John Doe I, Individually & as Administrator of the Estate of His Deceased Child Baby Doe I, & on Behalf of All Others Similarly Situated Jane Doe I, on Behalf of Herself, as Administratrix of the Estate of Her Deceased Child Baby Doe I, & on Behalf of All Others Similarly Situated John Doe II John Doe III John Doe IV John Doe v. Jane Doe II Jane Doe III John Doe Vi John Doe Vii John Doe Viii John Doe Ix John Doe X John Doe Xi, on Behalf of Themselves & All Others Similarly Situated & Louisa Benson on Behalf of Herself & the General Public v. Unocal Corporation, a California Corporation Total S.A., a Foreign Corporation John Imle, an Individual Roger C. Beach, an Individual, John Roe III John Roe Vii John Roe Viii John Roe X v. Unocal Corporation Union Oil Company of California

395 F.3d 932

• Court: Court of Appeals for the Ninth Circuit
• Decided: September 18, 2002
• Docket: 00-56603
• Status: Published

Opinion

395 F.3d 932

John DOE I, individually & as Administrator of the Estate of his deceased child Baby Doe I, & on behalf of all others similarly situated; Jane Doe I, on behalf of herself, as Administratrix of the Estate of her deceased child Baby Doe I, & on behalf of all others similarly situated; John Doe II; John Doe III; John Doe IV; John Doe V; Jane Doe II; Jane Doe III; John Doe VI; John Doe VII; John Doe VIII; John Doe IX; John Doe X; John Doe
XI, on behalf of themselves & all others similarly situated & Louisa Benson on behalf of herself & the general public, Plaintiffs-Appellants,
v.
UNOCAL CORPORATION, a California Corporation; Total S.A., a Foreign Corporation; John Imle, an individual; Roger C. Beach, an individual, Defendants-Appellees.
John Roe III; John Roe VII; John Roe VIII; John Roe X, Plaintiffs-Appellants,
v.
Unocal Corporation; Union Oil Company of California, Defendants-Appellees.

No. 00-56603.

No. 00-57197.

No. 00-56628.

No. 00-57195.

United States Court of Appeals, Ninth Circuit.

Argued and Submitted December 3, 2001.

Filed September 18, 2002.

COPYRIGHT MATERIAL OMITTED COPYRIGHT MATERIAL OMITTED Paul L. Hoffman, Schonbrun, Desimome, Seplow, Harris & Hoffman LLP, Venice, CA; Dan Stormer and Anne Richardson, Hadsell & Stormer, Inc., Pasadena, CA; William Goodman, Jennifer M. Green, and Beth Stephens, Center for Constitutional Rights, New York, NY; Katharine J. Redford and Richard Herz, Earthrights International, Washington, DC; Judith Brown Chomsky, Elkins Park, PA; Julie Shapiro, Tacoma, WA; Dilan Esper, Stein & Flugge, LLP, Los Angeles, CA, for plaintiffs-appellants Doe.

Terrence P. Collingsworth and Natacha Thys, International Labor Rights Fund, Washington, DC; Christopher E. Krafchak and Kenderton S. Lynch III, Krafchak & Associates, Los Angeles, CA; Martin J. D'Urso, Hilary Cohen, and Nadia Ezzelarab, Kohn, Swift & Graf, P.C., Philadelphia, PA; Christobal Bonifaz and John C. Bonifaz, Law Offices of Christobal Bonifaz, Amherst, MA, for plaintiffs-appellants Roe.

Edwin V. Woodsome, Jr., D. Barclay Edmundson, David G. Meyer, and Keri R. Curtis, Howrey Simon Arnold & White, LLP, Los Angeles, CA; Jerrold J. Ganzfried, Howrey Simon Arnold & White, LLP, Washington, DC, for defendants-appellees Unocal Corporation, Union Oil Company of California, John Imle, and Roger C. Beach.

William J. Aceves, California Western School of Law, San Diego, California, for Amici Curiae International Human Rights Organizations and International Law and Human Rights Law Scholars.

Appeal from the United States District Court for the Central District of California; Richard A. Paez and Ronald S.W. Lew, District Judges, Presiding1. D.C. Nos. CV-96-06959-RSWL, CV-96-06112-RSWL.

Before PREGERSON, REINHARDT and TASHIMA, Circuit Judges.

Opinion by Judge PREGERSON; Concurrence by Judge REINHARDT

PREGERSON, Circuit Judge.

This case involves human rights violations that allegedly occurred in Myanmar, formerly known as Burma. Villagers from the Tenasserim region in Myanmar allege that the Defendants directly or indirectly subjected the villagers to forced labor, murder, rape, and torture when the Defendants constructed a gas pipeline through the Tenasserim region. The villagers base their claims on the Alien Tort Claims Act, 28 U.S.C. § 1350, and the Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. § 1961 et seq., as well as state law.

The District Court, through dismissal and summary judgment, resolved all of Plaintiffs' federal claims in favor of the Defendants. For the following reasons, we reverse in part and affirm in part the District Court's rulings.

I.

FACTUAL AND PROCEDURAL BACKGROUND

A. Unocal's Investment in a Natural Gas Project in Myanmar.

Burma has been ruled by a military government since 1958. In 1988, a new military government, Defendant-Appellee State Law and Order Restoration Council ("the Myanmar Military"), took control and renamed the country Myanmar. The Myanmar Military established a state owned company, Defendant-Appellee Myanmar Oil and Gas Enterprise ("Myanmar Oil"), to produce and sell the nation's oil and gas resources.

In 1992, Myanmar Oil licensed the French oil company Total S.A. ("Total") to produce, transport, and sell natural gas from deposits in the Yadana Field off the coast of Myanmar ("the Project"). Total set up a subsidiary, Total Myanmar Exploration and Production ("Total Myanmar"), for this purpose. The Project consisted of a Gas Production Joint Venture, which would extract the natural gas out of the Yadana Field, and a Gas Transportation Company, which would construct and operate a pipeline to transport the natural gas from the coast of Myanmar through the interior of the country to Thailand.

Also in 1992, Defendant-Appellant Unocal Corporation and its wholly owned subsidiary Defendant-Appellant Union Oil Company of California, collectively referred to below as "Unocal," acquired a 28% interest in the Project from Total. Unocal set up a wholly owned subsidiary, the Unocal Myanmar Offshore Company ("the Unocal Offshore Co."), to hold Unocal's 28% interest in the Gas Production Joint Venture half of the Project.2 Similarly, Unocal set up another wholly owned subsidiary, the Unocal International Pipeline Corporation ("the Unocal Pipeline Corp."), to hold Unocal's 28% interest in the Gas Transportation Company half of the Project.3 Myanmar Oil and a Thai government entity, the Petroleum Authority of Thailand Exploration and Production, also acquired interests in the Project. Total Myanmar was appointed Operator of the Gas Production Joint Venture and the Gas Transportation Company. As the Operator, Total Myanmar was responsible, inter alia, for "determin[ing] ... the selection of ... employees [and] the hours of work and the compensation to be paid to all ... employees" in connection with the Project.

B. Unocal's Knowledge that the Myanmar Military Was Providing Security and Other Services for the Project.

It is undisputed that the Myanmar Military provided security and other services for the Project, and that Unocal knew about this. The pipeline was to run through Myanmar's rural Tenasserim region. The Myanmar Military increased its presence in the pipeline region to provide security and other services for the Project.4 A Unocal memorandum documenting Unocal's meetings with Total on March 1 and 2, 1995 reflects Unocal's understanding that "[f]our battalions of 600 men each will protect the [pipeline] corridor" and "[f]ifty soldiers will be assigned to guard each survey team." A former soldier in one of these battalions testified at his deposition that his battalion had been formed in 1996 specifically for this purpose. In addition, the Military built helipads and cleared roads along the proposed pipeline route for the benefit of the Project.

There is also evidence sufficient to raise a genuine issue of material fact whether the Project hired the Myanmar Military, through Myanmar Oil, to provide these services, and whether Unocal knew about this. A Production Sharing Contract, entered into by Total Myanmar and Myanmar Oil before Unocal acquired an interest in the Project, provided that "[Myanmar Oil] shall ... supply[ ] or mak[e] available ... security protection ... as may be requested by [Total Myanmar and its assigns]," such as Unocal. Unocal was aware of this agreement. Thus, a May 10, 1995 Unocal "briefing document" states that "[a]ccording to our contract, the government of Myanmar is responsible for protecting the pipeline." (Emphasis added.) Similarly, in May 1995, a cable from the U.S. Embassy in Rangoon, Myanmar, reported that Unocal On-Site Representative Joel Robinson ("Unocal Representative Robinson" or "Robinson") "stated forthrightly that the companies have hired the Burmese military to provide security for the project." (Emphasis added.)

Unocal disputes that the Project hired the Myanmar Military or, at the least, that Unocal knew about this. For example, Unocal points out that the Production Sharing Contract quoted in the previous paragraph covered only the off-shore Gas Production Joint Venture but not the Gas Transportation Company and the construction of the pipeline which gave rise to the alleged human rights violations. Moreover, Unocal President John Imle ("Unocal President Imle" or "Imle") stated at his deposition that he knew of "no ... contractual obligation" requiring the Myanmar Military to provide security for the pipeline construction. Likewise, Unocal CEO Roger Beach ("Unocal CEO Beach" or "Beach") stated at his deposition that he also did not know "whether or not Myanmar had a contractual obligation to provide ... security." Beach further stated that he was not aware of "any support whatsoever of the military[,] ... either physical or monetary." These assertions by Unocal President Imle and Unocal CEO Beach are called into question by a briefing book which Total prepared for them on the occasion of their April 1996 visit to the Project. The briefing book lists the "numbers of villagers" working as "local helpers hired by battalions," the monthly "amount paid in Kyats" (the currency of Myanmar) to "Project Helpers," and the "amount in Kyats" expended by the Project on "food rations (Army + Villages)."5

Furthermore, there is evidence sufficient to raise a genuine issue of material fact whether the Project directed the Myanmar Military in these activities, at least to a degree, and whether Unocal was involved in this. In May 1995, a cable from the U.S. Embassy in Rangoon reported:

[Unocal Representative] Robinson indicated ... Total/Unocal uses [aerial photos, precision surveys, and topography maps] to show the [Myanmar] military where they need helipads built and facilities secured.... Total's security officials meet with military counterparts to inform them of the next day's activities so that soldiers can ensure the area is secure and guard the work perimeter while the survey team goes about its business.

A November 8, 1995 document apparently authored by Total Myanmar stated that "[e]ach working group has a security officer ... to control the army positions." A January 1996 meeting document lists "daily security coordination with the army" as a "working procedure." Similarly, the briefing book that Total prepared for Unocal President Imle and Unocal CEO Beach on the occasion of their April 1996 visit to the Project mentions that "daily meeting[s]" were "held with the tactical commander" of the army. Moreover, on or about August 29, 1996, Unocal (Singapore) Director of Information Carol Scott ("Unocal Director of Information Scott" or "Scott") discussed with Unocal Media Contact and Spokesperson David Garcia ("Unocal Spokesperson Garcia" or "Garcia") via e-mail how Unocal should publicly address the issue of the alleged movement of villages by the Myanmar Military in connection with the pipeline. Scott cautioned Garcia that "[b]y saying we influenced the army not to move a village, you introduce the concept that they would do such a thing; whereas, by saying that no villages have been moved, you skirt the issue of whether it could happen or not." (Emphasis added.) This e-mail is some evidence that Unocal could influence the army not to commit human rights violations, that the army might otherwise commit such violations, and that Unocal knew this.

C. Unocal's Knowledge that the Myanmar Military Was Allegedly Committing Human Rights Violations in Connection with the Project.

Plaintiffs are villagers from Myanmar's Tenasserim region, the rural area through which the Project built the pipeline. Plaintiffs allege that the Myanmar Military forced them, under threat of violence, to work on and serve as porters for the Project. For instance, John Doe IX testified that he was forced to build a helipad near the pipeline site in 1994 that was then used by Unocal and Total officials who visited the pipeline during its planning stages. John Doe VII and John Roe X, described the construction of helipads at Eindayaza and Po Pah Pta, both of which were near the pipeline site, were used to ferry Total/Unocal executives and materials to the construction site, and were constructed using the forced labor of local villagers, including Plaintiffs. John Roes VIII and IX, as well as John Does I, VIII and IX testified that they were forced to work on building roads leading to the pipeline construction area. Finally, John Does V and IX, testified that they were required to serve as "pipeline porters" — workers who performed menial tasks such as such as hauling materials and cleaning the army camps for the soldiers guarding the pipeline construction.

Plaintiffs also allege in furtherance of the forced labor program just described, the Myanmar Military subjected them to acts of murder, rape, and torture. For instance, Jane Doe I testified that after her husband, John Doe I, attempted to escape the forced labor program, he was shot at by soldiers, and in retaliation for his attempted escape, that she and her baby were thrown into a fire, resulting in injuries to her and the death of the child. Other witnesses described the summary execution of villagers who refused to participate in the forced labor program, or who grew too weak to work effectively. Several Plaintiffs testified that rapes occurred as part of the forced labor program. For instance, both Jane Does II and III testified that while conscripted to work on pipeline-related construction projects, they were raped at knife-point by Myanmar soldiers who were members of a battalion that was supervising the work. Plaintiffs finally allege that Unocal's conduct gives rise to liability for these abuses.

The successive military governments of first Burma and now Myanmar have a long and well-known history of imposing forced labor on their citizens. See, e.g., Forced labour in Myanmar (Burma): Report of the Commission of Inquiry appointed under article 26 of the Constitution of the International Labour Organization to examine the observance by Myanmar of the Forced Labour Convention, 1930 (No. 29) Parts III.8, V.14(3) (1998) (describing several inquiries into forced labor in Myanmar conducted between 1960 and 1992 by the International Labor Organization, and finding "abundant evidence ... showing the pervasive use of forced labour imposed on the civilian population throughout Myanmar by the authorities and the military"), http:// www.ilo.org/public/english /standards/relm/gb/docs/gb273/myanmar.htm. As detailed below, even before Unocal invested in the Project, Unocal was made aware — by its own consultants and by its partners in the Project — of this record and that the Myanmar Military might also employ forced labor and commit other human rights violations in connection with the Project. And after Unocal invested in the Project, Unocal was made aware — by its own consultants and employees, its partners in the Project, and human rights organizations — of allegations that the Myanmar Military was actually committing such violations in connection with the Project.

Before Unocal acquired an interest in the Project, it hired a consulting company, Control Risk Group, to assess the risks involved in the investment. In May 1992, Control Risk Group informed Unocal that "[t]hroughout Burma the government habitually makes use of forced labour to construct roads."6 Control Risk Group concluded that "[i]n such circumstances UNOCAL and its partners will have little freedom of manoeuvre." Unocal's awareness of the risk at that time is also reflected in the deposition testimony of Unocal Vice President of International Affairs Stephen Lipman ("Unocal Vice President Lipman"):

[I]n our discussions between Unocal and Total[preceding Unocal's acquisition of an interest in the Project], we said that the option of having the [Myanmar] [M]ilitary provide protection[7] for the pipeline construction and operation of it would be that they might proceed in the manner that would be out of our control and not be in a manner that we would like to see them proceed, I mean, going to excess.

On January 4, 1995, approximately three years after Unocal acquired an interest in the Project, Unocal President Imle met with human rights organizations at Unocal's headquarters in Los Angeles and acknowledged to them that the Myanmar Military might be using forced labor in connection with the Project. At that meeting, Imle said that "[p]eople are threatening physical damage to the pipeline," that "if you threaten the pipeline there's gonna be more military," and that "[i]f forced labor goes hand and glove with the military yes there will be more forced labor." (Emphasis added.)

Two months later, on March 16, 1995, Unocal Representative Robinson confirmed to Unocal President Imle that the Myanmar Military might be committing human rights violations in connection with the Project. Thus, Robinson wrote to Imle that he had received publications from human rights organizations "which depicted in more detail than I have seen before the increased encroachment of [the Myanmar Military's] activities into the villages of the pipeline area." Robinson concluded on the basis of these publications that "[o]ur assertion that [the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny."8

Shortly thereafter, on May 10, 1995, Unocal Representative Robinson wrote to Total's Herve Madeo:

From Unocal's standpoint, probably the most sensitive issue is "what is forced labor" and "how can you identify it." I am sure that you will be thinking about the demarcation between work done by the project and work done "on behalf of" the project. Where the responsibility of the project ends is very important.

This statement is some evidence that Unocal knew that the Myanmar Military might use forced labor in connection with the Project.

In June 1995, Amnesty International also alerted Unocal to the possibility that the Myanmar Military might use forced labor in connection with the Project. Amnesty International informed Unocal that comments from a Myanmar Department of Industry official "could mean that the government plans to use `voluntary' labor in conjunction with the pipeline." Amnesty International went on to explain that "what they call `voluntary' labor is called forced labor in other parts of the world."9

Later that year, on December 11, 1995, Unocal Consultant John Haseman ("Unocal Consultant Haseman" or "Haseman"), a former military attache at the U.S. Embassy in Rangoon, reported to Unocal that the Myanmar Military was, in fact, using forced labor and committing other human rights violations in connection with the Project. Haseman told Unocal that "Unocal was particularly discredited when a corporate spokesman was quoted as saying that Unocal was satisfied with ... assurances [by the Myanmar Military] that no human rights abuses were occurring in the area of pipeline construction." Haseman went on to say:

Based on my three years of service in Burma, my continuous contacts in the region since then, and my knowledge of the situation there, my conclusion is that egregious human rights violations have occurred, and are occurring now, in southern Burma. The most common are forced relocation without compensation of families from land near/along the pipeline route; forced labor to work on infrastructure projects supporting the pipeline...; and imprisonment and/or execution by the army of those opposing such actions.... Unocal, by seeming to have accepted [the Myanmar Military]'s version of events, appears at best naive and at worst a willing partner in the situation.10

Communications between Unocal and Total also reflect the companies' shared knowledge that the Myanmar Military was using forced labor in connection with the Project. On February 1, 1996, Total's Herve Chagnoux wrote to Unocal and explained his answers to questions by the press as follows:

By stating that I could not guarantee that the army is not using forced labour, I certainly imply that they might, (and they might) but I am saying that we do not have to monitor army's behavior: we have our responsibilities; they have their responsibilities; and we refuse to be pushed into assuming more than what we can really guarantee. About forced labour used by the troops assigned to provide security on our pipeline project, let us admit between Unocal and Total that we might be in a grey zone.

And on September 17, 1996, Total reported to Unocal about a meeting with a European Union civil servant in charge of an investigation of forced labor in Myanmar: "We were told that even if Total is not using forced labor directly, the troops assigned to the protection of our operations use forced labour to build their camps and to carry their equipments." In reply, Total acknowledged that forced labor did indeed occur in connection with the pipeline: "We had to mention that when we had knowledge of such occurrences, the workers have been compensated." Unocal President Imle testified at his deposition that in Unocal's discussions with Total, "[s]urrounding the question of porters for the military and their payment was the issue of whether they were conscripted or volunteer workers." Imle further testified that "the consensus was that it was mixed," i.e., "some porters were conscripted, and some were volunteer." On March 4, 1997, Unocal nevertheless submitted a statement to the City Counsel of New York, in response to a proposed New York City select purchasing law imposed on firms that do business in Myanmar, in which Unocal stated that "no [human rights] violations have taken place" in the vicinity of the pipeline route.

D. Proceedings Below.

In September of 1996, four villagers from the Tenasserim region, the Federation of Trade Unions of Burma ("the Trade Unions"), and the National Coalition Government of the Union of Burma ("the Government in Exile") brought an action against Unocal and the Project. Nat'l Coalition Gov't of the Union of Burma v. Unocal, Inc., 176 F.R.D. 329, 334 (C.D.Cal.1997) ("Roe I"). Plaintiffs in Roe I alleged violations of the law of nations under the Alien Tort Claims Act ("the ATCA"), 28 U.S.C. § 1350, and violations of state law. One of the four individual Roe-Plaintiffs alleged that the Myanmar Military subjected him to forced labor, without compensation and under threat of death, along the pipeline route in connection with the Project. The other three individual Roe-Plaintiffs alleged they owned land located along the pipeline route, and were not compensated when the land was confiscated by the Myanmar Military in connection with the Project. The Trade Unions and the Government in Exile alleged similar injuries to their members and citizens, respectively.

In October of 1996, fourteen other villagers from the Tenasserim region brought another action against Unocal, Total, Myanmar Oil, the Myanmar Military, Unocal President Imle and Unocal CEO Beach. Doe I v. Unocal Corp., 963 F.Supp. 880, 883 (C.D.Cal.1997) ("Doe I"). Plaintiffs in Doe I alleged that the Defendants' conduct in connection with the Project had caused them to suffer death of family members, assault, rape and other torture, forced labor, and the loss of their homes and property. The Doe-Plaintiffs sought to represent a class of all residents of the Tenasserim region who have suffered or are or will be suffering similar injuries. As in the Roe case, liability in the Doe case was based on alleged violations of the ATCA and state law. In addition, liability in the Doe case was also based on alleged violations of the Racketeer Influenced and Corrupt Organizations Act ("RICO"), 18 U.S.C. § 1961 et seq.

On March 25, 1997, the District Court granted in part and denied in part Unocal's motion to dismiss the Doe action. See Doe I, 963 F.Supp. 880. The District Court dismissed the claims against the Myanmar Military and Myanmar Oil on the grounds that these defendants were entitled to immunity pursuant to the Foreign Sovereign Immunities Act, 28 U.S.C. §§ 1330, 1602 et seq. The District Court also determined, however, that the act of state doctrine did not require the dismissal of the claims against the other defendants, with the exception of the expropriation claims.11 Moreover, the District Court determined that subject matter jurisdiction was available under the ATCA and that the Doe-Plaintiffs had pled sufficient facts to state a claim under the ATCA. The District Court later denied the Doe-Plaintiffs' motion for class certification and dismissed their claims against Total for lack of personal jurisdiction. See Doe I v. Unocal Corp., 67 F.Supp.2d 1140 (C.D.Cal.1999); Doe I v. Unocal Corp., 27 F.Supp.2d 1174 (C.D.Cal.1998), aff'd 248 F.3d 915 (9th Cir.2001).

On November 5, 1997, the District Court similarly granted in part and denied in part Unocal's motion to dismiss the Roe action. See Roe I, 176 F.R.D. 329. The District Court determined that the Government in Exile (wholly) and the Trade Unions (in part) lacked standing to pursue their claims. The District Court's other determinations in the Roe action — regarding the act of state doctrine, subject matter jurisdiction under the ATCA, and failure to state a claim under the ATCA — were identical to its earlier determinations in the Doe action regarding the same issues.

On August 31, 2000, the District Court granted Unocal's consolidated motions for summary judgment on all of Plaintiffs' remaining federal claims in both actions. See Doe I v. Unocal Corp., 110 F.Supp.2d 1294 (9th Cir.2000) ("Doe/Roe II"). The District Court granted Unocal's motion for summary judgment on the ATCA claims based on murder, rape, and torture because Plaintiffs could not show that Unocal engaged in state action and that Unocal controlled the Myanmar Military. The District Court granted Unocal's motion for summary judgment on the ATCA claims based on forced labor because Plaintiffs could not show that Unocal "actively participated" in the forced labor. The District Court also determined that it did not have subject matter jurisdiction over the Doe-Plaintiffs' RICO claim. Finally, after having granted summary judgment on all of Plaintiffs' federal claims, the District Court declined to exercise its discretion to retain Plaintiffs' state claims and dismissed those claims without prejudice.

On September 5, 2000, the District Court granted Unocal's motion to recover costs in the amount of $125,846.07. On November 29, 2000, the District Court denied Plaintiffs' joint Fed.R.Civ.P. 54(d)(1) Motion to Retax, concluding that the motion actually constituted a time-barred Fed.R.Civ.P. 59(e) Motion to Alter or Amend Judgment.

The Doe-Plaintiffs appeal the District Court's dismissal of their claims against the Myanmar Military and Myanmar Oil and the District Court's grant of summary judgment in favor of Unocal on their ATCA and RICO claims against Unocal (No. 00-56603). The Roe-Plaintiffs appeal the District Court's grant of summary judgment in favor of Unocal on their ATCA claims against Unocal (No. 00-56628). Plaintiffs also appeal the District Court's denial of their motion to retax (Nos. 00-57195 & 00-57197). The four appeals have been consolidated. We have jurisdiction under 28 U.S.C. § 1291, and we reverse in part, affirm in part, and remand to the District Court for further proceedings consistent with this opinion.

II.

ANALYSIS

A. Liability Under the Alien Tort Claims Act.

1. Introduction

The Alien Tort Claims Act confers upon the federal district courts "original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations." 28 U.S.C. § 1350.12 We have held that the ATCA also provides a cause of action, as long as "plaintiffs ... allege a violation of `specific, universal, and obligatory' international norms as part of [their] ATCA claim." Papa v. United States, 281 F.3d 1004, 1013 (9th Cir.2002) (quoting In re Estate of Ferdinand E. Marcos, Human Rights Litig., 25 F.3d 1467, 1475 (9th Cir.1994) ("Marcos II")). See also Marcos II, 25 F.3d at 1474-75. Plaintiffs allege that Unocal's conduct gave rise to ATCA liability for the forced labor, murder, rape, and torture inflicted on them by the Myanmar Military.13

The District Court granted Unocal's motion for summary judgment on Plaintiffs' ATCA claims. We review a grant of summary judgment de novo. See Lopez v. Smith, 203 F.3d 1122, 1131 (9th Cir.2000) (en banc). We must determine whether, viewing the evidence in the light most favorable to the nonmoving party, there are any genuine issues of material fact and whether the district court correctly applied the relevant substantive law. See id.

One threshold question in any ATCA case is whether the alleged tort is a violation of the law of nations. We have recognized that torture, murder, and slavery are jus cogens violations and, thus, violations of the law of nations.14 See United States v. Matta-Ballesteros, 71 F.3d 754, 764 n. 5 (9th Cir.1995). Rape can be a form of torture. See Farmer v. Brennan, 511 U.S. 825, 852, 854, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994) (Blackmun, J., concurring) (describing brutal prison rape as "the equivalent of" and "nothing less than torture"); Kadic v. Karadzic, 70 F.3d 232, 242 (2d Cir.1995) (describing allegations of "murder, rape, forced impregnation, and other forms of torture" (emphasis added)); In re Extradition of Suarez-Mason, 694 F.Supp. 676, 682 (N.D.Cal.1988) (stating that "shock sessions were interspersed with rapes and other forms of torture" (emphasis added)); see also generally Evelyn Mary Aswad, Torture by Means of Rape, 84 Geo. L.J. 1913 (1996). Moreover, forced labor is so widely condemned that it has achieved the status of a jus cogens violation. See, e.g., Universal Declaration of Human Rights, G.A. Res. 217(A)III (1948) (banning forced labor); Agreement for the Prosecution and Punishment of the Major War Criminals of the European Axis, and Charter of the International Military Tribunal, Aug. 8, 1945, art. 6, 82 U.N.T.S. 280 (making forced labor a war crime). Accordingly, all torts alleged in the present case are jus cogens violations and, thereby, violations of the law of nations.15

Another threshold question in any ATCA case against a private party, such as Unocal, is whether the alleged tort requires the private party to engage in state action for ATCA liability to attach, and if so, whether the private party in fact engaged in state action. In his concurrence in Tel-Oren v. Libyan Arab Republic, 726 F.2d 774 (D.C.Cir.1984), Judge Edwards observed that while most crimes require state action for ATCA liability to attach, there are a "handful of crimes," including slave trading, "to which the law of nations attributes individual liability," such that state action is not required. Id. at 794-95 (Edwards, J., concurring) (emphasis added).16 More recently, the Second Circuit adopted and extended this approach in Kadic. The Second Circuit first noted that genocide and war crimes — like slave trading — do not require state action for ATCA liability to attach. See 70 F.3d at 242-43. The Second Circuit went on to state that although "acts of rape, torture, and summary execution," like most crimes, "are proscribed by international law only when committed by state officials or under color of law" to the extent that they were committed in isolation, these crimes "are actionable under the Alien Tort [Claims] Act, without regard to state action, to the extent that they were committed in pursuit of genocide or war crimes." Id. at 243-44 (emphasis added). Thus, under Kadic, even crimes like rape, torture, and summary execution, which by themselves require state action for ATCA liability to attach, do not require state action when committed in furtherance of other crimes like slave trading, genocide or war crimes, which by themselves do not require state action for ATCA liability to attach. We agree with this view and apply it below to Plaintiffs' various ATCA claims.

2. Forced Labor

a. Forced labor is a modern variant of slavery to which the law of nations attributes individual liability such that state action is not required.

Our case law strongly supports the conclusion that forced labor is a modern variant of slavery. Accordingly, forced labor, like traditional variants of slave trading, is among the "handful of crimes ... to which the law of nations attributes individual liability," such that state action is not required. Id. at 794-95 (Edwards, J., concurring). See supra section II.A.1.

Courts have included forced labor in the definition of the term "slavery" in the context of the Thirteenth Amendment.17 The Supreme Court has said that "[t]he undoubted aim of the Thirteenth Amendment ... was not merely to end slavery but to maintain a system of completely free and voluntary labor throughout the United States." Pollock v. Williams, 322 U.S. 4, 17, 64 S.Ct. 792, 88 L.Ed. 1095 (1944) (emphasis added).18 Accordingly, "[i]t has been held that forced labor of certain individuals amounts to involuntary servitude and therefore is violative of the thirteenth amendment." Weidenfeller v. Kidulis, 380 F.Supp. 445, 450 (E.D.Wis.1974) (citing Stone v. City of Paducah, 120 Ky. 322, 86 S.W. 531, 533 (1905)).

The inclusion of forced labor in the definition of the term "slavery" is not confined to the Thirteenth Amendment but extends, for example, to 18 U.S.C. § 1583. 18 U.S.C. § 1583 was introduced in 1866 to prevent the kidnaping of former slaves to countries which still permitted slavery.19 The Fourth Circuit has said that "[n]otwithstanding this limited purpose, the statute should be read as expressing the broad and sweeping intention of Congress during the Reconstruction period to stamp out the vestiges of the old regime of slavery and to prevent the reappearance of forced labor in whatever new form it might take." United States v. Booker, 655 F.2d 562, 565 (4th Cir.1981) (emphasis added).

In World War II Era Japanese Forced Labor Litig., 164 F.Supp.2d 1160, (N.D.Cal.2001), the District Court for the Northern District of California recently implicitly included forced labor in the definition of the term "slavery" for purposes of the ATCA. There, the district court concluded that "[g]iven the Ninth Circuit's comment in Matta-Ballesteros, 71 F.3d at 764 n. 5, that slavery constitutes a violation of jus cogens, this court is inclined to agree with the[District Court for the District of New Jersey's] conclusion [in Iwanowa v. Ford Motor Co., 67 F.Supp.2d 424 (D.N.J.1999)] that forced labor violates the law of nations." Id. at 1179.

In light of these authorities, we conclude that forced labor is a modern variant of slavery that, like traditional variants of slave trading, does not require state action to give rise to liability under the ATCA.

b. Unocal may be liable under the ATCA for aiding and abetting the Myanmar Military in subjecting Plaintiffs to forced labor.

Plaintiffs argue that Unocal aided and abetted the Myanmar Military in subjecting them to forced labor. We hold that the standard for aiding and abetting under the ATCA is, as discussed below, knowing practical assistance or encouragement that has a substantial effect on the perpetration of the crime. We further hold that a reasonable factfinder could find that Unocal's conduct met this standard.20

The District Court found that "[t]he evidence ... suggest[s] that Unocal knew that forced labor was being utilized and that the Joint Venturers benefitted from the practice." Doe/Roe II, 110 F.Supp.2d at 1310. The District Court nevertheless held that Unocal could not be liable under the ATCA for forced labor because Unocal's conduct did not rise to the level of "active participation" in the forced labor. Id. The District Court incorrectly borrowed the "active participation" standard for liability from war crimes cases before Nuremberg Military Tribunals involving the role of German industrialists in the Nazi forced labor program during the Second World War. The Military Tribunals applied the "active participation" standard in these cases only to overcome the defendants'"necessity defense."21 In the present case, Unocal did not invoke — and could not have invoked — the necessity defense. The District Court therefore erred when it applied the "active participation" standard here.22

We however agree with the District Court that in the present case, we should apply international law as developed in the decisions by international criminal tribunals such as the Nuremberg Military Tribunals for the applicable substantive law. "The law of nations `may be ascertained by consulting the works of jurists, writing professedly on public law; or by the general usage and practice of nations; or by judicial decisions recognizing and enforcing that law.'" Filartiga v. Pena-Irala, 630 F.2d 876, 880 (2d Cir.1980) (quoting United States v. Smith, 18 U.S. (5 Wheat.) 153, 160-61, 5 L.Ed. 57 (1820)) (emphasis added). It is "well settled that the law of nations is part of federal common law." Marcos I, 978 F.2d at 502.

In different ATCA cases, different courts have applied international law, the law of the state where the underlying events occurred, or the law of the forum state, respectively. See Wiwa v. Royal Dutch Petroleum Co., 226 F.3d 88, 105 n. 12 (2d Cir.2000). Unocal urges us to apply not international law, but the law of the state where the underlying events occurred, i.e., Myanmar. Where, as in the present case, only jus cogens violations are alleged — i.e., violations of norms of international law that are binding on nations even if they do not agree to them, see supra note 14 and accompanying text — it may, however, be preferable to apply international law rather than the law of any particular state, such as the state where the underlying events occurred or the forum state.23 The reason is that, by definition, the law of any particular state is either identical to the jus cogens norms of international law, or it is invalid. Moreover, "reading § 1350 as essentially a jurisdictional grant only and then looking to [foreign or] domestic tort law to provide the cause of action mutes the grave international law aspect of the tort, reducing it to no more (or less) than a garden-variety municipal tort," Xuncax v. Gramajo, 886 F.Supp. 162, 183 (D.Mass.1995), i.e., reducing it to a tort "relating to the internal government of a state of nation (as contrasted with international )," Black's Law Dictionary 1037 (7th ed.1999). Significantly, we have already held that the ATCA not only confers jurisdiction but also creates a cause of action. See Papa, 281 F.3d at 1013; Marcos II, 25 F.3d at 1474-75.

Application of international law — rather than the law of Myanmar, California state law, or our federal common law — is also favored by a consideration of the factors listed in the Restatement (Second) of Conflict of Laws § 6 (1969). First, "the needs of the ... international system[ ]" are better served by applying international rather than national law. Second, "the relevant policies of the forum" cannot be ascertained by referring — as the concurrence does — to one out-of-circuit decision which happens to favor federal common law and ignoring other decisions which have favored other law, including international law. See Wiwa, 226 F.3d at 105 n. 12. Third, regarding "the protection of justified expectations," the "certainty, predictability and uniformity of result," and the "ease in the determination and application of the law to be applied," we note that the standard we adopt today from an admittedly recent case nevertheless goes back at least to the Nuremberg trials and is similar to that of the Restatement (Second) of Torts. See infra note 26 and accompanying text.24 Finally, "the basic polic[y] underlying the particular field of law" is to provide tort remedies for violations of international law. This goal is furthered by the application of international law, even when the international law in question is criminal law but is similar to domestic tort law, as discussed in the next paragraph. We conclude that given the record in the present case, application of international law is appropriate.25

International human rights law has been developed largely in the context of criminal prosecutions rather than civil proceedings. See Beth Stevens, Translating Filartiga: A Comparative and International Law Analysis of Domestic Remedies for International Human Rights Violations, 27 Yale J. Int'l L. 1, 40 (2002). But what is a crime in one jurisdiction is often a tort in another jurisdiction, and this distinction is therefore of little help in ascertaining the standards of international human rights law. See id. at 44-46. Moreover, as mentioned above in note 23 and further discussed later in this section, the standard for aiding and abetting in international criminal law is similar to the standard for aiding and abetting in domestic tort law, making the distinction between criminal and tort law less crucial in this context. Accordingly, District Courts are increasingly turning to the decisions by international criminal tribunals for instructions regarding the standards of international human rights law under our civil ATCA. See, e.g., Cabello Barrueto v. Fernandez Larios, 205 F.Supp.2d 1325, 1333 (S.D.Fla.2002) (concluding on the basis of, inter alia, the statute of and a decision by the International Criminal Tribunal for the former Yugoslavia that defendants "may be held liable under the ATCA for ... aiding and abetting the actions taken by [foreign] military officials"); Mehinovic v. Vuckovic, 198 F.Supp.2d 1322 (N.D.Ga.2002) (noting that among "various contemporary sources" for ascertaining the norms of international law as they pertain to the ATCA, "the statutes of the [International Criminal Tribunal for the former Yugoslavia] and the International Criminal Tribunal for Rwanda ... and recent opinions of these tribunals are particularly relevant"). We agree with this approach. We find recent decisions by the International Criminal Tribunal for the former Yugoslavia and the International Criminal Tribunal for Rwanda especially helpful for ascertaining the current standard for aiding and abetting under international law as it pertains to the ATCA.

In Prosecutor v. Furundzija, IT-95-17/1 T (Dec. 10, 1998), reprinted in 38 I.L.M. 317 (1999), the International Tribunal for the former Yugoslavia held 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." Id. at ¶ 235. The Tribunal clarified that in order to qualify, "assistance need not constitute an indispensable element, that is, a conditio sine qua non for the acts of the principal." Furundzija at ¶ 209; see also Prosecutor v. Kunarac, IT-96-23-T & IT-96-23/1-T, ¶ 391 (Feb. 22, 2001), http: // www.un.org/icty/foca/trialc2/judgement/index.htm ("The act of assistance need not have caused the act of the principal."). Rather, it suffices that "the acts of the accomplice make a significant difference to the commission of the criminal act by the principal." Furundzija at ¶ 233. The acts of the accomplice have the required "[substantial] effect on the commission of the crime" where "the criminal act most probably would not have occurred in the same way [without] someone act[ing] in the role that the [accomplice] in fact assumed." Prosecutor v. Tadic, ICTY-94-1, ¶ 688 (May 7, 1997), http:// www.un.org/icty/tadic/trials2/judgement/ index.htm.26

Similarly, in Prosecutor v. Musema, ICTR-96-13-T (Jan. 27, 2000), http://www.ictr.org/, the International Criminal Tribunal for Rwanda described the actus reus of aiding and abetting as "all acts of assistance in the form of either physical or moral support" that "substantially contribute to the commission of the crime". Id. at ¶ 126.

As for the mens rea of aiding and abetting, the International Criminal Tribunal for the former Yugoslavia held that what is required is actual or constructive (i.e., "reasonabl[e]") "knowledge that [the accomplice's] actions will assist the perpetrator in the commission of the crime." Furundzija at ¶ 245. Thus, "it is not necessary for the accomplice to share the mens rea of the perpetrator, in the sense of positive intention to commit the crime." Id. In fact, it is not even necessary that the aider and abettor knows the precise crime that the principal intends to commit. See id. Rather, if the accused "is aware that one of a number of crimes will probably be committed, and one of those crimes is in fact committed, he has intended to facilitate the commission of that crime, and is guilty as an aider and abettor." Id.27

Similarly, for the mens rea of aiding and abetting, the International Criminal Tribunal for Rwanda required that "the accomplice knew of the assistance he was providing in the commission of the principal offence." Musema at ¶ 180. The accomplice does not have to have had the intent to commit the principal offense. See id. at ¶ 181. It is sufficient that the accomplice "knew or had reason to know" that the principal had the intent to commit the offense. Id. at¶ 182.

The Furundzija standard for aiding and abetting liability under international criminal law can be summarized as knowing practical assistance, encouragement, or moral support which has a substantial effect on the perpetration of the crime. At least with respect to assistance and encouragement, this standard is similar to the standard for aiding and abetting under domestic tort law. Thus, the Restatement of Torts states: "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...." Restatement (Second) of Torts § 876 (1979) (emphasis added). Especially given the similarities between the Furundzija international criminal standard and the Restatement domestic tort standard, we find that application of a slightly modified Furundzija standard is appropriate in the present case. In particular, given that there is — as discussed below — sufficient evidence in the present case that Unocal gave assistance and encouragement to the Myanmar Military, we do not need to decide whether it would have been enough if Unocal had only given moral support to the Myanmar Military. Accordingly, we may impose aiding and abetting liability for knowing practical assistance or encouragement which has a substantial effect on the perpetration of the crime, leaving the question whether such liability should also be imposed for moral support which has the required substantial effect to another day.28

First, a reasonable factfinder could conclude that Unocal's alleged conduct met the actus reus requirement of aiding and abetting as we define it today, i.e., practical assistance or encouragement which has a substantial effect on the perpetration of the crime of, in the present case, forced labor.

Unocal's weak protestations notwithstanding, there is little doubt that the record contains substantial evidence creating a material question of fact as to whether forced labor was used in connection with the construction of the pipeline. Numerous witnesses, including a number of Plaintiffs, testified that they were forced to clear the right of way for the pipeline and to build helipads for the project before construction of the pipeline began. For instance, John Doe IX testified that he was forced to build a helipad near the pipeline site in 1994 that was then used by Unocal and Total officials who visited the pipeline during its planning stages. Other Plaintiffs and witnesses, including John Doe VII and John Roe X, described the construction of helipads at Eindayaza and Po Pah Pta, both of which were near the pipeline site, were used to ferry Total/Unocal executives and materials to the construction site, and were constructed using the forced labor of local villagers, including Plaintiffs. Other Plaintiffs, such as John Roes VIII and IX, as well as John Does I, VIII and IX, testified that they were forced to work on building roads leading to the pipeline construction area. Finally, yet other Plaintiffs, such as John Does V and IX, testified that they were required to serve as "pipeline porters" — workers who performed menial tasks such as hauling materials and cleaning the army camps for the soldiers guarding the pipeline construction. These serious allegations create triable questions of fact as to whether the Myanmar Military implemented a policy of forced labor in connection with its work on the pipeline.

The evidence also supports the conclusion that Unocal gave practical assistance to the Myanmar Military in subjecting Plaintiffs to forced labor.29 The practical assistance took the form of hiring the Myanmar Military to provide security and build infrastructure along the pipeline route in exchange for money or food. The practical assistance also took the form of using photos, surveys, and maps in daily meetings to show the Myanmar Military where to provide security and build infrastructure.

This assistance, moreover, had a "substantial effect" on the perpetration of forced labor, which "most probably would not have occurred in the same way" without someone hiring the Myanmar Military to provide security, and without someone showing them where to do it. Tadic at ¶ 688. This conclusion is supported by the admission of Unocal Representative Robinson that "[o]ur assertion that[the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny," and by the admission of Unocal President Imle that "[i]f forced labor goes hand and glove with the military yes there will be more forced labor." (Emphasis added.)

Second, a reasonable factfinder could also conclude that Unocal's conduct met the mens rea requirement of aiding and abetting as we define it today, namely, actual or constructive (i.e., reasonable) knowledge that the accomplice's actions will assist the perpetrator in the commission of the crime. The District Court found that "[t]he evidence does suggest that Unocal knew that forced labor was being utilized and that the Joint Venturers benefitted from the practice." Doe/Roe II, 110 F.Supp.2d at 1310. Moreover, Unocal knew or should reasonably have known that its conduct — including the payments and the instructions where to provide security and build infrastructure — would assist or encourage the Myanmar Military to subject Plaintiffs to forced labor.

Viewing the evidence in the light most favorable to Plaintiffs, we conclude that there are genuine issues of material fact whether Unocal's conduct met the actus reus and mens rea requirements for liability under the ATCA for aiding and abetting forced labor. Accordingly, we reverse the District Court's grant of Unocal's motion for summary judgment on Plaintiffs' forced labor claims under the ATCA.30

3. Murder, Rape, and Torture

a. Because Plaintiffs testified that the alleged acts of murder, rape, and torture occurred in furtherance of forced labor, state action is not required to give rise to liability under the ATCA.

Plaintiffs further allege that the Myanmar military murdered, raped or tortured a number of the plaintiffs. In section II.A.1., we adopted the Second Circuit's conclusion that "acts of rape, torture, and summary execution," like most crimes, "are proscribed by international law only when committed by state officials or under color of law" to the extent that they were committed in isolation. Kadic, 70 F.3d at 243-44. We, however, also adopted the Second Circuit's conclusion that these crimes "are actionable under the Alien Tort[Claims] Act, without regard to state action, to the extent that they were committed in pursuit of genocide or war crimes," id. at 244 (emphasis added), i.e., in pursuit of crimes, such as slavery, which never require state action for ATCA liability to attach. According to Plaintiffs' deposition testimony, all of the acts of murder, rape, and torture alleged by Plaintiffs occurred in furtherance of the forced labor program.31 As discussed above in section II.A.2.a, forced labor is a modern variant of slavery and does therefore never require state action to give rise to liability under the ATCA. Thus, under Kadic, state action is also not required for the acts of murder, rape, and torture which allegedly occurred in furtherance of the forced labor program.32

b. Unocal may be liable under the ATCA for aiding and abetting the Myanmar Military in subjecting Plaintiffs to murder and rape, but Unocal is not similarly liable for torture.

In section II.A.2.b, we adopted "knowing practical assistance [or] encouragement ... which has a substantial effect on the perpetration of the crime," from Furundzija at ¶¶ 235, 245, as a standard for aiding and abetting liability under the ATCA. The same reasons that convinced us earlier that Unocal may be liable under this standard for aiding and abetting the Myanmar Military in subjecting Plaintiffs to forced labor also convince us now that Unocal may likewise be liable under this standard for aiding and abetting the Myanmar Military in subjecting Plaintiffs to murder and rape. We conclude, however, that as a matter of law, Unocal is not similarly liable for torture in this case.

Initially we observe that the evidence in the record creates a genuine question of material fact as to whether Myanmar soldiers engaged in acts of murder and rape involving Plaintiffs. For instance, Jane Doe I testified that after her husband, John Doe I, attempted to escape the forced labor program, he was shot at by soldiers, and in retaliation for his attempted escape, that she and her baby were thrown into a fire, resulting in injuries to her and the death of the child. Other witnesses described the summary execution of villagers who refused to participate in the forced labor program, or who grew too weak to work effectively. Several Plaintiffs testified that rapes occurred as part of the forced labor program. For instance, both Jane Does II and III testified that while conscripted to work on pipeline-related construction projects, they were raped at knife-point by Myanmar soldiers who were members of a battalion that was supervising the work. The record does not, however, contain sufficient evidence to establish a claim of torture (other than by means of rape) involving Plaintiffs. Although a number of witnesses described acts of extreme physical abuse that might give rise to a claim of torture, the allegations all involved victims other than Plaintiffs. As this is not a class action, such allegations cannot serve to establish the Plaintiffs' claims of torture here.

Next, a reasonable factfinder could conclude that Unocal's alleged conduct met the actus reus requirement of aiding and abetting as we define it today, i.e., practical assistance or encouragement which has a substantial effect on the perpetration of the crimes of murder and rape. As just discussed, the evidence supports the conclusion that the Myanmar Military subjected Plaintiffs to acts of murder and rape while providing security and building infrastructure for the Project. The evidence also supports the conclusion that Unocal gave "practical assistance" to the Myanmar Military in subjecting Plaintiffs to these acts of murder and rape. The practical assistance took the form of hiring the Myanmar Military to provide security and build infrastructure along the pipeline route in exchange for money or food. The practical assistance also took the form of using photos, surveys, and maps in daily meetings to show the Myanmar Military where to provide these services. This assistance, moreover, had a "substantial effect" on the perpetration of murder and rape, which "most probably would not have occurred in the same way" without someone hiring the Myanmar Military to provide security, and without someone showing them where to do it. Tadic at ¶ 688. This conclusion is supported by the admission of Unocal Representative Robinson that "[o]ur assertions that [the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny." (Emphasis added.) This conclusion is further supported by Unocal Consultant Haseman's comment to Unocal that "[t]he most common [human rights violations] are forced relocation without compensation of families from land near/along the pipeline route; forced labor to work on infrastructure projects supporting the pipeline ...; and ... execution by the army of those opposing such actions." (Emphasis added.)33

Finally, a reasonable factfinder could also conclude that Unocal's conduct met the mens rea requirement of aiding and abetting as we define it today, i.e., actual or constructive (i.e., reasonable) knowledge that the accomplice's actions will assist the perpetrator in the commission of the crime. The District Court found that "Plaintiffs present[ed] evidence demonstrating ... that the military, while forcing villagers to work ..., committed numerous acts of violence; and that Unocal knew or should have known that the military did commit, was committing, and would continue to commit these tortious acts." Doe/Roe II, 110 F.Supp.2d at 1306. Moreover, Unocal knew or should reasonably have known that its conduct — including the payments and the instructions where to provide security and build infrastructure — would assist or encourage the Myanmar Military to subject Plaintiffs to these acts of violence. Under Furundzija, it is not even necessary that the aider and abettor knows the precise crime that the principal intends to commit. See id. at ¶ 246. Rather, if the accused "is aware that one of a number of crimes will probably be committed, and one of those crimes is in fact committed, he has intended to facilitate the commission of that crime, and is guilty as an aider and abettor." Id. Thus, because Unocal knew that acts of violence would probably be committed, it became liable as an aider and abettor when such acts of violence — specifically, murder and rape — were in fact committed.

Viewing the evidence in the light most favorable to Plaintiffs, we conclude that there are genuine issues of material fact whether Unocal's conduct met the actus reus and mens rea requirements for liability under the ATCA for aiding and abetting murder and rape. Accordingly, we reverse the District Court's grant of Unocal's motion for summary judgment on Plaintiffs' murder and rape claims under the ATCA. By contrast, the record does not contain sufficient evidence to support Plaintiffs' claims of torture. We therefore affirm the District Court's grant of Unocal's motion for summary judgment on Plaintiffs' torture claims.

B. The Myanmar Military and Myanmar Oil are entitled to immunity under the Foreign Sovereign Immunities Act.

Under the Foreign Sovereign Immunities Act, 28 U.S.C. §§ 1330, 1602 et seq., a district court has jurisdiction over a civil action against a foreign state such as Myanmar — including its political subdivisions, agencies, or instrumentalities, such as the Myanmar Military or Myanmar Oil — only if one of several exceptions to foreign sovereign immunity applies. See 28 U.S.C. §§ 1330(a), 1603(a), & 1605-1607. Specifically,

A foreign state shall not be immune from the jurisdiction of courts of the United States or of the States in any case... (2) in which the action is based [1] upon a commercial activity carried on in the United States by the foreign state; or [2] upon an act performed in the United States in connection with a commercial activity of the foreign state elsewhere; or [3] upon an act outside the territory of the United States in connection with a commercial activity of the foreign state elsewhere and that act causes a direct effect in the United States....

28 U.S.C. § 1605(a). The District Court rejected the Doe-Plaintiffs' argument that the second and third of the above exceptions gave the District Court jurisdiction over their claims against the Myanmar Military and Myanmar Oil. The existence of subject matter jurisdiction under the Foreign Sovereign Immunities Act is a question of law which this court reviews de novo. See Holden v. Canadian Consulate, 92 F.3d 918, 920 (9th Cir.1996).

The Doe-Plaintiffs argue that their claims against the Myanmar Military and Myanmar Oil fall within the second exception to foreign sovereign immunity in § 1605(a)(2) because they are based "upon an act performed in the United States in connection with a commercial activity of the foreign state elsewhere." We have held that under this exception, a foreign state is not immune from the jurisdiction of the courts of the United States only if an act performed in the United States is an element of the plaintiff's claim against the foreign state. See Holden, 92 F.3d at 920. In the present case, the Doe-Plaintiffs' claims against the Myanmar Military and Myanmar Oil are based exclusively upon acts allegedly performed by these foreign state defendants in Myanmar (forced labor, murder, rape, torture). The Doe-Plaintiffs do not allege that the Myanmar Military or Myanmar Oil performed any acts in the United States. Any acts allegedly performed by Unocal in the United States (investments decisions, money transfers) are not elements of the Doe-Plaintiffs' claims against the Myanmar Military and Myanmar Oil. The Doe-Plaintiffs' claims against the Myanmar Military and Myanmar Oil therefore do not fall within the second exception to foreign sovereign immunity in § 1605(a)(2).

The Doe-Plaintiffs also argue that their claims against the Myanmar Military and Myanmar Oil fall within the third exception to foreign sovereign immunity in § 1605(a)(2) because they are based "upon an act outside the territory of the United States in connection with a commercial activity of the foreign state elsewhere and that act causes a direct effect in the United States." The Supreme Court has held that "a state engages in commercial activity ... where it exercises only those powers that can also be exercised by private citizens, as distinct from those powers peculiar to sovereigns." Saudi Arabia v. Nelson, 507 U.S. 349, 360, 113 S.Ct. 1471, 123 L.Ed.2d 47 (1993) (internal quotation marks omitted). The District Court noted that "[the Myanmar Military] and [Myanmar Oil] engaged in commerce in the same manner as a private citizen might do when they allegedly entered into the ... gas pipeline project." Doe I, 963 F.Supp. at 887. The District Court further noted that "[i]n addition, [the Myanmar Military and Myanmar Oil] engaged in the acts upon which the claims are based `in connection with' that commercial activity." Id. at 887-88. The District Court concluded, however, that "[b]ecause[the Doe Plaintiffs] essentially allege that [the Myanmar Military] and [Myanmar Oil] abused their police power" when they engaged in these additional acts upon which the claims are based, these acts were exercises of powers peculiar to sovereigns and, therefore, "do not come within the commercial activity exception to the FSIA." Id. at 888.

The problem with this reasoning is that neither Nelson, nor other case law, nor the legislative history of § 1605(a)(2) suggest that a foreign state's conduct "in connection with a commercial activity" must itself be a commercial activity to fall within the third exception to foreign sovereign immunity. In other words, there is no support for the proposition that the foreign state's conduct "in connection with a commercial activity" must be an "exercise[ ][of] only those powers that can also be exercised by private citizens" to fall within the third exception in § 1605(a)(2). Nelson, 507 U.S. at 360, 113 S.Ct. 1471. Rather, as the Supreme Court observed in Nelson,"Congress manifestly understood there to be a difference between a suit `based upon' commercial activity and one `based upon' acts performed `in connection with' such activity." Id. at 358, 113 S.Ct. 1471.

The District Court looked for support for its contrary conclusion in a different passage in Nelson, where the Supreme Court held that even if a foreign government often used detention and torture to resolve commercial disputes, this would "not alter the fact that the powers allegedly abused where those of police and penal officers." Id. at 363, 113 S.Ct. 1471. In that passage, however, Nelson held only that the use of detention and torture to resolve commercial disputes would not qualify as a commercial activity and, therefore, fall within the first exception to foreign sovereign immunity, which is not at issue here. See 507 U.S. at 356, 113 S.Ct. 1471. But Nelson did not hold that such use of detention and torture also would not qualify as an act performed in connection with a commercial activity and, therefore, fall within the third exception to foreign sovereign immunity, which is at issue here.34

The District Court's misreading of Nelson was, nevertheless, harmless, because the Court correctly concluded that the alleged acts of murder, torture, rape, and forced labor by the Myanmar Military and Myanmar Oil did not have the direct effect in the United States required by the third exception to foreign sovereign immunity in § 1605(a)(2). In Siderman, we approved of the definition of a "direct effect" as one that "occurs at the locus of the injury directly resulting from the sovereign defendant's wrongful acts." Siderman, 965 F.2d at 710 n. 11 (quoting Restatement (Third) of the Foreign Relations Law of the United States § 453 Reporter's Note 5 (1987)). The injuries directly resulting from the Myanmar Military and Myanmar Oil's alleged wrongful acts were the murder, rape, torture, and forced labor of the Doe-Plaintiffs. The locus of these injuries was Myanmar. Therefore, any effects — such as Unocal's profits — occurring in the United States were not "direct effects" of these acts within the meaning of § 1605(a)(2). Accordingly, the District Court did not err when it concluded that the Doe-Plaintiffs' claims against the Myanmar Military and Myanmar Oil did not fall within the third exception to foreign sovereign immunity in § 1605(a)(2).

C. Plaintiffs' claims against Unocal are not barred by the Act of State Doctrine.

Unocal also argues that Plaintiffs' claims against it are barred by the "act of state" doctrine. The act of state doctrine is a non-jurisdictional, prudential doctrine based on the notion that "the courts of one country will not sit in judgment on the acts of the government of another, done within its own territory." Underhill v. Hernandez, 168 U.S. 250, 252, 18 S.Ct. 83, 42 L.Ed. 456 (1897). "Act of state issues only arise when a court must decide — that is, when the outcome of the case turns upon — the effect of official action by a foreign sovereign." W.S. Kirkpatrick & Co., Inc. v. Envtl. Tectonics Corp., Int'l, 493 U.S. 400, 406, 110 S.Ct. 701, 107 L.Ed.2d 816 (1990). As long as this requirement is met, the act of state doctrine can be invoked by private parties such as Unocal. See, e.g., Credit Suisse v. United States Dist. Court, 130 F.3d 1342, 1348 (9th Cir.1997). In the present case, an act of state issue arises because the court must decide that the conduct by the Myanmar Military violated international law in order to hold Unocal liable for aiding and abetting that conduct. We review the applicability of the act of state doctrine de novo. See Liu v. Republic of China, 892 F.2d 1419, 1424 (9th Cir.1989).

The Second Circuit has said that "it would be a rare case in which the act of state doctrine precluded suit under [the ATCA]." Kadic, 70 F.3d at 250. We find that the present case is not that rare case, and that the act of state doctrine does not preclude suit under the ATCA here.

In Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964), the Supreme Court developed a three-factor balancing test to determine whether the act of state doctrine should apply:

"[1][T]he greater the degree of codification or consensus concerning a particular area of international law, the more appropriate it is for the judiciary to render decisions regarding it.... [2][T]he less important the implications of an issue are for our foreign relations, the weaker the justification for exclusivity in the political branches. [3] The balance of relevant considerations may also be shifted if the government which perpetrated the challenged act of state is no longer in existence...."

Id. at 428, 84 S.Ct. 923. We have added a fourth factor to this test: [4] "[W]e must [also] consider ... whether the foreign state was acting in the public interest." Liu, 892 F.2d at 1432. With the exception of the third factor, all of these factors weigh against application of the act of state doctrine in this case.

Regarding the first factor — international consensus — we have recognized that murder, torture, and slavery are jus cogens violations, i.e., violations of norms that are binding on nations even if they do not agree to them. See Matta-Ballesteros, 71 F.3d at 764 n. 5; Siderman, 965 F.2d at 714-15. As discussed supra in section II.A.1., rape can be a form of torture and thus also a jus cogens violation. Similarly, as discussed supra in section II.A.2.a, forced labor is a modern form of slavery and thus likewise a jus cogens violation. Accordingly, all torts alleged in the present case are jus cogens violations. Because jus cogens violations are, by definition, internationally denounced, there is a high degree of international consensus against them, which severely undermines Unocal's argument that the alleged acts by the Myanmar Military and Myanmar Oil should be treated as acts of state.

Regarding the second factor — implications for our foreign relations — the coordinate branches of our government have already denounced Myanmar's human rights abuses and imposed sanctions. It is also worth noting that in 1997, the State Department advised the District Court that "at this time adjudication of the claims based on allegations of torture and slavery would not prejudice or impede the conduct of U.S. foreign relations with the current government of Burma." Roe I, 176 F.R.D. at 362. This statement of interest at the dismissal stage is not conclusive at this later stage, especially in light of the fact that "[t]he Executive Branch ... cannot by simple stipulation change a political question into a cognizable claim." First Nat'l City Bank v. Banco Nacional de Cuba, 406 U.S. 759, 788-89, 92 S.Ct. 1808, 32 L.Ed.2d 466 (1972) (Brennan, J., dissenting). But the statement is also not irrelevant. See Banco Nacional de Cuba v. Chemical Bank N.Y. Trust Co., 594 F.Supp. 1553, 1563 (S.D.N.Y.1984) (stating that courts "may, as a matter of discretion, accept the views of the State Department"). We agree with the District Court's evaluation that "[g]iven the circumstances of the instant case, and particularly the Statement of Interest of the United States, it is hard to imagine how judicial consideration of the matter will so substantially exacerbate relations with [the Myanmar Military] as to cause hostile confrontations." Roe I, 176 F.R.D. at 354 n. 29 (internal quotation marks omitted).

Regarding the third factor — continued existence of the accused government — the Myanmar Military is still the government of Myanmar, although it changed its full name from State Law and Order Restoration Council to State Peace and Development Council following the events at issue here. That a condemnation of the alleged acts may offend the current government of Myanmar is the only factor that weighs in favor of applying the act of state doctrine.

Finally, regarding the fourth factor that we have imposed — public interest — it would be difficult to contend that the Myanmar Military and Myanmar Oil's alleged violations of international human rights were "in the public interest." Indeed, the District Court found at the summary judgment stage that "there is an issue of fact as to whether the forced labor was used to benefit the Project as opposed to the public's welfare." Doe/Roe II, 110 F.Supp.2d at 1308. This genuine issue of material fact precludes summary judgment in favor of Unocal on this basis.

Because the four factor balancing test weighs against applying the act of state doctrine, we find that Plaintiffs' claims are not barred by this doctrine.

D. The District Court lacked extraterritorial subject matter jurisdiction over the Doe-Plaintiffs' RICO claim against Unocal.

The Doe-Plaintiffs allege that Unocal's conduct violated the Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. § 1961 et seq. RICO makes it unlawful, inter alia,"for any person employed by or associated with any enterprise engaged in, or the activities of which affect, interstate or foreign commerce, to conduct or participate, directly or indirectly, in the conduct of such enterprise's affairs through a pattern of racketeering activity," or to conspire in such conduct. 18 U.S.C. § 1962(c),(d). "Racketeering activity" is partially defined as any act which is indictable under any one of a number of listed provisions of Title 18 of the United States Code. See 18 U.S.C. § 1961(1)(B). The Doe-Plaintiffs allege that Unocal engaged and conspired in a "pattern of extortion" that is indictable under the Hobbs Act, 18 U.S.C. § 1951, one of the provisions enumerated in RICO's definition of "racketeering activity." The Hobbs Act provides in relevant part:

Whoever in any way or degree obstructs, delays, or affects commerce or the movement of any article or commodity in commerce, by robbery or extortion [,] or attempts or conspires so to do, or commits or threatens physical violence to any person or property in furtherance of a plan or purpose to do anything in violation of this section[,] shall be fined under this title or imprisoned not more than twenty years, or both.

18 U.S.C. § 1951(a).

The District Court granted Unocal's motion for summary judgment on the Doe-Plaintiffs' RICO claim for lack of subject matter jurisdiction. We review the existence of subject matter jurisdiction under RICO de novo. See United States v. Juvenile Male, 118 F.3d 1344, 1346 (9th Cir.1997).

The Doe-Plaintiffs base their underlying Hobbs Act claim on the alleged "extortion" of their labor. The Hobbs Act defines "extortion" as the "obtaining of property from another, with his consent, induced by wrongful use of actual or threatened force, violence, or fear, or under color of official right." 18 U.S.C. § 1951(b)(2). We have observed that "[t]he concept of property under the Hobbs Act has not been limited to physical or tangible `things.'" United States v. Zemek, 634 F.2d 1159, 1174 (9th Cir.1980). Thus we have recognized the "right ... to solicit business free from wrongful coercion," id., and the "right to make personal and business decisions about the purchase of life insurance on [one's] own life free of threats," United States v. Hoelker, 765 F.2d 1422, 1425 (9th Cir.1985), as property rights that are protected by the Hobbs Act. More generally, the Second Circuit has held that "[t]he concept of property under the Hobbs Act ... includes, in a broad sense, any valuable right considered as a source or element of wealth." United States v. Tropiano, 418 F.2d 1069, 1075 (2d Cir.1969). The right to make personal and business decisions about one's own labor also fits this definition of "property." Forced labor allegations can, therefore, form the basis of a Hobbs Act claim, and this claim can, in turn, form the basis of a RICO claim.

The District Court nevertheless correctly granted summary judgment in favor of Unocal on the Doe-Plaintiffs' RICO claim for lack of extraterritorial subject matter jurisdiction. We agree with the Second Circuit that for RICO to apply extraterritorially, the claim must meet either the "conduct" or the "effect" test that courts have developed to determine jurisdiction in securities fraud cases. See North South Fin. Corp. v. Al-Turki, 100 F.3d 1046, 1051 (2d Cir.1996); see also Butte Mining PLC v. Smith, 76 F.3d 287, 291 (9th Cir.1996) (holding with respect to extraterritorial application of RICO that "[o]nce the securities fraud claim was dismissed [for lack of extraterritorial subject matter jurisdiction under the "conduct" or the "effect" test,] the wire and mail fraud and RICO claims that related to this fraud had to be dismissed as well"). The Doe-Plaintiffs do not challenge that they must meet one of these two test to succeed on their RICO claim. Instead, they challenge the District Court's conclusion that they cannot meet either test.

Under the "conduct" test, a district court has jurisdiction over securities fraud suits by foreigners who have lost money through sales abroad "[o]nly where conduct `within the United States directly caused' the loss." Psimenos v. E.F. Hutton & Co., 722 F.2d 1041, 1046 (2d Cir.1983) (emphasis added) (quoting Bersch v. Drexel Firestone, Inc., 519 F.2d 974, 993 (2d Cir.1975)). "Mere preparatory activities, and conduct far removed from the consummation of the fraud, will not suffice to establish jurisdiction." Id. (emphasis added).

Under the "effects" test, "[t]he anti-fraud laws of the United States may be given extraterritorial reach whenever a predominantly foreign transaction has substantial effects within the United States." Consol. Gold Fields PLC v. Minorco, S.A., 871 F.2d 252, 261-62 (2d Cir.1989). This test is met where the domestic effect is "a direct and foreseeable result of the conduct outside of the United States." Id. at 262 (emphasis added). By contrast, "courts have been reluctant to apply our laws to transactions that have only remote and indirect effects in the United States." Id. (emphasis added).

The "conduct" and the "effect" test appear to be two sides of one coin. The "conduct" test establishes jurisdiction for domestic conduct that directly causes foreign loss or injury. Conversely, the "effects" test establishes jurisdiction for foreign conduct that directly causes domestic loss or injury. The conduct involved in this case does not meet either of these two tests.

The Doe-Plaintiffs allege that in furtherance of an unlawful conspiracy, Unocal transferred significant financial and technical support for Project activity from the United States to Myanmar. Under the "conduct" test, the question is whether this transfer from the United States "directly caused" loss or injury in Myanmar. We conclude that it did not. In Butte Mining, the plaintiffs alleged that the defendants used domestic mail and wire "to further" a foreign securities fraud. 76 F.3d at 291. In that case, we found "no reason to extend the jurisdictional scope of RICO to make criminal the use of the mail and wire in the United States as part of an alleged fraud outside the United States." Id. Similarly, in the present case, we find no reason to extend the jurisdictional scope of RICO to create civil liability for the transfer of monies and technical support from the United States as part of an alleged "pattern of extortion" outside the United States. We therefore hold that the Doe-Plaintiffs' allegations do not satisfy the "conduct" test.

Nor have the Doe-Plaintiffs pointed to any evidence that Unocal's alleged conduct in Myanmar "directly caused" loss or injury in the United States and thus satisfied the "effects" test. "If the party moving for summary judgment meets its initial burden of identifying for the court the portions of the materials on file that it believes demonstrate the absence of any genuine issue of material fact, ... the nonmoving party may not rely on the mere allegations in the pleadings in order to preclude summary judgment," but instead "must set forth ... `specific facts showing that there is a genuine issue for trial.'" T.W. Elec. Serv., Inc. v. Pac. Elec. Contractors Ass'n, 809 F.2d 626, 630 (9th Cir.1987) (quoting Fed.R.Civ.P. 56(e)). The Doe-Plaintiffs assert in their Opening Brief that Unocal's actions in Myanmar gave Unocal an "unfair advantage over competitors" in the United States. The Doe-Plaintiffs, however, do not point to any "specific facts" in the record to support these conclusory allegations, as they are required to do by Fed.R.Civ.P. 56(e). These "mere allegations" are not enough to survive Unocal's motion for summary judgment on the Doe-Plaintiffs' RICO claim. We therefore hold that the Doe-Plaintiffs' allegations also do not meet the "effects" test.

III.

CONCLUSION

For the foregoing reasons, we REVERSE the District Court's grant of summary judgment in favor of Unocal on Plaintiffs' ATCA claims for forced labor, murder, and rape.35 We however AFFIRM the District Court's grant of summary judgment in favor of Unocal on Plaintiffs' ATCA claims for torture. We further AFFIRM the District Court's dismissal of all of the Doe-Plaintiffs' claims against the Myanmar Military and Myanmar Oil. We also AFFIRM the District Court's grant of summary judgment in favor of Unocal on the Doe-Plaintiffs' RICO claim against Unocal. We REMAND the case to the District Court for further proceedings consistent with this opinion.

Each party to bear its own costs.

REVERSED IN PART, AFFIRMED IN PART and REMANDED.

Notes:

1 Judge Paez initially authored the orders granting in part and denying in part Defendants' Motions to DismissSee Doe I v. Unocal Corp., 963 F.Supp. 880 (C.D.Cal.1997); Nat'l Coalition Gov't of the Union of Burma v. Unocal, Inc., 176 F.R.D. 329 (C.D.Cal.1997). Judge Lew later authored the order granting Defendants' consolidated Motions for Summary Judgment. See Doe I v. Unocal Corp., 110 F.Supp.2d 1294 (C.D.Cal.2000).

2 The Unocal Offshore Co. was originally owned by the Unocal International Corporation, a Delaware corporation and wholly-owned subsidiary of the Union Oil Company of California. In 1999, ownership of the Unocal Offshore Co. was transferred to Unocal Global Ventures, Ltd., a Bermuda corporation and wholly owned subsidiary of the Unocal International Corporation, "to achieve tax and cash management efficiencies."

3 The Unocal Pipeline Corp. was also originally owned by the Unocal International Corporation. In 1998, ownership of the Unocal Pipeline Corp. was transferred to Unocal Global Ventures, Ltd

4 Although anti-government rebels were active elsewhere in Myanmar, the record indicates that there was in fact little to no rebel activity in the region where the pipeline construction occurred, and that the center of the Myanmar civil war was 150-200 miles distant from the pipeline project

5 Moreover, in March 1996, a cable from the U.S. Embassy in Rangoon reflects the Embassy's understanding that "the consortium building the pipeline pays the Burmese military a hard-currency fee for providing security."

6 In the same year, the U.S. Department of State similarly reported that "[t]he military Government [in Myanmar] routinely employs corvee labor on its myriad building projects" and that "[t]he Burmese army has for decades conscripted civilian males to serve as porters." U.S. Department of State,Country Reports on Human Rights Practices for 1991 796-97 (1992).

7 As noted above, the Production Sharing Contract between Total Myanmar and Myanmar Oil provided that "[Myanmar Oil] shall... supply[ ] or mak[e] available ... security protectionas may be requested by [Total Myanmar and its assigns]," such as Unocal. (Emphasis added.)

8 Similarly, the briefing book that Total prepared for Unocal President Imle and Unocal CEO Beach on the occasion of their April 1996 visit to the Project listed the following "area[] of concern": "army = additional burden on the local population."

9 Also in 1995, Human Rights Watch informed Unocal that forced labor was so pervasive in Myanmar that Human Rights Watch could not condone any investment that would enrich the country's current regime. That same year, the General Assembly of the United Nations "strongly urge[d] the Government of Myanmar... to put an end to ... the practices of torture, abuse of women, forced labour ..., and ... disappearances and summary executions...."Situation of Human Rights in Myanmar, U.N. General Assembly, 50th Sess., Agenda Item 112(c), U.N. Doc. A/RES/50/194 (1995), http:www.un.org/documents/ga/res/50/ ares50-194.htm.

10 Similarly, on May 20, 1996, a State Department cable stated: "Forced labor is currently being channeled, according to[non-governmental organization] reports, to service roads for the pipeline to Thailand.... There are plans for a helicopter pad and airstrip in the area ... in part for use by oil company executives."

11 Plaintiffs in both actions subsequently filed amended complaints that do not contain claims based on expropriation of property

12 The "law of nations" is "the law of international relations, embracing not only nations but also ... individuals (such as those who invoke their human rights or commit war crimes)."Black's Law Dictionary 822 (7th ed.1999).

13 Plaintiffs' ATCA claims are timely under the ten-year statute of limitations we recently adopted for such claimsSee Papa, 281 F.3d at 1011-13.

14 Jus cogens norms are norms of international law that are binding on nations even if they do not agree to them. See Siderman de Blake v. Republic of Argentina, 965 F.2d 699, 714-15 (9th Cir.1992).

15 We stress that although ajus cogens violation is, by definition, "a violation of `specific, universal, and obligatory' international norms" that is actionable under the ATCA, any "violation of `specific, universal, and obligatory' international norms" — jus cogens or not — is actionable under the ATCA. Papa, 281 F.3d at 1013 (quoting Marcos II, 25 F.3d at 1475). Thus, a jus cogens violation is sufficient, but not necessary, to state a claim under the ATCA.

16 Our statement inIn re Estate of Ferdinand E. Marcos Human Rights Litig., 978 F.2d 493, 501-02 (9th Cir.1992) ("Marcos I"), that "[o]nly individuals who have acted under official authority or under color of such authority may violate international law," must be read like Judge Edwards' concurrence in Tel-Oren, on which this statement exclusively relied. Marcos I, like Tel-Oren, involved torture, a crime for which there is no purely private liability under international law. See Tel-Oren, 726 F.2d at 794-95 (Edwards, J., concurring); Kadic, 70 F.3d at 243.

17 The Thirteenth Amendment provides in part that "[n]either slavery nor involuntary servitude ... shall exist within the United States." U.S. CONST. amend. XIII, § 1. See also Tobias Barrington,The Thirteenth Amendment and Slavery in the Global Economy, 102 Colum. L. Rev. 973 (2002), for the proposition that "the knowing use of slave labor by U.S. based entities in their foreign operations constitutes the presence of `slavery' within the United States, as that term is used in the Thirteenth Amendment," id. at 978, and that "[i]f the allegations against it are true, then Unocal's participation in the Burma project makes out a strong case for a Thirteenth Amendment violation," id. at 1034.

18 The fact that the Thirteenth Amendment reaches private action,see Jones v. Alfred H. Mayer Co., 392 U.S. 409, 438-39, 88 S.Ct. 2186, 20 L.Ed.2d 1189 (1968), in turn supports the view that forced labor by private actors gives rise to liability under the ATCA.

19 The statute provides that anybody who kidnaps any other person, or induces such other person to go anywhere, with the intent that such other person be sold into involuntary servitude or held as a slave, shall be fined or imprisoned as specifiedSee 18 U.S.C. § 1581.

20 Plaintiffs also argue that Unocal is liable for the conduct by the Myanmar Military under joint venture, agency, negligence, and recklessness theories. The District Court did not address any of Plaintiffs' alternative theories. Because we reject the District Court's general reasons for holding that Unocal could not be liable under international law, and because we hold that Unocal may be liable under at least one of Plaintiffs' theories, i.e., aiding and abetting in violation of international law, we do not need to address Plaintiffs' other theories, i.e., joint venture, agency, negligence, and recklessness. Joint venture, agency, negligence, and recklessness may, like aiding and abetting, be viable theories on the specific facts of this ATCA case. Moreover, on the facts of other ATCA cases, joint venture, agency, negligence, or recklessness may in fact be more appropriate theories than aiding and abetting

21 The Military Tribunal in one of these case defined the necessity defense as follows: "Necessity is a defense when it is shown that the act charged was done to avoid an evil both serious and irreparable; that there was no other adequate means of escape; and that the remedy was not disproportionate to the evil."United States v. Krupp, 9 Trials of War Criminals Before the Nuremberg Military Tribunals Under Control Council Law No. 10, 1436 (1950) (["Krupp"]). (quoting 1 Wharton's Criminal Law 177 (12th ed.1932)).

22 A reasonable factfinder could moreover conclude that Unocal's conduct met the "active participation" standard erroneously applied by the District Court. For example, Unocal Representative Robinson stated that "[o]ur assertion that [the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny." Robinson is furthermore reported to have stated that "Total/Unocal uses [photos, maps, and surveys] to show the military where they need helipads built and facilities secured." In addition, Unocal President Imle stated that "[i]f forced labor goes hand in glove with the military yes there will be more forced labor" as the result of the Myanmar Military protecting the pipeline. Unocal thus resembles the defendants inKrupp, who "well knew that any expansion[of their business] would require the employment of forced labor," 9 Trials at 1442, and the defendants in United States v. Flick, 6 Trials of War Criminals Before the Nuremberg Military Tribunals Under Control Council Law No. 10 (1952), who sought to increase their production quota and thus their forced labor allocation, id. at 1198, 1202.

23 Because "the law of nations is part of federal common law,"Marcos I, 978 F.2d at 502, the choice between international law and the law of the forum state, which in the present case is California state law or our federal common law, is less crucial than the choice between international law and the law of the state where the underlying events occurred, which in the present case is the law of Myanmar. Moreover, as discussed later in this section, the standard for aiding and abetting in international criminal law is similar to the standard for aiding and abetting in domestic tort law, making the choice between international and domestic law even less crucial.

24 Because "moral support" is not part of the standard we adopt today, the concurrence's discussion in this context of "the international law regarding third party `moral support' "is beside the point. Concurrence at 967-68,see infra note 28.

25 We stress that our conclusion that application of international law is appropriate is based on the record in this case. In other cases with different facts, application of the law of the forum state — including federal common law — or the law of the state where the events occurred may be appropriate

26 TheFurundzija Tribunal based its actus reus standard for aiding and abetting on an exhaustive analysis of international case law and international instruments. See id. at ¶¶ 192-234. The international case law it considered consisted chiefly of decisions by American and British military courts and tribunals dealing with Nazi war crimes, as well as German courts in the British and French occupied zones dealing with such crimes in the aftermath of the Second World War. See id. at ¶¶ 195-97. The international instruments consisted of the Draft Code of Crimes Against the Peace and Security of Mankind adopted by the United Nations International Law Commission in 1996, as well as the Rome Statute of the International Criminal Court "adopted by an overwhelming majority of the States attending the Rome Diplomatic Conference and ... substantially endorsed by the General Assembly's Sixth Committee on 26 November 1998." Id. at 227. It is hard to argue with the Furundzija Tribunal's reliance on these sources.

27 TheFurundzija Tribunal based its mens rea standard for aiding and abetting on an analysis of the same international case law and international instruments mentioned above in note 26. See id. at ¶¶ 236-49. The Tribunal's reliance on these sources again seems beyond reproach.

28 We note, however, that there may be no difference between encouragement and moral supportSee Restatement (Second) of Torts § 876 cmt. d (stating that "encouragement to act operates as a moral support"). The concurrence claims: "Having declared ... that the Yugoslav Tribunal's standard constitutes the controlling international law, the majority cannot then escape the full implications of being bound by the law it has selected" and "has lost whatever opportunity it had to pick and chose the aspects of international law it finds appealing." Concurrence at 970 n. 9. But nowhere in this opinion have we declared that the Yugoslav Tribunal's standard "constitutes the controlling international law," id. (emphasis added), and as a result, we are also not "bound" by every aspect of that standard, the concurrence's protestations notwithstanding. In fact, we have merely declared that "[w]e find recent decisions by the International Criminal Tribunal for the former Yugoslavia and the International Criminal Tribunal for Rwanda especially helpful for ascertaining the current standard for aiding and abetting under international law as it pertains to the ATCA." Supra at 950. That is, we have done no more than declare that the decisions by these tribunals are one of the sources of international law, rather than the source of international law. Having done so, we then concluded that with respect to practical assistance and encouragement, these decisions accurately reflect "the current standard for aiding and abetting under international law as it pertains to the ATCA," and have left open the question whether this is also true with respect to moral support. This procedure is not particularly noteworthy, let alone improper. And the concurrence's repeated references to "the Yugoslav Tribunal's `moral support' standard," concurrence at 969, 970, are at best irrelevant and at worst intended to suggest that we, albeit unwittingly, adopted a standard which we, in fact, did not adopt, unwittingly or otherwise.

29 The evidence further supports the conclusion that Unocal gave "encouragement" to the Myanmar Military in subjecting Plaintiffs to forced labor. The daily meetings with the Myanmar Military to show it where to provide security and build infrastructure, despite Unocal's knowledge that the Myanmar Military would probably use forced labor to provide these services, may have encouraged the Myanmar Military to actually use forced labor for the benefit of the Project. Similarly, the payments to the Myanmar Military for providing these services, despite Unocal's knowledge that the Myanmar Military had actually used forced labor to provide them, may have encouraged the Myanmar Military to continue to use forced labor in connection with the Project

30 Unocal argues that "Unocal is not vicariously liable for the Myanmar military's torts because the pipeline was constructed by a separate corporation," i.e., the Gas Transportation Company, and because "[t]here is no basis to pierce the corporate veils of [the Unocal Pipeline Corp.] or [the Unocal Offshore Co.]" We initially observe that there is evidence allowing a reasonable factfinder to conclude that the Unocal Pipeline Corp. and the Unocal Offshore Co. were alter egos of Unocal, and that any actions by the Unocal Pipeline Corp. or the Unocal Offshore Co. are therefore attributable to Unocal. This evidence includes the Unocal Pipeline Corp.'s and the Unocal Offshore Co.'s undercapitalization and the direct involvement in and direction of the Unocal Pipeline Corp.'s and the Unocal Offshore Co.'s business by Unocal President Imle, Unocal CEO Beach, and other Unocal officers and employeesSee Wiwa v. Royal Dutch Petroleum Co., No. 96 Civ. 8386(KMW), 2002 WL 319887, *13 n. 14 (S.D.N.Y. Feb.28, 2002) (holding in the ATCA context that "[b]y involving themselves directly in [their subsidiary's] activities, and by directing these activities, [parent companies] made [their subsidiary] their agent with respect to the torts alleged in the complaint"). More importantly, we do not address — and neither did the District Court — whether a reasonable factfinder could hold Unocal "vicariously liable for the Myanmar military's torts." (Emphasis added.) See supra note 20. Rather, we find that there is sufficient evidence to hold Unocal liable based on its own actions and those of its alter ego subsidiaries which aided and abetted the Myanmar Military in perpetrating forced labor. These actions include the employment of the Myanmar Military to provide security and build infrastructure along the pipeline route, and the use of photos, surveys, and maps to show the Myanmar Military where to do this. Unocal took these actions with the knowledge that the Myanmar army was likely to use and did in fact use forced labor "on behalf of the Project."

31 In addition, some of the acts of murder, rape, and torture alleged by non-party witnesses apparently didnot occur in furtherance of the forced labor program. Because this is not a class action, the context in which tortious acts alleged by non-party witnesses took place is immaterial to this discussion.

32 Because state action is not required in the present case, the District Court erred when it required a showing that Unocal "controlled" the Myanmar Military's decision to commit the alleged acts or murder, rape, and torture to establish that Unocal proximately caused these actsSee Doe/Roe II, 110 F.Supp.2d at 1307. We require "control" to establish proximate causation by private third parties only in cases — under, e.g., 42 U.S.C. § 1983 — where we otherwise require state action. See, e.g., Arnold, 637 F.2d at 1356-57. In other cases — including cases such as this one — where state action is not otherwise required, we require no more than "forseeability" to establish proximate causation. See id. at 1355. This requirement is easily met in the present case, where Unocal Vice President Lipman testified that even before Unocal invested in the Project, Unocal was aware that "the option of having the [Myanmar] [M]ilitary provide protection for the pipeline construction ... would [entail] that they might proceed in the manner that would be out of our control and not be in a manner that we would like to see them proceed," i.e., "going to excess." (Emphasis added.)

33 The evidence also supports the conclusion that Unocal gave "encouragement" to the Myanmar Military in subjecting Plaintiffs to murder, rape, and torture. The daily meetings with the Myanmar Military to show it where to provide security and build infrastructure, despite Unocal's knowledge that the Myanmar Military would probably use murder, rape, and torture in the process, may have encouraged the Myanmar Military to actually use murder, rape, and torture. Similarly, the payments to the Myanmar Military for providing these services, despite Unocal's knowledge that the Myanmar Military had actually used murder, rape, and torture in the process, may have encouraged the Myanmar Military to continue to use murder, rape, and torture

34 For the same reason, and contrary to the District Court's conclusion,Nelson also does not "undermine" our holding in Siderman, 965 F.2d 699, another case involving the third — rather than the first — exception in § 1605(a)(2).

35 Even if we were to affirm the District Court's grant of summary judgment on Plaintiffs' ATCA claims for forced labor, murder, and rape, we would still reverse the District Court's denial of Plaintiffs' Fed.R.Civ.P. 54(d)(1) Motion to Retax. The District Court concluded that Plaintiffs' motion was, in actuality, a time-barred Fed.R.Civ.P. 59(e) Motion to Alter or Amend Judgment. The Supreme Court has observed, however, that Rule 59(e) covers only motions to reconsider "matters properly encompassed in a decision on the merits," and does not cover motions that raise "legal issues collateral to the main cause of action."White v. New Hampshire Dep't of Employment Sec., 455 U.S. 445, 451, 102 S.Ct. 1162, 71 L.Ed.2d 325 (1982). See also Whittaker v. Whittaker Corp., 639 F.2d 516, 520-21 (9th Cir.1981). In their motion, Plaintiffs asked the District Court "not [to] require the plaintiffs to pay any of this cost bill" because of their indigency and the chilling effect of an award of costs. Plaintiffs' indigency and the chilling effect of an award of costs are not "matters properly encompassed in a decision on the merits." Rather, they are "legal issues collateral to the main cause of action." Plaintiffs' motion, therefore, did not have to be brought as a Motion to Alter or Amend Judgment within ten days of the judgment on the merits under Rule 59(e). Instead, it could be — and in fact was — brought as a Motion to Retax within five days of the taxing of the costs under Rule 54(d)(1). Plaintiffs' motion was thus timely.

REINHARDT, Circuit Judge, concurring.

I agree with the majority opinion, except for Part II(A), in which the majority discusses the Alien Tort Claims Act. As to that Act, I agree with the majority that material factual disputes exist regarding plaintiffs' claims for forced labor used in connection with the Yadana Pipeline Project. I also agree with the majority that if plaintiffs prove their allegations, Unocal may be held liable under the Act for the use of forced labor as a part of the project. Where I differ from my colleagues is principally with respect to the standard of third-party liability under which Unocal may be held legally responsible for the human rights violations alleged. I do not agree that the question whether Unocal may be held liable in tort for the Myanmar military's alleged human rights violations should be resolved, as the majority holds, by applying a recently-promulgated international criminal law aiding-and-abetting standard that permits imposition of liability for the lending of moral support. In fact, I do not agree that the question of Unocal's tort liability should be decided by applying any international law test at all. Rather, in my view, the ancillary legal question of Unocal's third-party tort liability should be resolved by applying general federal common law tort principles, such as agency, joint venture, or reckless disregard. I also believe that there is no reason to discuss the doctrine of jus cogens in this case. Because the underlying conduct alleged constitutes a violation of customary international law, the violation was allegedly committed by a governmental entity, and Unocal's liability, if any, is derivative of that government entity's, jus cogens is irrelevant to any issue before us. Assuming the allegations to be true, the fact that the underlying conduct violated customary international law is sufficient to support liability not only on the part of the governmental actor, but also on the part of a third party whose liability is derivative thereof.

1. Forced Labor As A Violation of the Law of Nations

In order to bring an action under the Alien Tort Claims Act, an alien plaintiff must allege a tort committed in violation of the law of nations. Hilao v. Estate of Marcos, 25 F.3d 1467, 1475 (9th Cir.1994). I agree with the majority that the plaintiffs have alleged the requisite international law violation, and that a genuine issue of material fact exists regarding whether forced labor was used by the Myanmar government in connection with the Yadana Project. Because the majority opinion thoroughly sets forth the plaintiffs' serious allegations, and the evidence supporting those allegations, I do not repeat them here.

There can be little doubt that the use of forced labor violates widely-held international legal norms. Forced labor is banned by the Universal Declaration of Human Rights, the International Covenant of Civil and Political Rights, and the International Covenant of Economic, Social and Political Rights. See Universal Declaration of Human Rights, G.A. Res. 217(A)III (1948); International Covenant on Civil and Political Rights, art. 22, 999 U.N.T.S. 171, 173-74; 6 I.L.M. 368; International Covenant of Economic, Social and Political Rights, art. 8, 993 U.N.T.S. 3, 4; 6 I.L.M. 360. Forced labor was listed as a war crime in the charter of the Nuremberg Tribunal. See Agreement for the Prosecution and Punishment of the Major War Criminals of the European Axis, and Charter of the International Military Tribunal, Aug. 8, 1945, art. 6, 82 U.N.T.S. 280. Moreover, it is banned by two of the most widely-adopted international labor conventions. Convention Concerning the Abolition of Forced Labour (No. 105), June 25, 1957, 320 U.N.T.S. 291; Convention Concerning Forced or Compulsory Labour (No. 29), June 28, 1930, 39 U.N.T.S. 55. In light of these legal authorities, the allegations of forced labor practices, if true, constitute a violation of customary international law and, in any event, are sufficient to confer jurisdiction under the ATCA.

In contrast, the majority states that plaintiffs have alleged the necessary international law violation because forced labor is a modern variant of slavery, which is a jus cogens or "peremptory norm" of international law.1 In fact, whether or not forced labor is a modern variant of slavery is of no legal consequence in this case, because there is no requirement that plaintiffs state a jus cogens violation in order to obtain jurisdiction under the ATCA. It is true that a cause of action against non-state actors for conduct in which they engage directly exists only for acts that constitute jus cogens violations and that other conduct of private parties that would violate international law if engaged in by a governmental entity is not actionable under the ATCA. See Kadic v. Karadzic, 70 F.3d 232, 240 (2d Cir.1995). Here, however, if Unocal is held liable, it will be because the Myanmar military committed the illegal acts and Unocal is determined to be legally responsible for that governmental conduct under a theory of third-party liability — not because Unocal itself engaged in acts transgressing international law. Because the violations of customary international law, if they occurred, were committed by a governmental agency, third-party liability may attach regardless of whether the international law violated is jus cogens.2 Thus, I see no need to discuss whether forced labor is a modern variant of slavery, which would render it a jus cogens norm, or even whether the prohibition on forced labor is itself a jus cogens norm, which it may well be. See Princz v. Fed. Republic of Germany, 26 F.3d 1166, 1179-81 (D.C.Cir.1994) (Wald, J., dissenting on other grounds). The well-established principle that forced labor practices violate customary international law is sufficient in itself to confer jurisdiction in this case with respect to all parties, jus cogens or not.

2. The Appropriate Source of Law for Determining Third-Party Liability

If the plaintiffs can prove their allegations that the Myanmar military instituted a policy of forced labor, they would satisfy the Alien Tort Claims Act requirement of a violation of the law of nations. Then, in order to prevail on its claims against Unocal, plaintiffs would have to prove that the private entity may be held legally responsible for the Myanmar military's human rights violations. The latter requirement raises important questions of first impression: Under what circumstances may a private entity doing business abroad be held accountable in federal court for international law violations committed by the host government in connection with the business activities of the private entity; and to what body of law do we look in order to determine the answer? Logically, it is necessary to consider the second question first. In my view, the answer is that we look to traditional civil tort principles embodied in federal common law, rather than to evolving standards of international law, such as a nascent criminal law doctrine recently adopted by an ad hoc international criminal tribunal.

Plaintiffs allege that Unocal should be held liable for Myanmar's forced labor actions with respect to the pipeline under a number of international law theories, as well as under several theories based on federal common law principles. The text of the Alien Tort Claims Act states only that federal courts have jurisdiction over torts constituting a violation of "the law of nations." It is thus clear from the face of the statute that international law applies to determine whether a violation has occurred. The statute is silent, however, 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 governmental entity's violation of the law of nations. The majority elects to apply international law principles to resolve such issues. I strongly disagree. I believe that we are required to look to federal common law to resolve ancillary legal issues that arise in ATCA cases.

Following Erie Railroad Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938), federal courts ordinarily apply federal common law in limited circumstances, usually when authorized to do so by Congress. However, actions involving international relations constitute one category of cases in which federal common law is frequently applied. The Supreme Court has stated that even without congressional authority to develop federal common law, the federal courts should apply such 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 our relations with foreign nations, and admiralty cases." Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 641, 101 S.Ct. 2061, 68 L.Ed.2d 500 (1981); see also Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 425, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964) ("It seems fair to assume that the Court did not have rules like the act of state doctrine in mind when it decided Erie R. Co. v. Tompkins."). Because Alien Tort Claims Act cases involve the violation of international law, they almost always "implicat[e] ... our relations with foreign nations." Texas Industries, 451 U.S. at 641, 101 S.Ct. 2061.3 There are thus unique federal interests involved in Alien Tort Claims Act cases that support the creation of a uniform body of federal common law to facilitate the implementation of such claims.

There is another reason why the application of federal common law is appropriate here: we are required to resolve issues ancillary to a cause of action created by Congress. The Supreme Court has stated that in such cases, courts should apply federal common law "to fill the interstices of federal legislation." United States v. Kimbell Foods, Inc., 440 U.S. 715, 727, 99 S.Ct. 1448, 59 L.Ed.2d 711 (1979). In this regard, federal common law is applicable where courts are required to implement the policies underlying a federal statute by fashioning appropriate remedies. Illinois v. City of Milwaukee, Wisc., 406 U.S. 91, 100-04, 92 S.Ct. 1385, 31 L.Ed.2d 712 (1972) (holding that federal courts may fashion federal common law remedies to implement the policies of federal water pollution statutes, because interstate navigable waters are inherently a matter of federal concern, and the federal legislation did not address the specific legal issue presented.); see County of Oneida, N.Y. v. Oneida Indian Nation of New York State, 470 U.S. 226, 237, 105 S.Ct. 1245, 84 L.Ed.2d 169 (1985) (applying federal common law to a remedial question where the federal Non-Intercourse Act failed to address what legal standard to apply). It is precisely in order to implement the policies underlying Congress's decision to make the violation of international law a federal tort, that it is necessary to flesh out the statute and apply federal common law; here, we must do so in order to fashion a remedy with respect to the direct or indirect involvement of third parties in the commission of the underlying tort.4

Next, the question of when third-party liability arises is a straightforward legal matter that federal courts routinely resolve using common law principles. See cases cited Part 3, infra. It is not an issue of such rarity, so seldom broached and so puzzling that our domestic law offers inadequate guidance and we are compelled to look elsewhere. The fact that some of the acts at issue here may have taken place abroad does not militate in favor of applying international law; transnational matters are litigated in federal court, using federal legal standards, more and more frequently as the pace of globalization grows ever more rapid. Nor is there any reason to apply international law to the question of third-party liability simply because international law applies to the substantive violation; as discussed above, federal common law is properly invoked when the statute at issue leaves an ancillary question unanswered, regardless of the nature of the statute. In short, federal common law principles provide the traditional and time-tested method of filling in the interstices and resolving the type of ancillary legal questions presented by this case.

In my view, courts should not substitute international law principles for established federal common law or other domestic law principles, as the majority does here, unless a statute mandates that substitution, or other exceptional circumstances exist. Examples of when the substitution of international law is appropriate include interpreting the substantive provisions of the Torture Victims Protection Act, Pub.L. No. 102-256, 106 Stat. 73 (1992) (codified at 28 U.S.C. § 1350), certain provisions of the Foreign Sovereign Immunities Act, 28 U.S.C. § 1330 et seq., and the substantive component of the ATCA. In those cases, the statutes make it plain that certain provisions require the application of international law. In other instances, I believe it prudent to follow the general rules established by the Supreme Court regarding the use of federal common law. It is important to recognize that there is a distinction between substituting international law for federal common law and making proper use of international law as part of federal common law. Employing federal common law does not force courts to ignore a constructive or helpful rule adopted under international law, because in appropriate circumstances federal common law incorporates relevant principles of international law. The Paquete Habana, 175 U.S. 677, 700, 20 S.Ct. 290, 44 L.Ed. 320 (1900); see also n. 7, infra. Thus, the benefits of the vast experience embodied in federal common law as well as any useful international law principles are obtained when we employ the traditional common law approach ordinarily followed by federal courts. Those benefits are lost, however, when we substitute for the wide body of federal authority and reasoning, as the majority does here, an undeveloped principle of international law promulgated by a recently-constituted ad hoc international tribunal.5

Almost all of the factors that we are required to consider as part of a choice-of-law inquiry militate in favor of determining that the proper law to apply here is the federal common law.6 First, "ease in the determination and application of the law to be applied" is furthered by applying a well-developed body of law, as opposed to a standard announced in a criminal case only recently decided by an ad hoc international tribunal. Similarly, "certainty, predictability and uniformity of result" are more likely to be achieved when there exists extensive precedent upon which to draw, and the state of the law does not depend on the future decisions of some as-yet unformed international tribunal established to deal with other unique regional conflicts. Additionally, although the "justified expectations" of potential parties may be limited in the sense that no direct precedent exists on the question of third-party liability in ATCA cases, the federal common law principles of joint liability, agency, and reckless disregard that we regularly apply in other contexts are generally well-known. In contrast, the international law regarding third party "moral support" is of very recent origin, and our selection of that law would not lead to settled expectations in future cases; for, the standard may well change dramatically if and when it is applied by a different ad hoc tribunal appointed by future representatives of the nations that compose the General Assembly of the United Nations. Next, as noted earlier, the policy of the Alien Tort Claims Act is "to establish[ ] a federal forum where courts may fashion domestic common law remedies to give effect to violations of customary international law." Abebe-Jira v. Negewo, 72 F.3d at 848. Thus, the "relevant polic[y] of the forum" is to apply federal common law remedies such as the imposition of third-party liability in the case of violations of customary international law. Finally, "the basic polic[y] underlying the particular field of law" is to provide an appropriate tort remedy for certain international law violations. The application of third-party liability standards generally applicable to tort cases directly furthers the basic policy of using tort law to redress international wrongs, whereas the application of international criminal law doctrines does not advance that objective. The two remaining choice-of-law factors are neutral, at the least, and certainly do not support rejecting the use of federal common law.7

Moreover, although much of international human rights law has developed in the criminal context, as the majority notes, the question of how to establish third-party liability is not in any way unique to human rights cases. The fact that the substantive violation involves international prohibitions on forced labor rather than a more traditional tort does not present any different concerns with respect to the determination of third-party liability. I thus see no reason to look to international criminal law doctrines for a civil liability standard when a substantial body of federal common law already exists regarding third-party liability generally. In sum, because Supreme Court precedent concerning the application of federal common law dictates its application here, and because the accepted choice of law factors overwhelmingly militate in favor of applying federal common law, I would derive a third-party liability standard for ATCA cases from that body of law.

Finally, I would note that the majority's disclaimer in its opinion that its legal conclusion regarding the applicability of international law rather than federal common law "is based on the record before it", and that in cases with other facts federal common law may apply, see maj. op., n. 25, serves no apparent purpose other than to attempt to distance the majority from its choice of international law. Indeed the footnote undermines the opinion's fundamental holding. All appellate decisions are based on the record before the court. More important, in all cases in which a third party is alleged to be legally responsible for the acts of a host government, third part liability determinations must be based on a principled choice of law. We must decide whether either international law or federal common law is applicable to the category of cases at issue here. The choice of law in such cases does not depend on the facts of the particular case, nor does it vary with the particular circumstances of the case. A binding legal rule must apply, or to put it differently, a controlling legal principle must govern, the legal question involved, regardless of the particular facts of a case. Either international law applies to third-party liability issues in ACTA cases or it doesn't. Either the ancillary question of whether an American corporate entity may be held liable under ACTA for the conduct of a host government is governed by federal common law or it isn't. What varies from case to case is not the question of the governing law, but whether liability attaches in the particular instance. That determination is based on the facts and circumstances of the case and that determination must be made regardless of whether international law or federal common law has been held to be applicable when resolving the issue of third-party liability. Thus, I reiterate that, unlike the majority, I would hold that the ancillary issue of third-party liability in ACTA cases must be decided as a matter of federal common law. I would not have the choice of law depend on the facts of the particular case.

3. Application of Federal Common Law

Having determined that ancillary legal issues in ATCA cases are to be resolved in accordance with federal common law, the question remains, for me, as to the proper federal common law rule for third-party liability in this case. Federal common law has developed over time a number of principles under which courts determine whether third parties may be held liable for the wrongful acts of others. Specifically, as will be discussed infra, the principles of joint venture, agency, and reckless disregard have all been applied across a wide range of torts and other legal wrongs, and the overwhelming weight of federal authority supports their application here.

Before turning to the application of the three federal common law theories to the instant case, it is necessary first to consider briefly whether the international law principle adopted by the majority may be applicable as part of the federal common law. It plainly is not. As noted earlier, international law principles may, under appropriate circumstances, become a part of the federal common law. Specifically, when an international legal principle achieves sufficient international acceptance that it constitutes customary international law, it also becomes part of the federal common law. Filartiga v. Pena-Irala, 630 F.2d 876, 881 (2d Cir.1980).8 However, the Yugoslav Tribunal's "moral support" standard is far from such a settled rule. As I observed earlier, it is a novel standard that has been applied by just two ad hoc international tribunals. It does not constitute customary international law, and thus we are not free to apply it as part of federal common law.

Moreover, even if it were possible for this court to determine that the Yugoslav Tribunal's novel criminal standard constitutes a part of the federal common law, I would strongly doubt the wisdom of using that rule to override the well-established federal common law tort principles that would otherwise be applicable to resolving third-party tort liability questions. The Yugoslav Tribunal's standard provides that an individual may be liable for "practical assistance, encouragement, or moral support which has a substantial effect on the perpetration of the crime." Prosecutor v. Furundzija, IT-95-17/1-T (Dec. 10, 1998), reprinted in 38 I.L.M. 317, ¶ 235 (1999). It is, in my view, far too uncertain and inchoate a rule for us to adopt without further elaboration as to its scope by international jurists. Although it is of some comfort that the majority considers "moral support" to be equivalent to "encouragement" in domestic tort law, it is nevertheless far from clear what the practical implications would be of adopting the standard recently announced by the ad hoc tribunal on war crimes in the former Yugoslavia.9 Members of a future ad hoc tribunal elected by representatives of all of the nations that may then belong to the United Nations General Assembly might well define the term quite differently than does the majority here. Thus, the unintended consequences of adopting the ad hoc tribunal's "moral support" standard may be significant.10

Returning to the three federal common law theories of third-party liability, plaintiffs alleged all three in their complaint, all of the theories are well-established in the federal common law, and disputed questions of fact exist with respect to each. Thus, in my view, plaintiffs are entitled to proceed to trial on the basis of each of the three theories. I address the basis for each theory in federal common law, as well as the evidence in the record supporting each:

a. Joint Venture Liability

It is well-established as a federal common law principle that a member of a joint venture is liable for the acts of its co-venturers. Federal courts freely invoke this principle of liability when called upon to apply federal common law in a variety of contexts. See, e.g., Davidson v. Enstar Corp., 848 F.2d 574, 577-78 (5th Cir.1988) (applying federal common law of joint liability rather than the idiosyncratic Louisiana law of joint liability in determining whether a relationship constituted a joint venture for purposes of the Longshore and Harbor Workers' Compensation Act); United States v. United Pacific Insurance Co., 545 F.2d 1381, 1382-83 (4th Cir.1976) (applying the federal common law of joint venturer liability in interpreting the Miller Act, 40 U.S.C. § 270a-d). Moreover, "different jurisdictions generally adopt the same criterion for the establishment of a joint venture." United States v. USX Corp., 68 F.3d 811, 826 n. 30 (3d Cir.1995).11

The principle that a member of a joint venture is liable for the torts of its co-venturer is well-established in international law and in other national legal systems. International legal materials frequently refer to the principle of joint liability for co-venturers. See, e.g., United Nations Convention On the Law of the Sea, Art. 139, Oct. 21, 1982, 21 I.L.M. 1245, 1293 (establishing principle of joint liability in international maritime law for parties acting jointly in maritime ventures); Convention on International Liability for Damage Caused by Space Objects, Mar. 29, 1972, 961 U.N.T.S. 187 (establishing joint liability principles to harms caused by parties launching objects into space); see also John E. Noyes & Brian D. Smith, State Responsibility and the Principle of Joint and Several Liability, 13 YALE J. I NT'L. LAW 225, 249 (1988) (describing joint and several liability for co-venturers' actions as a general principle of international law). The status of joint liability as a general principle of law is supported not only by international law sources but also by the fact that it is fundamental to "major legal systems." See, e.g., N.Y. PARTNERSHIP LAW § 24 (McKinney 2002); Buckley v. Chadwick, 45 Cal.2d 183, 190, 288 P.2d 12 (1955); Caron v. Lynn Sand & Stone Co., 270 Mass. 340, 346, 170 N.E. 77 (1930); 2 LAWS AND R EGULATIONS OF THE PEOPLE'S REPUBLIC OF CHINA 71 (1984) (Chinese joint venture statute); AIB Group (UK) Plc. v. Martin, 2001 U.K.H.L. 63 (United Kingdom joint venture law).

The body of international law described above serves to confirm my view that federal common law regarding the liability of joint venturers applies in the Alien Tort Claims Act context in the same manner and to the same extent as it does in construing other federal statutes. I would therefore hold that plaintiffs may recover on a federal common law theory of joint liability if they can prove both that the forced labor violations occurred and that Unocal was a co-venturer with the Myanmar military, which perpetrated the violations.

As discussed above, there exists a question of fact requiring trial regarding the occurrence of the forced labor violations. There also exists a question of fact regarding whether Unocal and the Myanmar military were co-venturers. The corporate entity that oversaw the gas exploration project consisted of four partners: Unocal, Total, the Myanmar government (which is a military regime and thus indistinguishable from the military), and a Thai corporation. Thus, contrary to Unocal's contentions, the evidence supports more than the conclusion that Unocal simply chose to invest in a project that happened to take place in a nation in which human rights abuses were widespread. Rather, a reasonable jury could conclude that Unocal freely elected to participate in a profit-making venture in conjunction with an oppressive military regime — a regime that had a lengthy record of instituting forced labor, including forced child labor.

Unocal contends that the Myanmar regime was a partner only in the offshore drilling portion of the Yadana project and not in the pipeline construction portion of the project. The company argues that a Total affiliate and a Unocal subsidiary created a pipeline construction corporation (called the Moattama Gas Transportation Company, or "MGTC") which was independent of both the joint venture and the military. A factual dispute exists with respect to this contention. Significantly, however, one of Unocal's business managers stated that "the [Yadana] project is an entirety ... although there may appear to be two different businesses ... this is an illusion." There is substantial evidence in the record that MGTC was the alter ego of the joint venture, in which case Unocal would be responsible for torts committed by its co-venturer, the Myanmar military, in the course of the pipeline construction company's activities. Plaintiffs contend that despite the existence of MGTC, the only reasonable reading of the contract forming the joint venture is that the joint venture is also responsible for the pipeline construction. Plaintiffs also argue that MGTC is a shell corporation because it maintained no independent offices, was under-capitalized, and relied only on the employees of the joint venture. Finally, evidence in the record states that Unocal would share revenues and costs of both the drilling and transportation components of the Yadana project. In view of the above, I believe that plaintiffs ought to proceed to trial on their claim of joint venture liability.

b. Agency Liability

Plaintiffs contend that Unocal may also be held liable for the acts of the Myanmar military because the military acted as the company's agent. The theory of agency liability is also well-supported in the federal common law. The Supreme Court has observed in the context of the Copyright Act that "when we have concluded that Congress intended terms such as `employee,' `employer,' and `scope of employment' to be understood in light of agency law, we have relied on the general common law of agency, rather than on the law of any particular State, to give meaning to these terms." Community for Creative Non-Violence v. Reid, 490 U.S. 730, 740, 109 S.Ct. 2166, 104 L.Ed.2d 811 (1989); see also Gleason v. Seaboard Air Line Ry. Co., 278 U.S. 349, 356, 49 S.Ct. 161, 73 L.Ed. 415 (1929) ("[F]ew doctrines of the law are more firmly established or more in harmony with accepted notions of social policy than that of the liability of the principal without fault of his own.")

Plaintiffs' theory of agency liability is consistent with the substantial federal common law of agency developed in the context of the Labor-Management Relations Act and ERISA. See, e.g., Anderson v. International Union, United Plant Guard Workers of America, 150 F.3d 590, 592-93 (6th Cir.1998) (addressing an agency issue under ERISA and holding that "we are guided by the law of agency as developed and interpreted as a matter of federal common law."); National Football Scouting, Inc. v. Continental Assurance Co., 931 F.2d 646, 648 (10th Cir.1991) (examining whether "under the federal common law of agency" an agent of a plan fiduciary was acting within his actual or apparent authority). That federal common law should govern plaintiffs' claim that the Myanmar military acted as Unocal's agent.

The general principles of the federal common law of agency have been formulated largely based on the Restatement of Agency. Moriarty v. Glueckert Funeral Home, Ltd., 155 F.3d 859, 866 n. 15 (7th Cir.1998); Cilecek v. Inova Health Sys. Servs., 115 F.3d 256, 259-60 (4th Cir.1997). Under those general principles, an agency relationship may be express or implied; in addition, a jury may infer from the factual circumstances that apparent agency authority exists. See Hasbrouck v. Sheet Metal Workers Local 232, 586 F.2d 691, 693 (9th Cir.1978); see also 3 AM. JUR. 2D AGENCY § 21 (1986) ("The manner in which the parties designate the relationship is not controlling, and if an act done by one person in behalf of another is in its essential nature one of agency, the one is the agent of such other notwithstanding he is not so called." (footnote omitted)).

As is true of joint liability principles, agency liability principles are well-established in international law. They are frequently discussed in international legal materials. See, e.g., Case Concerning the Barcelona Traction Power & Light Co. (Bel. v. Spain), 1970 I.C.J. 3, 215 (discussing principal-agent liability as applicable to commercial relationships generally); Application for Review of Judgment No. 333 of the U.N. Administrative Tribunal (Yakimetz Case), May 27, 1987 (applying principal-agent liability to acts of United Nations employees).12 Principal-agent liability is also widely adopted by civil law and other common law systems. See, e.g., Bazley v. Curry, 2 S.C.R. LEXIS 134 (1999) (Canadian Supreme Court statement of vicarious liability principles); CIVIL CODE OF FRANCE, Art. 1384 (1994) ("[A person] is liable not only for the damage which he caused by his own act, but also for that which is caused by the act of persons for whom he is responsible, or by things which he has in his keeping."); CIVIL CODE OF GERMANY, § 831 (1975) ("A person who employs another to do any work is bound to compensate for any damage which the other unlawfully causes to a third party in the performance of his work."). Thus, the conclusion that plaintiffs' agency theory is cognizable under federal common law is further supported by the international legal authorities that establish agency as a general principle of international law.

A factual question requiring trial exists with respect to whether an agency relationship existed between Unocal and the Myanmar military. Some evidence in the record suggests that such a relationship existed. For instance, plaintiffs cite an internal Unocal briefing document regarding the Yadana Project, discussed by the majority, which states that"[a]ccording to our contract, the government of Myanmar is responsible for protecting the pipeline. There is military protection for the pipeline and, when we have work to do along the pipeline that requires security, the military people will, as a matter of course, be nearby." They also point to memoranda by various Total and Unocal employees recounting that oil company officials requested specific battalions to perform various tasks, including the construction of helipads for the convenience of corporate executives. Plaintiffs argue that the record supports either an implied or express agency relationship, based on the conduct of the parties. As the majority has also pointed out, the record contains evidence of daily meetings between Total and Unocal executives and Myanmar military commanders, so that the corporations could instruct the military leaders regarding specific security or infrastructure projects that were required for the pipeline construction. Moreover, Unocal stated publicly on several occasions that it controlled the Myanmar military's actions in connection with the pipeline project. In response to accusations of human rights abuses occurring by the Myanmar military with respect to the Yadana project, Unocal denied the existence of such abuses, and stressed its ability to prevent any wrongdoing due to its control of the military. Unocal's alleged actions directing the Myanmar military create a triable question of fact as to whether an agency relationship existed between Unocal and the Myanmar armed forces.

Significantly, there is evidence in the record that Unocal did not simply, by its inaction, take advantage of the routine presence of the Myanmar military in the Tenasserim region. The record indicates that there was in fact little to no rebel activity in the region where the pipeline construction occurred, and that the center of the Myanmar civil war was 150-200 miles distant from the pipeline project. A jury could reasonably conclude that the military was present in Tenasserim not merely to maintain order, as was its function in other parts of the nation. Rather, it could determine that military forces were brought to Tenasserim in order to support the pipeline project, that the military was performing duties for the pipeline project quite distinct from traditional military or state functions, and that it did so at the request of and in close coordination with Unocal and the other private entities. It is not essential that a formal contract have existed between Unocal and the Myanmar military in order for Unocal to be held liable for the government's actions under an agency theory. Nevertheless, should plaintiffs prove their allegation that such a contract existed, a jury might have considerable difficulty in accepting Unocal's denial of an agency relationship.

c. Reckless Disregard

Finally, the facts alleged by the plaintiffs, if proved, support a recovery against Unocal under an additional theory, that of the common law theory of recklessness or reckless disregard. Here, plaintiffs allege that Unocal had actual knowledge that the Myanmar military would likely engage in human rights abuses, including forced labor, if it undertook the functions Unocal and the other private parties desired it to perform in connection with the Yadana Pipeline Project. Nevertheless, according to plaintiffs, Unocal recklessly disregarded that known risk, determined to use and in fact did use the services of that military to perform pipeline-related tasks, and thereby set in motion international law abuses that were foreseeable to Unocal. Plaintiffs thus allege that Unocal acted with recklessness, which occurs when a party is aware of (or should be aware of) an unreasonable risk, yet disregards it, thereby leading to harm to another. Farmer v. Brennan, 511 U.S. 825, 836, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994).

Federal common law contains two variants of the theory of recklessness or reckless disregard. Plaintiffs include both in their complaint, and in my view the record contains evidence sufficient to require trial on both. The first is traditional civil-law recklessness, sometimes referred to as "objective recklessness"; the Supreme Court has stated that"[t]he civil law generally calls a person reckless who acts or (if the person has a duty to act) fails to act in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known." Farmer, 511 U.S. at 836-37, 114 S.Ct. 1970 (citing W. Keeton, D. Dobbs, R. Keeton, & D. Owen, PROSSER AND KEETON ON L AW OF TORTS § 34, pp. 213-214 (5th ed.1984); RESTATEMENT (SECOND) OF TORTS § 500 (1965)). The second version is "subjective recklessness," also referred to as "willful recklessness." This doctrine requires actual knowledge of a substantial risk which the defendant subsequently disregards. Ewolski v. City of Brunswick, 287 F.3d 492, 513 (6th Cir.2002).

The concept that one party may be held liable for a reckless disregard of the welfare of another pervades federal common law and has been applied in a variety of contexts. It is recognized, for instance, in admiralty law cases, see, e.g., Youell v. Exxon Corp., 48 F.3d 105, 110-11 (2d Cir.), vacated on other grounds, 516 U.S. 801, 116 S.Ct. 43, 133 L.Ed.2d 9 (1995), in cases arising under the Warsaw Convention, Saba v. Compagnie Nationale Air France, 78 F.3d 664, 668-69 (D.C.Cir.1996), as well as in constitutional tort cases brought pursuant to 42 U.S.C. § 1983. Medina v. City and County of Denver, 960 F.2d 1493, 1496 (10th Cir.1992). The Supreme Court has also adopted a willful recklessness standard in cases involving "deliberate indifference" under the Eighth Amendment. Farmer, 511 U.S. at 836, 114 S.Ct. 1970 ("With deliberate indifference lying somewhere between the poles of negligence at one end and purpose or knowledge at the other, the Courts of Appeals have routinely equated deliberate indifference with recklessness. It is, indeed, fair to say that acting or failing to act with deliberate indifference to a substantial risk of serious harm to a prisoner is the equivalent of recklessly disregarding that risk." (citations omitted)). On "a continuum that runs from simple negligence through gross negligence to intentional misconduct," Saba, 78 F.3d at 668, recklessness lies between gross negligence and intentional harm. The common law principle of recklessness has typically been applied to acts by a defendant that directly cause harm to a plaintiff. Nevertheless, I see no reason why the general principle that liability arises for one party's conscious disregard of unreasonable risks to another should not apply when a defendant consciously disregards the risks that arise from its decision to use the services of an entity that it knows or ought to know is likely to cause harm to another party.13

Proof of even willful recklessness does not require proof of intent; it requires only that a defendant have acted in conscious disregard of known dangers. City of Canton, Ohio v. Harris, 489 U.S. 378, 390 n. 10, 109 S.Ct. 1197, 103 L.Ed.2d 412 (1989). The doctrine of reckless disregard of another's rights has been well-developed in the § 1983 context; there, courts have held that a plaintiff need not prove that a defendant intended to cause harm to the specific plaintiff. "[R]eckless intent does not require that the actor intended to harm a particular individual; reckless intent is established if the actor was aware of a known or obvious risk that was so great that it was highly probable that serious harm would follow and he or she proceeded in conscious and unreasonable disregard of the consequences. Thus, reckless intent involves disregard of a particular risk rather than intent to cause a particularized harm." Medina, 960 F.2d at 1496. Plaintiffs' theory that by using the services of the Myanmar military in connection with the Yadana Project, Unocal recklessly disregarded the likelihood that their human rights would be violated is thus well grounded in federal common law.

Plaintiffs have presented sufficient evidence to proceed to trial on the reckless disregard claim. There exists a genuine question of material fact as to whether the Myanmar military caused the human rights abuses alleged, and if so, whether Unocal should be liable for those abuses because it acted in either a subjectively or an objectively reckless manner in choosing to make use of the services of the Myanmar military in connection with the Yadana gas exploration project. Plaintiffs allege that Unocal entered into an agreement to participate in the Yadana project knowing that the Myanmar military was also a participant, and that the Myanmar military would perform important pipeline tasks for the project, even though it had knowledge that the military engaged in widespread human rights abuses, including forced labor. Plaintiffs further allege that Unocal continued to use those services after obtaining knowledge that the military was conscripting forced labor in service of the Yadana project.

The record contains evidence that Unocal held several meetings with human rights groups both before and after it became a part of the Yadana venture, at which those organizations provided Unocal with documentation of the Myanmar military's brutal treatment of the Myanmar people. Moreover, as noted earlier, Unocal retained a risk management consulting firm prior to joining the Yadana project, and that firm completed a report informing the company that "the government habitually makes use of forced labor," and assessing the risk that the proposed joint venture would result in the use of forced labor as "high."

Additionally, as the majority has noted, Unocal Vice President Lipman testified that even before Unocal invested in the Project, it was aware that "the option of having the [Myanmar] [M]ilitary provide protection for the pipeline construction and operation of it would be that they might proceed in the manner that would be out of our control and not be in a manner that we would like to see them proceed, I mean, going to excess." (Emphasis added.) Unocal Representative Robinson later wrote to Unocal President Imle that "[o]ur assertions that [the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny." (Emphasis added.) And Unocal Consultant Haseman told Unocal that "[t]he most common [human rights violations] are forced relocation without compensation of families from land near/along the pipeline route; forced labor to work on infrastructure projects supporting the pipeline ...; and ... execution by the army of those opposing such actions." Based on this evidence, a question of fact exists with respect to whether consequences that were or should have been foreseeable to Unocal, in the form of violations of international law by the Myanmar military, resulted from Unocal's participation in the Yadana Pipeline Project, and from its continuing decision to use the services of the Myanmar military during the course of the pipeline's construction.14

d. Murder and Rape Claims15

Like the majority, I agree with the Second Circuit's holding in Kadic v. Karadzic, that under the Alien Tort Claims Act, a plaintiff may recover for wrongs that occur ancillary to a violation of international human rights law as part of the claim for the primary violation. 70 F.3d 232, 244 (2d Cir.1995) (holding acts of rape, torture, and summary execution committed in connection with genocide or war crimes to be actionable under the ATCA). Here, all of the acts of murder and rape of plaintiffs alleged in the complaint or otherwise in the record on summary judgment occurred in furtherance of the forced labor program. Following the Kadic rule, plaintiffs may pursue those claims as part of their forced labor claims.

As with the forced labor claims, however, I disagree with the majority regarding what it is necessary for plaintiffs to prove in order for Unocal to be held liable for acts of murder or rape. Specifically, I disagree with the majority's view that we must apply once again, independently, a third-party liability standard — whether international law or federal common law — this time with respect to the specific acts of murder and rape. In my opinion, if it is established that the alleged rapes and murder of plaintiffs occurred in furtherance of the forced labor program, and if Unocal is held liable for the forced labor practices of the Myanmar military, then plaintiffs need not again prove separately the elements of a third-party liability theory. In such case, they need prove only the additional facts supporting the rape and murder allegations. While I would not foreclose a possible foreseeability or ultra vires argument, I would not, as the majority does, require plaintiffs to make a second showing of third-party liability merely because specific acts conducted in furtherance of the primary tort are themselves tortious.

There is one final observation that I hope will clarify what the majority opinion does and does not do. Because the acts of murder and rape involving the plaintiffs all allegedly occurred in furtherance of a regime of forced labor, there is no need to address the question whether Unocal could be held liable if members of the Myanmar military had committed similar acts against plaintiffs separate and apart from a forced labor regime or a similarly well-established pattern of conduct violative of international law. Cf. Kadic, 70 F.3d at 244-45 (leaving open the related question of whether non-state actors who directly commit acts such as murder, rape, or torture may be held liable under the ATCA when those acts are not committed in furtherance of a jus cogens violation.). Accordingly, I read the majority's statements regarding the rapes and murders as leaving open the question whether a private entity could be held liable for such government conduct if it was unrelated to an underlying violation of international law.

4. Conclusion

In sum, I agree with the majority that disputed questions of fact exist with respect to whether human rights violations occurred during the construction of the Yadana Pipeline Project, and with respect to the nature of Unocal's involvement in such violations. Assuming the necessary evidence is introduced at trial, I would, however, direct that the jury be instructed to apply the three common law theories of third-party liability ordinarily applied in tort cases, rather than the international criminal law doctrine of aiding and abetting, including by means of moral support, recently announced by the ad hoc War Crimes Tribunal for the Former Yugoslavia.

1 As the majority notes, such a norm is one that is "accepted and recognized by the international community of states as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character." Vienna Convention on the Law of Treaties, art. 53, May 23, 1969, 1155 U.N.T.S. 332, 8 I.L.M. 679;see also Siderman de Blake v. Republic of Argentina, 965 F.2d 699, 714-15 (9th Cir.1992)

2 I do not read the majority opinion as holding otherwise. The opinion nowhere states that a third party can be liable for a governmental action only if that action constitutes ajus cogens violation. Rather, I view the majority's interesting academic dissertation on why forced labor is like slavery and why slavery constitutes a jus cogens violation as background historical material that may be of interest to legal scholars but is unnecessary to its decision.

3 It is the rare Alien Tort Claims Act case that does not involve a foreign state or official as a defendant. Most international law norms apply only to states; a private party will ordinarily violate international law by its own actions only if it transgresses a legal norm that has achievedjus cogens status. See discussion, supra.

4 That the principles discussed in the text apply not only to traditional domestic legislation but to the Alien Tort Claims Act as well is demonstrated by a statement by the Eleventh Circuit inAbebe-Jira v. Negewo, 72 F.3d 844, 848 (11th Cir.1996). There, that court observed that the purpose of the ATCA is "to establish[ ] a federal forum where courts may fashion domestic common law remedies to give effect to violations of customary international law."

5 The International Criminal Tribunal for the Former Yugoslavia was formed with the limited mandate of adjudicating allegations of human rights abuses that took place in the Balkans in the last decade. Established by Security Council Resolution 827 in May, 1993, it is a temporary body whose members are elected for four-year terms by the members of the United Nations General Assembly. The International Criminal Tribunal for Rwanda, which subsequently applied the Yugoslav Tribunal's test, is a similarly-constituted body

6 The Restatement (Second) of Conflict of Laws, § 6, provides that the following factors are to be considered as part of a choice of law analysis: (1) the needs of the interstate and international systems; (2) the relevant policies of the forum, (3) the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue; (4) the protection of justified expectations, (5) the basic policies underlying the particular field of law, (6) certainty, predictability and uniformity of result, and (7) ease in the determination and application of the law to be applied. "Federal choice of law rules follow the approach of the Restatement (Second) of Conflict of Laws."In re Vortex Fishing Sys., Inc., 277 F.3d 1057, 1069 (9th Cir.2002); see also Chuidian v. Philippine Nat'l. Bank, 976 F.2d 561, 564 (9th Cir.1992)

7 The two remaining Restatement factors are the first, "the needs of the interstate and international systems," and the third, "the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue." Restatement (Second) of Conflict of Laws, § 6

8 As the Second Circuit explained in the landmark case ofFilartiga, all international legal principles do not automatically become a part of the federal common law; only those that achieve the status of customary international law or are included in international treaties are incorporated as part of federal common law. A customary international law rule "results from a general and consistent practice of states followed by them from a sense of legal obligation." Restatement (Third) Foreign Relations Law, § 102. The Filartiga court observed, "[t]he requirement that a rule command the `general assent of civilized nations' to become binding upon them all is a stringent one. Were this not so, the courts of one nation might feel free to impose idiosyncratic legal rules upon others, in the name of applying international law." 630 F.2d at 881 (quoting The Paquete Habana, 175 U.S. at 694, 20 S.Ct. 290). Thus, the Filartiga court reached its conclusion that a claim for torture was cognizable under the ATCA in large part because the prohibition on torture had become part of customary international law. Id.

9 In an effort to minimize the damage caused by its unfortunate decision to apply international law to the third-party liability issue, and perhaps to make that choice of law more palatable to American courts generally, the majority disclaims an integral portion of the international law standard it adopts, purporting to leave "to another day" the question whether moral support alone (whatever that may mean) is sufficient to give rise to third-party liabilitySee maj. op., 951. However, by substituting international law standards for federal common law, rather than following federal common law and incorporating those portions of international law that attract sufficient legal support, the majority has lost whatever opportunity it had to pick and choose the aspects of international law that it finds appealing. Having declared that international law governs, and that the Yugoslav Tribunal's standard constitutes the controlling international law, the majority cannot then escape the implications of being bound by the law it has selected. Indeed, the majority, despite its disclaimer as to "moral support," is no more successful in avoiding the consequences of its choice of the Yugoslav Tribunal's criminal standard than it is in avoiding the consequences of its predicate decision to reject federal common law as the appropriate rule for ancillary issues in Alien Tort Claims Act cases. See maj. op., n. 25.

10 For instance, liability for moral support raises the question whether political advocacy not imminently causing violence that would otherwise be protected by the First Amendment could be the source of ATCA liability under the majority's standard

11 It is well-accepted that joint liability will exist where (1) parties intended to form a joint venture; (2) parties share a common interest in the subject matter of the venture; (3) the parties share the profits and losses of the venture; and (4) the parties have joint control or the joint right of control over the venture. W. Keeton, PROSSER & K EETON ON TORTS, § 72 at 518 (5th ed.1984)

12 In addition, the Foreign Sovereign Immunities Act, 28 U.S.C. § 1605, demonstrates that agency liability is a general principle applicable to international entities. In that statute, an exception exists to the general sovereign immunity accorded to foreign states in American courts for certain commercial activities of agents of foreign nationsId. § 1605(a)(3).

13 I reach this conclusion in part because at common law, a defendant may be liable for harms caused by an entity that it negligently employs, even if no respondeat superior or agency relationship existsSee, e.g., Bennett v. United States, 803 F.2d 1502, 1505 (9th Cir.1986) (holding that the U.S. government may be liable for damages arising from the kidnapping and raping of several children by a teacher hired by the government where the government knew or should have known that the teacher had a history of child molestation); see also Van Ort v. Estate of Stanewich, 92 F.3d 831, 837 (9th Cir.1996) (stating that under Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978), a municipality may be liable for the negligent hiring of municipal employees)."

14 The district court granted summary judgment for defendants in part because in its view plaintiffs produced insufficient evidence that Unocal proximately caused plaintiffs' injuries. However, under none of the three federal common law theories discussed in the text is proximate cause a necessary element of Unocal's liability. If proximate cause applies at all, it applies to the question whether the Myanmar military was responsible for the international law violations. As a practical matter, if the alleged violations occurred there can be little doubt that they were proximately caused by the Myanmar military. As explained in the text, under any of the three theories, Unocal's liability is indirect; proximate cause is therefore irrelevant to plaintiffs' claims against Unocal

15 I agree with the majority's conclusion that there is insufficient evidence in the record to permit plaintiffs to proceed on a claim of torture, and would limit their claims to those alleging forced labor, murder, and rape. I would also note that the record is replete with horrific accounts of physical abuse of Myanmar villagers by members of the military. Because the victims in those accounts are not plaintiffs in this action, and this suit does not constitute a class action, many substantial allegations of wrongdoing may not give rise to liability as a part of this case