Raymonde ABRAMS, Nicole B. Silberkleit, Janet Herman, Lily Redner, Bernard Caron, Ernest Haar, Harry Cybulski, Yvonne Litman, Cassandra Kirby Conahay Freund, Jean Jacques Fraenkel, Liliane Lichtenstein, Marie Weinrauch, Plaintiffs-Appellants, v. SOCIÉTÉ NATIONALE DES CHEMINS DE FER FRANCAIS, Defendant-Appellee.
The named plaintiffs have brought this action individually and on behalf of other similarly situated Holocaust victims and their heirs against the French railroad company Société Nationale des Chemins de Fer Francais (SNCF or railroad). The action arises out of SNCF's 1942-1944 operation of trains that transported tens of thousands of French civilians to the infamous Nazi death and slave labor camps. Plaintiffs allege that in so doing SNCF committed war crimes and crimes against humanity under customary international law and the law of nations. Customary international law, plaintiffs further allege, is enforceable in federal district court as federal common law.
Plaintiffs filed their complaint in the United States District Court for the Eastern District of New York before Judge David G. Trager on September 12, 2000. When SNCF moved to dismiss it for lack of subject matter jurisdiction, the district court granted the motion, ruling that SNCF was an “agency or instrumentality of a foreign state” as that term is defined in the Foreign Sovereign Immunities Act of 1976 (FSIA or Act), 28 U.S.C. § 1603(b), and that, because plaintiffs' claims did not fall within any of the Act's exceptions to foreign sovereign immunity, it was without jurisdiction to adjudicate them. Abrams v. Société Nationale des Chemins de Fer Francais, 175 F.Supp.2d 423, 428-29, 450 (E.D.N.Y.2001). On appeal, as in district court, plaintiffs urge that the FSIA does not apply to this case because it arises out of events predating the statute's 1976 enactment.
The following facts are alleged in the complaint and are accepted as true, as they must be at this stage of the litigation. SNCF, the national railway of France, was created in the late 1930s by consolidation of five then-existing French regional rail networks. Today it is operated as a separate legal entity wholly-owned by the French government. During the Nazi occupation of France the railroad remained under civilian control and preserved its independence by collaborating with the German authorities and by accommodating their transportation needs.
In March 1942 at the request of those authorities, SNCF began to operate trains deporting Jews and other so-called “undesirables” from France to Nazi concentration camps. In exchange for this assistance, the railroad was allowed to continue its operations and was paid for the transport it provided. The conditions inside the deportation trains were inhumane and frequently fatal to passengers who were often carried in cattle cars. Sanitation facilities were limited or nonexistent and passengers had to endure extreme heat and cold. Many did not live to the journey's end. By the time the Nazi occupation of France was over, SNCF had conveyed more than 72 deportation convoys, taking to concentration camps 75,000 Jews and tens of thousands of others. Fewer than three percent of those deported survived. The trains' ultimate destinations included unspeakable places like Dachau and Auschwitz, and these destinations and the conditions of travel were well known to those running the railroad.
The named plaintiffs are either survivors of those deportations or their heirs and descendants. They commenced suit against SNCF both in their individual capacities and on behalf of a putative class of all members of the civilian population of France transported by SNCF to the Nazi camps, and their respective heirs and beneficiaries. The complaint sought compensatory and punitive damages as well as disgorgement of wrongfully obtained profits.
The railroad moved to dismiss the complaint on two grounds: first, it asserted that federal courts in the United States had no subject matter jurisdiction over plaintiffs' causes of action; second, it contended it was entitled to sovereign immunity, both under the FSIA and under the laws in effect during World War II. In support of its motion, SNCF submitted affidavits from its attorneys containing information regarding the railroad's present organization and ownership. The railroad's attorneys confirmed that it is now organized as a separate legal entity, which is wholly-owned and controlled by the French government.
In opposing the motion to dismiss their complaint, plaintiffs contended that applying the Act to their claims would be impermissibly retroactive, and that questions of jurisdiction and immunity should be resolved based on laws in effect in the 1940s at the time the railroad's underlying conduct occurred. Under those laws, plaintiffs maintain, SNCF was not entitled to sovereign immunity because it was organized as a corporate entity separate and distinct from the French government. Plaintiffs also cross-moved for discovery, declaring that whether SNCF is entitled to immunity cannot be resolved based on the existing record.
In granting SNCF's motion to dismiss and denying plaintiffs' cross-motion for discovery the district court first found that the railroad fits on all fours into the definition of an agency or instrumentality of a foreign state under the terms of the Act. Substantially adopting the reasoning in Princz v. Federal Republic of Germany, 26 F.3d 1166 (D.C.Cir.1994), it further ruled that, to the extent that the Act defines the scope of federal courts' jurisdiction over claims against foreign states, the statute applies to actions commenced after its enactment regardless of when the underlying conduct occurred. The district court further ruled that the FSIA's jurisdictional grant does not encompass plaintiffs' cause of action. Abrams, 175 F.Supp.2d at 450. It found it unnecessary to decide whether plaintiffs could have brought their cause under laws in effect in the 1940s. Id. at 446.
Although we agree with the district court that SNCF is an agency or instrumentality of France under the FSIA, the existing record is insufficient for us to resolve the question of whether that Act's application to plaintiffs' causes of action would be impermissibly retroactive. Hence, we vacate and remand.
I Standard of Review
Upon reviewing a district court's determination with respect to its subject matter jurisdiction under the FSIA, we examine the court's legal conclusions de novo and its factual findings for clear error. Robinson v. Gov't of Malay., 269 F.3d 133, 138 (2d Cir.2001). Because the district court did not have to make any factual findings, and dismissed the complaint solely on its resolution of questions of law, our review is de novo.
II Foreign Sovereign Immunity in Federal Courts
A. Prior to the FSIA's Enactment
Foreign sovereign immunity has been a recognized doctrine of American law since the seminal Supreme Court case The Schooner Exchange v. McFaddon, 11 U.S. (7 Cranch) 116, 3 L.Ed. 287 (1812). In The Schooner Exchange, two American citizens claimed ownership of a ship they alleged had been wrongfully seized by the French navy. The United States Attorney for the District of Pennsylvania filed a suggestion of immunity with the district court, thereby raising a question of sovereign immunity. Id. at 117-18. The Supreme Court agreed that immunity existed and dismissed the Americans' claim to the ship. Id. at 147. In an opinion by Chief Justice Marshall, the Court explained that the “perfect equality and absolute independence” of sovereign nations required that United States courts refrain from exercising jurisdiction over claims against other states. Id. at 137.
From that beginning until the FSIA's enactment in 1976, the executive branch played a prominent role in deciding whether a foreign sovereign was immune from suit in American courts. See Verlinden B.V. v. Cent. Bank of Nig., 461 U.S. 480, 486-87, 103 S.Ct. 1962, 76 L.Ed.2d 81 (1983); Restatement (Third) of Foreign Relations Law of the United States [hereafter Restatement 3d] pt. IV, ch. 5, subch. A, introductory note (1987). Foreign states sued in the United States often requested that the Department of State ask the Department of Justice to file a suggestion of immunity with the courts. See Restatement 3d, pt. IV, ch. 5, subch. A, introductory note.
Courts, for their part, usually deferred to the decision of the executive, reasoning that the preferable method of resolving disputes with friendly foreign states is not litigation but diplomatic negotiation, a matter within the authority and expertise of the executive branch. See Ex parte Republic of Peru, 318 U.S. 578, 586-87, 63 S.Ct. 793, 87 L.Ed. 1014 (1943); see also Republic of Mexico v. Hoffman, 324 U.S. 30, 35, 65 S.Ct. 530, 89 L.Ed. 729 (1945) (reasoning that “the courts should not so act as to embarrass the executive arm in its conduct of foreign affairs”). In the 1940s, the Supreme Court expressly endorsed deference to the executive as “a guiding principle in determining whether a court should exercise or surrender its jurisdiction in such cases.” Hoffman, 324 U.S. at 35, 65 S.Ct. 530; see also Peru, 318 U.S. at 586-87, 63 S.Ct. 793 (accepting a claim of immunity where the Secretary of State had undertaken to settle the dispute through diplomatic channels).
Prior to 1952, the United States adhered to the absolute theory of foreign sovereign immunity. See Restatement 3d, pt. IV, ch. 5, subch. A, introductory note. Under that theory, a sovereign cannot be sued in the courts of another state without that sovereign's consent, regardless of the nature of the activity giving rise to the action. See Letter from Jack B. Tate, Acting Legal Adviser, Department of State, to Philip B. Perlman, Acting Attorney General of the United States (May 19, 1952) [hereafter Tate Letter], reprinted in Alfred Dunhill of London, Inc. v. Republic of Cuba, 425 U.S. 682, 711-15, 96 S.Ct. 1854, 48 L.Ed.2d 301 (1976). As foreign states increasingly began to participate in international commerce, an alternative to the theory of absolute immunity emerged called the restrictive theory of sovereign immunity. See Restatement 3d pt. IV, ch. 5, subch. A, introductory note. Under that theory, a foreign state is immune from claims arising out of the state's governmental activities, but not immune from claims arising out of its commercial activities. See id. § 451, cmt. a. The United States adopted the restrictive theory in 1952 when the Department of State announced its formal change of policy in a letter from Acting Legal Adviser Jack Tate to the Acting Attorney General. See Tate Letter, reprinted in Alfred Dunhill, 425 U.S. at 714, 96 S.Ct. 1854.
After the Tate Letter, the State Department continued to decide most of foreign states' immunity claims. In those cases where the foreign government did not request the State Department's intervention, the duty to resolve the immunity question in the first instance fell to the courts. Verlinden, 461 U.S. at 487, 103 S.Ct. 1962. Sovereign immunity determinations were thus made in two different branches of government resulting in rules that were neither clear nor uniform. Id. at 488, 103 S.Ct. 1962; see also Danny Abir, Foreign Sovereign Immunities Act: The Right to a Jury Trial in Suits Against Foreign Government-Owned Corporations, 32 Stan. J. Int'l L. 159, 165 (1996) (discussing difficulties in application of the restrictive theory between 1952 and 1976); William R. Dorsey, III, Reflections on the Foreign Sovereign Immunities Act After Twenty Years, 28 J. Mar. L. & Com. 257, 259-60 (1997) (same). Litigation of claims against foreign states was further complicated by the absence of effective procedures for service of process and by the lack of satisfactory standards for execution of judgments against foreign states. See Abir, supra, at 165; Dorsey, supra, at 260.
As a result of these inadequacies Congress in 1976 enacted the FSIA, a statute aimed “to free the Government from the case-by-case diplomatic pressures, to clarify the governing standards, and to ‘assur[e] litigants that ․ decisions are made on purely legal grounds and under procedures that insure due process.’ ” Verlinden, 461 U.S. at 488, 103 S.Ct. 1962 (quoting H.R.Rep. No. 94-1487, at 7 (1976), reprinted in 1976 U.S.C.C.A.N. 6604, 6606).
B. The FSIA
The Act was designed to codify the restrictive theory of sovereign immunity and to remove the subject from diplomatic pressures by transferring such decisions to the judiciary. Standards were imposed for serving process, obtaining personal jurisdiction, and executing judgment in an action against a foreign state. See H.R.Rep. No. 94-1487 [hereafter House Report], at 7-8 (1976), reprinted in 1976 U.S.C.C.A.N. 6604, 6605-06. For purposes of our analysis, the Act can be divided into three main parts: (1) statement of jurisdiction of the federal courts, (2) exclusions from jurisdiction and definition of immunities, and (3) other standards and rules. See Restatement 3d pt. IV, ch. 5, subch. A, introductory note. While the third group of provisions is largely irrelevant in the instant case, the first two parts warrant a brief overview.
1. Statement of Jurisdiction
The FSIA added to Title 28 of the United States Code a new jurisdiction-conferring provision, § 1330. Section 1330 granted federal district courts “original jurisdiction without regard to amount in controversy of any nonjury civil action against a foreign state as defined in section 1603(a) of this title as to any claim for relief in personam with respect to which the foreign state is not entitled to immunity either under sections 1605-1607 of this title [the immunity-defining provisions] or under any applicable international agreement.” Foreign Sovereign Immunities Act of 1976, Pub.L. No. 94-583, § 2, 1976 U.S.C.C.A.N. (90 Stat.) 2891, 2891 (codified at 28 U.S.C. § 1330). The statutory definition of a foreign state includes its agencies and instrumentalities. See 28 U.S.C. § 1603(a).
The FSIA also amended the diversity jurisdiction provision, § 1332. That amendment eliminated jurisdiction over actions against foreign states, and this section refers now only to suits where a foreign state is a plaintiff. See FSIA § 3, 1976 U.S.C.C.A.N. (90 Stat.) at 2891 (codified at 28 U.S.C. § 1332). The House Report explained that, because under the FSIA “jurisdiction in actions against foreign states is comprehensively treated by the new section 1330, a similar jurisdictional basis under section 1332 [became] superfluous.” House Report at 14, reprinted in 1976 U.S.C.C.A.N. at 6613.
The Act did not expressly amend other jurisdiction-granting statutes. In Argentine Republic v. Amerada Hess Shipping Corp., 488 U.S. 428, 109 S.Ct. 683, 102 L.Ed.2d 818 (1989), the Supreme Court rejected an argument that this silence indicated Congress' intent to retain claims against foreign states within the scope of jurisdictional provisions other than the new § 1330, such as the general admiralty and maritime jurisdiction statute, 28 U.S.C. § 1333(1), or the Alien Tort Claims Statute, 28 U.S.C. § 1350. Relying on the FSIA's comprehensiveness, as well as on its language and legislative history, the Court held that, after the FSIA's enactment, that statute “provides the sole basis for obtaining jurisdiction over a foreign state in federal court.” Id. at 439, 109 S.Ct. 683. Further, Congress did not have to amend all other existing jurisdictional statutes to ensure the Act's exclusivity. Unlike the diversity provision, other statutes conferring jurisdiction in general terms on district courts had not previously expressly provided for suits against foreign states. Id. at 437 n. 5, 109 S.Ct. 683. Thus, the Court concluded, Congress stated that “ ‘[c]laims of foreign states to immunity should henceforth be decided by courts of the United States in conformity with [the FSIA],’ and very likely it thought that should be sufficient.” Id. at 437-38, 109 S.Ct. 683 (quoting 28 U.S.C. § 1602).
2. Exclusions From Jurisdiction and Definition of Immunities
The Act added to Title 28 a new chapter, §§ 1602-1611, in which it set out a comprehensive set of standards for federal and state courts' determinations of foreign sovereigns' claims of immunity. See FSIA § 4, 1976 U.S.C.C.A.N. (90 Stat.) at 2891-97 (codified at 28 U.S.C. §§ 1602-1611); House Report at 14, reprinted in 1976 U.S.C.C.A.N. at 6613. The newly added § 1604 declares that, subject to existing international agreements to which the United States was a party at the time of the FSIA's enactment, foreign states are immune from the federal and state courts' jurisdiction unless the FSIA itself provides otherwise. See § 1604.
Section 1605 then lists general exceptions to immunity consistent with the restrictive theory. See id. § 1605. For example, it provides that foreign states are not immune from claims arising out of their commercial activities within the United States or out of their commercial activities elsewhere that cause a direct effect in the United States. See id. § 1605(a)(2). Sections 1606 and 1607 address the scope of foreign states' exposure to liability for punitive damages and to counterclaims. See id. §§ 1606-07.
III Role of the FSIA in This Appeal
In their complaint plaintiffs alleged federal subject matter jurisdiction under two statutes: the federal question statute, 28 U.S.C. § 1331, and the Alien Tort Claims Statute, 28 U.S.C. § 1350, which confers on district courts original jurisdiction over “any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States,” id. The Supreme Court's holding in Amerada Hess instructs that, if the FSIA applies, neither of the mentioned statutes can serve as a jurisdictional predicate for this action, and district courts may entertain the case based solely on the jurisdictional provision of the FSIA, 28 U.S.C. § 1330. 488 U.S. at 443, 109 S.Ct. 683. As noted, that section gives district courts jurisdiction over claims against foreign states only when such claims fall within the FSIA's exceptions to the general grant of immunity for foreign states.
Here, the trial court ruled that none of those exceptions applies. Plaintiffs do not contest this ruling, agreeing that if the FSIA applies, the district court has no jurisdiction. The thrust of their argument is that the FSIA does not apply to their claims. Thus, we turn now to the first of the two issues determinative of the Act's applicability-SNCF's status as a state actor.
IV Railroad's Status as French Agency or Instrumentality
Because the Act applies to claims brought against foreign states, their political subdivisions, and their agencies and instrumentalities, see 28 U.S.C. § 1603(a), deciding whether it applies in this case presents the threshold issue of whether SNCF is an agency or instrumentality of France. The district court concluded that it is. We agree.
Under the Act, an entity is an agency or instrumentality of a foreign state if it meets the following three requirements: first, it must be “a separate legal person, corporate or otherwise”; second, it must be “an organ of a foreign state or political subdivision thereof, or a majority of [its] shares or other ownership interest [must be] owned by a foreign state or political subdivision thereof”; third, it must be “neither a citizen of a State of the United States ․ nor created under the laws of any third country.” Id. § 1603(b).
In the case at hand, the complaint and the documents submitted by the parties clearly establish that SNCF has had the required characteristics of an agency or instrumentality of France throughout the course of this litigation. It is undisputed that SNCF is now-and was at the time the complaint was filed-a separate legal entity, wholly-owned by the French government, neither organized under the laws of any third country nor a citizen of any state of the United States.
The evidence in the record does not establish, however, that SNCF also had these three characteristics during World War II. For example, though the railroad's brief asserts that the French state owned 51 percent of the company between 1938 and 1982, no affidavits or documents in the record support this statement. This absence of proof regarding the railroad's status during World War II raises the question: Is the fact that the defendant entity fits the FSIA's definition of an agency or instrumentality of a foreign state at the time of the litigation sufficient to require the Act's application to the case, regardless of that entity's organization and ownership at the time of the alleged wrongdoing?
The issue was unresolved in our Circuit both at the time of the district court's decision in this case and at the time the parties briefed and argued the present appeal. After the oral argument, however, the Supreme Court decided Dole Food Co. v. Patrickson, 538 U.S. 468, 123 S.Ct. 1655, 155 L.Ed.2d 643 (2003), holding unequivocally that an entity's status as an instrumentality of a foreign state should be “determined at the time of the filing of the complaint.” Id. at 1663. Because the record clearly establishes that SNCF was an agency or instrumentality of France at the time the complaint was filed, it is an agency or instrumentality of a foreign state as defined in § 1603(b).
We pass now to the second question that we must resolve: whether the Act may be applied to this case even though the underlying events occurred before that statute's enactment. Citing the Supreme Court's decisions in Landgraf v. USI Film Products, 511 U.S. 244, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994), and Hughes Aircraft Co. v. United States, 520 U.S. 939, 117 S.Ct. 1871, 138 L.Ed.2d 135 (1997), plaintiffs declare Congress did not unequivocally express its aim that the Act apply to pre-enactment events. In the absence of such an unequivocal statement, they continue, application of the statute to their claims would be impermissibly retroactive as it would impair their antecedent rights and settled expectations. We agree there is no unequivocal statement, but find the existing record insufficient to assess the accuracy of plaintiffs' retroactivity argument.
A. Governing Legal Principles
1. Landgraf/Lindh Framework
In Landgraf, the Supreme Court established a two-step approach to determining whether a statute applies to events predating its enactment. First, a court must ask “whether Congress has expressly prescribed the statute's proper reach.” 511 U.S. at 280, 114 S.Ct. 1483. If Congress has done so, the inquiry ends. If not, the court must determine whether applying the statute to pre-enactment events “would have retroactive effect, i.e., whether it would impair rights a party possessed when he acted, increase a party's liability for past conduct, or impose new duties with respect to transactions already completed.” Id. If the statute's application would have such an effect, the court must decline to apply it. Id. This traditional presumption against retroactive legislation, Landgraf explained, is rooted in fundamental notions of fairness which dictate that “settled expectations should not be lightly disrupted” and “that individuals should have an opportunity to know what the law is and to conform their conduct accordingly.” Id. at 265, 114 S.Ct. 1483.
Lindh v. Murphy, 521 U.S. 320, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997), elaborated on Landgraf, holding that the normal rules of construction generally apply when a court determines the temporal reach of a statute. Id. at 326, 117 S.Ct. 2059. Absent a clear and express Congressional directive to apply a statute retroactively, the Court stated, applying a statute to pre-enactment events may be inappropriate for more than one reason. First, as discussed in Landgraf, such application could produce a retroactive effect, and therefore be barred by the customary presumption against retroactivity. In addition, other regular rules of statutory interpretation could “remove even the possibility of retroactivity,” by revealing that Congress planned solely for a prospective application. Id. For example, in Lindh, an amendment to the habeas corpus statute was held inapplicable to non-capital cases pending at the time of the amendment's enactment. The Court noted that a simultaneously enacted provision on capital cases expressly required application to then-pending cases, and reasoned that this express requirement, by negative implication, showed that the amendments pertinent to non-capital cases were meant to apply only to cases filed after the amendments' enactment. Id. at 326-37, 117 S.Ct. 2059.
After Lindh, therefore, a court faced with a retroactivity claim that survived the first step of the Landgraf analysis may not have to decide whether the statute produces a retroactive effect. Instead, by referring to other rules of statutory interpretation, the court may find the statute inapplicable to pre-enactment events.
2. Applicability of Landgraf to New Jurisdiction-Allocating Legislation
SNCF avers the Landgraf analysis does not apply to those aspects of the FSIA that govern federal courts' jurisdiction, that is, the enactment of the exclusive jurisdictional basis for claims against foreign states set out in § 1330 and the simultaneous exclusion of such claims from other general jurisdictional provisions. We disagree with the railroad's contention.
In Landgraf, the Supreme Court recognized that, “[e]ven absent specific legislative authorization, application of new statutes passed after the events in suit is unquestionably proper in many situations.” 511 U.S. at 273, 114 S.Ct. 1483. The Court noted jurisdiction-conferring and jurisdiction-ousting statutes as an example of statutes often properly applied to pre-enactment events. “Application of a new jurisdictional rule,” the Court instructed, “usually takes away no substantive right but simply changes the tribunal that is to hear the case.” Id. at 274, 114 S.Ct. 1483. Further, “[p]resent law normally governs in such situations because jurisdictional statutes speak to the power of the court rather than to the rights or obligations of the parties.” Id.
Landgraf did not suggest that all jurisdiction-defining statutes should be applied to all currently pending lawsuits, or that courts do not have to consider the effects of having such statutes cover pre-enactment events. To the contrary, the Court's use of the adverbs “usually” and “normally” suggests that it did not intend to create a categorical exception from the general retroactivity analysis for jurisdictional statutes.
Three years later, in Hughes Aircraft, the Court confirmed that the general presumption against retroactivity affects jurisdiction-allocating statutes to the same extent that it affects other legislation. 520 U.S. at 950-51, 117 S.Ct. 1871. At issue in Hughes Aircraft was the temporal reach of a 1986 amendment to the False Claims Act that expanded the range of circumstances in which private parties can bring suit “on behalf of the United States against anyone submitting a false claim to the Government.” Id. at 941, 117 S.Ct. 1871. After conducting the two-step analysis outlined in Landgraf, the Supreme Court concluded that the 1986 amendment did not apply where the defendant submitted the alleged false claims before 1986 and a private person could not have brought suit based on those claims under the pre-amendment version of the False Claims Act. Id. at 946-51, 117 S.Ct. 1871.
In rejecting plaintiff's argument that the 1986 amendment, as a jurisdictional statute, was exempt from the Landgraf presumption against retroactivity, the Supreme Court clarified Landgraf, stating
The fact that courts often apply newly enacted jurisdiction-allocating statutes to pending cases merely evidences certain limited circumstances failing to meet the conditions for our generally applicable presumption against retroactivity, not an exception to the rule itself․ As we stated in Landgraf:
“Application of a new jurisdictional rule usually ‘takes away no substantive right but simply changes the tribunal that is to hear the case.’ Present law normally governs in such situations because jurisdictional statutes ‘speak to the power of the court rather than to the rights or obligations of the parties.’ ”
Statutes merely addressing which court shall have jurisdiction to entertain a particular cause of action can fairly be said merely to regulate the secondary conduct of litigation and not the underlying primary conduct of the parties. Such statutes affect only where a suit may be brought, not whether it may be brought at all. The 1986 amendment, however, does not merely allocate jurisdiction among forums. Rather, it creates jurisdiction where none previously existed; it thus speaks not just to the power of a particular court but to the substantive rights of the parties as well. Such a statute, even though phrased in “jurisdictional” terms, is as much subject to our presumption against retroactivity as any other.
Id. at 951, 117 S.Ct. 1871.
In light of the quoted passage from Hughes Aircraft, the railroad's argument that we should apply the jurisdiction-allocating aspects of the Act to pre-1976 events without first engaging in the Landgraf analysis is misplaced. Nor are we persuaded by the railroad's view that the quoted passage pertains only to jurisdiction-creating statutes, and thus does not apply to the jurisdiction-ousting aspects of the FSIA, that is to say, the express amendment of the diversity statute and the implied limitation on other general jurisdictional grants recognized in Amerada Hess. As discussed in Part II B. above, the enactment of the Act's jurisdiction-conferring provision for claims against foreign states in § 1330 is the very reason that other jurisdictional grants no longer govern such claims. Hence, Congress plainly aimed to have the FSIA's jurisdictional ouster applied coextensively with its jurisdictional grant.
More importantly, the railroad's argument misses the central point of the quoted passage from Hughes Aircraft: although jurisdictional statutes are often applied to pre-enactment events, they are not categorically exempt from the Landgraf analysis. If the particular jurisdictional statute “affect[s] only where a suit may be brought, not whether it may be brought,” its application to all currently pending cases usually will not have the impermissible retroactive effect discussed in Landgraf and, therefore, will “fail[ ] to meet the conditions for our generally applicable presumption against retroactivity.” Hughes Aircraft, 520 U.S. at 951; cf. Scott v. Boos, 215 F.3d 940, 944-47 (9th Cir.2000) (rejecting as impermissibly retroactive application of new statute that, although phrased as an exception to existing federal jurisdictional provision, barred plaintiff from asserting in state or federal court a claim he could have pursued at the time of the alleged misconduct); Mathews v. Kidder, Peabody & Co., Inc., 161 F.3d 156, 163-66 (3d Cir.1998) (same).
In urging us to forgo the Landgraf analysis, SNCF relies heavily on Princz v. Federal Republic of Germany, 26 F.3d 1166 (D.C.Cir.1994). In Princz, the plaintiff, a Holocaust survivor, sued the Federal Republic of Germany for injuries suffered in Nazi concentration camps during World War II. Id. at 1168. Dismissing plaintiff's claims for lack of subject matter jurisdiction, the District of Columbia Circuit stated that it did not need to decide whether the FSIA applied to the case. Id. at 1168, 1171. If the statute applied, the court reasoned, plaintiff's claims had to be dismissed because they did not fall within any of the FSIA's exceptions to immunity. Id. at 1171. On the other hand, if the Act did not apply, plaintiff's claims still had to be dismissed because “there [was] no present basis for subject matter jurisdiction over Mr. Princz's claims in the district court regardless of the state of the circuit law concerning immunity in the period 1942-1945.” Id. at 1175-76. After the FSIA's enactment, the court continued, the diversity statute, § 1332, no longer encompasses claims against foreign states. At the same time, the new § 1330 jurisdictional provision of the Act, “asserts federal jurisdiction over a suit against a foreign state only insofar as the foreign state is not entitled to immunity under the FSIA,” and the court had already found no pertinent exception to immunity. Id. at 1176.
We are unable to agree with the Princz court's view that it did not have to and in fact did not decide the issue of the FSIA's applicability. To the contrary, in the second prong of its analysis the majority clearly applied both the FSIA's new jurisdiction-conferring provision, § 1330, and the amended, post-FSIA version of the diversity statute, 28 U.S.C. § 1332.
Further, Princz did so without answering either of the two questions posed by Landgraf: (1) has Congress clearly expressed its intent that these FSIA provisions be applied to lawsuits based on pre-1976 events, and, if not, (2) would the application of these provisions to the plaintiff's claims have a retroactive effect. Instead, Princz noted only that it was not aware of any case law “suggesting that a court can revive the pre-FSIA diversity jurisdiction” over claims against foreign states. 26 F.3d at 1171. Thus, it apparently accepted as axiomatic that the basis for federal subject matter jurisdiction must be found among the jurisdiction-conferring statutes in effect at the time of the lawsuit. As we explained above, in our view such an approach cannot be reconciled with the Supreme Court's discussion of jurisdiction-allocating statutes in Landgraf and, more recently, in Hughes Aircraft.
In consequence, we decline the railroad's invitation to adopt the reasoning of Princz and proceed instead to analyze the Act's applicability to this case under the analytical framework outlined in Landgraf and Lindh.
B. Landgraf/ Lindh Analysis Applied
1. Express Command Lacking to Apply FSIA Retroactively
Under Landgraf, the first question is whether Congress clearly expressed its aim that the statute apply to pre-enactment events. We conclude it did not. In I.N.S. v. St. Cyr, 533 U.S. 289, 121 S.Ct. 2271, 150 L.Ed.2d 347 (2001), the Supreme Court explained that the standard for finding the requisite “unambiguous direction” under Landgraf “is a demanding one.” Id. at 316, 121 S.Ct. 2271. “Cases where this Court has found truly ‘retroactive’ effect adequately authorized by statute have involved statutory language that was so clear that it could sustain only one interpretation.” Id. at 316-17, 121 S.Ct. 2271 (quoting Lindh, 521 U.S. at 328 n. 4, 117 S.Ct. 2059). For instance, we found this standard satisfied in Kuhali v. Reno, 266 F.3d 93 (2d Cir.2001), where a statute expanding the definition of the term “aggravated felony” expressly stated that the new definition applies “regardless of whether the conviction was entered before, on, or after” the date of the statute's enactment. Id. at 110.
In the case at hand, the rigorous requirement of an unambiguous Congressional direction is not satisfied. The closest Congress came to defining expressly the temporal reach of the FSIA was to state that “[c]laims of foreign states to immunity should henceforth be decided by courts of the United States and of the States in conformity with the principles set forth in [that statute],” 28 U.S.C. § 1602 (emphasis added). See Princz, 26 F.3d at 1170. The use of the word “henceforth” can reasonably sustain more than one construction. Courts have in fact reached diametrically opposite conclusions regarding Congressional purpose after pouring meaning into this word. See id. at 1178 (Wald, J., dissenting) (collecting cases). Some have interpreted the verbiage as suggesting that the Act should cover all cases decided after its enactment, regardless of the time of the underlying events. See id. at 1170 (majority opinion). Others have concluded that the word “henceforth” suggests prospective application, meaning application only to suits arising out of post-enactment events-i.e., to some but not all suits filed after the statute's enactment. See Jackson v. People's Republic of China, 794 F.2d 1490, 1497 (11th Cir.1986).
Thus, the statutory language obviously is not so clear as to sustain only one construction and thereby to resolve conclusively the question of Congressional purpose, making additional analysis unnecessary. Cf. id. at 1497-98 (proceeding to analyze the FSIA's legislative history, implications of its effective date provision and effect of its application on defendant's antecedent rights to decide whether the statute should apply to claim arising out of pre-1952 events).
The railroad also insists the FSIA's comprehensiveness shows Congress wanted the statute to apply to all suits filed after its enactment. This proposition is not persuasive in light of the Supreme Court's express pronouncement in St. Cyr that the comprehensiveness of a statute says nothing with respect to Congress' view regarding the “retroactivity of the enactment's individual provisions.” 533 U.S. at 317, 121 S.Ct. 2271.
2. Intent to Apply Prospectively
We next turn to the question added to the Landgraf analysis in Lindh: do ordinary methods of statutory construction establish that the Act's jurisdictional provisions are inapplicable to the pre-enactment events here at issue? We think the answer to that question is no.
As just discussed, Congress' statement that the Act should henceforth govern immunity determinations is ambiguous. Similarly unilluminating is the use of the verb “shall” in the phrase “district courts shall have original jurisdiction ․ of any nonjury civil action against a foreign state.” 28 U.S.C. § 1330(a). Just as the word henceforth does not necessarily indicate prospective application only, this language does not necessarily signal the statute's applicability only to cases arising out of post-enactment events. Consequently, the use of the cited statutory language may not eliminate the option of retroactive applicability of the statute.
Plaintiffs further maintain that Congress indicated the Act was not retroactive by postponing its effective date for 90 days after enactment “to give adequate notice of the act and its detailed provisions to all foreign states.” See FSIA § 8, 1976 U.S.C.C.A.N. (90 Stat.) at 2898; House Report at 33, reprinted in 1976 U.S.C.C.A.N. at 6632. We do not discern in this provision any suggestion of the intended temporal reach of the statute. Nor do we consider Congress' inclusion of express retroactivity directives in the 1996 and 1997 amendments to the Act a sound basis for construing a different legislature's silence on the same subject ten years earlier. See Antiterrorism and Effective Death Penalty Act of 1996, Pub.L. No. 104-132, § 221, 1996 U.S.C.C.A.N. (110 Stat.) 1214, 1241-43 (making amendments on lawsuits against “terrorist states” applicable “to any cause of action arising before, on, or after the date of the enactment of this Act”); Jurisdiction for Lawsuits Against Terrorist States: Technical Correction, Pub.L. No. 105-11, 1997 U.S.C.C.A.N. (111 Stat.) 22, 22 (same).
3. Retroactive Effect
Because other tools of statutory interpretation do not establish the FSIA's inapplicability to this case, we must consider whether its application here would have the retroactive effect described in Landgraf and, accordingly, would be barred by the general presumption against retroactivity. What is called for is “a commonsense, functional judgment about whether the new provision attaches new legal consequences to events completed before its enactment.” St. Cyr, 533 U.S. at 321, 121 S.Ct. 2271.
Plaintiffs declare that having the Act cover the instant case would be impermissibly retroactive because, under laws in effect during World War II, SNCF would not have been immune from this specific suit. Thus, plaintiffs assert, application of the Act to bar their claims deprives them of a right of action they previously had and upsets their settled and legitimate expectation of being able to sue SNCF in the United States.
In Hughes Aircraft, the Supreme Court distinguished between two types of nominally jurisdictional statutes. 520 U.S. at 951, 117 S.Ct. 1871. Those statutes that affect only where a suit may be brought can generally be applied to currently pending cases without impermissible retroactivity. Id. By contrast, statutes that affect whether a lawsuit may be brought at all can produce a retroactive effect if applied to pre-enactment events. Id.
Consistent with Hughes Aircraft, at least two other circuits have concluded that a new statute has a retroactive effect under Landgraf if the statute, although phrased in jurisdictional terms, does not merely change the forum available to plaintiff, but effectively deprives plaintiff of a claim. See Scott, 215 F.3d at 944-47; Mathews, 161 F.3d at 163-66. We adopt this view and hold that the FSIA's application to the present litigation would be retroactive in the Landgraf sense if, as plaintiffs contend, it fully barred claims that previously could have been adjudicated in the United States.
The question remains therefore whether the Act's application effectively extinguishes plaintiffs' causes of action. In our view, it does. We recognize that, unlike the immunity provisions, the jurisdictional provisions of the Act do not apply to state courts. See FSIA §§ 2, 3, 1976 U.S.C.C.A.N. (90 Stat.) at 2891 (codified at 28 U.S.C. §§ 1330, 1332). Hence, it could be argued that application of the jurisdictional provisions simply eliminates federal courts as possible forum choices for plaintiffs, but does not affect plaintiffs' ability to pursue their claims in state courts. There, in turn, plaintiffs could oppose the railroad's immunity defense under the FSIA on retroactivity grounds. Such an interpretation of the Act would expose the railroad to a possibility of being sued in state court, even though the same lawsuit could not be brought in federal court. SNCF appears to adopt this position, urging us to separate the analysis of the Act's applicability into two distinct inquiries: jurisdiction and immunity.
A foreign sovereign's exposure to suit in state courts may not be broader than its exposure to federal courts' jurisdiction under § 1330. To rule otherwise would be inconsistent with Congress' preference that actions involving foreign states be tried in federal courts. Such preference stems, of course, from the sensitivity of actions in American courts against foreign states and the importance of having a uniform, consistent law in this area. House Report at 32, reprinted in 1976 U.S.C.C.A.N. at 6631. The preference is most evident in the FSIA's liberal removal provision that allows a foreign state defendant-even over objections of co-defendants-to remove to federal court any civil action brought against it in state court. See FSIA § 6, 1976 U.S.C.C.A.N. (90 Stat.) at 2898 (codified at 28 U.S.C. § 1441(d)).
Moreover, an interpretation of the Act that allows for a state court action against a foreign government but excludes the same action from district courts' original jurisdiction under § 1330 leads to an obviously anomalous result. Because a foreign sovereign defendant may always remove a case to federal court, see § 1441(d), the action would then be removable to federal court by defendant even though it could not have been originally brought in that court by plaintiff. See 28 U.S.C. § 1441(a) (providing generally for removal of actions within federal district courts' original jurisdiction). Accordingly, applying the Act to bar plaintiffs' claims from federal courts does not simply eliminate one of plaintiffs' forum choices, but fully precludes the claims' adjudication in the United States. Cf. Verlinden, 461 U.S. at 489, 103 S.Ct. 1962 (noting in dictum that “any claim permitted under the [FSIA] may be brought from the outset in federal court” under § 1330(a)).
The final question is whether plaintiffs could have legitimately expected to have their claims adjudicated in the United States prior to the FSIA's enactment. Plaintiffs aver that before the FSIA expanded the definition of a foreign state to encompass state-owned corporations, courts treated such corporations as legal entities separate from their owners and did not recognize the corporations' claims of sovereign immunity. Some pre-FSIA case law from this Circuit supports plaintiffs' assertion. For example, in United States v. Deutsches Kalisyndikat Gesellschaft, 31 F.2d 199 (S.D.N.Y.1929), it was held that a French government-owned mining corporation was not entitled to immunity because the corporation was an entity distinct from its stockholders. Id. at 202-03. Similarly, in Kunglig Jarnvagsstyrelsen v. Dexter & Carpenter, 32 F.2d 195 (2d Cir.1929), we refused to recognize a Swedish railway company's claim of immunity because in its pleadings the company characterized itself as a corporation. Id. at 199-200. In rejecting the immunity claim, we reasoned that such a claim must be raised by an accredited representative of the foreign government. Id.
SNCF correctly points out that in other cases courts have treated national railroad companies as state instrumentalities. See Oliver Am. Trading Co. v. Gov't of the United States of Mex., 5 F.2d 659 (2d Cir.1924). For instance, in Oliver, we recognized the immunity claim of the National Railways of Mexico, on the ground that “the National Railways of Mexico” was simply the name given a system of railroads in the possession of the Mexican government, controlled and operated by Mexico for ten years for national purposes similar to its running the Post Office, Customs Service, or any other branch of the national government. Id. at 661; see also William C. Hoffman, The Separate Entity Rule in International Perspective: Should State Ownership of Corporate Shares Confer Sovereign Status for Immunity Purposes?, 65 Tul. L.Rev. 535, 545-47 (1991) (observing application of separate entity rule to state-owned corporations before FSIA not uniform).
Both sides overlook one crucial aspect of the pre-FSIA law and practice in the United States-the State Department's role in courts' foreign sovereign immunity determinations. As earlier noted, the State Department often intervened in litigation by filing a suggestion of immunity. Further, even in cases where the State Department did not intervene, courts looked to that agency for guidance and generally acted in accordance with its policies.
The relevance of the State Department's policies to our retroactivity analysis is twofold. First, it appears that the State Department sometimes recognized immunity claims of corporations owned by foreign governments. In Miller v. Ferrocarrill Del Pacifico De Nicaragua, 137 Me. 251, 18 A.2d 688 (1941), the State Department recognized a Maine corporation that operated railways within Nicaragua as an instrumentality of the Nicaraguan government, and the court, in turn, granted the company immunity. Id. at 690-91.
Second, and more importantly, the State Department's treatment of ordinary litigation with friendly foreign states in times of peace does not necessarily indicate the position the Department would have taken on claims closely related to war-time crimes of an enemy, such as plaintiffs' claims here. Thus, in Altmann v. Republic of Austria, 317 F.3d 954 (9th Cir.2002), the Ninth Circuit held that Austria could not have legitimately expected immunity from suit “for its alleged complicity in the pillaging and retention of treasured paintings from the home of a Jewish alien who was forced to flee for his life” during the Holocaust. Id. at 964. In reaching this result, the court relied on a 1949 State Department press release, which announced the Department's policy to “relieve American courts from any restraint upon the exercise of their jurisdiction” with respect to claims for the restitution of identifiable property wrongfully taken as a result of the Nazi persecution in Germany. Id. at 965-66.
The record contains no information with respect to the State Department's position during World War II on the significance of the corporate form in foreign sovereign immunity determinations. Nor is there any indication in the record whether the State Department would have recognized immunity in a case such as the one before us. As a consequence, without this information we cannot determine whether plaintiffs legitimately could have expected to litigate their claims in the United States.
Accordingly, for the reasons stated, we vacate the dismissal of the complaint and remand the case to the district court for further proceedings consistent with this opinion.
CARDAMONE, Circuit Judge.