78
issues. See generally Bi v. Union Carbide Chems. & Plastics Co., 984 F.2d 582 (2d Cir. 1993)

(Newman, J.). Perhaps the most significant factor is whether the foreign sovereign to which we

defer is a democratically elected government with an independent judiciary. Id. at 585-86.

In the present cases, the extent of the affront to the legitimate interests of the Republic of

South Africa, the steps it has taken to address the issue of reparations, and the depth of feeling

underlying its concern are amply reflected in the record. On April 15, 2003, in response to the

Report of the Truth and Reconciliation Commission (“TRC”), the President of the Republic, Thabo

Mbeki, announced the programs that would be implemented to assist the victims of apartheid:

[W]ith regard to specific cases of individual victims identified by
[the] TRC Act, the government has put in place and will intensify
programmes pertaining to medical benefits, educational assistance
and provision of housing and so on. From time to time, Ministers
have elaborated and will continue to expatiate on the implementation
of these and other related programmes.

The TRC has reported that about 22,000 individuals or surviving
families appeared before the Commission. Of these, about 19,000
required urgent reparations, and virtually all of them, where the
necessary information was available, were attended to as proposed by
the TRC with regard to interim reparations.

With regard to final reparations, government will provide a once-off
grant of R30,000 to those individuals or survivors designated by the
TRC. This is over and above other material commitments that we
have already mentioned.

We intend to process these payments as a matter of urgency, during
the current financial year. Combined with community reparations,
and assistance through opportunities and services we have referred to
earlier, we hope that these disbursements will help acknowledge the
suffering that these individuals experienced, and offer some relief.

79
A00745. President Mbeki then addressed the issue of “civil suits against corporations that benefitted

from the apartheid system,” specifically this litigation. A00747. After reiterating “that the South

African Government is not and will not be a party to such litigation,” he continued: “[W]e consider

it completely unacceptable that matters that are central to the future of our country should be

adjudicated in foreign courts which bear no responsibility for the well-being of our country and the

observance of the perspective contained in our Constitution of the promotion of national

reconciliation.” Id. While recognizing “the right of citizens to institute legal action,” President

Mbeki also emphasized that the government’s approach is informed by “the desire to involve all

South Africans, including corporate citizens, in a co-operative and voluntary partnership to

reconstruct and develop South African society.” Id. Accordingly, he also rejected the once-off

wealth tax on corporations suggested by the TRC. Id.

Likewise, in an address to Parliament on the same day, Alec Erwin, the Minister of Trade and

Industry, emphasized that the government’s rationale for its rejection of the once-off wealth tax

applied equally to its opposition to the cases at issue:

It is for all the above reasons that we are opposed to and indeed
contemptuous of attempts to use unsound extra-territorial legal
precepts in the [United States of America] to seek personal financial
gain in South Africa. Whilst individuals have an inviolate right to
recourse in law and this is firmly entrenched in South Africa it is an
abuse to use the law, unsound law at that, of another land to
undermine our sovereign right to settle our past and build our future
as we see fit. South Africans involved in this break that indefinable
collectivist identity that was the origin of our strength. The
government rejects the actions of legal practitioners in the USA to
exploit our history and will not allow any judgment made in the USA
or elsewhere to be carried out in South Africa.

A00754-55.

80
 Professor Kader Asmal, the Minister of Education, also assailed the prosecution of these

cases. He told the Parliament:

South Africa must settle this issue for themselves and does not need
the help of ambulance chasers and contingency fee operators, whether
in Switzerland, the Netherlands, or the United States of America. As
South Africans, we have effectively dealt with our own historical
challenges and we will continue to do so. It is part of our sovereign
right.

A00758.

On April 16, 2003, the day after the addresses quoted above, the Cabinet of the Republic of

South Africa resolved that “[i]t remains the right of the [G]overnment [of South Africa] to define

and finalise issues of reparations, both nationally and internationally.” A00803. Against this

backdrop, the then-Minister of Justice, Penuell Mpapa Maduna, filed a declaration in the district

court which set forth the Republic of South Africa’s view of various cases pending in the United

States against corporations that did business with and in South Africa during the apartheid period,

including the cases at issue here. The statement summarizes the actions undertaken by the Republic

of South Africa “to repair the damage caused by the apartheid system through a broad programme

of socioeconomic reparations which has at its heart, the betterment of the lives of the previously

disadvantaged.” A00801. Then, addressing the prosecution of these cases, he argued that the

remedies sought – including “the demand for billions of dollars in damages to be distributed by the

US courts” – are “inconsistent with South Africa’s approach to achieving its long term goals.”

A00805. Specifically, he continued:

Permitting this litigation to go forward will, in the government’s
view, discourage much-needed direct foreign investment in South
Africa and thus delay the achievement of our central goals. Indeed,

81
 the litigation could have a destabilising effect on the South African
economy as investment is not only a driver of growth, but also of
employment. One of the structural features of the South African
economy, and one of the terrible legacies of apartheid, is its high level
of unemployment and its by-product, crime. Foreign direct
investment is essential to address both these issues. If this litigation
proceeds, far from promoting economic growth and employment and
thus advantaging the previously disadvantaged, the litigation, by
deterring foreign direct investment and undermining economic
stability will do exactly the opposite of what it ostensibly sets out to
do.

A00805-06. The last paragraph of Minister Maduna’s declaration invoked the doctrine of

international comity. He observed

that under United States law, courts may abstain from adjudicating cases in deference
to the sovereign rights of foreign countries to legislate, adjudicate and otherwise
resolve domestic issues without outside interference, particularly where the relevant
government has expressed opposition to the actions proceeding in the United States,
and where adjudication in the United States would interfere with the foreign
sovereign’s efforts to address matters in which it has the predominant interest. The
government submits that its interest in addressing its apartheid past presents just such
a situation.

A00806.

Most recently, on this appeal, the Republic of South Africa filed an amicus curiae brief,

which is not addressed in the per curiam and concurring opinions and is not specifically listed among

the parties filing briefs on this appeal, arguing that these actions “fundamentally interfere with South

Africa’s independence and sovereignty and intervene in its internal affairs, including its right under

international law to address its apartheid past and to develop policies for its future in the manner it

deems most appropriate, subject to the support and approval of the democratic electorate.” Br. for

S. Afr. as Amicus Curiae at 1-2. Specifically, it emphasized that the continuation of this litigation

82
would discourage business investment and, thereby, disrupt the growth of the South African

economy. Id. at 3-4. In a statement appended to the brief, the current South African Minister of

Justice, Brigitte Sylvia Mabandla, repeated verbatim the earlier statement of former Minister

Maduna, adding that “another country’s courts should not determine how ongoing political processes

in South Africa should be resolved.” Id. annex at 1.

In light of the consistent and considered policy of the Republic of South Africa and the

programs it has put in place to provide reparations and assistance to the identifiable victims of

apartheid, it is difficult to conceive of any policy reason for exercising jurisdiction over these cases.

On the contrary, “[m]ost judges would not want, by asserting their own jurisdiction in a doubtful

case, to undermine efforts to rebuild national sovereignty, democratic institutions, and self-respecting

courts, in countries that have lately suffered from catastrophic events that include international

crimes.” Michael Kirby, Universal Jurisdiction and Judicial Reluctance: A New “Fourteen Points”,

in Universal Jurisdiction: National Courts and the Prosecution of Serious Crimes Under

International Law 240, 255 (Stephen Macedo, ed. 2004). Indeed, Professor William R. Casto,

whose scholarship was cited repeatedly by the Supreme Court in Sosa, has written that the apartheid

regime in South Africa may be an example of a situation “in which a nation’s prior government has

engaged in the most despicable misconduct, but the current government is making an honest, perhaps

faltering, but nevertheless real effort to rectify the past misdeeds . . . . If apartheid were determined

to be a violation of clearly established international law, a tort remedy for the victims of that evil

system might nevertheless be inappropriate.” William R. Casto, The New Federal Common Law of

Tort Remedies for Violations of International Law, 37 Rutgers L.J. 635, 656 (2006).

83
 Particularly apposite here is the characteristically thoughtful opinion of Judge Newman in

Bi v. Union Carbide Chemicals & Plastics Co, 984 F.2d 582. The case arose out of the devastating

Bhopal industrial accident, in which deadly gas from a plant operated by Union Carbide India Ltd.

(“UCIL”) blew into a densely populated part of India. Some 145 class actions were filed in federal

district courts across the United States and consolidated in the Southern District of New York. After

the case was dismissed on the grounds of forum non conveniens, the Indian government, to which

its parliament granted the exclusive authority to represent the victims of the disaster in India and

elsewhere, filed suit in India on behalf of all claimants. India and Union Carbide ultimately agreed

to a court-approved settlement, by which the latter agreed to pay $470 million to the Indian

government for the benefit of all victims of the disaster.

After the settlement of the class actions in India, two cases were filed in Texas state courts

on behalf of other victims of the Bhopal disaster. The cases were transferred to the Southern District

of New York, where Judge Keenan dismissed both actions on the ground of forum non conveniens.

We affirmed the dismissal, although on different grounds. Judge Newman described the appeal as

presenting “an interesting issue of comity among nations in the resolution of claims arising from

torts occurring within a foreign country.” Bi, 984 F.2d at 583. The precise issue he identified

is whether the federal and state courts of this country should defer to the judgment
of a democratic foreign government that disputes arising from a mass tort occurring
within its borders can be best resolved by according the foreign government
exclusive standing to represent the victims of the disaster in the courts of the world.

Id. We declined to permit the plaintiffs to prosecute the class actions in federal or state court,

notwithstanding their arguments that settlement of their claims by the Indian government was unfair

and improper because, among other things,

84
 the Indian Government had an unacceptable conflict of interest as part owner of
UCIL, most of the victims oppose the settlement as grossly inadequate, and their due
process rights were violated because they received inadequate notice and inadequate
representation in the proceedings and because they could not opt out of the
settlement.

Id. at 584.

After describing in detail the structure of the Indian government, pursuant to a Constitution

that “provides for a republican form of parliamentary government and guarantees the fundamental

rights of the people, including equal protection and procedural due process,” id. at 585, Judge

Newman observed that India had chosen “to represent exclusively all the victims in a suit against

Union Carbide and to use the money it received in settlement of that suit to fund a plan framed . . .

to process the claims of all the victims.” Id. at 586. Under these circumstances, he wrote, “[t]o grant

the victims of the Bhopal disaster, most of whom are citizens of India, access to our courts where

India has set up what it believes to be the most effective method of dealing with a difficult problem

would frustrate India’s efforts.” Id. “[W]ere we to pass judgment on the validity of India’s response

to a disaster that occurred within its borders,” Judge Newman continued, “it would disrupt our

relations with that country and frustrate the efforts of the international community to develop

methods to deal with problems of this magnitude in the future.” Id.

Moreover, it was not relevant whether “under our constitutional standards, our Government

could pass an act similar to the Bhopal Act” pursuant to which the Indian government settled the

claims that the plaintiffs sought to prosecute. Id. Instead, he wrote that:

when a recognized democracy determines that the interests of the
victims of a mass tort that occurred within its borders will be best
served if the foreign government exclusively represents the victims
in courts around the world, we will not pass judgment on that

85
 determination, and we will permit only the foreign government access
to our courts to litigate those claims, subject of course to our own
requirements for standing. This conclusion is especially compelling
in a case such as this where almost all of the victims are Indian
citizens.

Id.

The present cases involve a stronger basis for dismissal under the doctrine of international

comity. Like India, the Republic of South Africa is a sovereign democratic state whose policies

broadly reflect the interests of its people. Moreover, it has adopted policies and programs designed

to address the effects of apartheid and compensate its victims. However, unlike the plaintiffs in Bi,

whose claims were wiped out under Indian law and who had no alternative forum, the plaintiffs here

may have a competent alternative forum in their home country that, by all accounts, is prepared to

hear their grievances. Indeed, the plaintiffs have themselves acknowledged that relief would be

available in the courts of the Republic of South Africa. See Khulumani Reply Br. at 27 (“No relief

from civil or criminal liability was enacted for those who did not apply for and obtain amnesty from

the TRC, . . . South Africa’s approach leaves open the possibility for individual citizens to take up

any grievance related to human rights violations with the courts . . . .”) (internal quotation marks

omitted); Ntsebesa Reply Br. at 28 (“The acts alleged in plaintiffs’ complaints are illegal and would

subject the defendants to suit in both the United States and South Africa . . . .”).

Significantly, President Mbeki has stated, “[the] Government recognizes the right of citizens

to institute legal action.” A00747. Likewise, in criticizing the prosecution of these cases in the

United States, Minister Erwin noted that the plaintiffs “have an inviolate right to recourse in law”

– just not in the United States. A00754. Thus, unlike the victims of the Bhopal disaster, there is

86
every indication they may have their day in court, in their home country, before an independent

judiciary, in a manner consistent with traditional notions of due process. As Minister Maduna stated

in his declaration to the district court:

Under the Constitution, the judicial authority of the Republic is
vested in the courts, which are independent and subject only to the
Constitution and the law, which they must apply impartially and
without fear, favour or prejudice. No person or organ of state may
interfere with the functioning of the courts, while all other organs of
the state, through legislative and other measures, must assist and
protect the courts to ensure their independence, impartiality, dignity,
accessibility and effectiveness. An order or decision of a court binds
all persons to whom and organs of state to which it applies. South
Africa has a well developed judicial system, with the Constitutional
Court at its apex and the Supreme Court of Appeal as the final court
of appeal in non-constitutional matters. Judgments of the
Constitutional Court and, indeed, the Supreme Court of Appeal, are
widely admired for their independence and incisiveness and are
frequently referred to in judgments of other final courts of appeal
internationally.

A00798-A00799. Similarly, as one scholar of South African law has observed,

[t]he new constitutional law not only introduced the ethos of a democratic
Rechtsstaat, but also enacted an extensive array of socioeconomic rights that are
designed to change South African society fundamentally. These can be and indeed
already have been used to secure, through litigation, the interests of the traditionally
marginalised groups and to ensure their place on the political agenda.

François du Bois, Introduction: History, System and Sources, in Introduction to the Law of South

Africa 1, 3-4 (C.G. van der Merwe & Jacques E. du Plessis eds., 2004).

While I have assumed the availability of a forum in the Republic of South Africa, Bi makes

clear that such an alternative forum is not always required before the doctrine of comity may be

invoked. See also Jota, 157 F.3d at 160 (“[C]ases might be imagined where a foreign sovereign’s

87
interests were so legitimately affronted by the conduct of litigation in a United States forum that

dismissal is warranted without regard to the defendant’s amenability to suit in an adequate foreign

forum . . . .”). Nevertheless, even if the availability of an alternative forum were an absolute

prerequisite to dismissal on the ground of international comity, and one were not available in these

cases, deference to South African law would be justified under a traditional choice-of-law analysis.

Indeed, although he did not in haec verba invoke such analysis, Judge Newman’s opinion in Bi is

entirely consistent with it.

We have previously held that the assumption of subject matter jurisdiction over a cause of

action under the ATCA does not preclude a choice-of-law analysis resulting in dismissal of the case.

Filartiga v. Pena-Irala, 630 F.2d 876, 889 (2d Cir. 1980). This inquiry is separate from the

jurisdictional analysis. Indeed, we have read Filartiga as “requiring the district court to perform a

traditional choice-of-law analysis to determine whether international law, [the] law of forum state,

or [the] law of state where events occurred should provide substantive law in such an action.” Wiwa

v. Royal Dutch Petroleum Co., 226 F.3d 88, 105 n.12 (2d Cir. 2000).

I engage in a hypothetical choice-of-law analysis here, because it focuses on the

governmental interests of South Africa and the United States and serves to reinforce the argument

against exercising jurisdiction over these cases. As we have observed,

[t]he federal common law choice-of-law rule is to apply the law of the jurisdiction
having the greatest interest in the litigation . . . . The goal of this analysis is to
evaluate the various contacts each jurisdiction has with the controversy, and
determine which jurisdiction’s laws and policies are implicated to the greatest extent.

In re Koreag, Controle et Revision S.A., 961 F.2d 341, 350 (2d Cir. 1992). In this case, the Republic

of South Africa has the greatest interest, if not the only interest, in the application of its laws. The

88
conduct alleged took place in South Africa and the victims were its own citizens. The remedies that

its citizens should have as a result of the injuries they suffered at the hands of the apartheid regime

are matters exclusively for its democratically elected post-apartheid government.

Moreover, the tenuous interest of the United States in the issues raised by these cases is also

reflected in the fact that, under customary international law, we could not exercise subject matter

jurisdiction over a cause of action against the primary tortfeasor – the officials of the Union of South

Africa – or the foreign corporate defendants. This is so, because apartheid, however abhorrent it may

have been, has not been regarded as an offense subject to the exercise of universal jurisdiction. This

concept, as its name implies, “recognize[s] that international law permits any state to apply its laws

to punish certain offenses although the state has no links of territory with the offense, or of

nationality with the offender (or even the victim).” Restatement (Third) of Foreign Relations Law

§ 404 cmt. a (1987); see also Matter of Extradition of Demjanjuk, 612 F. Supp. 544, 555-58 (N.D.

Ohio 1985). Universal jurisdiction is dependent not only on “substantive agreement as to certain

universally condemned behavior,” which transforms the behavior into a violation of customary

international law, but also “procedural agreement that universal jurisdiction exists to prosecute a

subset of that behavior.” Sosa, 542 U.S. at 762 (Breyer, J., concurring in part and concurring in the

judgment).

There is no agreement with respect to the latter issue. Although the Restatement (Third) of

Foreign Relations Law cites racial discrimination, “when [ ] practiced systematically as a matter of

state policy, e.g., apartheid,” as a violation of customary international law, id. § 702 cmt. i (emphasis

deleted), it omits apartheid from the list of offenses subject to universal jurisdiction. Instead, the

Restatement states that universal jurisdiction exists only for “certain offenses recognized by the

89
community of nations as of universal concern, such as piracy, slave trade, attacks on or hijacking of

aircraft, genocide, war crimes, and perhaps certain acts of terrorism.” Id. § 404. Indeed, while the

Reporters of the Restatement observed that the International Convention on the Suppression and

Punishment of the Crime of Apartheid, adopted Nov. 30, 1973, 1015 U.N.T.S. 243 ( the “Apartheid

Convention”), provided for universal jurisdiction, they stated that it did so only “[a]mong [the]

parties to the Convention.” Restatement (Third) of Foreign Relations Law § 702 reporters’ note 7.

See also Antonio Cassese, Crimes Against Humanity, in 1 The Rome Statute of the International

Criminal Court: A Commentary 353, 376 (Antonio Cassese et al., eds., 2002) (observing that the

Rome Statute, enacted in 1998, is broader than customary international law and “expands general

international law” insofar as it, inter alia, “broadens the classes of conduct amounting to crimes

against humanity” to include “the crime of apartheid”).1 Likewise, the European Commission, the

executive body of the European Union, has stated explicitly that, while “apartheid[] is widely

condemned by states . . . at least at present, it does not give rise to universal jurisdiction because,

among other reasons, the [Apartheid Convention] . . . has not been widely ratified.” Br. for the

European Commission as Amicus Curiae Supporting Neither Party, Sosa v. Alvarez-Machain, 542

U.S. 692 (No. 03-339), 2004 WL 177036, at *16 n.35. Another reason is that jurisdiction is limited

to certain universally condemned crimes which “by their nature occur either outside of a State or

1
The Rome Statute does not apply to “conduct prior to the entry into force of the Statute.”
Rome Statute of the International Criminal Court art. 24(1), open for signature July 17, 1998, 37 I.L.M.
999, 1016 (entered into force July 1, 2002). As one commentator explained, “[t]o be acceptable to the
broadest possible majority of countries, many of which had dark experiences in the past and had to resort
to amnesties or similar solutions to achieve national reconciliation, criminal responsibility under the
Rome Statute could not be made retroactive . . . .” Per Saland, International Criminal Law Principles, in
The International Criminal Court: The Making of the Rome Statute – Issues, Negotiations, Results 189,
196 (Roy S. Lee ed., 1999). This may explain why the Republic of South Africa, which never ratified the
Apartheid Convention, ratified the Rome Statute.

90
where there is no State capable of punishing, or competent to punish, the crime (as in a time of

war).” United States v. Yousef, 327 F.3d 56, 105 (2d Cir. 2003). Of course, with respect to the

conduct alleged here, there is now a State capable of providing redress in a competent manner.

Under these circumstances, particularly the failure of the United States to sign or ratify the

Apartheid Convention, denying a forum for these cases at this point cannot be said to undermine any

interest of the United States. To the contrary, as the United States argues in its amicus brief, the

prosecution of these cases harms significant interests of the United States. Obviously, the

congressional interest in making available a forum under the ATCA would be undermined if we

denied a forum where the regime, which violated the norm, was the same one that decided that such

a cause of action should not be pursued. See Tachiona v. Mugabe, 234 F. Supp. 2d 401, 415

(S.D.N.Y. 2002) (declining deference to law of the foreign state in action alleging actionable conduct

by a sitting government). But that is not the case here, for as the State Department has observed, the

laws of the Republic of South Africa reflect policy judgments made by a government that “is broadly

representative of the victims of the apartheid regime” and “uniquely charged with a popular mandate

to deal with the legacy of apartheid.” A01090. We have no interest in undermining its law or policy.

In sum, to quote again from one learned commentary:

A court decision to hear an Alien Tort Statute claim over action in
South Africa reflects the worst sort of “judicial imperialism.” It
would send the message that the United States does not respect the
ability of South African society to administer justice by implying that
U.S. courts are better placed to judge the pace and degree of South
Africa’s national reconciliation. In contrast, U.S. intervention to
block such a suit sends a different signal to South Africa and other
countries struggling through difficult political transitions. It would
communicate our recognition of the respected position that the justice

91
 system holds in South Africa and reinforce the importance of having
those claims judged in that country.

Schrage, supra, at 166.2

-----------------------------

The majority declines to address the aspect of Sosa directing courts to determine the

deference owed to the Executive Branch’s view of the case, notwithstanding the facts that both

parties and amici devote many pages of argument to this issue and that the issue goes directly to

whether subject matter jurisdiction should be exercised over the cause of action alleged by the

plaintiffs. The principal reason given for this silence is that “the district court explicitly refrained

from addressing the defendants’ arguments that the ATCA claim presented a non-justiciable political

question.” Per Curiam Op. ante at 12 (citing In re S. African Apartheid Litig., 346 F. Supp. 2d at

543 n.4) (parenthetical omitted). The footnote in Judge Sprizzo’s opinion, from which the per

curiam opinion quotes an incomplete excerpt, reads as follows:

Defendants also argue that there is no case or controversy for this
Court to hear under Article III of the Constitution because plaintiffs
cannot establish that they have standing to bring this action and
because the matter is a non-justiciable political question. See
Memorandum of Law in Support of Defendants’ Joint Motion to
Dismiss at 3-4. Given the Court’s finding that defendants are entitled

2
The Khulumani plaintiffs also argue that, “[t]he ATS aside, there are independent grounds for
jurisdiction in federal court”, namely, diversity of citizenship pursuant to 28 U.S.C. § 1332. Br. for App.
at 56. These causes of action, however, are also subject to dismissal based on deference to the Executive
Branch’s view of the foreign policy impact of allowing them to proceed, and on the doctrine of
international comity. See Bi., 984 F.2d at 586-87; Doe v. Exxon Mobil Corp., 473 F.3d 345, 357 (D.C.
Cir. 2007) (Kavanaugh J., dissenting from denial of writ of mandamus.) The point of disagreement in
Exxon , as the majority acknowledged, id., turned on the availability of a writ of mandamus to review the
denial of the defendant’s motion to dismiss where the Executive Branch did not unequivocally seek such
relief.

92
 to relief on other grounds, the Court need not address these remaining
grounds for defendants’ motion.

In re S. African Apartheid Litig., 346 F. Supp. 2d at 543 n.4.

Contrary to the per curiam opinion, a careful reading of the record shows that Judge Sprizzo

did not explicitly refrain from addressing the issue of the deference to be accorded to the position

of the State Department under Sosa. Indeed, the Memorandum of Law in Support of the Defendants’

Joint Motion to Dismiss (“Joint Motion”), which Judge Sprizzo expressly referenced in the part of

his opinion declining to address the argument that the ACTA claim presented a non-justiciable

political question, argued (in relevant part) that the “adjudication of plaintiffs’ claim would require

a court to pass on the merits of political questions that were resolved by the executive and legislative

branches of the United States government in favor of commerce with South Africa.” Id. at 3. The

defendants did not argue that the case should be dismissed in deference to the Statement of Interest

filed by the United States, nor did they rely on the advice of the Supreme Court in Sosa that the

views of the United States in this case “certainly deserve[d] great weight.” Sosa, 542 U.S. at 733

n.21. Indeed, the defendants’ motion was filed before the Statement of Interest and a year before the

Supreme Court suggested that it be afforded deference.

More significantly, an objective reading of Judge Sprizzo’s opinion demonstrates that he did

rely on the Supreme Court’s advice in Sosa. See In re S. African Apartheid Litig., 346 F. Supp. 2d

at 547, 551, 553-54. Nevertheless, I am unwilling to burden this opinion any further on this score

because, even if Judge Sprizzo did not rely expressly on the Supreme Court’s guidance in Sosa as

a basis for his holding, this would not be an excuse for avoiding the issue. While cases may be

found where we declined to reach an issue not addressed by the district court, see, e.g., Bigio v.

93
Coca-Cola Co., 239 F. 3d 440 (2d Cir. 2000), we have held expressly that we are “free to affirm a

district court decision . . . even [on] grounds not relied upon by the district court.” In re Methyl

Tertiary Butyl Ether (“MTBE”) Prods. Liab. Litig., 488 F.3d 112, 134 (2d Cir. 2007) (quoting

Gmurzynska v. Hutton, 355 F.3d 206, 210 (2d Cir. 2004)). Thus, in Bi, 984 F.2d 582, a mirror image

of Bigio, the district court had dismissed a complaint on the grounds of forum non conveniens – a

ground on which the defendant had not moved to dismiss, presumably because of the unavailability

of another forum. See In re Union Carbide Corp. Gas Plant Disaster, MDL No. 626, 1992 WL

36135 (S.D.N.Y. Feb. 18, 1992). We affirmed the dismissal, applying the principle of international

comity, even though that issue had not been addressed by the district court. Bi, 984 F.2d at 586.

Similarly, in Kadic v. Karadzic, 70 F.3d 232 (2d Cir. 1995) (Newman, J.), the district court

dismissed the two related cases before it solely on the ground of lack of subject matter jurisdiction

under the ATCA. See Doe v. Karadzic, 866 F. Supp. 734, 736 (S.D.N.Y. 1994). On appeal,

however, the parties briefed three different grounds for dismissal, even though two of them had not

been addressed by the district court, and Judge Newman considered each of them in turn. Kadic, 70

F.3d at 238.

There is a good reason to reach the issue here. Prosecution of these cases undermines

significant interests of the United States and our relations with the Republic of South Africa, as well

as allies of the United States, who have expressed concern over the exercise of jurisdiction over these

cases. Under these circumstances, it is an issue that should be resolved at the threshold. Particularly

apposite here are the words of the Court of Appeals for the Seventh Circuit addressing a motion to

dismiss a Sherman Act claim on the ground that the conduct alleged did not have a substantial effect

on commerce within the United States. (The “effect on commerce” requirement was enacted as part

94
of the Foreign Trade Antitrust Improvements Act of 1982 (“FTAIA”).) In holding that subject

matter jurisdiction was implicated, the Seventh Circuit observed:

There are good policy reasons for [this conclusion]. The
extraterritorial scope of our antitrust laws touches our relations with
foreign governments, and so, it seems, it is prudent to tread softly in
this area. If FTAIA sets out an issue on the merits, resolution of the
issue could be delayed until late in the case, and the potential for a
lawsuit to have an effect on foreign markets would exist while the
case remained pending. . . . Treating the matter as one of subject
matter jurisdiction reduces the potential for offending the economic
policies of other nations. In short, FTAIA limits the power of the
United States courts (and private plaintiffs) from nosing about where
they do not belong. And the power of the courts is precisely what
subject matter jurisdiction is about.

United Phosphorus, Ltd. v. Angus Chem. Co., 322 F.3d 942, 952 (7th Cir. 2003); see also In re BDC

56 LLC, 330 F.3d 111, 118 (2d Cir. 2003) (finding that a particular issue was jurisdictional because

it implicated “a threshold determination that should be made at the earliest possible stage of the

proceedings”).

While I rely on these cases simply to demonstrate the force of the policy favoring the exercise

of our discretion to decide this issue, they also support the argument that the issue here should be

treated as one going to subject matter jurisdiction – which we must resolve – even if it would not

necessarily be so treated in other contexts. This argument derives from the language of the ATCA

and the Supreme Court’s analysis of the manner in which the threshold jurisdictional question should

be resolved. Specifically, Sosa stated that “[a]ll Members of the Court agree that § 1350 is only

jurisdictional” and “that the jurisdiction was originally understood to be available to enforce a small

number of international norms that a federal court could properly recognize as within the common

law enforceable without further statutory authority.” 542 U.S. at 729.

95
 The Supreme Court then observed that “it would be unreasonable to assume that the First

Congress would have expected federal courts to lose all capacity to recognize enforceable

international norms . . . .” Id. at 730. Sosa went on to discuss “the ultimate criteria for accepting

a cause of action subject to jurisdiction under § 1350,” id. at 732, while making it clear that the

criteria it was setting forth were not exclusive. One consideration “involve[s] an element of

judgment about the practical consequences of making that cause available to litigants in the federal

courts.” Id. at 732-33. Since subject matter jurisdiction under the ATCA depends on whether the

defendants have violated an international law norm which federal courts are prepared to recognize,

accept and make available to litigants, the application of the criteria for making that determination

is one that by definition goes to the issue of subject matter jurisdiction.

The per curiam opinion argues that subject matter jurisdiction under the ATCA is established

simply by virtue of an allegation that the defendant’s conduct violated a norm of international law

and that the decision to provide a remedy for the alleged violation simply relates to the issue of

“whether a cause of action exists.” Per Curiam Op. ante at 19. The latter issue, this argument

continues, does not implicate subject matter jurisdiction. Id. ante at 19-20. While this may be so

with respect to 28 U.S.C. § 1331, which confers jurisdiction “of all civil actions arising under the

Constitution, laws, or treaties of the United States,” the per curiam opinion ignores the language of

the ATCA, which is one of a number of statutes in which “Congress has exercised its prerogative

to restrict the subject-matter jurisdiction of federal district courts based on a wide variety of factors,

some of them also relevant to the merits of a case.” Arbaugh v. Y & H Corp., 546 U.S. 500, 516 n.11

(2006). Where Congress has done so, subject matter jurisdiction turns on whether the complaint

states a cause of action. See id. at 515-16 (“If the Legislature clearly states that a threshold limitation

96
on a statute’s scope shall count as jurisdictional, then courts and litigants will be duly instructed and

will not be left to wrestle with the issue.”).3 Indeed, in Members For a Better Union v. Bevona, 152

F.3d 58 (2d Cir. 1998), we vacated a judgment entered after trial for failure of the complaint to state

a cause of action, id. at 61, without regard to the trial record and the assertion of the parties “that the

district court had subject matter jurisdiction.” Id. at 67 n.1.4

More significantly, even before Sosa reaffirmed the jurisdiction-conferring nature of the

ATCA, we “distinguish[ed] [the] Alien Tort Claims Act, with its jurisdictional pleading

requirement, from general federal question jurisdiction, which is ‘not defeated by the possibility that

the averments in the complaint may fail to state a cause of action.’” Bigio, 239 F.3d at 447 (citing

Filartiga v. Pena-Irala, 630 F.2d 876, 887-88 (2d Cir. 1980) (citing Bell v. Hood 327 U.S. 678

(1946))). Indeed, as early as the very “birth of the modern line of cases” construing the ATCA, Sosa,

542 U.S. at 724-25, we acknowledged that, because of “the statute’s requirement of alleging a

‘violation of the law of nations’ . . . at the jurisdictional threshold[,] [c]ourts have . . . engaged in a

more searching preliminary review of the merits than is required, for example, under the more

3
The jurisdictional statutes cited by the Supreme Court as examples in which Congress has
exercised its prerogative to restrict the subject matter jurisdiction of federal district courts include
statutes that contain jurisdiction-conferring language identical to that in 28 U.S.C. § 1350. See, e.g., 28
U.S.C. §§ 1345 & 1348 (cited in Arbaugh v. Y & H Corp., 546 U.S. 500, 516 n.11 (2006)). Indeed, in
Robinson v. Government of Malaysia, 269 F.3d 133 (2d Cir. 2001), in which Judge Katzmann concurred,
we said that the “Alien Tort Claims Act (“ATCA”), 28 U.S.C. § 1350, which provides that ‘[t]he 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,’ similarly identifies certain substantive
law violations as jurisdictional pleading requirements.” Id. at 144 n.13.
4
The statute at issue in Bevona, section 102 of the Labor-Management Reporting and Disclosure
Act, 28 U.S.C. § 412, did not by its terms restrict the exercise of subject matter jurisdiction.
Nevertheless, we held that it had been so construed by the Supreme Court. Bevona, 152 F.3d at 62
(citing Calhoon v. Harvey, 379 U.S. 134, 138 (1964)).

97
flexible ‘arising under’ formulation.” Filartiga, 630 F.2d at 887-88 (comparing O’Reilly de Camara

v. Brooke, 209 U.S. 45, 52 (1907) (question of ATCA jurisdiction disposed of “on the merits”)

(Holmes, J.), with Bell, 327 U.S. 678 (general federal question jurisdiction not defeated by the

possibility that the averments in the complaint may fail to state a cause of action)) (emphasis in

original); accord Kadic, 70 F.3d at 238; Bigio, 239 F.3d at 447. After Sosa, “a more searching

preliminary review of the merits” necessarily includes determining whether the plaintiffs have

adequately pled a violation of the law of nations over which it is appropriate for the district court to

exercise subject matter jurisdiction.

Again, even if I am wrong on this score, we should consider the issue of case-specific

deference at this juncture because “the continued adjudication of the . . . matters risks potentially

serious adverse consequences for significant interests of the United States,”, A01090, threatens our

relations with the Republic of South Africa, and poses “continuing tensions in our relations” with

the other countries where many of the defendants are incorporated A01091. Just as the Seventh

Circuit treated the FTAIA as a jurisdictional statute in order to avoid “offending the economic

policies of other nations” during the pendency of the case, United Phosphorous, 322 F.3d at 952, we

should do so here.

Moreover, the per curiam opinion offers no good reason for failing to do so. It suggests that

its “approach is particularly appropriate here because plaintiffs have indicated that, if given the

opportunity, they would narrow their claims and clarify the nature of their allegations against various

defendants, changes that may affect how the district court ultimately decides to resolve these issues.”

Per Curiam Op. ante at 13-14 (citing Tr. 7-8, 14-15). The Khulumani plaintiffs, however, did not

seek leave to file an amended complaint and Judge Sprizzo has previously granted the Ntsebesa and

98
Digwamaje plaintiffs leave to file a second amended complaint alleging that the defendants aided

and abetted the violation of norms of customary international law.5 Moreover, the per

curium opinion does not identify a single flaw in the current complaints that would need to be cured

by an amended complaint, and the transcript of the oral argument does not reflect any offer by the

plaintiffs to amend their complaints in any significant way. Indeed, the plaintiffs expressly argued

against the standard that Judge Katzmann adopts in section II.B of his concurring opinion (which I

have joined) for imposing liability on the defendants for aiding-and-abetting in violation of the norm

of international law.

Nevertheless, relying on the amended complaints that the plaintiffs may be permitted to file,

the per curiam opinion declines “to determine whether the plaintiffs have adequately pled a violation

of international law sufficient to avail themselves of jurisdiction under the ATCA . . . .” Per Curiam

Op. ante at 10. The per curiam opinion does so in the face of our holding in Bigio that the failure

to properly plead a violation of the law of nations means that “neither [the district court] nor we may

consider the matter further . . . .” 239 F.3d at 447. Instead of upholding its obligation to determine

whether subject matter jurisdiction exists, see In re MTBE, 488 F.3d at 121-22, the per curium

opinion offers up an advisory opinion that a plaintiff may allege that a private party aided and abetted

a violation of a norm of customary international law without so much as a reference to the factual

allegations in the complaint or citation to the specific norm in which the defendants were allegedly

complicit.

5
On April 28, 2003, before Sosa was decided, Judge Sprizzo permitted the Ntsebeza plaintiffs to
file a third amended complaint. In re S. African Apartheid Litig., 346 F. Supp. 2d at 544 n.8. While
some defendants received a copy of such a complaint on May 19, 2003, Joint Motion at 5 n.2, “[t]hat
Complaint was never properly filed.” In re S. African Apartheid Litig., 346 F. Supp. 2d at 544 n.8.
Nevertheless, it was virtually identical to the one previously filed. Id.

99
 Under these circumstances, it is difficult to understand the disposition of the appeal, a

judgment that “vacate[s] the district court’s order denying plaintiffs’ motion for leave to amend.”

Per Curiam Op. ante at 9. The majority does say that, “[i]n denying this motion, the district court

relied, in part, on the erroneous premise that subject mater jurisdiction did not inhere and reasoned

that any additional amendments to the pleadings would be futile.” Id. I pass over the incomplete

and inaccurate description of the grounds for the denial of the motion to amend.6 Nevertheless, I do

not understand how the majority can say that Judge Sprizzo denied the motion to amend the

complaint on the “erroneous premise” that subject matter jurisdiction was lacking, while declining

to rule on the basic jurisdictional issue of whether the plaintiffs “have adequately pled a violation

of international law sufficient to avail themselves of jurisdiction under the ATCA.” Id. at 10.

I decline to join in this peculiar disposition, by which my colleagues seek desperately to avoid

the easiest ground on which to resolve this appeal – that of deference to the judgment of the Republic

of South Africa, supported by our State Department, that these cases are none of our business – and

go on to grapple unnecessarily with difficult issues relating to the ATCA and customary international

law without being able to agree on the rationale for the result they reach. Nevertheless, since the

majority has chosen to do so, I address the issue whether the scope of liability for violations of the

international law norms at issue here extends to private actors, such as corporations, for aiding-and-

abetting. I then turn to the separate opinions of Judges Hall and Katzmann.

6
Judge Sprizzo denied the motion to amend because (1) he had previously provided the plaintiffs
leave to replead; (2) the plaintiffs’ request came “four months after this Court dismissed these actions
and more than three months after plaintiffs noticed their appeal”; (3) allowing them to replead would
interfere with “ensuring a speedy appeal”; and (4) the proposed amendments would have been “fruitless”
in light of his ruling that aiding-and-abetting is not recognized under the ATCA. A01139.

100
2. The Scope of Liability Under the ATCA

Sosa indicates that this inquiry requires consideration of two separate principles. First, the

Court held that the ATCA should not be read to create jurisdiction over “violations of any

international law norm with less definite content and acceptance among civilized nations than the

historical paradigms familiar when § 1350 was enacted.” 542 U.S. at 732. Second, the Court

observed that “[a] related consideration is whether international law extends the scope of liability

for a violation of a given norm to the perpetrator being sued, if the defendant is a private actor such

as a corporation or individual.” Id. at 732 n.20. To that end, immediately thereafter, the Court

contrasted two opinions, one finding no consensus that torture by a private actor violated

international law, Tel-Oren v. Libyan Arab Republic, 726 F.2d 774, 791-95 (D.C. Cir. 1984)

(Edwards, J., concurring), another finding such a consensus that genocide by a private actor did

violate international law, Kadic, 70 F.3d at 239-41. Justice Breyer, in a separate opinion, agreed

with the Court that the international law norm invoked “must extend liability to the type of

perpetrator (e.g., a private actor) the plaintiff seeks to sue.” Sosa, 542 U.S. at 760 (Breyer, J.,

concurring in part and concurring in the judgment) (citing id. at 732 n.20).

As did the Supreme Court in Sosa, we have held that, “[i]n order to determine whether the

offenses alleged by the [plaintiffs] in this litigation are violations of the law of nations that may be

the subject of Alien Tort Act claims against a private individual, we must make a particularized

examination of these offenses . . . .” Kadic, 70 F.3d at 241 (emphasis added). In that case, Judge

Newman then proceeded to engage in such an exercise. Id. at 241-44 (discussing independently

genocide, war crimes, and torture and summary execution). More recently, in Wiwa, we reiterated

our holding in Kadic that “the ATCA reaches the conduct of private parties provided that their

101
conduct is undertaken under the color of state authority or violates a norm of international law that

is recognized as extending to the conduct of private parties.” 226 F.3d at 104.

These holdings are consistent with the position of the European Commission, which clearly

influenced Justice Souter’s majority opinion and Justice Breyer’s concurrence in Sosa. Specifically,

in its amicus brief in Sosa, the Commission urged the Supreme Court to hold, as it ultimately did,

that in determining whether a non-state actor was complicit in a violation of customary international

law by a state actor, courts should apply international, rather than domestic, legal standards. 2004

WL 177036, at *4. Significantly, and again consistent with our holding in Kadic, the Commission

argued that only a subset of norms recognized as customary international law applies to non-state

actors and “hence only that subset may form the basis of liability against such actors. For example,

non-state actors may be liable for genocide, war crimes, and piracy, while torture, summary

execution, and prolonged arbitrary detention do not violate the law of nations unless they are

committed by state officials or under color of law.” Id. at *11 (citations omitted).

Consistent with this analysis, I first address whether the complaint here makes allegations

against defendants sufficient to hold them liable as acting under the color of law. I next turn to

whether, at the time the alleged crimes were committed, there was a well established and universally

recognized international norm providing for liability of private parties who aid and abet apartheid.

Finally, while officers and employees of a corporation may be held responsible for using the entity

as the vehicle for the commission of crimes against humanity, I address the issue whether the entities

themselves may be held responsible.

(a) Private Party Liability under Color of Law

102
 Judge Sprizzo held that the allegations against the defendants were not sufficient to “elevate

them to the status of state actors . . . .” In re S. African Apartheid Litig., 346 F. Supp. 2d at 548-49.

This aspect of his decision is challenged in the briefing of only one group of appellants. The state

action element derives from the principle that violations of international human rights norms,

including apartheid, are violations of customary international law only if practiced, encouraged, or

condoned by the government of a state as a matter of state policy. Restatement (Third) of Foreign

Relations Law § 702 & cmt. b (discussed in Sosa, 542 U.S. at 737). See United States v. Josef

Altstoetter (“The Justice Case”), 3 Trials of War Criminals Before the Nuernberg Military Tribunals

Under Control Council Law No. 10 954, 982 (William S. Hein & Co., Inc. 1997) (1947) (Crimes

against humanity committed against German nationals are subject to punishment “only where there

is proof of conscious participation in systematic government organized or approved procedures . .

. .”).7 Indeed, this explains why Sosa specifically held that, before jurisdiction is accepted over an

ATCA cause of action against a private actor, it must be determined “whether international law

extends the scope of liability for a violation of a given norm to . . . a private actor. . . .” 542 U.S. at

732 n.20. Cf. Am. Mfrs. Mut. Ins. Co. v. Sullivan, 526 U.S. 40, 49-50 (1999) (linking the color-of-

law element of 42 U.S.C. § 1983 to the underlying state-action requirement of the Fourteenth

Amendment, and holding that both exclude from their reach “merely private conduct, no matter how

discriminatory or wrongful”) (internal quotation marks and citation omitted).

7
The ICTY held recently that “[w]hile this may have been the case during the Second World
War . . . the law in relation to crimes against humanity has developed to take into account forces which,
although not those of the legitimate government, have de facto control over, or are able to move freely
within, defined territory.” Prosecutor v. Tadiƒ, 36 I.L.M. 908, 945, Case No. IT-94-1-T, Opinion and
Judgment ¶ 654 (Trial Chamber May 7, 1997); see also Antonio Cassese, Crimes Against Humanity, in
Cassese et al., supra, at 353, 357.

103
 We have held that case law construing 42 U.S.C. § 1983 is “a relevant guide to whether a

defendant has engaged in official action for purposes of jurisdiction under the Alien Tort Act.”

Kadic, 70 F.3d at 245. As Judge Newman observed,“[a] private individual acts under color of law

within the meaning of section 1983 when he acts together with state officials or with significant state

aid.” Id. (citing Lugar v. Edmondson Oil Co., 457 U.S. 922, 937 (1982)). Subsequently, the

Supreme Court has cautioned that the language from Lugar, which Judge Newman paraphrased,

“must not be torn from the context out of which it arose,” Sullivan, 526 U.S. at 58, namely, an ex

parte application for a writ of attachment. This admonition is a useful reminder that the “under color

of law” cases arise in varying contexts. Many, including cases on which plaintiffs rely, have been

limited to their facts. Id. at 55-59.

Perhaps the largest single category of cases under color of law are those in which the primary

offender or tortfeasor is a private party, and the issue turns on whether “seemingly private behavior

‘may be fairly treated as that of the state itself.’” Brentwood Acad. v. Tenn. Secondary Sch. Athletic

Ass’n, 531 U.S. 288, 295 (2001) (quoting Jackson v. Met. Edison Co., 419 U.S. 345, 351 (1974)).

Because the cases on this appeal do not fit in this category, I am reluctant to discuss them in detail.

Nevertheless, because plaintiffs rely on general language from the cases “torn from the context out

of which it arose,” Sullivan, 526 U.S. at 59, I borrow from Professor Schwartz’s treatise, first, one

significant caveat, and second, his useful synthesis of the Supreme Court’s holdings in this area.

The caveat is that, in analyzing the large body of decisional law concerning state action, “it

is important to consider the era in which the decision was rendered.” 1 Martin A. Schwartz, Section

1983 Litigation: Claims and Defenses § 5.12, at 5-85 (4th ed. 2003). Professor Schwartz explains

that the Warren Court took an expansive view of state action in its effort to combat racial

104
discrimination in society, but that the subsequent Burger and Rehnquist courts reversed this trend

in order to shield private behavior from the reach of the Constitution. Id. § 5.12, at 5-85–5-86. This

reversal was undertaken not by explicitly overturning any earlier state-action decisions, but through

a series of decisions finding no state action despite heavy government involvement in private

conduct. Id. § 5.12, at 5-86. The unmistakable messages sent by these decisions are that the concept

of state action is to be given only limited berth, and that federal courts must assess the continued

vitality of earlier state-action precedents in light of more recent decisional law. Id.8

Professor Schwartz then supplemented this perceptive analysis by stating the principles that

have emerged as the Supreme Court has narrowed the scope of its state-action jurisprudence:

1. Mere state regulation of private conduct, even if extensive, is
insufficient to support a finding of state action.

2. State authorization of private conduct does not make the private
party a state actor; to find state action, the state must participate in,
coerce, or significantly encourage the contested activity.

3. State assistance to a private party, even if substantial, will not
support a finding of state action, whether that assistance is in the form

8
Burton v. Wilmington Parking Authority, 365 U.S. 715 (1961), which provides the basis for the
so-called “symbiotic relationship test” between public and private actors and upon which plaintiffs rely,
was one of the Warren Court cases that took an expansive view of state action. The Supreme Court did
not employ the “symbiotic relationship” language in Burton, though in a subsequent case, it referred to
“the symbiotic relationship between lessor and lessee . . . present in Burton.” Moose Lodge No. 107 v.
Irvis, 407 U.S. 163, 175 (1972). “Although neither Burton nor the symbiotic relationship doctrine has
been overruled, they have been severely narrowed in scope and diminished as precedent. Supreme Court
decisional law has given Burton a very narrow interpretation and, with the possible exception of its
decision that a civil litigant’s exercise of a racially based peremptory challenge constitutes state action
[Edmondson v. Leesville Concrete Co., 500 U.S. 614 (1991)], has rejected every attempt to establish state
action on the basis of Burton.” 1 Martin A. Schwartz, Section 1983 Litigation: Claims and Defenses §
5.13[A], at 5-90–5-91 (4th ed. 2003); see also Sullivan, 526 U.S. at 57-58; Laurence Tribe, American
Constitutional Law § 18-3, at 1701 n.13 (2d ed. 1988).

105
 of direct financial aid, tax exemptions, monopoly power, government
mortgage insurance, or the grant of a license.

4. The mere importance of the function carried out by the private
sector is an insufficient basis upon which to find state action; for state
action to be found, the function must be historically, traditionally, and
exclusively governmental.

Id. § 5.12, at 5-87–5-88 (footnotes omitted). Professor Schwartz continues,

[t]he Supreme Court in some cases has found no state action, even when all four of
these government involvements coalesced in the same case. Thus, the Court has
found on more than one occasion that an entity was not engaged in state action even
though it was extensively regulated, obtained governmental approval, received
substantial governmental assistance, and performed an important societal function.

Id. § 5.12, at 5-88 (citing San Francisco Arts & Athletics v. United States Olympic Comm., 483 U.S.

522 (1987); Rendell-Baker v. Kohn, 457 U.S. 830 (1982); Blum v. Yaretsky, 457 U.S. 991 (1982);

Jackson v. Metro. Edison Co., 419 U.S. 345 (1974)).

In analyzing the subgroup of state-action cases of which the instant cases are representative

– those where the primary offender or tortfeasor is a state actor, and the question is whether a private

party who is alleged to be an accessory may be deemed a state actor – it is useful to discuss the facts

of the leading Supreme Court cases because they illustrate the kind of showing that must be made

before a private party may be deemed a state actor in this context. The case most directly on point

is Dennis v. Sparks, 449 U.S. 24 (1980), in which a private party bribed a judge to issue an order

enjoining the production of minerals from certain oil leases. In holding that the private actor was

acting “under color of law,” the Supreme Court observed that “[p]rivate persons, jointly engaged

with state officials in a challenged action, are acting ‘under color’ of law for purposes of § 1983

actions.” Id. at 24. Specifically, “the allegations were that an official act of the defendant judge was

106
the product of a corrupt conspiracy involving bribery of the judge. Under these allegations, the

private parties conspiring with the judge were acting under color of state law; and it is of no

consequence in this respect that the judge himself is immune from damages liability.” Id. at 28. On

the other hand, if the private actor had merely resorted to the courts and prevailed in the lawsuit, he

would not have been acting under color of law.

United States v. Price, 383 U.S. 787 (1966), which arose out of one of the saddest episodes

of the civil rights movement, provides another compelling example of this subgroup. Price involved

the arrest and subsequent murder of three civil rights workers in Mississippi. The facts are described

as follows in the opinion of the Court:

On June 21, 1964, Cecil Ray Price, the Deputy Sheriff of Neshoba
County, Mississippi, detained Michael Henry Schwerner, James Earl
Chaney and Andrew Goodman in the Neshoba County jail located in
Philadelphia, Mississippi. He released them in the dark of that night.
He then proceeded by automobile on Highway 19 to intercept his
erstwhile wards. He removed the three men from their automobile,
placed them in an official automobile of the Neshoba County Sheriff's
office, and transported them to a place on an unpaved road.

These acts, it is alleged, were part of a plan and conspiracy whereby
the three men were intercepted by the 18 defendants, including
Deputy Sheriff Price, Sheriff Rainey and Patrolman Willis of the
Philadelphia, Mississippi, Police Department. The purpose and intent
of the release from custody and the interception, according to the
charge, were to “punish” the three men. The defendants, it is alleged,
“did wilfully assault, shoot and kill” each of the three. And, the
charge continues, the bodies of the three victims were transported by
one of the defendants from the rendezvous on the unpaved road to the
vicinity of the construction site of an earthen dam approximately five
miles southwest of Philadelphia, Mississippi.

107
Id. at 790. The Supreme Court held that private individuals who participated directly with state

actors pursuant to a common design were acting under color of law. “In effect, if the allegations are

true, they were participants in official lawlessness, acting in willful concert with state officers and

hence under color of law.” Id. at 795. Price is a classic example of the rule that private parties act

under color of law when they are “engaged in a conspiracy with state officials to violate the

Fourteenth Amendment.” Conway v. Vill. of Mount Kisco, 750 F.2d 205, 214 n.12 (2d Cir. 1985)

(Oakes, J.); accord Ginsberg v. Healey Car & Truck Leasing, Inc., 189 F.3d 268, 272-73 (2d Cir.

1999); Alexis v. McDonald’s Rests. of Mass. Inc., 67 F.3d 341, 352 (1st Cir. 1995); Mershon v.

Beasley, 994 F.2d 449, 451 (8th Cir. 1993); Annunziato v. The Gan, Inc., 744 F.2d 244, 251 (2d Cir.

1984).

There is good reason to require direct participation with a state actor pursuant to a conspiracy

or common design before holding a private party, who is not the primary wrongdoer, liable as a state

actor. Where the issue is whether a public official can be held liable for the primary wrongdoing of

a private actor, the public office he holds suffices to satisfy the state-actor requirement; the only issue

is whether his conduct in that capacity constitutes a proximate cause of the plaintiff’s injury. Thus,

“a state actor may be subject to liability for an action physically undertaken by private actors in

violation of the plaintiff’s liberty or property rights if the state actor directed or aided and abetted the

violation.” Dwares v. City of New York, 985 F.2d 94, 98 (2d Cir. 1993). On the other hand, where

the primary actor is a public official, for a private actor to be deemed a state actor, he must jointly

participate in the wrongful conduct, pursuant to a common design or plan. As the Supreme Court

explained in Adickes v. S.H. Kress & Co.,“[t]he involvement of a state official in such a conspiracy

plainly provides the state action essential to show a direct violation of petitioner’s Fourteenth

108
Amendment equal protection rights . . . . Moreover, a private party involved in such a conspiracy,

even though not an official of the State, can be liable under § 1983.” 398 U.S. 144 , 152 (1970).

This holding finds support in the law of agency. “When two persons engage jointly in a

partnership for some criminal [or legitimate] objective, the law deems them agents for one another.

Each is deemed to have authorized the acts and declarations of the other undertaken to carry out their

joint objective.” United States v. Russo, 302 F.3d 37, 45 (2d Cir. 2002). Thus, a private party who

conspires with a state actor becomes his agent, and his actions in that capacity are sufficient to make

him an actor under color of law. Similarly where a private party induces an otherwise innocent

public official to commit an offense, he has effectively made that person his or her agent.

This doctrine is an outgrowth of common law principles of criminal responsibility
dating at least as far back as Regina v. Saunders, 2 Plowd. 473 (1575); and of
principles of civil responsibility established, by force of the maxim qui facit per
alium facit per se [he who acts thorough another acts for himself], at least as early as
the 14th century.

United States v. Lester, 363 F.2d 68, 72 (6th Cir. 1966). On the other hand, there is no legal

framework or precedent for holding that a private party, merely selling goods or materials to a state

actor, is himself acting under color of law.

Consistent with this precedent, we have held that a conspiracy to violate an international law

norm is necessary to render a private party liable for conduct in which the state is the primary actor.

Bigio, 239 F.3d at 447. Indeed, there we rejected a cause of action alleging that Coca-Cola Co. was

acting “under color of law” where there was “no allegation in the complaint, let alone any hint of

evidence . . . that Coca-Cola had any role as a participant or co-conspirator in the confiscation of the

Bigios’ property.” Id. at 449; cf. In re Brooklyn Navy Yard Asbestos Litig., 971 F.2d 831, 841 (2d

109
Cir. 1992) (holding that joint participation pursuant to an agreement – express or implied – to

commit a tort is necessary to impose liability on those acting in concert); Restatement (Second) of

Torts § 876(a) (imposing liability on one who “does a tortious act in concert with the other or

pursuant to a common design with him”).

Neither Judge Katzmann nor Judge Hall takes issue with my analysis of section 1983 cases

defining circumstances under which a private party may be deemed to have acted under color of law.

Judge Katzmann, however, relying on cases applying the general criminal aiding-and-abetting statute

applicable to all federal crimes, 18 U.S.C. § 2, argues that a private person may be liable for aiding-

and-abetting a public official in the commission of a crime even though he may be incapable of

committing the crime himself and even though his conduct would not satisfy the standard set out in

§ 1983 for treating him as a state actor. Op. of Judge Katzmann ante at 47 (citing In re Nofziger,

956 F.2d 287, 290 (D.C. Cir. 1992); United States v. Tannenbaum, 934 F.2d 8, 14 (2d Cir. 1992)).

The fact that a person may be convicted of aiding-and-abetting an offense of which he could be

found guilty as a principal - because Congress has so provided - does not provide a basis for

imposing such liability in the present context. Because Congress has not enacted a comparable civil

aiding-and-abetting statute, a private party could not be subject to such liability in a civil action

under 42 U.S.C. § 1983. Central Bank of Denver v. First Interstate Bank of Denver, 511 U.S. 164,

181-82 (1994).

Significantly, when Congress enacted the Torture Victims Protection Act of 1991 (“TVPA”)

to comply to with the Torture Convention, it provided for civil causes of action only against “[a]n

individual who, under actual or apparent authority, or color of law, of any foreign nation,”

undertakes conduct prohibited by the Act. TVPA § 2(a), Pub. L. No. 102-256, 106 Stat. 73 (1992).

110
Moreover, the Senate Report accompanying the legislation provides that “[c]ourts should look to

principles of liability under U.S. civil rights laws, in particular section 1983 of title 42 of the United

States Code, in construing ‘under color of law’ as well as interpretations of ‘actual or apparent

authority’ derived from agency theory in order to give the fullest coverage possible.” S. Rep. 102-

249, at 8 (Nov. 26, 1991). To the extent that the Senate Report speaks to the issue of aiding-and-

abetting, it addresses the issue of supervisory liability and makes the point that “a higher [public]

official need not have personally ordered the abuses in order to be held liable. Under international

law, responsibility for torture, summary execution, or disappearances extends beyond the person or

persons who actually committed those acts [to] anyone with higher authority who authorized,

tolerated, or knowingly ignored those acts.” Id. at 9 (citing In re Yamashita, 327 U.S. 1; Forti v.

Suarez-Mason, 672 F. Supp. 1531 (N.D. Cal. 1987)).

Nevertheless, even if 18 U.S.C. § 2 applied in this context, the standard for imposing liability

for aiding-and-abetting pursuant to this section is not satisfied by the complaints in this case.

Specifically, the Supreme Court has held that, “[i]n order to aid and abet another to commit a crime

it is necessary that a defendant in some sort associate himself with the venture, that he participate

in it as in something that he wishes to bring about, that he seek by his action to make it succeed.”

Nye & Nissen v. United States, 336 U.S. 613, 619 (1949) (internal quotation marks and citation

omitted). This element is also essential to a cause of action for aiding-and-abetting. See, e.g., E.

Trading Co. v. Refco, 229 F.3d 617, 623 (7th Cir. 2000) (Posner, J.) (observing “that one who aids

and abets a fraud, in the sense of assisting the fraud and wanting it to succeed, is himself guilty of

fraud”). Indeed, in United States v. Blankenship, Judge Easterbrook, after citing to the Model Penal

Code for the proposition that a “supplier [is] culpable only if he has ‘the purpose of promoting or

111
facilitating’ the crime,” observed that a similar approach has been adopted to aiding-and-abetting in

civil cases. 970 F.2d 283, 286 (7th Cir. 1992). Moreover, in Boim v. Quranic Literacy Institute, the

same court held that 18 U.S.C. § 2339A, which created a civil cause of action for those injured by

terrorist acts against individuals who provided support to terrorist groups, survived a First

Amendment challenge “so long as the plaintiffs are able to prove that the defendants knew about the

organization’s illegal activity, desired to help that activity succeed and engaged in some act of

helping.” 291 F.3d 1000, 1028 (7th Cir. 2002). Cf. Halberstam v. Welch, 705 F.2d 472, 488 (D.C.

Cir. 1983) (holding that “desire to make the venture succeed” is a factor in determining aiding-and-

abetting liability).

There is no allegation here that the defendants acted with the intent to make the violation

succeed. On the contrary, plaintiffs strenuously argued in their briefs and at oral argument that they

were not required to make such a showing. Ntsebesa Reply Br. at 19; Khulumani Reply Br. at 11;

Tr. at 13. They argue that a private party who knowingly facilitates the commission of an

international law violation by a state actor acts under color of law. I use the term “facilitates”

because “[o]ne who merely sells goods to a buyer is not an aider and abettor of crimes that the buyer

might commit, even if the seller knows that the buyer is likely to use the goods unlawfully, because

the seller does not share the specific intent to further the buyer’s venture.” Corrie v. Caterpillar,

Inc., 403 F. Supp. 2d 1019, 1027 (W.D. Wash. 2005) (citing Blankenship, 970 F.2d at 285-87), aff’d

on other grounds, Corrie v. Caterpillar, Inc., _F.3d_, 2007 WL 2694701 (9th Cir. Sept. 17, 2007);

see also N.Y. Penal Code § 115.05 (criminal facilitation). Indeed, such conduct does not violate

customary international law. See The Ministries Case, 14 Trials of War Criminals Before the

Nuernberg Military Tribunals Under Control Council Law No. 10 621-22.

112
 Nevertheless, plaintiffs suggest that the relevant inquiry is not whether the relationship

between the State and the private actor is of a commercial nature, but rather – much like aiding-and-

abetting under the substantial assistance prong of section 876 of the Restatement (Second) of Torts

– whether the private actor had sufficient prior knowledge of the government’s plans to commit

violations of international law and of the ways in which the goods or services procured by the

government would aid in those illicit actions that the private actor can be said to have become party

to the government’s actions. The issue, however, is not whether the private actor has become “a

party to the government’s actions” – whatever that means. Instead, it is whether by engaging in this

conduct he acted “under color of law.” Adickes, 398 U.S. at 150; Ginsberg, 189 F.3d at 273. A

holding that he does, in the circumstances described in the complaints, strains both the phrase and

the concept to the breaking point.

Indeed, notwithstanding the division on the panel with respect to other issues, the per curiam

opinion, reflecting the agreement of all members of the panel, holds that the allegations of criminal

facilitation alleged in the complaint are insufficient to render the defendants liable under the color

of law for violation of the TVPA. Specifically, it rejects the cause of action under the TVPA,

alleging that the defendants “aided and abetted the apartheid regime’s subjecting the Plaintiffs to

torture and extra judicial killing within the meaning of the [TVPA] . . . under actual or apparent

authority, or under color of law.” A-00505. Applying principles of liability under section 1983, the

per curiam opinion properly concludes that the complaint “failed to link any defendants to state aid

or the conduct of state officials.” Per Curiam Op. ante at 8. The same is true with respect to the

other causes of action.

(b) Aiding-and-abetting Liability of Private Parties

113
 Ordinarily, the fact that the defendants’ conduct is not sufficient to hold them liable for acting

under color of law would compel the dismissal of the causes of action alleging various crimes against

humanity that were not committed during war or an armed conflict. See Kadic, 70 F.3d at 244.

Nevertheless, Judge Sprizzo identified two potentially applicable international agreements that

contemplated liability for private actors. The first and most obvious source of such liability is the

Apartheid Convention, which was expressly drafted to cover all of the conduct of the Union of South

Africa that the defendants allegedly aided and abetted. Of this Convention, the Reporters of the

Restatement write:

The Convention [ ] creates obligations beyond those imposed by
customary international law. It attaches personal criminal
responsibility to all individuals who commit, participate in, incite,
abet, encourage or co-operate in the crime. Art. III. The Convention
also requires states to suppress, prevent any encouragement of, and
punish apartheid. Among parties to the Convention, apartheid is also
effectively made a subject of universal jurisdiction. Art. IV.

Restatement (Third) of Foreign Relations Law § 702 reporters’ note 7. Thus, the Reporters

recognize that the Apartheid Convention’s imposition of liability on private parties not acting under

color of law represented a departure from customary international law.9 Moreover, while it may be

an overstatement to describe the list of countries that have ratified the Convention as a rogues’

gallery of human rights violators, the Convention has not been ratified by the United States, most

other mature democracies, or other states that play a significant role in international affairs, including

three of the five permanent members of the United Nations Security Council – the United States, the

9
The Chief Reporter of the Restatement was Professor Louis Henkin of Columbia University.
Associated with Professor Henkin as Reporters were Professors Andreas F. Lowenfeld of New York
University, Louis B. Sohn of the University of Georgia, and Detlev F. Vagts of Harvard University.

114
United Kingdom, and France. Indeed, it has not been ratified by the post-apartheid Republic of

South Africa. Thus, it is not a persuasive source of customary international law on this point. See

Flores v. S. Peru Copper Corp., 414 F.3d 233, 257 (2d Cir. 2003) (Cabranes, J.) (“[T]he more States

that have ratified a treaty, and the greater the relative influence of those States in international affairs,

the greater the treaty’s evidentiary value.”). Dean Koh made a similar point in an analogous context

when he wrote:

[T]hose who advocate the use of international and foreign sources in
U.S. constitutional interpretation [as he does] are not urging U.S.
courts to defer automatically to some kind of global “nose count.”
Instead, they are suggesting that the practices of other mature
democracies – not those that lag behind developmentally – constitute
the most relevant evidence of . . . the “evolving standards of decency
that mark the progress of a maturing society.”

Harold Hongju Koh, International Law as Part of Our Law, 98 Am. J. Int’l L. 43, 56 (2004). Under

these circumstances, for the same reason that the Apartheid Convention is not a source for the

exercise of universal jurisdiction, Judge Sprizzo properly rejected it as a source for extending the

scope of liability to private actors who aided and abetted apartheid in South Africa.

The other international agreement identified by Judge Sprizzo that expressly recognizes

liability for private persons is the Convention on the Prevention and Punishment of the Crime of

Genocide, opened for signature Dec. 9, 1948, 102 Stat. 3045, 78 U.N.T.S. 277 (“Genocide

Convention”). The near-unanimous adoption of the convention constitutes concrete evidence of a

peremptory norm that has become part of customary international law. The problem with the

invocation of the Genocide Convention – aside from the issue whether the complaints here

sufficiently allege activity constituting genocide – is that when Congress enacted legislation to

115
implement the agreement, see Genocide Convention Implementation Act of 1987, Pub. L. 100-606,

Nov. 4, 1988, 102 Stat. 3045, codified at 18 U.S.C. § 1091 et seq., it expressly provided that the Act

shall not “be construed as creating any substantive or procedural right enforceable by law by any

party in any proceeding.” 18 U.S.C. § 1092.

In Kadic, which was decided prior to Sosa, we held that “the legislative decision not to create

a new private remedy does not imply that a private remedy is not already available under the Alien

Tort Act.” 70 F.3d at 242. As Judge Newman explained:

Nothing in the Genocide Convention Implementation Act or its
legislative history reveals an intent by Congress to repeal the Alien
Tort Act insofar as it applies to genocide, and the two statutes are
surely not repugnant to each other. Under these circumstances, it
would be improper to construe the Genocide Convention
Implementation Act as repealing the Alien Tort Act by implication.

Id.

After Sosa, however, the issue for us to decide is not whether the Genocide Convention

Implementation Act implicitly repealed the ATCA insofar as it applies to genocide. As Judge

Katzmann acknowledges, “Sosa . . . deviate[d] from our case law in one crucial respect. We had

allowed cases to proceed under the ATCA on the assumption that when plaintiffs alleged violations

of well-established international law, their ‘causes of action are statutorily authorized.’” Op. of Judge

Katzmann ante at 17 (quoting Kadic, 70 F.3d at 246). Sosa “flatly rejected this notion.” Id.

Instead, we are now required to determine “whether to recognize a common-law cause of action to

provide a remedy for the alleged violation” of international law. Id. ante at 19. Thus, we have to

address the threshold question of whether to recognize a civil cause of action under the ATCA for

116
genocide where Congress has quite plainly indicated its intention that such a cause of action should

not be available.

Judge Katzmann appears to acknowledge that, notwithstanding Kadic, the issue whether to

recognize a cause of action for genocide based on the ATCA must be considered anew in light of

Sosa. Nevertheless, he would “leave it to the district court to address in the first instance should it

undertake to decide whether to recognize a cause of action.” Op. of Judge Katzmann ante at 51

n.32. While leaving it to Judge Sprizzo to resolve the issue, however, Judge Katzmann goes on to

suggest to him that such a cause of action should be recognized. Id. Instead of deciding an issue that

we should and must resolve, Judge Katzmann’s disposition of this issue invites a game of ping-pong

with Judge Sprizzo, in which I decline to participate.

The Genocide Convention Implementation Act should shut the door to the recognition of a

cause of action under the ATCA for violation of the international law norm that it implements,

because it would be an abuse of our discretion to expand the scope of ATCA liability to causes of

action that Congress indicated its intent to proscribe. We should not “override a clear indication

from the political branches that a ‘specific, universal, and obligatory’ norm against genocide is not

to be enforced through a private damages action[.]” Sosa, 542 U.S. at 749 (Scalia, J., concurring in

part and concurring in the judgment).

(c) Liability of Corporations

In Sosa, the Supreme Court directed federal courts to determine “whether international law

extends the scope of liability for a violation of a given norm to the perpetrator being sued, if the

defendant is a private actor such as a corporation or an individual.” Id. at 732 n.20. Judge

117
Katzmann, who agrees with me that the issue of the scope of liability is governed by international

law, draws no distinction between a corporation and an individual. Nor do our decisions in cases

invoking jurisdiction under ACTA. Nevertheless, the specific issue of corporate liability under

customary international law was not discussed, nor was it raised by the parties, in any of those cases.

See, e.g., Bigio, 239 F. 3d 440; Flores, 414 F.3d 233. Consequently, they are simply not apposite

here, because “[q]uestions which merely lurk in the record, neither brought to the attention of the

court nor ruled upon, are not to be considered as having been so decided as to constitute precedents.”

Webster v. Fall, 266 U.S. 507, 511 (1925).

There is a significant basis for distinguishing between personal and corporate liability.

Where the private actor is an individual, he is held liable for acts which he has committed and for

which he bears moral responsibility. On the other hand, “legal entities, as legal abstractions can

neither think nor act as human beings, and what is legally ascribed to them is the resulting harm

produced by individual conduct performed in the name or for the benefit of those participating in

them or sharing in their benefits.” M. Cherif Bassiouni, Crimes Against Humanity in International

Criminal Law 378 (2d ed. 1999). Thus, the issue here is whether an artificial entity that is allegedly

used as a vehicle for the commission of a crime against humanity may be held vicariously liable.

The sources evidencing the relevant norms of international law at issue plainly do not

recognize such liability. While the London Charter, which created the International Military

Tribunal (the “IMT”) at Nuremberg, does not explicitly limit its jurisdiction to “natural persons,”

it seems clear from its language and context that it conferred jurisdiction only over the prosecution

of individuals. Specifically, the Nuremberg Tribunal was empowered “to try and punish persons

who, acting in the interests of the European Axis countries, whether as individuals or as members

118
of organizations,” committed any of several enumerated crimes. Charter of the IMT art. 6, Aug. 8,

1945, 59 Stat. 1544, 82 U.N.T.S. 279 (the “London Charter” or the “Charter”) (emphasis added).10

Moreover, in rejecting the defense argument that “international law is concerned with the actions of

sovereign states, and provides no punishment for individuals,” the Nuremberg Tribunal held that

“[c]rimes against international law are committed by men, not by abstract entities, and only by

punishing individuals who commit such crimes can the provisions of international law be enforced.”

The Nuremberg Trial, 6 F.R.D. 69, 110 (1946); 1 Trial of the Major War Criminals 223 (William

S. Hein & Co., Inc. 1995) (1947). See also Ernst Schneeberger, The Responsibility of the Individual

under International Law, 35 Geo. L. J. 481, 489 (1947) (“[I]n the last resort responsibility under

international law can only be responsibility of an individual . . . .”).

Indeed, the distinction between corporate and individual responsibility is illustrated by The

I.G. Farben Case in which corporate officers were charged with “acting through the instrumentality

of Farben” in committing various crimes against humanity. United States v. Krauch (“The I.G.

Farben Case”), 7 Trials of War Criminals Before the Nuernberg Military Tribunals Under Control

Council Law No. 10 14, 39, 50, 59 (1952). Nevertheless, the notorious I.G. Farben was not named

as a defendant. Id. at 11-14. Nor were any other corporations charged with crimes at Nuremberg.

“In fact, in the Nuremberg trials, this point of lack of corporate liability appeared to have been

10
Article 9 of the London Charter provided that “[a]t the trial of any individual member of any
group or organization the Tribunal may declare (in connection with any act of which the individual may
be convicted) that the group or organization of which the individual was a member was a criminal
organization.” 1 Trial of Major War Criminals 255-57. The consequence of a finding in the main trial
that organizations, such as the Schultzstaffeln (more commonly referred to as the S.S.) and the Gestapo,
were criminal in nature was that it permitted a finding of guilt against other members of the organization
in subsequent trials. See The Nuremberg Trial, 6 F.R.D. at 131-33.

119
explicitly stated.” In re Agent Orange Product Liab. Litig., 373 F. Supp. 2d 7, 57 (E.D.N.Y. 2005)

(Weinstein, J.).

In 1948, the United Nations General Assembly asked the International Law Commission (the

“ILC”), a United Nations body, to study the possibility of creating an international judicial tribunal,

which ultimately emerged in the form of the International Criminal Court (the “ICC”), to prosecute

genocide and other crimes. Under the auspices of the ILC, the Committee on International Legal

Jurisdiction began to study the issue in 1951, and it released its report in 1953. See U.N. GAOR, 9th

Sess., Supp. No. 12, U.N. Doc. A/2645 (1954). The Committee considered a proposal by Australia

to grant the international court jurisdiction over corporations. Id. ¶ 85. The proposal was soundly

defeated because “it was undesirable to include so novel a principle as corporate criminal

responsibility in the draft statute.” Id.

More than forty years later, during negotiations for the Rome Statute, pursuant to which the

ICC was created, France proposed bringing corporations and other juridical persons (though not

States) within the jurisdiction of the ICC. That proposal was again rejected, for three principal

reasons: (1) “from a pragmatic point of view it was feared that the ICC would be faced with

tremendous evidentiary problems when prosecuting legal entities”; (2) “from a more normative-

political point of view it was emphasized that the criminal liability of corporations is still rejected

in many national legal orders, and international disparity which could not be brought in concord with

the principle of complementarity”; and (3) “it was felt morally obtuse for States to insist on the

criminal responsibility of all entities other than themselves.” Albin Eser, Individual Criminal

Responsibility, in 1 Antonio Cassese et al., The Rome Statute of the International Criminal Court:

A Commentary 767, 778-79 (2002) (footnotes omitted). Thus, the Rome Statute provides for

120
jurisdiction over only “natural persons.” The Rome Statute of the ICC art. 25(1), opened for

signature July 17, 1998, 37 I.L.M. 999, 1016 (entered into force July 1, 2002) (“The Rome Statute”).

Similarly, Article III of the Apartheid Convention provides that “[i]nternational criminal

responsibility shall apply, irrespective of the motive involved, to individuals, members of

organizations and institutions and representatives of the State . . . .” Article 4 of the Genocide

Convention is to the same effect. Specifically, it provides that “[p]ersons committing genocide or

any of the other acts enumerated in [a]rticle 3 shall be punished, whether they are constitutionally

responsible rulers, public officials or private individuals.” The statutes for the International

Criminal Tribunal for the former Yugoslavia (“ICTY”) and International Criminal Tribunal for

Rwanda (“ICTR”) also confer jurisdiction only over “natural persons.” See Statute of the

International Tribunal for the Prosecution of Persons Responsible for Serious Violations of

International Humanitarian Law Committed in the Territory of the Former Yugoslavia Since 1991

art. 25(1), adopted May 25, 1993, S.C. Res. 827, U.N. Doc. S/RES/827 (the “ICTY Statute”); Statute

of the International Criminal Tribunal for the Prosecution of Persons Responsible for the Genocide

and Other Serious Violations of International Humanitarian Law Committed in the Territory of

Rwanda and Rwandan Citizens Responsible for Genocide and Other Such Violations Committed

in the Territory of Neighboring States Between 1 January 1994 and 31 December 1994 art. 5,

adopted Nov. 8, 1994, S.C. Res. 955, U.N. Doc. S/RES/955 (the “ICTR Statute”).

Likewise, the Convention Against Torture and Other Cruel, Inhuman, or Degrading

Treatment or Punishment (“Torture Convention”), contemplates violations by “persons,” which in

the agreement’s textual context can only mean natural persons. See Torture Convention arts. 4(1),

6(1), 6(3), adopted Dec. 10, 1984, 1465 U.N.T.S. 85, 23 I.L.M. 1027. The language of the TVPA,

121
which executed in part the Torture Convention, Flores, 414 F.3d at 247 n.20, provides further

support for the fact that customary international law has not recognized corporate liability. Under

the TVPA, the term “individual” describes both those who can violate its proscriptions against

torture, as well as those who can be victims of torture. Specifically, the TVPA provides that “[a]n

individual who . . . subjects an individual to torture shall . . . be liable for damages to that

individual,” and it defines “torture” as “any act, directed against an individual . . . by which severe

pain or suffering . . . is intentionally inflicted on that individual.” Id. §§ 2(a)(1), 3(b)(1) (emphasis

added). As Judge Weinstein recently wrote:

[b]oth from context and common sense only natural persons can be
the “individual” victims of acts that inflict “severe pain and
suffering.” See [TVPA § 3(b)1]. Because the TVPA uses the same
term “individual” to identify offenders, the definition of “individual”
within the statute appears to refer to a human being, suggesting that
only natural persons can violate the Act. See Desert Palace, Inc. v.
Costa, 539 U.S. 90, 101, 123 S.Ct. 2148, 156 L.Ed.2d 84 (2003)
(noting that “[a]bsent some congressional indication to the contrary,
[courts] decline to give the same term in the same Act a different
meaning depending on whether the rights of the plaintiff or the
defendant are at issue”); see also Beanal v. Freeport-McMoRan, Inc.,
969 F.Supp. 362, 381-82 (E.D. La. 1997) (“[T]he plain meaning of
the term ‘individual’ does not ordinarily include a corporation.”),
aff’d by 197 F.3d 161 (5th Cir. 1999).

In re Agent Orange Product Liab. Litig., 373 F. Supp. 2d at 56.

The same is true with respect to the statute making torture committed outside the United

States a federal offense. 18 U.S.C. § 2340 et seq. Section 2340 provides that, “[a]s used in this

chapter . . . ‘torture’ means an act committed by a person acting under the color of law specifically

intended to inflict severe physical or mental pain or suffering . . . upon another person within his

custody or physical control . . . .” Id. § 2340(1). Because here, as in the TVPA, the context

122
indicates that the word “person” refers to natural persons, the general rule that the word “person”

when used in any Act of Congress includes corporations, unless the context indicates otherwise, 1

U.S.C. §1, is not applicable here. See Rowland v. Cal. Men’s Colony, Unit II Men’s Advisory

Council, 506 U.S. 194, 199 (1993) (“‘Context’ here means the text of the Act of Congress

surrounding the word at issue, or the texts of other related congressional Acts . . . .”).

These sources are not undermined by Professor Henkin’s suggestion that, “[a]t this

juncture[,] the Universal Declaration [of Human Rights (G.A. Res. 217A (III), U.N. Doc. A/180

(1948)] may also address multinational companies.” Louis Henkin, The Universal Declaration at

50 and the Challenge of Global Markets, 25 Brook. J. Int’l L. 17, 24 (1999). Relying on the

Preamble to the Declaration, Professor Henkin argued that it “is ‘a common standard for all people

and all nations.’ It means that ‘every individual and every organ of society shall strive – by

progressive measures . . . to secure their universal and effective recognition and observance among

the people of member states.’ Every individual includes juridical persons.” Id. at 24-25. The

Supreme Court, however, has held that the Declaration “does not of its own force impose obligations

as a matter of international law.” Sosa, 542 U.S. at 734. Moreover,

it is [also] important to keep in mind what exactly the Preamble expects of such
individuals and organs: that they “promote” respect for the rights set forth in the
Declaration by ‘teaching and education’ and by supporting “progressive national and
international measures.” The language is thus consistent with the idea that legal
obligations bind corporations only to the extent that further “national and
international measures” are taken.

Carlos M. Vázquez, Direct vs. Indirect Obligations of Corporations Under International Law, 43

Colum. J. Transnat’l L. 927, 942 (2005).

123
 Nor are the sources rejecting corporate liability undermined by two international conventions

and a Security Council resolution, all of which post-date the dismantling of apartheid. See The

United Nations Convention Against Transnational Organized Crime art. 5(1)(b), Jan. 8, 2001, U.N.

GAOR, 55th sess., Supp. No. 49, U.N. Doc. A/45/49; Convention on Combating Bribery of Foreign

Public Officials in International Business Transactions arts. 1(2) & 2, Dec. 18, 1997, 37 I.L.M. 1

(1998); S.C. Res. 1566, U.N. Doc. S/RES/1566 (Oct. 8, 2004); Mandatory Action to Fight

Terrorism, S.C. Res. 1373 ¶ 1(d), U.N. Doc. S/RES/1373 (Sept. 28, 2001); International Convention

for the Suppression of the Financing of Terrorism, adopted Dec. 9, 1999, G.A. Res. 109, U.N.

GAOR 54th Sess., 76th plen. mtg., U.N. Doc. A/RES/54/109, 39 I.L.M. 270 (2000). While each of

these sources call upon parties to enact measures to impose some kind of liability on legal persons

in accordance with their own domestic legal principles, none of them directly impose any liability

on corporations. Without an analysis of the laws enacted by the various signatories, it is impossible

to draw any conclusions regarding the practice of states with regard to the issue of the degree of

corporate liability for violation of international law norms.11 Indeed, as Professor Bassiouni has

observed, although “contemporary international efforts to deal with organized crime, corruption, and

drug trafficking” are moving in the direction of corporate liability “these new concepts of corporate

criminal responsibility have not yet found their way into [customary international law].” Bassiouni,

supra, at 377 (internal quotation marks and footnotes omitted). See also Steven R. Ratner & Jason

S. Abrams, Accountability for Human Rights Atrocities in International Law: Beyond the Nuremberg

11
As a general rule, treaties, like contracts, “are legally binding on States that become parties to
them by consenting to be bound”, and may constitute evidence of a norm of customary international law
only if “an overwhelming majority of States have ratified the treaty and those States uniformly and
consistently act in accordance with its principles.” Flores, 414 F.3d at 256.

124
Legacy 16 (2d ed. 2001) (“It remains unclear [ ] whether international law generally imposes

criminal responsibility on groups and organizations.”).

In sum, the issue here is not whether policy considerations favor (or disfavor) corporate

responsibility for violations of international law. Cf. Carlos M. Vásquez, Direct v. Indirect

Obligations of Corporations Under International Law, 43 Colum. J. Transnat’l L. 927, 932-959

(2005) (acknowledging that very few norms apply directly to corporations under “international law

as it exists today” and outlining arguments for why “international law should move in the direction

of generally extending human rights obligations of states to private corporations”). Instead, it

involves a determination of what the law was during the relevant period. Indeed, the Supreme Court

has held that the retroactive application of statutes enacting civil liability is presumptively

inappropriate even where the conduct giving rise to liability was already proscribed and the newly

enacted legislation simply increased the scope of liability for damages. Landgraf v. USI Film Prods.,

511 U.S. 244, 281-86 (1994).

“[T]he presumption against retroactive legislation is [not only] deeply rooted in our

jurisprudence, and embodies a legal doctrine centuries older than our Republic,” id. at 265, but is

also reflected in customary international law. As one scholar has noted, “[a] further traditional

corollary of the nullum crimen sine lege principle – namely, the notion of non-retroactivity – views

the principle of legality from a temporal perspective: conduct may be punished only on the basis of

a norm that came into force prior to when the conduct occurred.” Susan Lamb, Nullum Crimen,

Nulla Poena Sine Lege in International Criminal Law, in Cassese et al., supra, at 733, 751

(emphasis in the original). Indeed, “[t]he applicability of the nullum crimen principle to serious

breaches of international humanitarian law was by 1998 sufficiently well-accepted that its inclusion

125
within the Rome Statute was seen as necessary . . . .” Id. at 755. See also Prosecutor v. Tadiƒ, 36

I.L.M. 908, 945, Case No. IT-94-1-T, Opinion and Judgment ¶ 654 (Trial Chamber May 7, 1997)

(the International Tribunal “must apply customary international law as it stood at the time of the

offences”).

Although the precise date on which the heart of the apartheid regime stopped beating may

be difficult to pinpoint, the executive order by President George H.W. Bush on July 10, 1991,

terminating United States sanctions against the country, provides one guidepost. Exec. Order No.

12,769, 56 Fed. Reg. 31855 (July 10, 1991), reprinted in 22 U.S.C. § 5061 (2004). By that order,

the President concluded as of that date that the Government of South Africa had: (1) released all

persons persecuted for their political beliefs or detained unduly without trial and Nelson Mandela

from prison; (2) repealed the state of emergency in effect on the date of enactment of [the

“Comprehensive Anti-Apartheid Act of 1986”] and released all detainees held under such state of

emergency; (3) unbanned democratic political parties and permitted the free exercise by South

Africans of all races of the right to form political parties, express political opinions, and otherwise

participate in the political process; (4) repealed the Group Areas Act and the Population Registration

Act and instituted no other measures with the same purposes; and (5) agreed to enter into good faith

negotiations with truly representative members of the black majority without preconditions. See

Comprehensive Anti-Apartheid Act of 1986, Pub. L. No. 99-440 § 311(a) (1986). Because the only

sources of customary international law that suggest some movement toward the recognition of

corporate liability post-date the collapse of the apartheid regime, and because the established norm

during the apartheid era was that corporations were not responsible legally for violations of norms

126
proscribing crimes against humanity, the complaints are subject to dismissal on this ground alone.

3. The Concurring Opinions

(a) Judge Hall’s Concurring Opinion

Judge Hall flatly ignores the holding of the Supreme Court that the second consideration in

deciding whether to accept jurisdiction over a cause of action alleging a violation of the law of

nations “is whether international law extends the scope of liability for a violation of a given norm

to the perpetrator being sued, if the defendant is a private actor such as a corporation or individual.”

Sosa, 542 U.S. at 732 n.20. Instead of undertaking an analysis of a given norm of international law

to determine the scope of liability of a private actor, Judge Hall concludes that international law is

irrelevant and that once a violation of a given norm is alleged, ATCA subjects a private party to

liability if he aided-and-abetted that violation.

Judge Hall’s concurring opinion is premised on the assumption that, even though the ATCA

does not by its terms encompass aiding-and-abetting liability, it should be construed as if it contains

such language. This aspect of Judge Hall’s opinion not only contradicts Sosa and Kadic, it does

disservice to the holding of the leading Supreme Court case addressing the issue of accessorial

liability for violations of statutes that create civil causes of action, Central Bank of Denver v. First

Interstate Bank of Denver, 511 U.S. 164 (1994). In that case, the Supreme Court observed that:

Congress has not enacted a general civil aiding and abetting statute –
either for suits by the Government (when the Government sues for
civil penalties or injunctive relief) or for suits by private parties.
Thus, when Congress enacts a statute under which a person may sue
and recover damages from a private defendant for the defendant’s
violation of some statutory norm, there is no general presumption that
the plaintiff may also sue aiders and abettors.

127
Id. at 182. Central Bank, to be sure, only stands for the proposition that when Congress enacts a law

creating a private right of action for violation of a statute, “there is no general presumption that the

plaintiff may also sue aiders and abettors.” Id. That decision, however, does not preclude civil

liability for aiding-and-abetting in cases “where Congress’ intent is clear from the language and

structure of the statute itself as well as from the legislative history.” Boim, 291 F.3d at 1019.

Boim, upon which Judge Hall relies, concerns the issue of liability for aiding-and-abetting

international terrorism in violation of 18 U.S.C. § 2333(a). While the statute, creating the civil cause

of action, 18 U.S.C. § 2339A, did not in haec verba provide for aiding-and-abetting liability,

accomplice liability was implicit in the definition of international terrorism. Thus, in distinguishing

Central Bank, the Seventh Circuit held that:

Unlike section 10(b) [the statute at issue in Central Bank], Congress
also expressed an intent in section 2333 to make civil liability at least
as extensive as criminal liability. The statute defining “international
terrorism” includes activities that “involve violent acts or acts
dangerous to human life that are a violation of the criminal laws of
the United States or of any State, or that would be a criminal violation
if committed within the jurisdiction of the United States or of any
State.” 18 U.S.C. § 2331(1). This language, embracing activities that
“involve” violent acts, taken at face value would certainly cover
aiding and abetting violent acts. Remember, too, the criminal laws
include 18 U.S.C. § 2, which creates liability for aiding and abetting
violations of any other criminal provisions. By incorporating
violations of any criminal laws that involve violent acts or acts
dangerous to human life, Congress was expressly including aiding
and abetting to the extent that aiding and abetting “involves”
violence.

Id. at 1020.

The language of the ATCA, like the language of the statute at issue in Boim, can support the

recognition of a cause of action for aiding-and-abetting. Nevertheless, it does so only if the plaintiff

128
invokes an international law norm that provides for such liability. Nothing in the language of the

ATCA supports the broad holding that Congress intended to provide a forum to adjudicate causes

of action for aiding-and-abetting the violation of a norm when such conduct did not fall within the

scope of the norm. Indeed, the same Congress that enacted the ATCA, without reference to aiding-

and-abetting liability, explicitly made it a crime to aid-and-abet acts of piracy, a violation of the law

of nations. See Act of April 30, 1790, ch. 9, § 10, 1 Stat. 112, 114 (1790). This shows that, despite

Judge Hall’s suggestion to the contrary, Op. of Judge Hall ante at 59 n.37, the First Congress “knew

how to impose aiding and abetting liability when it chose to do so,” id. (quoting Central Bank, 511

U.S. at 174), with respect to violations of the law of nations.

In the absence of either statutory language or legislative history to support his interpretation

of the ATCA, Judge Hall’s opinion argues that “the Founding Generation nevertheless understood

that ATCA encompassed aiding and abetting liability.” Op. of Judge Hall ante at 59 n.37. This

contention is beside the point, for the reasons discussed above, and it is also wrong. In Central

Bank, the plaintiff argued that “Congress legislated with an understanding of general principles of

tort law and that aiding and abetting liability was ‘well established in both civil and criminal actions

by 1934,’” the year the statute at issue was adopted. 511 U.S. at 181 (citation omitted). The Court’s

opinion suggested that the plaintiff there relied on the principle derived from the Restatement

(Second) of Torts (1979), which provides that a person is liable for another’s torts if he “knows that

the other’s conduct constitutes a breach of duty and gives substantial assistance or encouragement

to the other so to conduct himself. . . .” Id. § 876(b); see Central Bank, 511 U.S. at 181. The

principle of “substantial assistance,” which Judge Hall adopts, initially appeared in the Restatement

(First) of Torts in 1939, of which Prosser wrote:

129
 The form of the Restatement is perhaps unfortunate, in that it seeks
to reduce the law to a definite set of black-letter rules or principles,
ignoring all contrary authority – since the law of torts in its present
stage of development does not lend itself at all readily to such
treatment. There is room for suspicion that the courts have tended to
cite the Restatement when they are already in agreement with it, and
to ignore it when they are not, so that the impressive list of references
to it in the cases may be somewhat misleading; and there are those
who have disagreed with many of its conclusions, and even
denounced the whole project.

William L. Prosser, Law of Torts § 4, at 20-21 (4th ed. 1971).12 Perhaps for this reason, the Supreme

Court could write fifty-five years after the publication of the first Restatement that the doctrine of

aiding-and-abetting in civil cases “has been at best uncertain in application . . . with the common-law

precedents ‘largely confined to isolated acts of adolescents in rural society.’” Central Bank, 511

U.S. at 181 (quoting Halberstam, 705 F.2d at 489). If there was no general understanding that civil

aiding-and-abetting liability was “well established in . . . civil . . . actions by 1934,” id. at 181

(citation omitted), it is difficult to understand how it can be asserted that such an understanding was

present in 1789 when Congress enacted the ATCA.

The historical support for the proposition that “[t]he Founding Generation nevertheless

understood that civil liability for aiding and abetting international law violations was contemplated

under the ATCA,” Op. of Judge Hall ante at 59 n.37, is ambiguous at best. The lynchpin for this

12
Judge Learned Hand shared Prosser’s concern. Hand “repeatedly . . . urged that the
Restatements confine themselves to articulating the law ‘as it is,’ not as it should be; repeatedly, he
insisted that the ALI’s task was to ‘restate, not legislate’ . . . .” Gerald Gunther, Learned Hand: The Man
and the Judge 411 (1994) (internal citations omitted). Indeed, although Hand was a founding member of
the ALI and held major positions in it for the rest of his life, he rarely cited the Restatements. Id. at 413.
The most that he was willing to say about them was that he found them to be useful “ ‘on questions
which were not controversial and [otherwise took] a long time to look up.’ ” Id. (internal citations
omitted).

130
argument is Attorney General William Bradford’s 1795 opinion, Breach of Neutrality, 1 Op. Atty.

Gen. 57 (1795). In his opinion, Bradford specifically addressed whether American citizens who

“voluntarily joined, conducted, aided, and abetted a French fleet in attacking the settlement, and

plundering or destroying the property of British subjects on that coast” could be subject to criminal

prosecution in the United States. Id. at 58. Bradford expressed some doubt about a criminal

prosecution, but he stated that

there can be no doubt that the company or individuals who have been injured by
these acts of hostility have a remedy by a civil suit in the courts of the United States;
jurisdiction being expressly given to these courts in all cases where an alien sues for
a tort only, in violation of the law of nations, or a treaty of the United States . . . .

Id. at 59.

Because Bradford did not distinguish between primary and secondary liability, it is not

possible to discern that, when he said a cause of action would lie for “these acts of hostility,” he was

focusing on aiding-and-abetting as opposed to direct participation in the conduct that violated a

treaty of the United States. Indeed, because the conduct involved direct participation by American

citizens, who acted with the intent to make the attack succeed, it seems likely that Bradford

recognized all of the perpetrators as joint tortfeasors, as that term was understood at the time. As

the D.C. Circuit explained in Halberstam:

Prosser notes that “[t]he original meaning of ‘joint tort’ was that of
vicarious liability for concerted action. All persons who acted in
concert to commit a trespass, in pursuance of a common design, were
held liable for the entire result.” W. Prosser, Law of Torts § 46, at
291 (4th ed. 1971). His illustration portrays a standard situation that
involved this “joint tort”: combined action by tortfeasors on the scene
together – “one might have battered the plaintiff, while another
imprisoned him, and a third stole his silver buttons.” Id. (footnote
omitted). Each was responsible for the others’ actions.

131
705 F.2d at 476-77. Moreover, Prosser goes on to explain that the essential element of a joint tort

was “the pursuance of a common plan or design to commit a tortious act.” Prosser, supra, at 292.

Two early cases cited in Judge Hall’s opinion, Talbot v. Jansen, 3 U.S. (3 Dall.) 133 (1795),

and Henfield’s Case, 11 F. Cas. 1099 (C.C.D. Pa. 1793), neither of which involved the ATCA, bear

out Prosser’s analysis. Both cases arose from the seizure of a vessel not by pirates, but by privateers,

“armed vessels that are owned, equipped and officered by one or more private persons, but sailing

under a commission, usually called letters of marque, from a belligerent state, which empowers the

person or persons to whom it is granted to attack and seize, at sea, vessels or other property of its

enemy.” 3 James Fairbanks Conley, Privateering, in Cyclopædia of Political Science, Political

Economy, and the Political History of the United States 361 (John J. Lalor ed., 1899). The use of

privateers in this way was recognized under international law. Id. Nevertheless, Talbot’s and

Henfield’s legal problems in these cases derived from the fact that they were United States citizens

engaged in acts of war against nations with which the United States was at peace.

Talbot was an admiralty case. Talbot himself was the captain of a privateer commissioned

by the Republic of France. Henfield’s Case was a criminal prosecution, and a most unusual one at

that, of the prize-master of a privateer also commissioned by France. Henfield was not charged with

conduct proscribed by an act of Congress. The undelivered grand jury charge of Chief Justice Jay,

upon which Judge Hall relies, Op. of Judge Hall ante at 59 n.37, appears to reflect the practice of

most of the early justices “to create crimes by judicial fiat[, which] was particularly infuriating to the

vast majority of citizens who believed the Constitution had established a government of limited

powers.” Jean Edward Smith, John Marshall: Definer of a Nation 284 (1996). The Supreme Court

132
repudiated the practice in United States v. Hudson, 11 U.S. (7 Cranch) 32 (1812). Under these

circumstances, it is not surprising that Henfield, who was not charged with aiding-and-abetting, was

acquitted because “he was ignorant of the unlawfulness of his undertaking.” Henfield’s Case, 11 F.

Cas. at 1122 n.7. Nor is it surprising, given Henfield’s role as a principal (or as a joint tortfeasor in

the civil context), that Jay’s grand jury charge, which was written to apply “to the class of offenders,

of whom Henfield was one,” id. at 1105 n.2, did not define aiding-and-abetting, and that the judge

who actually charged the Henfield grand jury did not allude to the concept altogether. Id. at 1105-09.

In sum, Bradford’s opinion and the cases cited by Judge Hall do not support the extraordinary

proposition that Congress intended the ATCA to permit jurisdiction to be exercised over claims of

aiding-and-abetting without regard to whether the conduct at issue violated an international law

norm. Moreover, Judge Hall compounds his flawed discussion of this issue by adopting a standard

for aiding-and-abetting that is vague and inappropriate in the present context. As if the language of

section 876(b) of the Restatement (Second) of Torts imposing liability for “substantial assistance”

was not vague enough, he endorses a five-factor test, first suggested by the Restatement and then

adopted in Halberstam, to “assess whether the defendant’s encouragement or assistance was

sufficiently substantial to support liability.” Op. of Judge Hall ante at 58. These factors are “the

nature of the act encouraged, the amount of assistance given by the defendant, his presence or

absence at the time of the tort, his relation to the other tortfeasor and his state of mind.” Id.

The “state of mind” factor, as applied in Halberstam, includes a “desire to make the venture

succeed.” 705 F.2d at 488. Nevertheless, despite his reliance on Halberstam, Judge Hall’s opinion

fails to even allude to this critical state-of-mind element. Moreover, even under Halberstam’s

formulation, it is only one factor among five for determining whether “substantial assistance” was

133
provided. Such a five-prong test for determining whether assistance was substantial hardly provides

the clear guidance necessary to those engaging in commercial transactions.

By incorporating a vague “substantial assistance” standard, this newly minted theory of

aiding-and-abetting liability will create many practical problems harmful to the political and

economic interests of the United States. As the United States observes in its amicus brief, the

decision to embrace this broader scope of liability under the ATCA will generate tremendous

uncertainty for private corporations, who will be reluctant to operate in countries with poor human

rights records for fear of incurring legal liability for those regimes’ bad acts. Br. for the U.S.A. as

Amicus Curiae at 13. This uncertainty, in turn, will undermine efforts by the United States to

encourage reform in those countries through active economic engagement, id., and will deter the free

flow of trade and investment more generally. Id. at 18.

(b) Judge Katzmann’s Concurring Opinion

Unlike Judge Hall, Judge Katzmann looks to international law to resolve the issue of whether

a private party or a corporation (he draws no distinction between the two) can be held liable for

aiding-and-abetting a violation of international law otherwise applicable to state actors or to private

parties acting under color of law. While Judge Katzmann properly looks to international law, he

disregards the holding in Sosa and our own holdings in Kadic and Wiwa, which I have already

discussed at some length, ante at 101-02, that require a norm-by-norm analysis to determine whether

“international law extends the scope of liability for a violation of a given norm to the perpetrator

being sued, if the defendant is a private actor such as a corporation or individual.” Sosa, 542 U.S.

at 732 n.20 (emphasis added). Judge Katzmann declines to undertake this analysis because that “is

134
not how the inquiry is undertaken by international tribunals whose jurisdiction is limited by

customary international law.” Op. of Judge Katzmann ante at 43.

Contrary to Judge Katzmann’s suggestion, the jurisdiction of the international tribunals – he

relies principally on the ICTY – are conferred by the Security Council resolutions creating them –

or in the case of the ICC by the Treaty of Rome. Indeed, in one significant respect the statute

creating the ICTY is inconsistent with the evolving standards of customary international law. Post

at 140. C.f. Rome Statute art. 25(3)(e) (expanding the liability for genocide beyond that provided

for other crimes against humanity). More significantly, in the cases on which Judge Katzmann relies

– those of the ICTY – the tribunal was not required to make any inquiry regarding the issue of

whether “international law extends the scope of liability for a given norm to the perpetrator being

sued if the defendant is a private actor such as a corporation.” Sosa, 542 U.S. at 732 n.20.

The Statute of the ICTY, article 5, followed the London Charter, the Judgment of the

Nuremberg Tribunal, construing it, and the tribunals empaneled pursuant to CCL 10, all of which

required a connection between crimes against humanity and war crimes.13 Because private parties

are individually responsible for crimes committed in the course of a war, this connection made it

unnecessary for the CCL 10 tribunals to address the scope of liability of private actors as aiders-and-

abetters. See Kadic, 70 F.3d at 239-41; see also id. at 242-43 (“Plaintiffs also contend that the acts

of murder, rape, torture, and arbitrary detention of civilians, committed in the course of hostilities,

13
See The Nuremberg Trial, 6 F.R.D. 69, 131, 1 Trial of the Major War Criminals 253-54,
Brigadier General Telford Taylor, Final Report to the Secretary of the Army on the Nuerenberg War
Crimes Trials under Control Council Law No. 10, app. B at 224, 228 (William S. Hein & Co., Inc. 1997)
(1949). A useful discussion of art. 6(c) of the London Charter, which provided the basis for these
holdings, may be found in Bassiouni, supra, 19-30.

135
violate the law of war. Atrocities of the types alleged here have long been recognized in

international law as violations of the law of war . . . . The liability of private individuals for

committing war crimes has been recognized since World War I and was confirmed at Nuremberg

after World War II.”). Thus, ICTY cases provide no support for failing to follow the instruction of

the Supreme Court, one that is consistent with our own holding in Kadic and Wiwa, which requires

an analysis of the particular norm the defendant is accused of violating to determine whether a

private party may be held responsible as an aider-and-abettor.14

While Judge Katzmann erroneously concludes that there was – during the period when the

crimes alleged here took place – an independent norm of customary international law, making private

actors legally responsible as aiders-and-abetters without regard to whether the particular norm they

allegedly violated imposed such liability, he correctly rejects the “substantial assistance with

knowledge” standard for this newly minted norm that Judge Hall finds in section 876(b) of the

Restatement (Second) of Torts. Instead, he “conclude[s] that a defendant may be held liable under

international law for aiding and abetting the violation of that law by another when the defendant (1)

provides practical assistance to the principal which has a substantial effect on the perpetration of the

crime, and (2) does so with the purpose of facilitating the commission of that crime.” Op. of Judge

Katzmann ante at 38-39. The basis for this formulation is article 25 of the Rome Statute. Rome

Statute of the ICC, opened for signature July 17, 1998, 37 I.L.M. 999, 1016 (entered into force July

1, 2002).

14
Another reason the ICTY cases are silent on this score is that, while the defendants in these
cases may not have been officials of formally recognized states, they were state actors. As we have held,
“the state action concept . . . requires merely the semblance of official authority. The inquiry, after all, is
whether a person purporting to wield official power has exceeded internationally recognized standards of
civilized conduct, not whether statehood in all its formal aspects exists.” Kadic, 70 F.3d at 245.

136
 Specifically, article 25 of the Rome Statute provides that a person shall be criminally

responsible and liable for punishment for specified crimes against humanity if that person:

(c) For the purpose of facilitating the commission of such a crime,
aids, abets or otherwise assists in its commission or its attempted
commission, including providing the means for its commission;

(d) In any other way contributes to the commission or attempted
commission of such a crime by a group of persons acting with a
common purpose. Such contribution shall be intentional and shall
either:

(i) Be made with the aim of furthering the criminal activity
or criminal purpose of the group, where such activity or
purpose involves the commission of a crime within the
jurisdiction of the Court; or

(ii) Be made in the knowledge of the intention of the group
to commit the crime;

Id. art. 25(3)(c), (d).

This article is significant because it makes clear that, other than assistance rendered to the

commission of a crime by a group of persons acting with a common purpose, a defendant is guilty

of aiding-and-abetting the commission of a crime only if he does so “[f]or the purpose of facilitating

the commission of such a crime . . . including providing the means for its commission.” Id. The

same standard was adopted by the United Nations Transitional Administration in East Timor

(“UNTAEAT”). See UNTAET, June 6, 2000, Reg. No. 2000/15 § 14.3(c). As one commentator has

observed:

With regard to facilitating the commission of a crime, the aider and
abettor must act with ‘purpose.’ . . . This means more than the mere
knowledge that the accomplice aids the commission of the offence,

137
 as would suffice for complicity according to the ICTR and ICTY
Statutes, rather he must know as well as wish that his assistance shall
facilitate the commission of the crime.

Albin Eser, Individual Criminal Responsibility, in 1 The Rome Statute of the International Criminal

Court 767, 801 (Antonio Cassese et al., eds. 2002) (emphasis in original).

The Rome Statute has been signed by 139 countries and ratified by 105, including most of

the mature democracies of the world. I agree that it reflects an international consensus on the issue

of the appropriate standard for determining liability for aiding-and-abetting, it is consistent with our

own domestic law, see ante at 111-12, and addresses the concern over the adoption of a “substantial

assistance with knowledge” standard raised in the amicus brief filed by the United States. Br. for

the U.S.A. as Amicus Curiae at 13. Indeed, the amicus brief appears to endorse the elements set out

in article 25(3). Id. at 26 (arguing that “the standard the plaintiffs propose differs materially from

the most recent formulations adopted in international practice”) (citing article 25(3) of the Rome

Statute). Perhaps more significantly, the standard in article 25(3)(c) of the Rome Statute is

consistent with the Ministries Case, discussed at the outset, that holds that the conduct which the

defendants allegedly engaged in here was not a violation of customary international law. The

Ministries Case, 14 Trials of War Criminals Before the Nuernberg Military Tribunals Under Control

Council Law No. 10 622. These considerations also obviate any concern regarding the failure of the

United States to ratify the Treaty of Rome for reasons unrelated to the definition of aiding-and-

abetting. See Yousef, 327 F.3d at 92 n.25.

My point of disagreement with Judge Katzmann relates to the narrow issue of whether there

was any “‘well established[] [and] universally recognized” definition of aiding-and-abetting

sufficient to be considered customary international law for the purposes of the ATCA during the

138
apartheid era when the defendants allegedly violated customary international law. Op. of Judge

Katzmann ante at 39 (quoting Kadic, 70 F.3d at 239). Nevertheless, I concur in section II.B of his

opinion that articulates the customary international law standard for aiding-and-abetting based on

the Rome Statute. I do so because it provides a clear standard, adopted by a majority of the panel,

for Judge Sprizzo to apply, in deciding whether to grant the plaintiffs’ motion to file amended

complaints.15 Moreover, as applied to the facts in this case, the standard Judge Katzmann adopts is

consistent with the Ministries Case – the validity of which has never been questioned.

Judge Katzmann’s opinion, however, does not end with section II.B. Thus, while he does

not rely on post-apartheid decisions of the ICTY or the ICTR to support the standard that he

enunciates for determining aiding-and-abetting liability, he does rely on those cases to suggest that

the “substantial assistance with knowledge” standard for aiding-and-abetting may provide a

foundation for future development of the law in this area. I discuss the cases on which he relies to

demonstrate why, contrary to Judge Katzmann’s gratuitous suggestion, they do not provide a reliable

basis for a broader definition than the one proscribed in the Rome Statute.

(a) The ICTY

By way of background, the ICTY was created by United Nations Security Council Resolution

827, which was passed on May 25, 1993, to prosecute individuals responsible for serious violations

of international humanitarian law committed in the former Yugoslavia beginning on January 1, 1991.

Under the ICTY Statute, the tribunal has jurisdiction over four clusters of crimes: grave breaches of

15
In deciding this motion, Judge Sprizzo should also consider applying the pleading standard
enunciated in Bell Atlantic Corp. v. Twombly, 127 S. Ct. 1955 (2007). The standard seems particularly
appropriate to the class actions in this case, which are intended to coerce a settlement rather than provide
the framework for a trial which we all know will never take place.

139
the 1949 Geneva Conventions, violations of the laws or customs of war, genocide, and crimes

against humanity. See ICTY Statute arts. 2-5. Significantly, as previously noted, the ICTY

criminalizes crimes against humanity only “when committed in armed conflict.” While this

connection with armed conflict is consistent with the London Charter and the Judgement of the

Nuremberg Tribunal, ante at 135, the ICTY Statute undermines “subsequent developments de-

coupling ‘crimes against humanity’ from the initiation and waging of an aggressive war . . . .”

Bassiouni, supra, at 194. Indeed, if we applied this retrograde statute here, it would require the

dismissal of the complaints.

Against this backdrop and my earlier discussion of the ICTY statute, I turn to the specific

ICTY decisions relied upon by Judge Katzmann. The gruesome facts in the Furundzija case, the

principal case which Judge Katzmann cites, are illustrative of the circumstances under which the

ICTY chose to address aiding-and-abetting liability. There, two victims, identified as A and D, were

apprehended and were brutally interrogated. The ICTY made the following finding with respect to

the defendant and his co-defendant, who it identified as Accused B: “There is no doubt that the

accused and Accused B, as commanders, divided the process of interrogation by performing different

functions. The role of the accused was to question, while Accused B’s role was to assault and

threaten in order to elicit the required information from Witness A and Witness D.” Prosecutor v.

Furundzija, 38 I.L.M. 317, 363, Case No. IT-95-17/1-T, Judgment ¶ 130 (Trial Chamber Dec. 10,

1998). While this was an open-and-shut case of joint participation in a violation of international law

by members of an armed group in an international conflict, the ICTY panel entered into a confused

and rambling discussion of aiding-and-abetting.

140
 This theme animates the other ICTY cases cited by Judge Katzmann. In Tadiƒ, the Trial

Chamber found that the defendant, as part of a group proclaiming the creation of an independent

Bosnian Serb republic, played an active role in the attack on a majority Muslim town in a

predominantly Serb region by, inter alia, beating civilians and forcibly transferring them to

concentration camps. E.g., id., Judgement ¶¶ 397, 455, 461. He was also among a group of men

who beat, stabbed, and/or killed a number of individuals at such concentration camps in the context

of the armed conflict. E.g., id. ¶¶ 235-36, 261, 303, 316, 435, 448. Again, this is a classic case of

joint participation in criminal conduct by state actors in an armed conflict who provided direct and

substantial assistance to the “common purpose” of effecting a violation of an international law norm.

Id. ¶¶ 730, 735, 738. Indeed, Tadiƒ was expressly cited in Furundzija, as holding that “the accused

‘intentionally assisted directly and substantially in the common purpose of the group’ to commit the

offence.” Id. Judgment ¶ 226 (internal citations omitted). See also Prosecutor v. Vasiljeviƒ, Case

No. IT-98-32-A, Judgment ¶ 134 (App. Chamber Feb. 25, 2004) (The evidence showed that

“Appellant knew that the seven Muslim men were to be killed; that he walked armed with the group

from the place where they had parked the cars to the Drina River; that he pointed his gun at the seven

Muslim men; and that he stood behind the Muslim men with his gun together with the other three

offenders shortly before the shooting started.”).

Similarly, in Prosecutor v. Kvo…ka, Case No. IT-98-30/1-T, Judgment (Trial Chamber Nov.

2, 2001), the defendants were employees of a camp created by Serb Forces in Bosnia Herzegovina,

where non-Serbs were detained, killed, and otherwise gravely mistreated. The administrative aide

to the commander of the camp was found to have participated in the joint criminal enterprise because

his “administrative duties constituted one of many integral cogs in the wheel of a system of gross

141
mistreatment.” Id. Judgment ¶ 460. The other two defendants held individually responsible were

both guard shift leaders at the camp. One defendant was convicted on the basis of “substantial

evidence presented that the guards on [his] shift beat detainees, sometimes in his presence, and he

not only failed to object, but participated on occasion.” Id. ¶ 497. The other guard leader also

permitted serious crimes to be committed by those under his command and he personally

“perpetrated a number of serious crimes, particularly sexual violence.” Id. ¶ 566. This case is the

mirror image of the Dachau Concentration Camp case in which

there was in the camp a general system of cruelties and murders of the
inmates (most of whom were allied nationals) and that this system
was practiced with knowledge of the accused, who were members of
the staff, and with their active participation. Such a course of
conduct, then, was held by the court in this case to constitute “acting
in pursuance of a common design to violate the laws and usages of
war.” Everybody who took any part in such common design was held
guilty of a war crime, though the nature and extent of the participation
may vary.

Trial of Martin Gottfried Weiss and thirty-nine others, 11 Law Reports of Trials of War Criminals

5, 14 (William S. Hein & Co., Inc. 1997) (1945).16

These cases hardly provide a foundation upon which a norm of international law, of the kind

Judge Katzmann suggests, can be constructed. Nor do the opinions of the ICTR, to which I now

turn.

16
The citation is to the United Nations War Crimes Commission summary of the case based on
the indictment, the evidence and arguments of counsel. See Foreword 1 Law Reports of Trials of War
Criminals, at x. The General Military Government Court of the United States Zone itself reached a
guilty verdict without any opinion.

142
 (b) The ICTR

The ICTR was created on November 8, 1994, by the United Nations Security Council

Resolution 995 in order to judge those individuals responsible for serious violations of international

law in Rwanda and nearby states, between January 1 and December 31, 1994. Specifically, the

defendants there were accused of genocide, crimes against humanity, and violations of Article 3

common to the Geneva Conventions and of Additional Protocol II. The ICTR Statute did not require

a connection between a crime against humanity and an armed conflict. While the ICTR and ICTY

are separate tribunals they shared the same Appeals Chamber, Op. of Judge Katzmann ante at 40

n.28. Not surprisingly, as Judge Katzmann observes, there was little to distinguish judgments of one

from the other. Id. These judgments, as I proceed to show, are equally useless precedent on the

issue of liability of private parties for violations of customary international law.

In Prosecutor v. Akayesu, Case No. ICTR-96-4-T, Judgement (Trial Chamber Sept. 2, 1998),

the defendant, serving as the local “bourgmestre,” was charged with executing the laws adopted by

the communal legislature, had control over the appointment and removal of communal employees,

and had ultimate authority over the communal police and any gendarmes put at the communes

disposal. Judgement ¶¶ 61-71. Bourgmestres also had significant de facto powers over their

constituents, like those of a tribal chief. Id. ¶¶ 72-74. Indeed, the indictment only charges Akayesu

with violations performed during his time as a public official. Given the widespread use of public

employees and facilities in this conduct, it is clear that Akayesu was a state actor and a joint

participant with other state actors.

143
 Similarly, in Prosecutor v. Bagilishema, Case No. ICTR-95-1A-T (Trial Chamber June 7,

2001), and Prosecutor v. Musema, Case No. ICTR-96-13-T (Trial Chamber Jan. 27, 2000), the

defendants were not only charged with acts of direct criminal participation in crimes in which they

were joint participants, they were significant public officials. Bagilishema, the Mayor of a Rwandan

commune, was accused of, inter alia, arming individuals and directing them to attack, personally

attacking and killing persons, as well as other official acts encouraging mass killings. See

Bagilishema, Judgment annex A. Musema was the director of a state-owned tea factory, under the

authority of the Rwandan Ministry of Agriculture, who led and participated in attacks on non-Hutus

using factory employees and equipment, personally shot at refugees, ordered the rape of one Tutsi

woman, and personally raped another woman. See Musema, Judgment § 5.

Finally, in Prosecutor v. Ntakirutimana, Case No. ICTR 96-10-A, ICTR-96-17 (App.

Chamber Dec. 13, 2004), the Appeals Chamber summarized the Trial Chamber’s findings of facts,

in part, as follows:

The Trial Chamber found 1) in relation to the Mugonero Indictment
that, in addition to killing Charles Ukobizaba and shooting at Tutsi
refugees at the Complex, Gérard Ntakirutimana’s participation in the
attacks included procuring ammunition and gendarmes for the attack
on the Complex and participation in the attack on Witness SS; and 2)
in relation to the Bisesero Indictment that, in addition to killing
Esdras and the wife of Nzamwita, pursing and shooting at the
refugees, he transported attackers at Kidashya, headed a group of
armed attackers at Muyira Hill in June 1994, was at Mutiti Hill in
June 1994 with Interahamwe where they shot at refugees in a forest
by a church, and participated in attacks in Bisesero during the period
April to June 1994.

144
Id. ¶ 491. This is a classic example of joint participation of persons acting pursuant to a common

design. Moreover, the defendants were also direct participants in the offenses for which they were

convicted.

Judge Katzmann concedes that the defendants in the ICTY and ICTR cases he cites involved

joint participants in violations of international law. Indeed, virtually all of the defendants in these

cases were state actors. They hold, at most, that “substantial assistance with knowledge” satisfies

the participation necessary for the imposition of liability on joint participants sharing the common

purpose of violating a norm of customary international law. This is entirely consistent with the

Rome Statute.

Moreover, to the extent that any language in these opinions suggest more than that, it rises

only to the level of dicta, of which peremptory norms of international law are not made. Indeed, the

leading treatise Judge Katzmann cites explains that “decisions of international tribunals . . . exercise

considerable influence as an impartial and considered statement of the law by jurists of authority in

light of actual problems which arise before them.” 1 Oppenheim’s International Law: Peace 31

(Robert Y. Jennings & Arthur Watts eds. 9th ed. 1962) (emphasis added). Dicta unrelated to the

actual problems which arise before them do not warrant such deference. This is all the more so

where dicta are inserted by judges who are using their positions to make law with respect to matters

with which they have preconceived views. Yet, a leading jurist and commentator has observed that

the practice of “deal[ing] through obiter dicta with legal issues that were incidental, as it were, to

the main questions,” was a regular practice of ICTY panels. Antonio Cassese, The ICTY: A Living

and Vital Reality, 2 J. Int’l Crim. Just. 585, 589 (2004). The excuse given for this practice is that

it provided “clarification [that] might have some value for the future development of international

145
criminal law[.]” Id. As Antonio Cassese, the President of the ICTY and Presiding Judge of the

ICTY Trial Chamber, who was one of the judges on the ICTY panel in Furundzija, goes on to

explain:

In this connection, one may find much merit in the witty remark
reportedly made by a senior member of the ICTY Office of the
Prosecutor in 1995, after the Appeals Chamber handed down its
lengthy judgment in [Prosecutor v. Tadiƒ, Case No. IT-94-1 (App.
Chambers Oct. 2, 1995) (Cassese, Presiding J.)]: ‘We had gone for a
steak,’ he said, ‘and have got a whole cow.’ However, international
courts operate in a legal system that is notably lacking in many
respects. Among other things, the absence of an international
law-maker and an international court with compulsory universal
jurisdiction entails that many rules are not clear, particularly when
they are of customary origin, and are thus open to differing
interpretations-hence the need for courts gradually to spell out the
contents of those rules, if need be through obiter dicta.

Id. at 590.

Nevertheless, Judge Cassese cautioned that “one should not be oblivious of the possible flaws

or abuses. Courts may indulge in legal discussions that not only are peripheral to the ratio decidendi,

but may also prove simply academic and sometimes also misleading for future courts pronouncing

on the same matters.” Id. This accurately describes the aiding-and-abetting dicta in the ICTY

opinions. They do not constitute a foundation strong enough to provide a future basis for rejecting

the international consensus achieved by the Rome Statute on the elements of aiding-and-abetting.

146
 Conclusion

I dissent from judgment reversing the dismissal of the complaint for the reasons I have

elaborated above. Nevertheless, I concur in section II.A of Judge Katzmann’s opinion that rejects

Judge Hall’s argument that the scope of liability for the violation of the norm of international law

must be decided by reference to our own domestic law. I concur as well in section II.B of Judge

Katzmann’s opinion that articulates the elements of aiding-and-abetting liability as defined in article

25(b) & (c) of the Rome Statute. These elements are consistent with, if not mandated by, customary

international law. I also concur in section II of the per curiam opinion dismissing the TVPA cause

of action for the reasons stated in the per curium opinion and for the additional reason that only

natural persons are subject to liability under it.

147