Christopher M. Curran argued the cause for appellants. With him on the briefs were Nicole Erb, Claire A. DeLelle, and Celia A. McLaughlin. Bruce E. Fein, Washington, DC, entered an appearance.

Stuart H. Newberger and Matthew D. McGill argued the causes for appellees James Owens, et al. With them on the brief were Clifton S. Elgarten, Aryeh S. Portnoy, Emily Alban, John L. Murino, Jonathan C. Bond, Michael R. Huston, Steven R. Perles, Edward B. MacAllister, John Vail, Thomas Fortune Fay, Jane Carol Norman, Michael J. Miller, and David J. Dickens. Annie P. Kaplan, John D. Aldock, Washington, DC, and Stephen A. Saltzburg, entered appearances.

Before: Henderson and Rogers, Circuit Judges, and Ginsburg, Senior Circuit Judge.

Ginsburg, Senior Circuit Judge:

Table of Contents

I. Background 762

A. The FSIA Terrorism Exception 763

B. History of this Litigation 765

II. Extrajudicial Killings 769

A. Textual Arguments 770

1. State action requirements under international law 770

2. International law and the TVPA 772

3. State action requirements in the TVPA and the FSIA terrorism exception 773

B. Statutory Purpose 775

C. Statutory History 777

III. Sufficiency of the Evidence Supporting Jurisdiction 778

A. The Evidentiary Hearing 779

1. The sources of evidence presented 779

2. The district court's findings of fact 781

B. Standard of Review 784

C. Admissibility of the Evidence 786

1. The expert testimony 787

2. The State Department reports 792

D. Sufficiency of the Evidence 793

1. Proximate causation 794

2. Sudan's specific intent 798

IV. Timeliness of Certain Claims 799

V. Jurisdiction and Causes of Action for Claims of Third Parties 804

A. Jurisdiction 805

B. Causes of Action 807C. Intentional Infliction of Emotional Distress 809

VI. Punitive Damages 812

A. Whether to Review the Awards of Punitive Damages 812

B. Retroactivity of Punitive Damages Under § 1605A(c) 814

1. Section 1605A operates retroactively 815

2. Clear statement of retroactive effect 816

C. Retroactivity of Punitive Damages Under State Law 817

VII. Vacatur Under Rule 60(b) 818

A. Excusable Neglect Under Rule 60(b)(1) 819

B. Extraordinary Circumstances Under Rule 60(b)(6) 824

On August 7, 1998 truck bombs exploded outside the United States embassies in Nairobi, Kenya and in Dar es Salaam, Tanzania. The explosions killed more than 200 people and injured more than a thousand. Many of the victims of the attacks were U.S. citizens, government employees, or contractors.

As would later be discovered, the bombings were the work of al Qaeda, and only the first of several successful attacks against U.S. interests culminating in the September 11, 2001 attack on the United States itself. From 1991 to 1996, al Qaeda and its leader, Usama bin Laden, maintained a base of operations in Sudan. During this time, al Qaeda developed the terrorist cells in Kenya and Tanzania that would later launch the embassy attacks. This appeal considers several default judgments holding Sudan liable for the personal injuries suffered by victims of the al Qaeda embassy bombings and their family members.

I. Background

Starting in 2001 victims of the bombings began to bring suits against the Republic of Sudan and the Islamic Republic of Iran, alleging that Sudan, its Ministry of the Interior, Iran, and its Ministry of Information and Security materially supported al Qaeda during the 1990s. Specifically, the plaintiffs contended Sudan provided a safe harbor to al Qaeda and that Iran, through its proxy Hezbollah, trained al Qaeda militants. In bringing these cases, the plaintiffs relied upon a provision in the Foreign Sovereign Immunity Act (FSIA) that withdraws sovereign immunity and grants courts jurisdiction to hear suits against foreign states designated as sponsors of terrorism. 28 U.S.C. § 1605(a)(7). This provision and its successor are known as the "terrorism exception" to foreign sovereign immunity.

Initially, neither Sudan nor Iran appeared in court to defend against the suits. In 2004 Sudan secured counsel and participated in the litigation. Within a year, its communication with and payment of its attorneys ceased but counsel continued to litigate until allowed to withdraw in 2009. In the years that followed, several new groups of plaintiffs filed suits against Sudan and Iran. The sovereign defendants did not appear in any of these cases, and in 2010 the district court entered defaults in several of the cases now before us. After an evidentiary hearing in 2010 and the filing of still more cases, the court in 2014 entered final judgments in all pending cases. Sudan then reappeared, filing appeals and motions to vacate the judgments. The district court denied Sudan's motions to vacate, and Sudan again appealed.

Today we address several challenges brought by Sudan on direct appeal of the default judgments and collateral appeal from its motions to vacate. Most of Sudan's contentions require interpretation of the FSIA terrorism exception, to which we now turn.

A. The FSIA Terrorism Exception

Enacted in 1976, the FSIA provides the sole means for suing a foreign sovereign in the courts of the United States. Argentine Republic v. Amerada Hess Shipping Corp. , 488 U.S. 428, 439, 109 S.Ct. 683, 102 L.Ed.2d 818 (1989). A foreign state is presumptively immune from the jurisdiction of the federal and state courts, 28 U.S.C. § 1604, subject to several exceptions codified in §§ 1605, 1605A, 1605B, and 1607.

When first enacted, the FSIA generally codified the "restrictive theory" of sovereign immunity, which had governed sovereign immunity determinations since 1952. Under the restrictive theory, states are immune from actions arising from their public acts but lack immunity for their strictly commercial acts. Verlinden B.V. v. Cent. Bank of Nigeria , 461 U.S. 480, 487-88, 103 S.Ct. 1962, 76 L.Ed.2d 81 (1983). Thus, the original exceptions in the FSIA withdrew immunity for a sovereign's commercial activities conducted in or causing a direct effect in the United States, 28 U.S.C. § 1605(a)(2), and for a few other activities not relevant here. See 28 U.S.C. § 1605(a)(1)-(6).

None of the original exceptions in the FSIA created a substantive cause of action against a foreign state. Rather, the FSIA provided "the foreign state shall be liable in the same manner and to the same extent as a private individual under like circumstances" except that it prohibited the award of punitive damages against a sovereign. 28 U.S.C. § 1606. As a result, a plaintiff suing a foreign sovereign typically relied upon state substantive law to redress his grievances. In this way, the FSIA "operate[d] as a 'pass-through' to state law principles," Pescatore v. Pan Am. World Airways, Inc. , 97 F.3d 1, 12 (2d Cir. 1996), granting jurisdiction yet leaving the underlying substantive law unchanged, First Nat'l City Bank v. Banco Para El Comercio Exterior de Cuba , 462 U.S. 611, 620, 103 S.Ct. 2591, 77 L.Ed.2d 46 (1983).

Until 1996 the FSIA provided no relief for victims of a terrorist attack. Courts consistently rebuffed plaintiffs' efforts to fit terrorism-related suits into an existing exception to sovereign immunity. See, e.g. , Saudi Arabia v. Nelson , 507 U.S. 349, 113 S.Ct. 1471, 123 L.Ed.2d 47 (1993) ; Cicippio v. Islamic Republic of Iran , 30 F.3d 164 (D.C. Cir. 1994) ; Smith v. Socialist People's Libyan Arab Jamahiriya , 886 F.Supp. 306 (E.D.N.Y. 1995). This changed with the passage of the Antiterrorism and Effective Death Penalty Act (AEDPA) of 1996, Pub. L. No. 104-132, 110 Stat. 1214, which added a new exception to the FSIA withdrawing immunity and granting jurisdiction over cases in which

money damages are sought against a foreign state for personal injury or death that was caused by an act of torture, extrajudicial killing, aircraft sabotage, hostage taking, or the provision of material support or resources ... for such an act if such act or provision of material support is engaged in by an official, employee, or agent of such foreign state while acting within the scope of his or her office, employment, or agency.

Id. at § 221, 110 Stat. at 1241-43 (codified at 28 U.S.C. § 1605(a)(7) (2006) (repealed)).

This new "terrorism exception" applied only to (1) a suit in which the claimant or the victim was a U.S. national, 28 U.S.C. § 1605(a)(7)(B)(ii), and (2) the defendant state was designated a sponsor of terrorism under State Department regulations at or around the time of the act giving rise to the suit, § 1605(a)(7)(A) (referencing 50 U.S.C. App. § 2405(j) and 22 U.S.C. § 2371 ). The AEDPA also set a filing deadline for suits brought under the new exception at ten years from the date upon which a plaintiff's claim arose. 28 U.S.C. § 1605(f).

Initially, there was some confusion about whether the new exception created a cause of action against foreign sovereigns. See In re Islamic Republic of Iran Terrorism Litig. , 659 F.Supp.2d 31, 42-43 (D.D.C. 2009). Within five months of enacting the AEDPA, the Congress clarified the situation with an amendment, codified as a note to the FSIA, Pub. L. No. 104-208, § 589, 110 Stat. 3009, 3009-172 (1996) (codified at 28 U.S.C. § 1605 note), which provides:

[A]n official, employee, or agent of a foreign state designated as a state sponsor of terrorism ... while acting within the scope of his or her office, employment, or agency shall be liable to a United States national or the national's legal representative for personal injury or death caused by acts of that official, employee, or agent for which the courts of the United States may maintain jurisdiction under section 1605(a)(7) of title 28, United States Code, for money damages which may include economic damages, solatium, pain, and suffering, and punitive damages if the acts were among those described in section 1605(a)(7).

This amendment was known as the Flatow Amendment after Alisa Flatow, a Brandeis University student mortally wounded in a suicide bombing in the Gaza Strip. The Flatow Amendment, which the Congress intended to deter state support for terrorism, (1) provided a cause of action against officials, employees, or agents of a designated state sponsor of terrorism and (2) authorized the award of punitive damages against such a defendant. These two changes marked a departure from the other FSIA exceptions, none of which provided a cause of action or allowed for punitive damages. See 28 U.S.C. § 1606.

Although it referred in terms only to state officials, for a time some district courts read the Flatow Amendment and § 1605(a)(7) to create a federal cause of action against foreign states themselves. See, e.g. , Kilburn v. Republic of Iran , 277 F.Supp.2d 24, 36-37 (D.D.C. 2003). But see Roeder v. Islamic Republic of Iran , 195 F.Supp.2d 140, 171 (D.D.C. 2002). In Cicippio-Puleo v. Islamic Republic of Iran , we rejected this approach, holding that "neither 28 U.S.C. § 1605(a)(7) nor the Flatow Amendment, nor the two considered in tandem, creates a private right of action against a foreign government." 353 F.3d 1024, 1033 (D.C. Cir. 2004). We based this conclusion upon the plain text of the Flatow Amendment—which applied only to state officials—and upon the function of all the other exceptions to the FSIA, which withdraw immunity but leave the substantive law of liability unchanged. Id. at 1033-34 (noting the "settled distinction in federal law between statutory provisions that waive sovereign immunity and those that create a cause of action"). Because there was no federal cause of action, we remanded the case "to allow plaintiffs an opportunity to amend their complaint to state a cause of action under some other source of law, including state law." Id. at 1036. Hence, a plaintiff proceeding under the terrorism exception would follow the same pass-through process that governed an action under the original FSIA exceptions.

The pass-through approach, however, produced considerable difficulties. In cases with hundreds or even thousands of claimants, courts faced a "cumbersome and tedious" process of applying choice of law rules and interpreting state law for each claim. See Iran Terrorism Litig. , 659 F.Supp.2d at 48. Differences in substantive law among the states caused recoveries to vary among otherwise similarly situated claimants, denying some any recovery whatsoever. See Peterson v. Islamic Republic of Iran , 515 F.Supp.2d 25, 44-45 (D.D.C. 2007) (denying recovery for intentional infliction of emotional distress to plaintiffs domiciled in Pennsylvania and Louisiana while permitting recovery for plaintiffs from other states).

The Congress addressed these problems in 2008. Section 1083 of the National Defense Authorization Act for Fiscal Year 2008 (NDAA) repealed § 1605(a)(7) and replaced it with a new "Terrorism exception to the jurisdictional immunity of a foreign state." Pub. L. No. 110-181, § 1083, 122 Stat. 3, 338-44 (2008) (hereinafter NDAA) (codified at 28 U.S.C. § 1605A ). The new exception withdrew immunity, granted jurisdiction, and authorized suits against state sponsors of terrorism for "personal injury or death" arising from the same predicate acts—torture, extrajudicial killing, aircraft sabotage, hostage taking, and the provision of material support—as had the old exception. 28 U.S.C. § 1605A(a)(1). Jurisdiction for suits under the new exception extended to "claimants or victims" who were U.S. nationals, and for the first time, to members of the armed forces and to government employees or contractors acting within the scope of their employment. 28 U.S.C. § 1605A(a)(2)(A)(ii). Most important, the new exception authorized a "[p]rivate right of action" against a state over which a court could maintain jurisdiction under § 1605A(a). 28 U.S.C. § 1605A(c). By doing so, the Congress effectively abrogated Cicippio-Puleo and provided a uniform source of federal law through which plaintiffs could seek recovery against a foreign sovereign. Iran Terrorism Litig. , 659 F.Supp.2d at 59. A claimant who was a U.S. national, military service member, government employee or contractor acting within the scope of his employment, and the claimant's legal representative could make use of this cause of action. As with the Flatow Amendment but unlike § 1605(a)(7), the NDAA authorized awards of punitive damages under the new federal cause of action. The exception also provided claimants a host of other new benefits not relevant here.

Like its predecessor, the new exception contained a ten-year limitation period on claims brought under § 1605A. Notwithstanding the limitation period, the NDAA provided two means of bridging the gap between the now-repealed § 1605(a)(7) and the new § 1605A. Claimants with claims "before the courts in any form" who had been adversely affected by the lack of a federal cause of action in § 1605(a)(7) could move to convert or refile their cases under § 1605A(c). NDAA § 1083(c)(2). Furthermore, "[i]f an action arising out of an act or incident has been timely commenced under section 1605(a)(7) or [the Flatow Amendment]," then a claimant could bring a "related action" "arising out of the same act or incident" within 60 days of the entry of judgment in the original action or of the enactment of the NDAA, whichever was later. NDAA § 1083(c)(3). Each of these provisions is examined below in greater detail as they relate to Sudan's arguments.

B. History of this Litigation

This appeal follows 15 years of litigation against Sudan arising from the 1998 embassy bombings. In October 2001 plaintiff James Owens filed the first lawsuit against Sudan and Iran for his personal injuries. Other plaintiffs joined the Owens action in the following year. These included individuals (or the legal representatives of individuals) killed or injured in the bombings, who sought recovery for their physical injuries (or deaths), and the family members of those killed or injured, who sued for their emotional distress. The Owens complaint alleged that the embassy bombings were "extrajudicial killings" under the FSIA and that Sudan provided material support for the bombings by sheltering and protecting al Qaeda during the 1990s.

When Sudan failed to appear, the district court entered an order of default in May 2003. The default was translated into Arabic and sent to Sudan in accordance with 28 U.S.C. § 1608(e). In February 2004 Sudan secured counsel and in March 2004 moved to vacate the default and to dismiss the Owens action. Sudan argued, among other things, it remained immune under the FSIA because the plaintiffs had not adequately pleaded facts showing it had materially supported al Qaeda or that its support had caused the bombings. Sudan attached to its motion declarations from a former U.S. Ambassador to Sudan and a former FBI agent stating that it neither assisted al Qaeda nor knew of the group's terrorist aims during the relevant period.

In March 2005 the district court granted, in part, Sudan's motion to dismiss and vacated the order of default. Owens v. Republic of Sudan , 374 F.Supp.2d 1, 9-10 (D.D.C. 2005) ( Owens I ). The court, however, allowed the plaintiffs to amend their complaint in order to develop more fully their allegations of material support. Id. at 15. The court further noted that although "the Sudan defendants severed ties to al Qaeda two years before the relevant attacks," this timing did not necessarily foreclose the conclusion that Sudan had "provided material support within the meaning of the statute and that this support was a proximate cause of the embassy bombings." Id. at 17.

The plaintiffs then amended their complaint, and Sudan again moved to dismiss. Sudan once again argued the complaint had not sufficiently alleged material support and that any support it provided was not a legally sufficient cause of the embassy bombings. Assuming the truth of the plaintiffs' allegations, the district court denied Sudan's motion in its entirety. Owens v. Republic of Sudan , 412 F.Supp.2d 99, 108, 115 (D.D.C. 2006) ( Owens II ).

While the motions to dismiss were pending, difficulties arose between Sudan and its counsel. After filing the first motion to dismiss, Sudan's initial counsel withdrew due to a conflict of interest with the Iranian codefendants. Sudan retained new counsel, but their relationship soon deteriorated. Starting in January 2005 new counsel filed several motions to withdraw, citing Sudan's unresponsiveness and failure to pay for legal services. Sudan's last communication with counsel was in September 2008. The district court eventually granted a final motion to withdraw in January 2009, leaving Sudan without representation.

Despite these difficulties, counsel for Sudan continued to defend their client until the court granted the motion to withdraw in January 2009. Following the denial of its second motion to dismiss, Sudan pursued an interlocutory appeal to this court. Its appeal, in part, challenged the legal sufficiency of the plaintiffs' allegations that Sudan's material support had caused the embassy bombings. In July 2008 we affirmed the district court's decision, holding that "[a]ppellees' factual allegations and the reasonable inferences that can be drawn therefrom show a reasonable enough connection between Sudan's interactions with al Qaeda in the early and mid-1990s and the group's attack on the embassies in 1998" to maintain jurisdiction under the FSIA. Owens v. Republic of Sudan , 531 F.3d 884, 895 (D.C. Cir. 2008) ( Owens III ). We then remanded the case to allow the plaintiffs to pursue the merits of their claims.

Shortly after our decision, several new groups of plaintiffs filed actions against Sudan and Iran arising from the embassy bombings. These actions—brought by the Wamai, Amduso, Mwila, and Osongo plaintiffs—were filed after the enactment of the new terrorism exception and before the expiration of its limitation period. This brought the total number of suits against Sudan to six, including the original Owens action and a suit filed by the Khaliq plaintiffs under § 1605(a)(7).

From that point on, neither Sudan nor its counsel participated in the litigation again until after the 2014 entry of final judgment in Owens . After entering new orders of defaults against Sudan in several of the pending actions, the court held a consolidated evidentiary hearing in order to satisfy a requirement in the FSIA that "the claimant establish[ ] his claim or right to relief by evidence satisfactory to the court." 28 U.S.C. § 1608(e). Without considering this evidence, the court could not transform the orders of default into enforceable default judgments establishing liability and damages against Sudan.

For three days, the district court heard expert testimony and reviewed exhibits detailing the relationship between both Iran and Sudan and al Qaeda during the 1990s. Shortly after this hearing the district court held both defendants liable for materially supporting the embassy bombings. Owens v. Republic of Sudan , 826 F.Supp.2d 128, 157 (D.D.C. 2011) ( Owens IV ). More specifically, the district court found Sudan had provided al Qaeda a safe harbor from which it could establish and direct its terrorist cells in Kenya and Tanzania. Id. at 139-43, 146. The court further found Sudan provided financial, military, and intelligence assistance to the terrorist group, which allowed al Qaeda to avoid disruption by hostile governments while it developed its capabilities in the 1990s. Id. at 143-46. These findings established both jurisdiction over and substantive liability for claims against Sudan and Iran.

The court also addressed the claims of non-American family members of those killed or injured in the bombings. Although those plaintiffs could not make use of the federal cause of action in § 1605A(c), the court concluded they could pursue claims under state law, as was the practice under the previous terrorism exception. Id. at 153. The court's opinion was translated into Arabic and served upon Sudan in September 2012.

The district court then referred the cases to special masters to hear evidence and recommend the amounts of damages to be awarded. While this process was ongoing, two new sets of plaintiffs entered the litigation. In July 2012 the Opati plaintiffs filed suit against Sudan, claiming their suits were timely as a "related action" with respect to the original Owens litigation. In May 2012 the Aliganga plaintiffs sought to intervene in the Owens suit. Notwithstanding the expiration of the ten-year limitation period starting from the date of the bombings, the district court allowed both groups of plaintiffs to proceed against Sudan and to rely upon the court's factual findings of jurisdiction and liability. The court then referred the Aliganga and Opati claims to the special masters.

In 2014 the district court entered final judgments in favor of the various plaintiffs. All told, the damages awarded against Sudan came to more than $10.2 billion. Family members, who outnumbered those physically injured by the bombing, received the bulk of the award—over $7.3 billion. Of the total $10.2 billion, approximately $4.3 billion was punitive damages. See, e.g. , Opati v. Republic of Sudan , 60 F.Supp.3d 68, 82 (D.D.C. 2014).Within a month of the first judgments, Sudan retained counsel and reappeared in the district court. Sudan appealed each case and in April 2015 filed motions in the district court to vacate the default judgments under Federal Rule of Civil Procedure 60(b). We stayed the appeals pending the district court's ruling on the motions.

In those motions, Sudan raised a number of arguments for vacatur, most of them challenging the district court's subject matter jurisdiction. As before, Sudan also attacked the plaintiffs' evidence. It argued the judgments were void because they rested solely upon inadmissible evidence to prove jurisdictional facts, which Sudan argued was impermissible under § 1608(e). It also argued the evidence did not show it proximately caused the bombings because al Qaeda did not become a serious terrorist threat until after Sudan had expelled bin Laden in 1996.

Sudan raised a host of new arguments as well. In its most sweeping challenge, Sudan argued it did not provide material support for any predicate act that would deprive it of immunity under the FSIA. In making this argument, Sudan contended the embassy bombings, carried out by al Qaeda, were not "extrajudicial killings" because that term requires the involvement of a state actor in the act of killing. Sudan also contended the claims brought by the Opati, Aliganga, and Khaliq plaintiffs were barred by the statute of limitation in § 1605A(b) which, it argued, deprived the court of jurisdiction to hear their suits.

Sudan's last jurisdictional challenge took aim at the family members of those physically injured or killed by the bombings. Sudan argued that the court could hear claims only from a person who was physically harmed or killed by the bombings or the legal representative of that person. And even if jurisdiction was proper, Sudan contended, foreign (i.e., non-U.S.) family members could not state a claim under either the federal cause of action or state law.

Finally, Sudan raised two nonjurisdictional arguments: First, it urged the district court to vacate its awards of punitive damages to the plaintiffs proceeding under state law, contending § 1605A(c) is the sole means for obtaining punitive damages against a foreign state. Second, Sudan argued the court should vacate the default judgments under Federal Rule of Civil Procedure 60(b) for "extraordinary circumstances" or "excusable neglect" on Sudan's part. In support of the latter argument, Sudan submitted a declaration from the Sudanese Ambassador to the United States detailing the country's troubled history of civil unrest, natural disaster, and disease, which allegedly impeded Sudan's participation in the litigation.

After a consolidated hearing, the district court denied the motions to vacate in all respects. Owens v. Republic of Sudan , 174 F.Supp.3d 242 (D.D.C. 2016) ( Owens V ). Sudan appealed and its appeal was consolidated with its earlier appeals from the final judgments. Sudan's briefs before this court are directed primarily to the district court's jurisdiction, and present novel questions of law, which we review de novo. See Jerez v. Republic of Cuba , 775 F.3d 419, 422 (D.C. Cir. 2014). Ordinarily, all of Sudan's nonjurisdictional arguments would be forfeited by reason of its having defaulted in the district court. See Practical Concepts, Inc. v. Republic of Bolivia , 811 F.2d 1543, 1547 (D.C. Cir. 1987). In this case, however, due to the size of the judgments against Sudan, their possible effects upon international relations, and the likelihood that the same arguments will arise in future litigation, we exercise our discretion to consider some, but not all, of Sudan's nonjurisdictional objections. See Acree v. Republic of Iraq , 370 F.3d 41, 58 (D.C. Cir. 2004) ("while we will ordinarily refrain from reaching non-jurisdictional questions that have not been raised by the parties ... we may do so on our own motion in 'exceptional circumstances' ").

At the end of the day, we affirm the judgments in most respects, holding the FSIA grants jurisdiction over all the claims and claimants present here. We hold also that those plaintiffs ineligible to proceed under the federal cause of action may continue to press their claims under state law. We also vacate all the awards of punitive damages and certify a question of local tort law to the District of Columbia Court of Appeals.

We turn first to Sudan's challenges to the district court's subject matter jurisdiction, starting with those that would dispose of the entire case. In Part II we address Sudan's challenge to the meaning of "extrajudicial killings" under the FSIA. In Part III we review the sufficiency of the evidence supporting the conclusions that Sudan provided material support to al Qaeda and that this support was a jurisdictionally sufficient cause of the embassy bombings.

We then proceed to Sudan's jurisdictional challenges that would eliminate the claims of particular plaintiffs. In Part IV we consider whether some of the plaintiffs' claims are barred by the statute of limitation in the FSIA terrorism exception, which Sudan contends is jurisdictional. In Part V we address both jurisdictional and nonjurisdictional arguments opposing the claims of the family members of victims physically injured or killed by the embassy bombings. Finally, we address Sudan's purely nonjurisdictional arguments in Part VI—whether the new terrorism exception authorizes punitive damages for a sovereign's pre-enactment conduct—and Part VII—addressing Sudan's arguments for vacatur under Rule 60(b)(1) and 60(b)(6).

II. Extrajudicial Killings

Sudan first argues the 1998 embassy bombings were not "extrajudicial killings" within the meaning of the FSIA terrorism exception. As noted above, § 1605A divests a foreign state of immunity and grants courts jurisdiction over cases

in which money damages are sought against a foreign state for personal injury or death that was caused by ... extrajudicial killing ... or the provision of material support or resources for such an act if such act or provision of material support or resources is engaged in by an official, employee, or agent of such foreign state while acting within the scope of his or her office, employment, or agency.

Because this argument poses a challenge to the court's subject matter jurisdiction, it was not forfeited by Sudan's failure to appear in the district court. See Practical Concepts , 811 F.2d at 1547. This is Sudan's most sweeping challenge, and, if correct, then the claims of all the plaintiffs must fail. The district court rejected Sudan's jurisdictional argument based upon the plain meaning of "extrajudicial killing." Owens V , 174 F.Supp.3d at 259-66. Reviewing de novo this question of law relating to our jurisdiction, we agree that "extrajudicial killings" include the terrorist bombings that gave rise to these cases.

Section 1605A(h)(7) of the FSIA provides that the term "extrajudicial killing" has the meaning given to it in § 3(a) of the Torture Victim Protection Act of 1991, which defines an extrajudicial killing as:

a deliberated killing not authorized by a previous judgment pronounced by a regularly

constituted court affording all the judicial guarantees which are recognized as indispensable by civilized peoples. Such term, however, does not include any such killing that, under international law, is lawfully carried out under the authority of a foreign nation.

Pub. L. No. 102-256, § 3(a), 106 Stat. 73, 73 (1991) (codified at 28 U.S.C. § 1350 note) (hereinafter TVPA).

On its face, this definition contains three elements: (1) a killing; (2) that is deliberated; and (3) is not authorized by a previous judgment pronounced by a regularly constituted court. The 1998 embassy bombings meet all three requirements and do not fall within the exception for killings carried out under the authority of a foreign nation acting in accord with international law. First, the bombings caused the death of more than 200 people in Kenya and Tanzania. The bombings were "deliberated" in that they involved substantial preparation, meticulous timing, and coordination across multiple countries in the region. See Mamani v. Berzain , 654 F.3d 1148, 1155 (11th Cir. 2011) (defining "deliberated" under the TVPA as "being undertaken with studied consideration and purpose"). Finally, the bombings themselves were neither authorized by any court nor by the law of nations. Therefore, on its face, the FSIA would appear to cover the bombings as extrajudicial killings.

Sudan offers a host of reasons we should ignore the plain meaning of "extrajudicial killing" in the TVPA and exclude terrorist bombings like the 1998 embassy attacks from jurisdiction under the FSIA terrorism exception. Sudan's arguments draw upon the text and structure, the purpose, and the legislative history of the TVPA and of the FSIA terrorism exception. Each of Sudan's arguments shares the same basic premise: Only a state actor, not a nonstate terrorist, may commit an "extrajudicial killing."

A. Textual Arguments

We begin, as we must, with the text of the statute. First, Sudan contends the text of the TVPA, and, by extension of the FSIA, defines an "extrajudicial killing" in terms of international law, specifically the Geneva Conventions. According to Sudan, international law generally and the Geneva Conventions specifically prohibit only killings carried out by a state actor. The plaintiffs vigorously contest both propositions.

1. State action requirements under international law

Sudan bases its argument that principles of international law supply the meaning of "extrajudicial killing" in the FSIA upon similarities between the TVPA and the prohibition on "summary executions" in Common Article 3 of the Geneva Conventions of 1949, which condemns "the passing of sentences and the carrying out of executions without previous judgment pronounced by a regularly constituted court, affording all the judicial guarantees which are recognized as indispensable by civilized peoples." Geneva Convention for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field, art. 3(1)(d), Aug. 12, 1949, 6 U.S.T. 3114, 75 U.S.T.S. 85. The similarities between the two definitions, Sudan contends, shows the Congress intended to define an "extrajudicial killing" in the TVPA with reference to principles of international law adopted in the Geneva Conventions.

To Sudan, this is of critical importance because the Geneva Conventions and international law, it argues, proscribe killings only when committed by a state agent, not when perpetrated by a nonstate actor. Three pieces of evidence are said to demonstrate this limitation. First, Sudan notes, the United Nations adopted a resolution in 1980 condemning as inconsistent with international law "[e]xtra-legal executions" carried out by "armed forces, law enforcement or other governmental agencies." Congress on the Prevention of Crime and the Treatment of Offenders Res., A/Conf.87/L.11 (Sep. 5, 1980). Second, Sudan cites a United Nations annual report, S. Amos Wako (Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions), Summary or Arbitrary Executions , ¶¶ 74-85, U.N. Doc. E/CN.4/1983/16 (Jan. 31, 1983), which describes "extralegal executions" and "summary executions" in terms suggesting state involvement. And third, Sudan references an online database of the United Nations, which links the term "extrajudicial killing" to the definition of "extralegal execution." U.N. Terminology Database, http://untermportal.un.org/UNTERM/display/Record/UNHQ/extra-legal_execution/c253667 (last visited July 19, 2017).

Each of these references to international law is both inapposite and rebutted by the plaintiffs. If Sudan means to say the TVPA incorporates the prohibition against a "summary execution" in the Geneva Conventions, then it must show what was meant by that term in the Geneva Conventions themselves. In doing so, however, Sudan principally relies upon U.N. documents published more than a quarter century after the ratification of the Geneva Conventions in 1949, rather than the deliberations over the proposed Conventions, which Sudan does not cite at all. Odder still, none of these documents (or the terminology database) actually says the Geneva Conventions proscribe only "summary executions" committed by a state actor. See Summary or Arbitrary Executions , supra p. 22, ¶¶ 35-36 (noting Article 3 of the Geneva Conventions prohibits "murder" in general and "also specifically prohibits the passing of sentences and the carrying out of executions without previous judgement pronounced by a regularly constituted court"). Indeed, the plaintiffs present reasons to doubt whether the Geneva Conventions in specific, or international law in general, prohibit only killings by a state actor. As the plaintiffs note, Article 3 of the First Convention prohibits "violence to life and person, in particular murder of all kinds." Geneva Convention, art. 3(1)(a), Aug. 12, 1949, 6 U.S.T. 3114, 75 U.S.T.S. 85. Likewise, the U.N. Terminology Database lists "[k]illings committed by vigilante groups" as an example of an "extrajudicial killing." And finally, a "Handbook" published by the U.N. Special Rapporteur on Summary or Arbitrary Executions contains a full chapter on "killings by non-state actors and affirmative state obligations," which states that "Human rights and humanitarian law clearly apply to killings by non-State actors in certain circumstances." Project on Extrajudicial Executions, UN Special Rapporteur on Extrajudicial Executions Handbook , ¶ 45, http://www.extrajudicialexecutions.org/application/media/Handbook% 20Chapter% 203-Responsibility% 20of% 20states% 20for% 20non-state% 20killings.pdf (last visited July 19, 2017).

This does not mean Sudan's interpretation of international law as it pertains to summary executions (as opposed to extrajudicial killings) is wrong or that direct state involvement is not needed for certain violations of international law. Rather, the point is that the role of the state in an extrajudicial killing appears less clear under international law than Sudan would have us believe; indeed it appears less clear than the definition of an "extrajudicial killing" in the TVPA itself. Accordingly, we doubt the Congress intended categorically to preclude state liability for killings by nonstate actors by adopting a definition of "extrajudicial killing" similar to that of a "summary execution" in the Geneva Conventions.

2. International law and the TVPA

More important, even if Sudan's interpretation of the Geneva Conventions and international law is correct, its argument would fail because the TVPA does not appear to define an "extrajudicial killing" coextensive with the meaning of a "summary execution" (or any similar prohibition) under international law. For example, the TVPA does not adopt the phrasing of the Geneva Conventions wholesale. Rather, as the plaintiffs point out, the TVPA substitutes the term "deliberated killing" for "the passing of sentences and the carrying out of executions" in the Geneva Conventions. While "the passing of sentences and the carrying out of executions" strongly suggests at least some level of state involvement, a nonstate party may commit a "deliberated killing" as readily as a state actor. Indeed, several other statutes contemplate "deliberate" attacks by nonstate entities, including terrorist groups. See, e.g. , 6 U.S.C. § 1169(a) (requiring the Secretary of Transportation to assess vulnerability of hazardous materials in transit to a "deliberate terrorist attack"); 42 U.S.C. § 16276 (mandating research on technologies for increasing "the security of nuclear facilities from deliberate attacks"). Due to the substitution of "deliberated" killings for "the passing of sentences and the carrying out of executions," the inference of direct state involvement is much less strong in the TVPA than in the Geneva Conventions. The difference between the definition in the TVPA and the prohibition in the Geneva Conventions also signals the Congress intended the TVPA to reach a broader range of conduct than just "summary executions." For the court to rely upon the narrower prohibition in the Geneva Conventions would contravene the plain text of the TVPA, which is, after all, the sole "authoritative statement" of the law. See Exxon Mobil Corp. v. Allapattah Servs., Inc. , 545 U.S. 546, 568, 125 S.Ct. 2611, 162 L.Ed.2d 502 (2005).

Resisting this conclusion, Sudan points to two phrases that, it contends, impose a state actor requirement upon the definition of an extrajudicial killing in the TVPA. First, Sudan notes that an extrajudicial killing must not be one "authorized by a previous judgment pronounced by a regularly constituted court." As Sudan would have it, the "only killings that can be reasonably be imagined to be authorized by a 'previous judgment' are those by state actors." Regardless whether Sudan is right on this point, the argument does not imply what Sudan intends. If only a state actor may lawfully kill based upon a "previous judgment," then all killings committed by a nonstate actor are, by definition, not "authorized by a previous judgment." Therefore, only a killing committed by a state actor might not be an "extrajudicial killing," that is, if it was "authorized by a previous judgment pronounced by a regularly constituted court." Accepting Sudan's premise, no other outcome can "reasonably be imagined."

Similarly, Sudan argues the second sentence in the definition of an "extrajudicial killing" in the TVPA anchors the meaning of the first sentence in international law which, in Sudan's view, prohibits only summary executions by state actors. Even accepting Sudan's view of international law, we are not persuaded. In the first sentence of § 3(a), the Congress defined the proscribed conduct (i.e., a "deliberated killing") in terms that extended beyond the prohibition on a "summary execution" under international law. The second sentence excludes from the definition of "extrajudicial killing" "any ... killing that, under international law, is lawfully carried out under the authority of a foreign nation." This ensured that the more expansive prohibition of the first sentence would not reach the traditional prerogatives of a sovereign nation. Were "extrajudicial killings" no broader than "summary executions," the limitation in international law of what constitutes an "extrajudicial killing" would be unnecessary because, by Sudan's own argument, a "summary execution" always violates international law. Therefore, Sudan's interpretation would make superfluous the reference to killings "lawfully carried out" "under international law," contrary to the "cardinal principle of statutory construction that we must give effect, if possible, to every clause and word of a statute." See Williams v. Taylor , 529 U.S. 362, 404, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (internal quotation marks and citation omitted).

Moreover, the reference to international law in the second sentence of § 3(a) of the TVPA highlights its omission in the first sentence. Had the Congress intended the definition of an "extrajudicial killing" to track precisely with that of a "summary execution" under international law, § 3(a) could have expressly referenced international law in both the prohibition and its limitation. That approach is found elsewhere in the FSIA, see 28 U.S.C. § 1605(a)(3) (authorizing jurisdiction where "rights in property [are] taken in violation of international law"), as well as in other statutes, see 18 U.S.C. § 1651 (proscribing "the crime of piracy as defined by the law of nations"). Indeed, the Congress specifically defined other predicate acts in § 1605A by reference to international treaties, see 28 U.S.C. § 1605A(h)(1),(2) (defining "aircraft sabotage" and "hostage taking" with reference to international treaties), but referenced only a U.S. statute, the TVPA, in its definition of "extrajudicial killing." That the Congress incorporated international law expressly into other jurisdictional provisions undermines the inference that it intended implicitly to do so here. See Dep't of Homeland Sec. v. MacLean , ––– U.S. ––––, 135 S.Ct. 913, 919, 190 L.Ed.2d 771 (2015) ("Congress generally acts intentionally when it uses particular language in one section of a statute but omits it in another").

3. State action requirements in the TVPA and the FSIA terrorism exception

The plaintiffs provide another persuasive reason Sudan's textual arguments are flawed. The TVPA authorizes an action only for harms arising from the conduct of a state actor. See TVPA § 2(a) (providing a cause of action against an "individual who, under actual or apparent authority, or color of law, of any foreign nation" engages in torture or extrajudicial killing). Sudan argues the state actor requirement for a suit under the TVPA is "necessarily incorporated" in § 3(a) and therefore applies to those actions arising from "extrajudicial killings" under the FSIA. The limitation of actions to state actors, however, is found not in § 3(a) but in § 2(a) of the TVPA. As the plaintiffs note, when passing the current and prior FSIA terrorism exceptions, the Congress each time incorporated the section of the TVPA that defined an "extrajudicial killing" but not the section that limited the cause of action under the TVPA to state actors. If the Congress had wanted to limit extrajudicial killings to state actors, then it could have incorporated both sections of the TVPA into the FSIA terrorism exception. That it did not compels us to conclude the state actor limitation in the TVPA does not transfer to the definition of an "extrajudicial killing" in the FSIA. Cf. Sebelius v. Cloer , 569 U.S. 369, 133 S.Ct. 1886, 1894, 185 L.Ed.2d 1003 (2013) (declining to apply limitations from one section of a statute when the text of another section does not cross-reference the first section).

Indeed, the reason the Congress declined to incorporate the state-actor limitation in the TVPA is plain on the face of the FSIA terrorism exception. As the plaintiffs observe, the TVPA and the FSIA share a similar structure. Each statute defines the predicate acts that give rise to liability in one section—TVPA § 3 and FSIA § 1605A(h) —and then limits who may be subjected to liability in another—TVPA § 2 and FSIA §§ 1605A(a)(1) and (c). Both statutes also require a plaintiff to show a certain type of nexus to a foreign sovereign. In the TVPA, a state official must act "under actual or apparent authority, or color of law" of a foreign sovereign. In the FSIA, liability arises when the state official, employee, or agent acting within the scope of his authority either directly commits a predicate act or provides "material support or resources" for another to commit that act. If the more stringent state-actor limitation in the TVPA traveled with the definition of an "extrajudicial killing" in that statute, then it would all but eliminate the "material support" provision of § 1605A(a), at least with respect to extrajudicial killings. For example, § 1605A(a) would extend jurisdiction over a sovereign that did not directly commit an extrajudicial killing only if an official of the defendant state materially supported a killing committed by a state actor from a different state. We seriously doubt the Congress intended the exception to immunity for materially supporting an extrajudicial killing to be so narrow.

Sudan attempts to avoid the conclusion that the FSIA does not adopt the state-actor limitation in the TVPA in two ways. First, Sudan contends the introductory clause of § 3(a) implicitly incorporates the state actor limitation of § 2(a). This clause states that an "extrajudicial killing" is defined "[f]or the purposes of this Act." That supposedly indicates the Congress intended to import the state actor limitation of § 2(a) into the definition of an extrajudicial killing in § 3(a). But Sudan's reading of this phrase leads to an illogical conclusion. A statutory definition made expressly "[f]or the purposes of this Act" informs our understanding of the entire statute. In other words, the definitions in TVPA § 3 govern the use of those defined terms elsewhere in the Act. Under Sudan's interpretation, however, the reverse would occur: in order to understand the meaning of a defined term, we would have to look to the remainder of the statute, and not to the definition itself. What then, we wonder, would the definition contribute to the statute? Would it be wholly redundant, a conclusion that conflicts with our usual interpretive presumptions? See Nat'l Ass'n of Home Builders v. Defs. of Wildlife , 551 U.S. 644, 669, 127 S.Ct. 2518, 168 L.Ed.2d 467 (2007). Or, if not redundant, how would a court then apply the definition to terms used in the remainder of the statute if the remainder of the statute, in turn, gave meaning to the definition? Given these paradoxes, the phrase "[f]or the purposes of this Act" cannot mean what Sudan contends. Instead, that phrase simply means that the definition of an "extrajudicial killing" in TVPA § 3(a) informs the remainder of the TVPA (and, by extension, the FSIA), and not the reverse.

Second, Sudan contends the definition of an "extrajudicial killing" in the TVPA implicitly incorporates international law (and the supposed state-actor limitation therein) even without reference to the state-actor limitation in § 2(a). Here Sudan relies principally upon a dictum in a Second Circuit opinion discussing the TVPA in a case arising under the Alien Tort Claims Act (ATCA), which expressly incorporates international law: "torture and summary execution—when not perpetrated in the course of genocide or war crimes—are proscribed by international law only when committed by state officials or under color of law." Kadic v. Karadžic, 70 F.3d 232, 243 (2nd Cir.1995). The court further noted that "official torture is prohibited by universally accepted norms of international law, and the Torture Victim Act confirms this holding and extends it to cover summary execution." Id. at 244 (citation omitted). This, Sudan contends, shows the TVPA definition of an "extrajudicial killing" (and not just the TVPA in general) draws upon international law. The court's discussion in that case, however, relied not only upon the definition of an "extrajudicial killing" in TVPA § 3(a) but also upon the limitation of the cause of action to state actors in TVPA § 2(a). Id. at 243. Indeed, the court later separately summarized the two provisions of the TVPA, distinguishing § 2(a), which "provides a cause of action" against an individual acting under state authority, from § 3, which "defines the terms 'extrajudicial killing' and 'torture.' " Id. at 245.

Sudan's argument that the definitions in the TVPA incorporate international law is flawed as a matter of statutory interpretation. If the definition of an "extrajudicial killing" (and "torture") in TVPA § 3(a) already had a state actor limitation from international law, then the additional state actor limitation in § 2(a) would be surplusage. See Gustafson v. Alloyd Co. , 513 U.S. 561, 574, 115 S.Ct. 1061, 131 L.Ed.2d 1 (1995) (instructing courts in interpreting a statute to "avoid a reading which renders some words altogether redundant"). That the Congress included § 2(a) in the TVPA therefore implies either that the definition of extrajudicial killing in § 3(a) of the FSIA does not incorporate international law or that international law contains no state actor limitation. Either way, Sudan is out of luck.

In sum, Sudan's textual arguments that an extrajudicial killing requires a state actor all fail. Even if international law contained such a limitation—a proposition we doubt but do not decide—the TVPA does not incorporate international law (or any limitations therein) into its definition of an "extrajudicial killing." Because the FSIA terrorism exception references only the definitions in TVPA § 3, and not the limitation to state actors in TVPA § 2(a), nothing in the text of the FSIA makes a state actor a prerequisite to an extrajudicial killing.

B. Statutory Purpose

Without a viable textual basis for its position, Sudan argues the purpose of the TVPA and the FSIA extend only to an "extrajudicial killing" committed by a state actor. Even if we could ignore the statutory text in pursuit of its supposed purpose, Sudan's arguments from the purpose of the statutes would still not be convincing.

With respect to the purpose of the TVPA, Sudan pursues a line of reasoning parallel to that of its textual arguments: Because the TVPA was intended to "carry out obligations of the United States under the United Nations Charter and other international agreements ... by establishing a civil action for recovery of damages from an individual who engages in torture or extrajudicial killing," Pub. L. No. 102-256, 106 Stat. at 73 (preamble), Sudan contends the supposed state-actor requirement for a killing to violate international law also limits the definition of an "extrajudicial killing" in the TVPA and hence the jurisdictional requirements of the FSIA. Even if international law both motivated enactment of the TVPA and limits extrajudicial killing to a killing by state actor, Sudan's argument about the purpose of the TVPA still would fail. The TVPA may well be intended to carry out certain international obligations, but this purpose is reflected in the TVPA as a whole, not in each individual provision viewed in isolation. One would struggle to find a distinct purpose in the definition section of the TVPA, which neither creates rights nor imposes duties, divorced from the broader statute. When one statute, such as the FSIA, incorporates a definition from another statute, here the TVPA, it imports only the specified definition and not the broader purpose of the statute from which it comes.

In any event, the different purposes of the TVPA and the FSIA are plain on the face of those statutes. The TVPA targets individual state officials for their personal misconduct in office, while the terrorism exception to the FSIA targets sovereign nations in an effort to deter them from engaging, either directly or indirectly, in terrorist acts.

Sudan's own arguments tacitly admit the FSIA serves a different purpose than the TVPA, but it again frames this purpose in terms of international law. To Sudan, the FSIA serves to withdraw sovereign immunity only for "certain universally defined and condemned acts" that are "firmly grounded in international law." Once again Sudan contends, this excludes killings committed by nonstate terrorists because international law proscribes killings only when committed by a state actor. Furthermore, § 1605A, Sudan contends, should be read to exclude acts of terrorism because terrorism lacks "universal condemnation, or even [an] accepted definition ... under international law." Other predicate acts included in § 1605A, particularly aircraft sabotage and hostage taking, are inconsistent with this reading of the FSIA. As the plaintiffs and the district court recognized, "[f]or the past fifteen years it has been hard to think of a more quintessential act of terrorism than the purposeful destruction of a passenger aircraft in flight—yet such an act is manifestly covered by § 1605A." Owens V , 174 F.Supp.3d at 264. Indeed, both aircraft sabotage and hostage taking are more often committed by a nonstate terrorist than by a state actor, and both often result in extrajudicial killings. Moreover, the definitions of these acts in the FSIA clearly do not require state action. 28 U.S.C. §§ 1605A(h)(1) (referencing the Convention for the Suppression of Unlawful Acts Against the Safety of Civil Aviation, art. 1, Sept. 23, 1971, 24 U.S.T. 564, 974 U.N.T.S. 177 (proscribing aircraft sabotage committed by "[a]ny person")); 1605A(h)(2) (referencing the International Convention Against the Taking of Hostages, art. 1, Dec. 17, 1979, 1316 U.N.T.S. 205 (proscribing hostage taking by "[a]ny person")). It would be more than odd if a provision designed to sanction acts "firmly grounded in international law"—but not international terrorism—included only acts synonymous with international terrorism while excluding other violations of international law, such as genocide, not closely associated with terrorist groups. Against this backdrop, it also strains belief that the Congress would assert jurisdiction over claims against a state that materially supports nonstate terrorists who kill via aircraft sabotage or hostage taking, yet deny jurisdiction for similarly supported killings caused by a truck bombing or a kidnapping. It is far more likely the Congress intended to penalize a state's provision of material support for terrorist killings in general, rather than to codify broad principles of international law or to regulate the specific way state-supported terrorists go about their horrific deeds. Were the law otherwise, designated state sponsors of terrorism could effectively contract out certain terrorist acts and avoid liability under the FSIA.

As the district court correctly recognized, § 1605A strives to hold designated state sponsors of terrorism accountable for their sponsorship of terror, regardless whether they commit atrocities themselves or aid others in doing so. Owens V , 174 F.Supp.3d at 262. Therefore, the purpose of the statute clearly embraces liability for the embassy bombings here in question.

C. Statutory History

Sudan next resorts to the legislative history of the FSIA and the TVPA to explain why an "extrajudicial killing" requires state involvement. The short answer to its long and winding argument through the characteristically inconclusive background materials is that when the meaning of a statute is clear enough on its face, "reliance on legislative history is unnecessary." See Mohamad v. Palestinian Auth. , 566 U.S. 449, 132 S.Ct. 1702, 1709, 182 L.Ed.2d 720 (2012) (citation omitted).

Subsequent legislation, on the other hand, because it is enacted and not just compiled, may inform our understanding of a prior enactment with which it should be read in harmony. In this instance, the Congress made clear that an extrajudicial killing includes a terrorist bombing when, in 1996, it enacted the Flatow Amendment to the FSIA to provide a federal cause of action against state officials who had committed or materially supported one of the predicate acts listed in § 1605(a)(7), including an extrajudicial killing. See Pub. L. No. 104-208, § 589, 110 Stat. at 3009-172. The Flatow Amendment responded to a suicide bombing in Israel, carried out by a nonstate terrorist group supported by Iran; it aimed to deter terrorism by making officials of states that sponsor terrorism liable for punitive damages. We do not believe the Congress would provide a cause of action aimed at killings over which it had not authorized jurisdiction.

Subsequent events in the Flatow saga reinforce this conclusion. Immediately following passage, relatives of the victim sued Iran under the Amendment, and the district court asserted jurisdiction based upon this "extrajudicial killing." Flatow , 999 F.Supp. at 18. The plaintiffs won a default judgment but could not collect due to Iran's lack of attachable assets. In 2000 the Congress again responded, passing a compensation scheme to pay individuals who "held a final judgment for a claim or claims brought under section 1605(a)(7) of title 28," including the Flatows. See Victims of Trafficking and Violence Protection Act of 2000, Pub. L. No. 106-386, § 2002(a)(2)(A), 114 Stat. 1464, 1541-43 (authorizing payment to claimants with judgments against Iran, which included the Flatows); H.R. REP. NO. 106-939, at 116 (2000). This legislation too would make little sense if the judgments themselves were void because no extrajudicial killing had occurred.

Finally, after courts had applied the FSIA terrorism exception to terrorist bombings for over a decade, the Congress reenacted the same predicate acts in § 1605(a)(7) when authorizing the new FSIA exception under § 1605A. The Congress thereby ratified the Flatow court's understanding—and those of every other court since then—that a nonstate actor may commit an extrajudicial killing. See Lorillard v. Pons , 434 U.S. 575, 580, 98 S.Ct. 866, 55 L.Ed.2d 40 (1978) ("Congress is presumed to be aware of an administrative or judicial interpretation of a statute and to adopt that interpretation when it re-enacts a statute without change"). Now, after more than two decades of consistent judicial application of the FSIA, narrowing the term "extrajudicial killing" to include only killings committed by a state actor would contravene the Congress's revealed intent in repeatedly authorizing judicial remedies for victims of terrorist bombings.

To summarize, the plain meaning of § 1605A(a) grants the courts jurisdiction over claims against designated state sponsors of terrorism that materially support extrajudicial killings committed by nonstate actors. Contrary to Sudan's contention, the purpose and statutory history of the FSIA terrorism exception confirm this conclusion. Therefore, this court may assert jurisdiction over claims arising from al Qaeda's bombing of the U.S. embassies in 1998 if the plaintiffs have adequately demonstrated Sudan's material support for those bombings.

III. Sufficiency of the Evidence Supporting Jurisdiction

Sudan's weightiest challenge to jurisdiction relates to the admissibility and sufficiency of the evidence that supported the district court's finding of jurisdiction. As discussed above, § 1605A(a)(1) of the FSIA grants jurisdiction and withdraws immunity for claims "caused by an act of ... extrajudicial killing ... or the provision of material support or resources for such an act."

In order to establish the court's jurisdiction, the plaintiffs in this case must show (1) Sudan provided material support to al Qaeda and (2) its material support was a legally sufficient cause of the embassy bombings. See Kilburn v. Socialist People's Libyan Arab Jamahiriya , 376 F.3d 1123, 1127 (D.C. Cir. 2004) (treating causation as a jurisdictional requirement). Sudan challenges the district court's factual findings on both accounts. Because the elements of material support and causation are jurisdictional, Sudan may contest them on appeal even though it forfeited its right to contest the merits of the plaintiffs' claims. See Practical Concepts , 811 F.2d at 1547. This does not mean, however, that the plaintiffs on appeal must offer the same quantum of evidence needed to show liability in the first instance. Establishing material support and causation for jurisdictional purposes is a lighter burden than proving a winning case on the merits. See Agudas Chasidei Chabad of U.S. v. Russian Federation , 528 F.3d 934, 940 (D.C. Cir. 2008).

In its opinion rejecting Sudan's motion to vacate the default judgments, the district court identified two bases upon which the plaintiffs established material support and causation for the purpose of jurisdiction. For plaintiffs proceeding under the federal cause of action in § 1605A(c), the court—following then-binding Circuit precedent—held the plaintiffs had established jurisdiction by making a "non-frivolous" claim that Sudan materially supported al Qaeda and that such support proximately caused their injuries. Owens V , 174 F.Supp.3d at 272-75. Since that decision, the Supreme Court has overruled the precedent upon which the district court relied, requiring a plaintiff to prove the facts supporting the court's jurisdiction under the FSIA, rather than simply to make a "non-frivolous" claim to that effect. Bolivarian Republic of Venezuela v. Helmerich & Payne Int'l Drilling Co. , ––– U.S. ––––, 137 S.Ct. 1312, 1316, 197 L.Ed.2d 663 (2017). The Court's decision eliminates the first basis for the district court's jurisdictional holding.

The decision in Helmerich , however, left intact the district court's second basis for concluding the plaintiffs had sufficiently shown material support and causation in this case. For reasons no longer relevant, the district court concluded that plaintiffs who are ineligible to use the federal cause of action in § 1605A(c) —namely, victims or claimants who were not U.S. nationals, military service members, or government employees or contractors—could not establish jurisdiction simply by making a non-frivolous claim of material support and causation. Owens V , 174 F.Supp.3d at 275. Consequently, the court required those plaintiffs to offer evidence proving these jurisdictional elements. Id. First in its 2011 opinion on liability and again in its 2016 opinion denying vacatur, the district court weighed the plaintiffs' evidence of material support and causation and concluded it satisfied the jurisdictional standard. Owens V , 174 F.Supp.3d at 276 ; Owens IV , 826 F.Supp.2d at 150-51. Because the court's finding of Sudan's material support for the 1998 embassy bombings plainly applies to all claimants and all claims before this court, Sudan can prevail in its challenge to material support and causation only if the district court erred in its factual findings of jurisdiction. We conclude it did not.

In each of the cases, the plaintiffs' evidence was received at the three-day evidentiary hearing held by the district court in October 2010. The court held that hearing to satisfy the FSIA requirement that, in order to secure a default judgment, a claimant must "establish[ ] his claim or right to relief by evidence satisfactory to the court." 28 U.S.C. § 1608(e). At the hearing, the court received evidence of both Iran's and Sudan's support for al Qaeda in advance of the embassy bombings, but we limit our discussion here to the evidence pertaining to Sudan.

In evaluating Sudan's evidentiary arguments, we proceed in three steps. First, we summarize the proceedings at the 2010 evidentiary hearing and the facts presented by the plaintiffs and their expert witnesses. Then we consider Sudan's two challenges to this evidence. In the first, Sudan argues the district court relied upon inadmissible evidence to conclude that it materially supported al Qaeda. In the second, Sudan contends that, even if admissible, the evidence presented could not establish material support and causation as a matter of law.

A. The Evidentiary Hearing

At the October 2010 evidentiary hearing the plaintiffs presented evidence from a variety of sources. Reviewing this evidence as a whole, the district court concluded it sufficed both to establish jurisdiction and to prove Sudan's liability on the merits. We first describe the sources of evidence the court received and then briefly summarize the factual findings the court drew from this evidence.

1. The sources of evidence presented

As is apparent from the opinions of the district court, the testimony of expert witnesses and al Qaeda operatives was of critical importance to its factual findings. For this reason, we discuss the experts' and operatives' testimony first and in greatest detail. The plaintiffs produced three expert witnesses and prior recorded testimony from three former members of al Qaeda.

First, the plaintiffs called terrorism consultant Evan Kohlmann to testify about the relationship between Sudan and al Qaeda in the 1990s. Kohlmann advised government and private clients on terrorist financing, recruitment, and history. He has authored a book and several articles on terrorism and has testified as an expert in multiple criminal trials. Kohlmann based his opinions regarding Sudan's support for al Qaeda upon a review of secondary source materials, including but not limited to the exhibits introduced at the hearing, testimony from criminal trials, and firsthand interviews he conducted with al Qaeda affiliates over the past decade. Kohlmann testified that this information was of the type routinely relied upon by experts in the counterterrorism field.

Next, the court received a written expert report from Dr. Lorenzo Vidino on "Sudan's State Sponsorship of al Qaeda." Dr. Vidino was a fellow at the Belfer Center for Science and International Affairs, Kennedy School of Government, at Harvard University. Like Kohlmann, Vidino has authored books and articles on terrorism and has previously testified in federal court on Sudan's support for al Qaeda. Vidino based his report upon open source materials initially gathered around 2004, which he reviewed and updated for the present case.

The district court also received live testimony and a written report from Steven Simon, a security consultant and Special Advisor for Combatting Terrorism at the Department of State. From 1995 to 1999, during which time al Qaeda bombed the embassies, Simon served on the National Security Council (NSC) as Senior Director for Transnational Threats. His responsibilities at the NSC included directing counterterrorism policy and operations on behalf of the White House. After his government service, Simon published a book and several articles on international terrorism and taught graduate courses on counterterrorism.

The court also heard recorded trial testimony from three former al Qaeda operatives. In particular, the plaintiffs' star witness, Jamal al Fadl, cast a long shadow over the proceedings. al Fadl was a Sudanese national and former senior al Qaeda operative turned FBI informant. Now in the witness protection program, in 2001 he testified at the criminal trial of Usama bin Laden and other terrorists arising from the African embassy bombings. Al Fadl was particularly well-suited to address the relationship between al Qaeda and the government of Sudan in the 1990s because he served then as a principal liaison between the terrorist group and Sudanese intelligence. He had also been instrumental in facilitating al Qaeda's relocation from Afghanistan to Sudan in 1991 and had assisted the group in acquiring properties there. Although al Fadl did not testify at the evidentiary hearing, his prior testimony provided much of the factual basis for the expert witnesses' opinions.

The court also received transcripts of prior testimony from two other al Qaeda operatives: Essam al Ridi and L'Houssaine Kherchtou. Both al Ridi and Kherchtou were members of al Qaeda when the terrorist group was based in Sudan, and both testified at the bin Laden trial. They testified, based upon firsthand knowledge, about the Sudanese government and military facilitating al Qaeda's movement throughout East Africa and protecting al Qaeda leadership. The plaintiffs also submitted a deposition from al Ridi prepared for the instant case.

In addition to this witness testimony, the court viewed videos produced by al Qaeda describing its move to Sudan and its terrorist activities thereafter. And finally, the court considered reports from the U.S. Department of State and the Central Intelligence Agency describing Sudan's relationship with al Qaeda in the 1990s.

2. The district court's findings of fact

From the plaintiffs' evidence, the district court found that Sudan had provided material support to al Qaeda and that such support caused the embassy bombings. This support was provided in several ways, which we recount in a much abbreviated form.

First, the district court found Sudan provided al Qaeda a safe harbor from which it could direct its operations. Owens IV , 826 F.Supp.2d at 139-43. This began with the overthrow of the Sudanese government in 1989 by Omar al Bashir, leader of the Sudanese military, and Hassan al Turabi, head of the National Islamic Front (NIF), Sudan's most powerful political party. Kohlmann and Simon testified that al Turabi initiated contact with al Qaeda and other extremist groups, encouraging them to relocate to Sudan. Al Bashir formalized this initial outreach with a 1991 letter of invitation to Usama bin Laden. According to all three experts, Sudan's outreach to al Qaeda was part of a broader strategy of inviting radical Islamist groups to establish bases of operations in the country, which is confirmed by the State Department Patterns of Global Terrorism reports. See U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1991, at 3 (1991) ("The government reportedly has allowed terrorist groups to train on its territory and has offered Sudan as a sanctuary to terrorist organizations"). Sudan's extensive ties to terrorist groups prompted the Department of State to designate Sudan as a state sponsor of terrorism in August 1993. U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1993, at 25 (1994).

In 1991 al Qaeda accepted Sudan's invitation. According to Kohlmann and Simon, the invitation benefited both bin Laden and the Sudanese government. For bin Laden, it allowed al Qaeda to depart an increasingly unstable Afghanistan and relocate closer to its strategic interests in the Middle East. For Sudan, outreach to terrorist groups provided leverage against the government's enemies at home and abroad and advanced al Turabi's ideological ambition for Sudan to become "the new haven for Islamic revolutionary thought." Sudan also viewed al Qaeda as a source of domestic investment as bin Laden was rumored to be extremely wealthy and was well-known as a financier of the mujahedeen insurgency in Afghanistan.

Once bin Laden had determined Sudan was a trustworthy partner, al Qaeda moved its operations there. All three experts described al Qaeda purchasing several properties in Sudan, including a central office and a guesthouse in Khartoum, and starting terrorist training camps on farms throughout the country. Al Fadl personally participated in some of these transactions. For a time, according to Kohlmann, al Qaeda even shared offices with the al Turabi's NIF party in Khartoum. The close relationship between al Qaeda and the Sudanese government continued throughout the early 1990s, according to Kohlmann and Vidino, even after bin Laden publicized his intent to attack American interests in a series of fatwas and after al Qaeda members claimed responsibility for the killing of U.S. soldiers in Mogadishu, Somalia. For example, bin Laden appeared in multiple television broadcasts with al Bashir and al Turabi celebrating the completion of infrastructure projects financed, in part, by bin Laden. Sudanese intelligence officials also worked hand-in-glove with al Qaeda operatives to screen purported al Qaeda volunteers entering the country in order "to ensure that they were not seeking to infiltrate bin Laden's organization on behalf of a foreign intelligence service." Al Fadl personally took part in these efforts.

Sudan also helped al Qaeda develop contacts with other terrorist organizations. In 1991 the NIF organized an unprecedented gathering of terrorist organizations from around the world in Khartoum at the Popular Arab and Islamic Congress. Several of these groups, including the Egyptian Islamic Jihad (EIJ), whose membership would later overlap with that of al Qaeda, and the Iranian-backed Hezbollah, which later provided training to al Qaeda operatives, also established bases in Sudan. According to Kohlmann and Simon, Sudanese intelligence actively assisted al Qaeda in forming contacts with these groups, allowing the nascent organization to acquire skills and to recruit members from the more experienced groups that it would later use with devastating effect.

Although Sudan expelled bin Laden in 1996 under international pressure, Kohlmann, Vidino, and one other expert testified that some al Qaeda operatives remained in the country thereafter. They based this conclusion, in part, upon an unclassified report of the CIA, dated December 1998. A State Department report from 1998, published after the embassy bombings, reinforced the conclusion that "Sudan continued to serve as a meeting place, safe haven, and training hub for a number of international terrorist groups, particularly Usama Bin Ladin's al-Qaida organization." U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1998 (1999). Although expelling bin Laden was a "positive step [ ]," the CIA concluded Sudan continued to send "mixed signals about cutting its terrorist ties" after his expulsion but before the embassy bombings. CENT. INTEL. AGENCY, SUDAN: A PRIMER ON BILATERAL ISSUES WITH THE UNITED STATES , at 4 (May 12, 1997). Notably, Sudan remains a designated state sponsor of terrorism today.

The district court also found Sudan had provided financial, governmental, military, and intelligence support to al Qaeda. Owens IV , 826 F.Supp.2d at 143-46. During its time in Sudan, al Qaeda operated several business and charities. All three experts explained that these enterprises provided legitimate employment for al Qaeda operatives as well as cover for the group's illicit activities throughout the region. The Sudanese government actively promoted al Qaeda's businesses in several ways. As described by al Fadl, Sudan partnered with al Qaeda-affiliated businesses in major infrastructure projects, allowing al Qaeda to gain access to and experience with explosives. Sudan also granted al Qaeda businesses "customs exemptions" and "tax privileges" which, according to Vidino, enabled al Qaeda nearly to monopolize the export of several agricultural products. Sudan offered al Qaeda the services of its banking system, which helped the organization in "laundering money and facilitating other financial transactions that stabilized and ultimately enlarged Bin Laden's presence in the Sudan."

From the very beginning Sudan also aided al Qaeda's movement throughout the region. Relying upon al Fadl's testimony, Kohlmann testified that al Qaeda circulated copies of President al Bashir's letter of invitation among its operatives. Al Qaeda agents could present these copies to Sudanese officials in order to "avoid having to go through normal immigration and customs controls" and to head off any "problems with the local police or authorities." According to Kohlmann, Sudanese intelligence also transported weapons and equipment for al Qaeda from Afghanistan to Sudan via the state-owned Sudan Airways. On at least one occasion, Sudan allowed al Qaeda operative Kherchtou to smuggle $10,000 in currency—an amount above that permitted by law—to an al Qaeda cell in Kenya. This Kenyan cell ultimately carried out the bombing of the U.S. embassy in Nairobi in 1998.

In addition to aiding al Qaeda's movements directly, all three experts testified that the government provided al Qaeda members hundreds of passports and Sudanese citizenship. Al Qaeda operatives needed these passports because they were "de facto stateless individuals" who could no longer safely travel on passports from their countries of origin. Upon returning from abroad, Sudanese officials allowed al Qaeda operatives to bypass customs and immigration controls. As al Fadl testified, this allowed militants to avoid having their passport stamped by a nation that had come under increasing scrutiny for its ties to terrorist organizations.

Finally, the district court identified several instances in which Sudan provided security to al Qaeda leadership. Owens IV , 826 F.Supp.2d at 145. In his prior testimony, al Fadl recounted an occasion when Sudanese intelligence intervened to prevent the arrest of al Qaeda operatives by local police. Al Ridi also testified that Sudan assigned 15 to 20 uniformed soldiers to act as personal bodyguards for bin Laden and other al Qaeda members. In 1994, according to Kohlmann, Sudanese intelligence even foiled an assassination attempt against bin Laden in Khartoum. On another occasion, Sudanese intelligence thwarted a plot against al Qaeda's second-in-command, Ayman al-Zawahiri. Even as international pressure mounted on Sudan to expel bin Laden, Simon—who covered terrorism matters for the NSC during the events in question—explained that the Sudanese government refused to provide actionable intelligence on al Qaeda's plans throughout the region or to hand bin Laden over to the United States. Simon echoed the State Department's conclusion that bin Laden's eventual expulsion was nothing more than a "symbolic gesture designed to placate the international community" that changed little in the day-to-day reality of Sudan's support for terrorism. See U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1998.From this evidence, all three experts concluded Sudan provided material support to al Qaeda. Moreover, the experts viewed this support as "indispensable" to the success of the 1998 embassy bombings. Without "a country that not only tolerated, but actually actively assisted ... al Qaeda terrorist activities," Vidino asserted, "al Qaeda could not have achieved its attacks on the US Embassies." Noting that "the vast majority of planning and preparation [for the attacks] took place between the years of 1991 and 1997," Kohlmann opined "without the base that Sudan provided, without the capabilities provided by the Sudanese intelligence service, without the resources provided, none of this would have happened." Simon likewise surmised "it's difficult to see how ... the attacks could have been carried out with equal success" without Sudan's "active support" and safe haven.

From the expert testimony, trial transcripts, and government reports, the district court concluded that the plaintiffs had met their burden of demonstrating "to the satisfaction of the court" that Sudan had provided material support to al Qaeda and that such support was a legally sufficient cause of the embassy bombings. Owens IV , 826 F.Supp.2d. at 150. As such, the plaintiffs both established jurisdiction and prevailed on the merits of liability. When faced with Sudan's Rule 60(b)(4) motion to vacate the default judgments as void, the district court reaffirmed that its findings of material support and causation satisfied the standard for jurisdiction under § 1605A(a). Owens V , 174 F.Supp.3d at 276.

On this appeal, Sudan contends the record contains insufficient evidence of material support and causation to give the court jurisdiction under the FSIA. Its attack comes in two forms. First, Sudan disputes the admissibility of much of the evidence introduced to support the district court's factual findings. It does so despite having failed to participate in the evidentiary hearing, where such challenges would have been properly raised. Second, even assuming the evidence was admissible, Sudan contends the district court's factual findings on material support and causation were clearly erroneous and insufficient to sustain jurisdiction as a matter of law. As we shall see, neither argument has merit.

B. Standard of Review

Sudan faces an uphill battle with its evidentiary challenges for two reasons. First is the burden of proof applicable to a FSIA case. The FSIA "begins with a presumption of immunity" for a foreign sovereign. Bell Helicopter Textron, Inc. v. Islamic Republic of Iran , 734 F.3d 1175, 1183 (D.C. Cir. 2013). The plaintiff bears an initial burden of production to show an exception to immunity, such as § 1605A, applies. Id. Then, "the sovereign bears the ultimate burden of persuasion to show the exception does not apply," id. , by a preponderance of the evidence. See Simon v. Republic of Hungary , 812 F.3d 127, 147 (D.C. Cir. 2016). Therefore, if a plaintiff satisfies his burden of production and the defendant fails to present any evidence in rebuttal, then jurisdiction attaches.

Although a court gains jurisdiction over a claim against a defaulting defendant when a plaintiff meets his burden of production, the plaintiff must still prove his case on the merits. This later step, however, does not affect the court's jurisdiction over the case, and a defaulting defendant normally forfeits its right to raise nonjurisdictional objections. See Practical Concepts , 811 F.2d at 1547. Thus, the only question before this court is whether the plaintiffs have met their rather modest burden of production to establish the court's jurisdiction.This brings us to Sudan's second obstacle on appeal. When assessing whether a plaintiff has met his burden of production, appellate review of the district court's findings of fact and evidentiary rulings is narrowly circumscribed. With respect to a defaulting sovereign, the FSIA requires only that a plaintiff "establish[ ] his claim or right to relief by evidence satisfactory to the court." 28 U.S.C. § 1608(e). This standard mirrors a provision in Federal Rule of Civil Procedure 55(d) governing default judgments against the U.S. Government. Commercial Bank of Kuwait v. Rafidain Bank , 15 F.3d 238, 242 (2d Cir. 1994). While both § 1608(e) and Rule 55(d) give an unresponsive sovereign some protection against an unfounded default judgment, see Jerez , 775 F.3d at 423, neither provision "relieves the sovereign from the duty to defend cases," Rafidain Bank , 15 F.3d at 242. Moreover, § 1608(e) does not "require the court to demand more or different evidence than it would ordinarily receive," cf. Marziliano v. Heckler , 728 F.2d 151, 158 (2d Cir. 1984) (applying Rule 55(d) ); indeed, "the quantum and quality of evidence that might satisfy a court can be less than that normally required." Alameda v. Sec'y of Health, Ed. & Welfare , 622 F.2d 1044, 1048 (1st Cir. 1980) (applying Rule 55(d) ).

Unlike the court's conclusions of law, which we review de novo, we review for abuse of discretion the district court's satisfaction with the evidence presented. Hill v. Republic of Iraq , 328 F.3d 680, 683 (D.C. Cir. 2003). A district court abuses its discretion when it relies upon a clearly erroneous finding of fact. Amador County v. U.S. Dep't of the Interior , 772 F.3d 901, 903 (D.C. Cir. 2014). In a FSIA default proceeding, a factual finding is not deemed clearly erroneous if "there is an adequate basis in the record for inferring that the district court ... was satisfied with the evidence submitted." Rafidain Bank , 15 F.3d at 242 (quoting Marziliano , 728 F.2d at 158 ). That inference is drawn when the plaintiff shows "her claim has some factual basis," cf. Giampaoli v. Califano , 628 F.2d 1190, 1194 (9th Cir. 1980) (applying Rule 55(d) ), even if she might not have prevailed in a contested proceeding. Provided "the claimant's district court brief and reference to the record appear[ ] relevant, fair and reasonably comprehensive," we will not set aside a default judgment for insufficient evidence. Alameda , 622 F.2d at 1049. This lenient standard is particularly appropriate for a FSIA terrorism case, for which firsthand evidence and eyewitness testimony is difficult or impossible to obtain from an absent and likely hostile sovereign.

The district court also has an unusual degree of discretion over evidentiary rulings in a FSIA case against a defaulting state sponsor of terrorism. For example, we have allowed plaintiffs to prove their claims using evidence that might not be admissible in a trial. See Han Kim v. Democratic People's Republic of Korea , 774 F.3d 1044, 1048-51 (D.C. Cir. 2014) (noting "courts have the authority—indeed, we think, the obligation—to adjust evidentiary requirements to differing situations" and admitting affidavits in a FSIA default proceeding) (internal alterations and quotation marks removed). This broad discretion extends to the admission of expert testimony, which, even in the ordinary case, "does not constitute an abuse of discretion merely because the factual bases for an expert's opinion are weak." Joy v. Bell Helicopter Textron, Inc. , 999 F.2d 549, 567 (D.C. Cir. 1993). Section 1608(e) does not require a court to step into the shoes of the defaulting party and pursue every possible evidentiary challenge; only where the court relies upon evidence that is both clearly inadmissible and essential to the outcome has it abused its discretion. This is part of the risk a sovereign runs when it does not appear and alert the court to evidentiary problems. Cf. Bell Helicopter Textron , 734 F.3d at 1181.

In this case, the district court has already undertaken to weigh the plaintiffs' evidence and determine its admissibility without any assistance from Sudan. Under these circumstances, we accord even more deference to the district court's factual findings and evidentiary rulings in a FSIA case than in reviewing default judgments to which the strictures of § 1608(e) (or Rule 55(d) ) do not apply.

Deference is especially appropriate when considering the lengthy history of the proceedings in the district court. The same learned judge has presided over this litigation since 2001. Over that time, the court has gained considerable familiarity with the plaintiffs' evidence and, during the periods when Sudan participated, with its objections to that evidence. The court has issued four lengthy and detailed opinions that directly address many of Sudan's challenges to the evidence of material support and jurisdictional causation. Through its opinions and actions, it is abundantly clear that the district court both appreciated and carried out is obligation under § 1608(e). Cf. Compania Interamericana Exp.-Imp., S.A. v. Compania Dominicana de Aviacion , 88 F.3d 948, 951 (11th Cir. 1996) (vacating default judgment when "the record does not reflect that the court considered the differing standard required by § 1608(e)"). Only if we found the record wholly lacking an "adequate basis" for the district court's conclusions would we overturn its jurisdictional findings.

C. Admissibility of the Evidence

Sudan first challenges the admissibility of evidence supporting the district court's findings of material support and jurisdictional causation. In order to issue a default judgment under § 1608(e), a court must base its findings of fact and conclusions of law upon evidence admissible under the Federal Rules of Evidence. Kim , 774 F.3d at 1049. If inadmissible evidence alone substantiates an essential element of jurisdiction, then the court abuses its discretion in concluding the claimant has established his case "by evidence satisfactory to the court." 28 U.S.C. § 1608(e).

Reviewing the admissibility of evidence supporting a default judgment presents significant challenges, which color our treatment of Sudan's arguments. The adversarial process gives the parties an incentive to raise evidentiary challenges at the earliest opportunity because failure to do so ordinarily results in their forfeiture. Raising evidentiary challenges early on also provides the proponent of the evidence the opportunity to respond by offering an alternative theory of admissibility or different, admissible evidence on the same point. Thus, the adversarial process properly places the burden of admissibility upon the interested party, allocates the original determination of admissibility to the district court, which is more familiar with the evidence, and preserves evidentiary disputes for appellate review with the aid of a full trial record. Furthermore, allowing a defaulting defendant to benefit from sandbagging the plaintiff with an admissibility objection on appeal would be unfair and would encourage gamesmanship. When the defendant defaults, therefore, we do not consider its evidentiary challenges on appeal.

These principles do not map neatly to a FSIA case because a defaulting defendant may challenge the factual basis for the court's jurisdiction for the first time on appeal. And because a FSIA plaintiff must produce evidence that is both admissible, Kim , 774 F.3d at 1049, and "satisfactory to the court," 28 U.S.C. § 1608(e), in order to obtain a default judgment, we presume a defendant may also challenge for the first time on appeal the admissibility of evidence supporting a jurisdictional fact. As previously noted, however, a defendant sovereign that defers its challenge until appealing a default judgment makes the district court's decision less fully informed and deprives the reviewing court of a fully developed record; it also handicaps the non-defaulting plaintiff in filling out the evidentiary record. For these reasons, we will not accept a belated challenge to admissibility raised by a defaulting sovereign unless the contested evidence is clearly inadmissible and we seriously doubt the plaintiff could have provided alternative evidence that would have been admissible. Those circumstances are not present here.

In this case, Sudan principally challenges the admissibility of two types of evidence: (1) the plaintiffs' expert testimony and (2) reports from the Department of State and the CIA. We find no error in the district court's reliance upon either.

1. The expert testimony

In its opinions on liability and on Sudan's Rule 60(b) motion, the district court discussed the experts' testimony in great detail and concluded it sufficed to establish jurisdiction. Owens V , 174 F.Supp.3d at 276. Because it may be dispositive, we, too, start with the expert testimony.

The testimony of expert witnesses is of crucial importance in terrorism cases, see, e.g. , Kilburn , 376 F.3d at 1132 (jurisdiction satisfied based solely upon the declaration of an expert witness); Boim v. Holy Land Found. for Relief & Dev. , 549 F.3d 685, 704 (7th Cir. 2008) ; United States v. Damrah , 412 F.3d 618, 625 (6th Cir. 2005), because firsthand evidence of terrorist activities is difficult, if not impossible, to obtain. Victims of terrorist attacks, if not dead, are often incapacitated and unable to testify about their experiences. Perpetrators of terrorism typically lie beyond the reach of the courts and go to great lengths to avoid detection. Eyewitnesses in a state that sponsors terrorism are similarly difficult to locate and may be unwilling to testify for fear of retaliation. The sovereigns themselves often fail to appear and to participate in discovery, as Sudan did here. With a dearth of firsthand evidence, reliance upon secondary materials and the opinions of experts is often critical in order to establish the factual basis of a claim under the FSIA terrorism exception.

Sudan raises three challenges to the expert testimony presented at the evidentiary hearing. First, despite conceding that expert testimony is "doubtless admissible" in a FSIA default proceeding, Sudan contends that experts alone are insufficient to establish jurisdiction in the absence of other direct, admissible evidence. Second, Sudan objects that the plaintiffs' experts merely served as conduits for inadmissible hearsay, upon which the district court relied. Finally, Sudan quarrels with the inferences drawn by the experts and by the district court from the underlying factual background. None of these arguments is persuasive.

a. Need for direct evidence

The recent case of Han Kim v. Democratic People's Republic of Korea demonstrates the importance of expert testimony in FSIA proceedings and forecloses Sudan's first argument. In Kim , relatives of a pastor who was a U.S. citizen sued the Democratic People's Republic of Korea (DPRK) under the FSIA terrorism exception, alleging the regime abducted, tortured, and killed the cleric for his ministry to DPRK refugees. 774 F.3d at 1046. Because the DPRK refused to participate in the litigation and intimidated potential eyewitnesses, the plaintiffs could offer no direct evidence of their relative's torture and killing by the DPRK. Instead, two experts submitted declarations stating that North Korea invariably tortured and killed its political prisoners. Id . The court in Kim found these declarations "doubtless admissible" under Federal Rule of Evidence 702 and refused categorically to require eyewitness testimony or direct evidence on both practical and policy grounds:

In these circumstances, requiring that the Kims prove exactly what happened to the Reverend and when would defeat the Act's very purpose: to give American citizens an important economic and financial weapon to compensate the victims of terrorism, and in so doing to punish foreign states who [sic] have committed or sponsored such acts and deter them from doing so in the future. This is especially true in cases of forced disappearance, like this one, where direct evidence of subsequent torture and execution will, by definition, almost always be unavailable, even though indirect evidence may be overwhelming. Were we to demand more of plaintiffs like the Kims, few suits like this could ever proceed, and state sponsors of terrorism could effectively immunize themselves by killing their victims, intimidating witnesses, and refusing to appear in court.

Id. at 1048-49 (internal citations and quotation marks omitted).

Here, as in Kim , the plaintiffs face a state sponsor of terrorism that has refused to participate in the litigation. By skipping discovery and the evidentiary hearing, Sudan made it virtually impossible for the plaintiffs to get eyewitness accounts of its activities in the 1990s. Nor can the plaintiffs ordinarily subpoena members of al Qaeda, many of whom are dead or in hiding, to testify regarding the actions of the regime. The Congress originally enacted the terrorism exception in the FSIA because state sponsors of terrorism "ha[d] become better at hiding their material support" and misdeeds. Kilburn , 376 F.3d at 1129 (internal quotation marks omitted). Just as requiring firsthand evidence of the DPRK's covert atrocities in Kim would "effectively immunize" the regime from responsibility for its crimes, requiring that a victim of a state-supported bombing offer direct evidence of material support would shield state sponsors of terrorism from liability for the very predicate act—material support—that gives the court jurisdiction.

Nevertheless, Sudan persists that expert testimony alone cannot establish jurisdiction and liability under the FSIA. To wit, Sudan complains that the plaintiffs did not offer "any admissible factual evidence" or "call any percipient witnesses competent to testify about relevant facts in Sudan in the 1990s." In particular, Sudan would have us distinguish Kim as having turned solely upon a piece of non-expert evidence.

Sudan's argument is both legally and factually flawed. Neither § 1608(e) nor any other provision of the FSIA requires a court to base its decision upon a particular type of admissible evidence. As long as the evidence itself is admissible, as expert testimony certainly may be, and the court finds it satisfactory, its form or type is irrelevant. Cf. Holland v. United States , 348 U.S. 121, 140, 75 S.Ct. 127, 99 L.Ed. 150 (1954) (refusing to distinguish between different types of evidence in a criminal prosecution). Indeed, cases in this Circuit and in others have repeatedly sustained jurisdiction or liability or both under the terrorism exception to the FSIA and in other terrorism cases based solely upon expert testimony. Kilburn , 376 F.3d at 1132 ; Boim , 549 F.3d at 705 ("[W]ith [the plaintiff's expert report] in the record and nothing on the other side the [district] court had no choice but to enter summary judgment for the plaintiffs with respect to Hamas's responsibility for the Boim killing"). Therefore the plaintiffs' "failure" to present eyewitness testimony or other direct evidence is of no moment as to whether they have satisfied their burden of production.

Sudan's attempt to distinguish Kim on its facts is similarly unpersuasive. True, in Kim , we placed great weight upon a single piece of admissible non-expert evidence: the conviction of a DPRK agent who had kidnapped the victim, of which the district court took judicial notice. Kim , 774 F.3d at 1049. This conviction placed the victim at the scene of the crime and allowed the court to conclude he had been subjected to the torture and killing that the DPRK "invariably" inflicts upon its prisoners. Id . at 1051. Without this conviction, we noted, "[o]ur conclusion would no doubt differ" because there was no other evidence linking the DPRK to the victim's disappearance. Id.

Our conclusion, however, turned upon the specific facts of that case; we did not announce a categorical requirement of direct evidence in FSIA cases. Whereas the conviction in Kim linked the defendant sovereign to the plaintiff's disappearance, in the present case there is no missing link between Sudan's actions and the embassy bombings. It is undisputed that al Qaeda came to Sudan in the early 1990s and maintained its headquarters there. It is also beyond question that al Qaeda perpetrated the embassy bombings in 1998. As in Kim , expert testimony supplies the predicate act (here material support, in Kim torture and extrajudicial killing) linking these two events and conferring jurisdiction upon the court. But here, unlike in Kim , we need no further evidence beyond the expert testimony to connect the defendant sovereign to the extrajudicial killings. The expert testimony therefore suffices to meet the plaintiffs' burden of production on jurisdiction.

b. Reliance upon inadmissible hearsay

Sudan next contends the experts recited facts based upon inadmissible hearsay and the district court improperly relied upon those facts to establish jurisdiction and to hold Sudan liable.

Under Federal Rule of Evidence 703, a properly qualified expert may base his opinion upon otherwise inadmissible sources of information as long as those sources are reasonably relied upon in his field of expertise. Further, the expert may disclose to the factfinder otherwise inadmissible "underlying facts or data as a preliminary to the giving of an expert opinion." See, e.g. , FED. R. EVID . 705 advisory committee's note. Indeed, disclosure is often necessary to enable the court to "decid[e] whether, and to what extent, the person should be allowed to testify." Id. ; 2 MCCORMICK ON EVIDENCE § 324.3 (7th ed. 2016) ("otherwise the opinion is left unsupported with little way for evaluation of its correctness"). Nevertheless, "the underlying information" relied upon by a qualified expert "is not admissible simply because the [expert's] opinion or inference is admitted."See FED. R. EVID . 703 advisory committee's note. Thus, as Sudan points out, "a party cannot call an expert simply as a conduit for introducing hearsay under the guise that the testifying expert used the hearsay as the basis of his testimony." Marvel Characters, Inc. v. Kirby , 726 F.3d 119, 136 (2d Cir. 2013) (internal quotation marks omitted).

Applying these standards to the case at hand, we see that the district court properly distinguished the experts' clearly admissible opinions from the potentially inadmissible facts underlying their testimony. Sudan principally objects to the district court's recitation of those underlying facts in its 2011 opinion on liability, which facts it claims are inadmissible even if the experts' opinions were properly admitted. The district court acknowledged this complication in its 2016 opinion on Sudan's motion to vacate: "Sudan may have plausible arguments" that not "every factual proposition in the Court's 2011 opinion can be substantiated by record evidence admissible under the Federal Rules of Evidence." Owens V , 174 F.Supp.3d at 275. But even if "particular statements in that opinion may not be adequately supported," the experts' opinions "nonetheless" provided "sufficient evidence in the record of the necessary jurisdictional facts." Id. We agree with this conclusion.

At the outset, we note the district court did not err—much less prejudicially err—in reciting potentially inadmissible facts in its 2011 opinion on liability. For their conclusions to be admissible and credible, the plaintiffs' experts needed to disclose the factual basis for their opinions. See, e.g. , Fox v. Taylor Diving & Salvage Co. , 694 F.2d 1349, 1356 (5th Cir. 1983) ("An expert is permitted to disclose hearsay for the limited purpose of explaining the basis for his expert opinion"). Without that disclosure, the district court would have been at a loss to determine whether the opinions were admissible as reliable expert testimony. See FED. R. EVID . 702 (requiring court to determine whether expert's knowledge "is based on sufficient facts or data," and is "the product of reliable principles and methods" that have been "reliably applied ... to the facts of the case"). Therefore, the court did not err in allowing the plaintiffs' experts to recount potentially inadmissible facts in order to establish the basis for their admissible opinions.

The district court also needed to engage with the underlying facts in order to explain why it admitted and credited the experts' opinions. Without those facts, we too would struggle to evaluate Sudan's evidentiary challenges to the opinion testimony. Hence, some discussion of the potentially inadmissible underlying facts was unavoidable in the 2011 opinion in order to admit, to credit, and to enable our review of the experts' opinions.

More important, the district court properly based its findings upon the experts' "undoubtedly admissible" opinions and not upon any arguably inadmissible facts. The district court's 2011 and 2016 opinions extensively quote the experts' opinions in reaching the conclusion that Sudan's material support caused the embassy bombings. See Owens V , 174 F.Supp.3d at 277-79 (quoting the opinions of Kohlmann, Simon, and Vidino); Owens IV , 826 F.Supp.2d at 146 (quoting Simon and Kohlmann to conclude "Sudanese government support was critical to the success of the 1998 embassy bombings"). We therefore see no error in the court's conclusion that the expert testimony satisfied the plaintiffs' burden of production on jurisdictional causation.

In a supplemental filing, Sudan compares the experts' opinions in this case to those held inadmissible in Gilmore v. Palestinian Interim Self-Government Authority , 843 F.3d 958 (D.C. Cir. 2016), but the gulf between the two cases is wide. In Gilmore , the plaintiff's expert neither stated nor applied "a reliable methodology" from which he had derived his opinions. Id. at 972-73. Instead, "his analysis consist[ed] entirely of deductions and observations that flow directly from the content of the hearsay statements and would be self-evident to a layperson." Gilmore v. Palestinian Interim Self-Gov't Auth. , 53 F.Supp.3d 191, 213 (D.D.C. 2014). Indeed, the Gilmore expert's opinion derived solely from materials that had been proffered at trial but excluded as inadmissible hearsay. Id. at 212-13. In this case, the plaintiffs' experts relied upon their own extensive research into terrorist organizations to conclude that Sudan provided material support that caused the embassy bombings. In doing so, the experts—unlike the expert in Gilmore —drew upon both materials admitted at the evidentiary hearing and sources encountered in their research and professional experience. A "layperson" could not reliably have reached the same conclusions as the experts in this case.

Finally, Sudan belatedly challenges the reliability of the factual bases for the experts' testimony. Of course, "the decision whether to qualify an expert witness is within the broad latitude of the trial court and is reviewed for abuse of discretion." Haarhuis v. Kunnan Enters. , 177 F.3d 1007, 1015 (D.C. Cir. 1999) (citing Kumho Tire Co. v. Carmichael , 526 U.S. 137, 152, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) ). As previously stated, experts may rely upon hearsay evidence in forming their admissible, professional opinions. Indeed, it is hard to imagine what other than hearsay an expert on terrorism could use to formulate his opinion. See Boim , 549 F.3d at 704 ("Biologists do not study animal behavior by placing animals under oath, and students of terrorism do not arrive at their assessments solely or even primarily by studying the records of judicial proceedings"). All the Federal Rules require is that the "facts or data in the particular case upon which an expert bases an opinion or inference ... [are] of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." FED. R. EVID . 703 (2010) (amended without substantive change 2011).

Here, the plaintiffs' experts used, among other things, trial testimony of al Qaeda informants, intelligence reports from the U.S. Government, and their exhaustive review of secondary sources to reach their conclusions. Courts have consistently held these sorts of materials provide an adequate basis for expert testimony on terrorism. See Damrah , 412 F.3d at 625 & n.4 (approving an expert's reliance upon books, press releases, newspaper articles, and the State Department's Patterns of Global Terrorism reports); Boim , 549 F.3d at 704-05 (approving reliance upon terrorist websites and observations from prior criminal trials). In light of the general acceptance of the plaintiffs' experts' sources and methodologies, we conclude the district court did not abuse its discretion in qualifying the experts, summarizing their testimony, or crediting their conclusions.

c. Reliability of the experts' conclusions

Sudan's third objection attacks the reliability of the experts' opinions in this case as inconsistent with the underlying facts. In other words, Sudan asks this court to hold the expert opinions are inadmissible because the plaintiffs' witnesses have not "reliably applied [their] principles and methods to the facts of the case." See FED. R. EVID . 702(d). This challenge also implies the district court based its findings of jurisdiction upon clearly erroneous facts. See Price , 389 F.3d at 197 (reviewing for clear error jurisdictional findings of fact in a FSIA terrorism case); see also Duke Power Co. v. Carolina Envtl. Study Grp. , 438 U.S. 59, 74-77, 98 S.Ct. 2620, 57 L.Ed.2d 595 (1978).

The problem with this argument is that Sudan has not explained—either at the evidentiary hearing or on appeal—why these expert opinions are unreliable or clearly erroneous. By refusing to participate in the evidentiary hearing, Sudan gave up its opportunity to challenge the fit between the experts' opinions and the underlying facts. At the hearing, the witnesses described the general bases of their expertise, and the district court found them qualified to give opinions on Sudan's material support for al Qaeda. In doing so, the experts said they had relied upon multiple sources of information, including but not limited to those presented at the hearing. But the experts did not—and did not need to—provide the specific basis for their knowledge for each factual proposition they advanced. See FED. R. EVID . 705 ("an expert may state an opinion—and give the reasons for it—without first testifying to the underlying facts or data"). Therefore, we cannot know with certainty whether the experts' opinions were consistent or in conflict with the underlying facts upon which they relied. Had Sudan participated in the hearing, it could have challenged the experts to substantiate each and every factual proposition they asserted. Cf. Bryan v. John Bean Div. of FMC Corp. , 566 F.2d 541, 545 (5th Cir. 1978) (noting "the onus of eliciting the bases of the opinion is placed on the" party opposing admission). That would have allowed this court to determine whether the experts' opinions reliably reflected the more developed factual record. By deferring its attack until this appeal, Sudan has deprived the experts of an opportunity to respond, and instead asks this court to rule on an incomplete record. We decline the invitation. See Boim , 549 F.3d at 704-05 (rejecting a challenge to the reliability of an expert's inferences first brought on appeal).

2. The State Department reports

Of course, the district court did not rely solely upon expert testimony to establish jurisdiction and liability. Of particular importance, the plaintiffs marshaled nearly a decade of State Department reports that speak directly to Sudan's support for terrorist groups, including al Qaeda. See, e.g. , U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1993 ("Despite several warnings to cease supporting radical extremists the Sudanese government continued to harbor international terrorist groups in Sudan"); U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 1998 ("Sudan provides safe haven to some of the world's most violent terrorist groups, including Usama Bin Ladin's al-Qaida"); U.S. DEP'T OF STATE, PATTERNS OF GLOBAL TERRORISM : 2000 (2001) ( "Sudan ... continued to be used as a safe haven by members of various groups, including associates of Osama bin Laden's al-Qaeda organization"). These reports both bolster the experts' conclusions about Sudan's material support for the al Qaeda embassy bombings and independently show the plaintiffs' claims "ha[ve] some factual basis," as required by § 1608(e). Giampaoli , 628 F.2d at 1194.

As with the expert testimony, Sudan contends these reports are inadmissible hearsay. The plaintiffs urge the State Department reports were admissible under the hearsay exception for public records. See FED. R. EVID . 803(8). That exception allows the admission of "a record or statement of a public office if" it: (1) contains factual findings (2) from a legally authorized investigation. Id at 803(8)(iii). Pursuant to the "broad approach to admissibility" under Rule 803(8), a court may also admit "conclusion[s] or opinion[s]" contained within a public record. Beech Aircraft Corp. v. Rainey , 488 U.S. 153, 170, 109 S.Ct. 439, 102 L.Ed.2d 445 (1988). Once proffered, a public record is presumptively admissible, and the opponent bears the burden of showing it is unreliable. Bridgeway Corp. v. Citibank , 201 F.3d 134, 143 (2d Cir. 2000).

The State Department's Patterns of Global Terrorism reports fit squarely within the public records exception. First, the reports contain both factual findings and conclusions on Sudan's support for terrorism in general and al Qaeda in particular. Second, the reports were created pursuant to statute, see 22 U.S.C. § 2656f(a) (requiring annual reports on terrorism), and are therefore the product of a "legally authorized investigation." See Bridgeway , 201 F.3d at 143 (holding State Department reports required by statute are public records). Indeed, in contested FSIA proceedings we have previously approved admission of the very reports Sudan challenges, Simpson , 470 F.3d at 361 ; Kilburn , 277 F.Supp.2d at 33, aff'd 376 F.3d at 1131, as have other courts, Damrah , 412 F.3d at 625 n.4.

Sudan objects on appeal to the "trustworthiness" of these reports, but that objection should have been made in the district court. See FED. R. EVID . 803(8)(B) (providing for the admission of public records if "the opponent does not show that the possible source of the information or other circumstances indicate a lack of trustworthiness"). Even now, Sudan does not present any reason, beyond their reliance upon hearsay, to deem these reports unreliable. See Kehm v. Procter & Gamble Mfg. Co. , 724 F.2d 613, 618 (8th Cir. 1983) (holding inclusion of hearsay is not a sufficient ground for excluding a public record as unreliable). Although the reports lack the details that the expert witnesses provided concerning Sudan's material support, they are competent, admissible evidence, which together with the plaintiffs' admissible opinion evidence satisfy the burden of production on material support and jurisdictional causation. Because Sudan, by defaulting in the district court, has not carried its burden of persuasion, the district court properly asserted jurisdiction over the cases.

D. Sufficiency of the Evidence

This brings us to Sudan's second major challenge to the plaintiffs' evidence. In addition to disputing the admissibility of the evidence, Sudan argues the totality of the evidence cannot establish material support and jurisdictional causation as a matter of law. First, Sudan contends the plaintiffs cannot show its actions caused the plaintiffs' injuries because its conduct neither substantially nor foreseeably provided material support for the embassy bombings. Second, Sudan argues the plaintiffs cannot recover because its support, if any, was not intended to cause the bombings.

1. Proximate causation

Sudan's first challenge to the sufficiency of the evidence rests upon the standard for jurisdictional causation, viz. , proximate cause. In Kilburn , we held a plaintiff must show proximate cause to establish jurisdiction under § 1605(a)(7), the predecessor of the current FSIA terrorism exception. 376 F.3d at 1128. Because § 1605A(a) restates the predicate acts of § 1605(a)(7), it stands to reason that proximate cause remains the jurisdictional standard.

Proximate cause requires "some reasonable connection between the act or omission of the defendant and the damage which the plaintiff has suffered." Id. (quoting PROSSER & KEETON ON THE LAW OF TORTS 263 (5th ed. 1984)). It "normally eliminates the bizarre," Jerome B. Grubart, Inc. v. Great Lakes Dredge & Dock Co. , 513 U.S. 527, 536, 115 S.Ct. 1043, 130 L.Ed.2d 1024 (1995), "preclud[ing] liability in situations where the causal link between conduct and result is so attenuated that the consequence is more aptly described as mere fortuity." Paroline v. United States , ––– U.S. ––––, 134 S.Ct. 1710, 1719, 188 L.Ed.2d 714 (2014). As Sudan points out, the inquiry into proximate cause contains two similar but distinct elements. First, the defendant's actions must be a "substantial factor" in the sequence of events that led to the plaintiff's injury. Rothstein v. UBS , 708 F.3d 82, 91 (2d Cir. 2013). Second, the plaintiff's injury must have been "reasonably foreseeable or anticipated as a natural consequence" of the defendant's conduct. Id. Sudan contends that its support satisfies neither element of the inquiry into proximate cause with respect to the 1998 embassy bombings here at issue.

a. Substantial factor

Sudan offers two reasons its actions were not a "substantial factor" in al Qaeda's embassy bombings. Most basically, Sudan contends it did not provide any material support at all to al Qaeda during the 1990s, making proximate causation impossible. Much of this argument reprises Sudan's objections to the inferences drawn by the experts from al Fadl's testimony, which objections we have considered and rejected.

Nevertheless, Sudan points to a number of events as to which it contends the district court erroneously found material support for al Qaeda. For example, Sudan criticizes the district court's discussion of al Qaeda purchasing properties, starting businesses, and establishing terrorist training camps in Sudan. Owens IV , 826 F.Supp.2d at 141, 143-44. Viewed in isolation, none of these events necessarily evinces a Sudanese hand in al Qaeda's activities. That view, however, like Nelson at the Battle of Copenhagen, turns a blind eye to the broader picture. The record shows that after al Qaeda started its businesses, Sudan fostered their growth through tax exceptions and customs privileges. This allowed al Qaeda nearly to monopolize the export of several agricultural commodities, plowing its profits back into its broader organization. Again, after al Qaeda opened its training camps, Sudanese intelligence shielded their operations from the local police despite complaints from nearby residents. This preferential treatment certainly qualifies as material support, even if Sudan played no role in creating the underlying businesses and training camps.

Sudan also disputes the district court's finding that it provided financial support to al Qaeda. To the contrary, Sudan argues, al Qaeda financially supported Sudan by investing in Sudanese infrastructure. Sudan is correct—bin Laden did provide financial assistance to Sudan—but it ignores record evidence of Sudan's reciprocal aid. For example, as the district court noted, bin Laden's $50 million investment in the partially state-owned al Sharmal Islamic Bank gave al Qaeda "access to the formal banking system," which proved useful for "laundering money" and "financing terrorist operations." Id. at 144. Al Qaeda operatives, including bin Laden himself, held accounts in their real names in al Sharmal bank, demonstrating the impunity with which the group operated in Sudan. Thus, although Sudan did not directly fund al Qaeda or its business, the court reasonably concluded its in-kind assistance had the same practical effect.

Finally, Sudan invokes the testimony of Simon, the former NSC staffer overseeing counterterrorism activities, that Sudan provided no "useful information on bin Laden's" activities that "might have helped the U.S. unravel the plots to attack the two East African U.S. embassies." Id. at 145. The district court's finding of material support, Sudan argues, is unsustainable "without a showing that Sudan had useful intelligence and nonetheless elected not to share it." Although the district court did not say what Sudan knew about al Qaeda or when it knew it, Sudan's claims of ignorance regarding al Qaeda's aims defies both reason and the record. After all, Sudan invited "literally every single jihadist style group," including al Qaeda, to relocate to Sudan in the early 1990s. At the time, bin Laden was known as a wealthy Islamist financier and a leader in the Afghani mujahedeen. As soon as al Qaeda took up residence in Sudan, bin Laden began issuing fatwas denouncing the United States and calling for attacks upon U.S. interests. And after the Battle of Mogadishu in 1993, al Qaeda operatives publically boasted about killing U.S. soldiers in Somalia. According to Kohlmann, bin Laden himself took to the Arab press and U.S. cable television to claim responsibility for this attack. Sudanese intelligence officers would have been privy to all this information because they frequented al Qaeda's guesthouses, and al Turabi's NIF shared offices with al Qaeda for a time.

Sudan's own actions also gave it knowledge of al Qaeda's capabilities and aims. For example, Sudanese intelligence must have known that al Qaeda operated training camps where explosives were used because it shielded those camps from interference by the local police. Sudan also knew al Qaeda was transporting large, undeclared sums of money to Kenya because Sudanese agents shepherded operatives with this money past airport inspections. Likewise, Sudan knew something of al Qaeda's arsenal because its own planes transported al Qaeda's weapons from Afghanistan to Sudan. Indeed, on one occasion, a Sudanese official even assisted al Qaeda in an ultimately unsuccessful bid to obtain nuclear weapons from a smuggler in South Africa. Contrary to Sudan's contention, all this information would have aided the United States in appreciating the threat of al Qaeda and attempting to disrupt its operations. Sudan's refusal to divulge any of this information—even after a specific request from the United States in 1996—certainly qualifies as material support. Cf. Estate of Parsons v. Palestinian Auth. , 651 F.3d 118, 125-26 (D.C. Cir. 2011) (security officers who, with knowledge, failed to intervene in ongoing bomb plot provided material support).

Sudan's second argument that its actions were not a "substantial factor" causing the plaintiffs' injuries focuses upon the temporal distance between Sudan's support for al Qaeda and the embassy bombings. Principally, Sudan argues that by expelling bin Laden in 1996 it broke the chain of causation leading to the 1998 embassy bombings. We confronted and rejected the same objection in our 2008 opinion affirming the district court's denial of Sudan's motion to dismiss. Owens III , 531 F.3d at 895. Although we there recognized the "[p]laintiffs' allegations are somewhat imprecise as to the temporal proximity of Sudan's actions to and their causal connection with the terrorist act," we held "this imprecision [was] not fatal for purposes of jurisdictional causation." Id. (quoting Rux v. Republic of Sudan , 461 F.3d 461, 474 (4th Cir. 2006) ) (internal quotation marks omitted). In order to bridge the gap, we noted the plaintiffs' "allegations, and the reasonable inferences drawn therefrom" need only "demonstrate a reasonable connection between the foreign state's actions and the terrorist act." Id. In other words, provided the plaintiffs demonstrated proximate cause, the temporal remoteness between Sudan's material support and the embassy bombings was irrelevant. See Grubart , 513 U.S. at 536, 115 S.Ct. 1043 (proximate cause "normally eliminates the bizarre" without "the need for further temporal or spatial limitations"). And at that stage in the litigation, we concluded, the plaintiffs had more than met their burden of pleading facts sufficient to establish proximate causation. Owens III , 531 F.3d at 895.

Fast-forwarding to the present day, the plaintiffs have substantiated their allegations of material support and jurisdictional causation with admissible evidence, which Sudan did not challenge at the evidentiary hearing. Once again, the district court found the evidence established a "reasonable connection" between Sudan's actions and the embassy bombings. As in our 2008 decision, we see nothing erroneous with this conclusion for two reasons.

First, we do not believe Sudan broke the chain of proximate causation by completely disassociating itself from al Qaeda in or after 1996. A declassified CIA President's Daily Brief in December 1998—months after the embassy bombings—reports a "Bin Laden associate in Sudan" sending materials to al Qaeda in Afghanistan. The State Department's 1998 Patterns of Global Terrorism further reports that "Sudan continued to serve as a meeting place, safehaven, and training hub for a number of international terrorist groups, particularly Usama Bin Ladin's al-Qaida organization" even after the embassy bombings. Although counterterrorism cooperation between the United States and Sudan improved after the bombings, the 2000 Patterns of Global Terrorism report reiterates "Sudan continued to serve as a safehaven for members of al-Qaida, the Lebanese Hizballah, al-Gama'a al-Islamiyya, Egyptian Islamic Jihad, the PIJ, and HAMAS." In addition, both Kohlmann and Simon testified that al Qaeda operatives remained in Sudan after 1996. Sudan insists that a gap remained between its expulsion of bin Laden and the government reports detailing al Qaeda's presence in Sudan in late 1998, but it strains credulity that Sudan would immediately resume relations with al Qaeda following bombings for which the group claimed credit after completely cutting ties two years earlier. Rather, as the district court inferred, it is far more likely that Sudan, despite having expelled bin Laden in 1996, continued to harbor al Qaeda terrorists until and after the bombings.

Second, even if Sudan were correct on this factual point, severing ties with al Qaeda would not preclude a finding that its material support remained a substantial factor in the embassy bombings. See Boim , 549 F.3d at 699-700 (holding a "two year[ ]" interval between the defendant's material support and the plaintiff's injury was far from the point at which "considerations of temporal remoteness might ... cut off liability").

Sudan counters by selectively quoting the 9/11 Commission Report, stating "Bin Ladin left Sudan ... significantly weakened." Perhaps so if viewed in isolation, but bin Laden's expulsion did not undo the support Sudan provided in the previous years. Sudan's invitation, after all, allowed al Qaeda to extricate itself from a war-torn Afghanistan and organize its terrorist enterprise in a stable safe haven. During al Qaeda's stay, Sudan sheltered the group from foreign intelligence and facilitated its movement throughout the region. It also put al Qaeda in contact with other, more experienced terrorist groups residing in Sudan. These actions allowed al Qaeda to grow its membership, to develop its capabilities, and to establish the cells in Kenya and Tanzania, which ultimately launched the 1998 bombings. Indeed, "the vast majority of the planning and preparation [for the embassy attacks] took place between the years of 1991 and 1997" when Bin Laden, for the most part, remained in the Sudan. According to one expert, Sudan's expulsion of bin Laden may have even "accelerated the bomb plot" by allowing al Qaeda to militarize its African cells without fear of reprisal against him by the United States, which had known of his presence in Sudan. Id. at 310-11. As Sudan notes, al Qaeda had not committed "any terrorist attacks predating" its arrival in the country, and indeed "the idea that al-Qaeda was capable of anything significant" in the early 1990s "was laughable." Yet in a few short years, al Qaeda progressed from mounting small-scale, often-unsuccessful attacks to orchestrating the near-simultaneous bombings of American embassies in two different countries. Although the expulsion of bin Laden may have marked a temporary setback for Al Qaeda, on balance, the organization benefited greatly from Sudan's aid during the 1990s. Therefore, the district court's conclusion that Sudan's support was a "substantial factor" in the chain of causation leading to the embassy bombings was far from clearly erroneous.

b. Reasonable foreseeability

Sudan contends even if its support was a "significant factor" in the embassy bombings, the attacks were not "reasonably foreseeable or anticipated as a natural consequence" of that support. Principally, Sudan argues it was not foreseeable in 1991—when Sudan invited bin Laden to relocate—that al Qaeda would engage in terrorist activities. As evidence, Sudan points out that bin Laden was not yet infamous for acts of terrorism and the United States had not yet designated al Qaeda a terrorist organization or bin Laden a terrorist and did not do so until after the embassy bombings. Designation of Foreign Terrorist Organizations, 64 Fed. Reg. 55,112, 55,112/1 (Oct. 8, 1999); Exec. Order No. 13099, 63 Fed. Reg. 45,167, 45,167 (Aug. 20, 1998). That bin Laden and al Qaeda "may have abused their opportunities" in the country, Sudan urges, does not mean it should be held accountable when "its residents later turn out to be terrorists."

Once again Sudan ignores the broader context of its actions. In the early 1990s the Sudanese government reached out to numerous terrorist groups, including the "Palestinian HAMAS movement, the Palestinian Islamic Jihad, Hezbollah, ... al Qaeda, the Egyptian Islamic Jihad, the Libyan Islamic Fighting Group, dissident groups from Algeria, Morocco, the Eritrean Islamic Jihad movement." Owens IV , 826 F.Supp.2d at 141 (quoting Kohlmann). "[L]iterally every single jihadist style group, regardless of what sectarian perspective they had, was invited to take a base in Khartoum" during this period. Id. That al Qaeda was included in this list of renowned terrorist organizations supports an inference that its terrorist aims were foreseeable—indeed, foreseen—at the time of Sudan's invitation.Sudan's own briefs implicitly concede the foreseeability of al Qaeda's aims in the early 1990s. To wit, Sudan reiterates the district court's finding that "Bin Laden 'was a famous mujahedeen fighter who had successfully fought the Soviet Union' and 'was thought to be fabulously wealthy.' " See Owens IV , 826 F.Supp.2d at 140-41. Yet it argues "the idea that al-Qaeda was capable of anything significant was laughable." True, al Qaeda was then a fledgling terrorist organization, but one led by a "famous ... fighter" and a "fabulously wealthy" fundamentalist jihadi who had "successfully fought" a world superpower. Any impartial observer could see the group's future potential for mayhem far outstripped its then already substantial capabilities. Sudan cannot bury its head in the sand and contend otherwise.

Furthermore, as its relationship with al Qaeda deepened, Sudan undoubtedly became aware of al Qaeda's hostility to the United States and its intention to launch attacks against American interests. Starting in 1991, bin Laden issued a series of fatwas against the United States from Khartoum, and al Qaeda operatives publically boasted about attacking American soldiers in Somalia in 1993. Despite this, Sudan continued to assist the group in moving people and resources throughout the region. Sudan's claimed ignorance of al Qaeda's specific aim to bomb American embassies focuses too narrowly upon those events; Sudan could not help but foresee that al Qaeda would attack American interests wherever it could find them.

In sum, Sudan's actions in the 1990s were undoubtedly a "substantial factor in the sequence of responsible causation" that led to the embassy bombings. Rothstein , 708 F.3d at 91. Moreover, the bombings were a "reasonably foreseeable or anticipated as a natural consequence" of its material support. Id. Therefore, the district court correctly held that the plaintiffs had demonstrated proximate cause, establishing jurisdiction under the FSIA.

2. Sudan's specific intent

Sudan resists this conclusion by attempting to graft an additional requirement onto the proximate cause analysis. The FSIA terrorism exception, Sudan argues, requires something more than proximate causation: "The foreseeability aspect of proximate causation" it says, "is reinforced by § 1605A(a)(1)'s requirement that material support be provided 'for' the predicate act." Sudan's point is that the use of "for" with reference to "the provision of material support" indicates that the FSIA "requires a showing of intent" on the part of the foreign sovereign to achieve the predicate act, for which it refers us to Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc. , 455 U.S. 489, 502, 102 S.Ct. 1186, 71 L.Ed.2d 362 (1982) (prohibition on selling merchandise "marketed for use" with illegal drugs requires a showing of intent on the defendant's behalf). But see Posters 'N' Things, Ltd. v. United States , 511 U.S. 513, 519, 114 S.Ct. 1747, 128 L.Ed.2d 539 (1994) (prohibition in the same statute on selling a product "intended or designed for use" with illegal drugs looks only to the objective features of the product, not to a defendant's intent). Under this reading, Sudan's material support could not give rise to jurisdiction unless Sudan specifically intended its support to cause the embassy bombings.

Although the record contains much evidence of Sudan's support for al Qaeda and its general awareness of the group's terrorist aims, nothing suggests that Sudan specifically knew of or intended its support to cause the embassy bombings. Nothing in the FSIA, however, requires a greater showing of intent than proximate cause. Indeed, we dispatched a similar argument in Kilburn , along with a hypothetical raised by the sovereign defendant:

A terrorist organization is supported by two foreign states. One specifically instructs the organization to carry out an attack against a U.S. citizen. Can the state which only provides general support, but was not involved with the act giving rise to the suit, also be stripped of its immunity?

376 F.3d at 1128. Yes, we said. Because material support "is difficult to trace," requiring more than proximate cause "could absolve" a state from liability when its actions significantly and foreseeably contributed to the predicate act. Id.

Further, we rejected the related argument that the "provision of material support or resources ... for such an act " required that "a state's material support must go directly for the specific act." Id. at 1130. That limitation, we explained, "would likely render § 1605(a)(7)'s material support provision ineffectual" because material support "is fungible" and "terrorist organizations can hardly be counted on to keep careful bookkeeping records." Id. Indeed, in other situations, courts have required neither specific intent nor direct traceability to establish the liability of material supporters of terrorism. See Boim , 549 F.3d at 698 (approving liability for donors to terrorist organizations whose donations were made for non-terrorism purposes). As Judge Posner has aptly said, "[t]o require proof that [a defendant] intended that his contribution be used for terrorism ... would as a practical matter eliminate ... liability except in cases in which the [defendant] was foolish enough to admit his true intent." Id. at 698-99. The same holds true for a state sponsor of terrorism under the FSIA; it may not avoid liability for supporting known terrorist groups by professing ignorance of their specific plans for attacks. In sum, that the evidence failed to show Sudan either specifically intended or directly advanced the 1998 embassy bombings is irrelevant to proximate cause and jurisdictional causation.

* * * * *

In short, the plaintiffs have offered sufficient admissible evidence that establishes that Sudan's material support of al Qaeda proximately caused the 1998 embassy bombings. The district court, therefore, correctly held the plaintiffs met their burden of production under the FSIA terrorism exception. Because Sudan failed to participate in the litigation, it did not rebut that its material support caused these extrajudicial killings. Therefore, this court has jurisdiction to hear claims against Sudan arising from the 1998 embassy bombings.

IV. Timeliness of Certain Claims

The remainder of Sudan's jurisdictional arguments apply only to certain groups of plaintiffs. Even if we rule for Sudan on all these matters, many of the judgments—and the district court's 2011 holding on liability—will therefore remain intact.

One such argument is that the claims of certain plaintiffs are barred by the statute of limitation in the FSIA, which Sudan views as a jurisdictional limit on the court's power to hear a case. Like its predecessor, the current version of the FSIA terrorism exception contains a limitation period on personal injury claims against a state sponsor of terrorism. Application of the limitation period requires analysis of three components of the 2008 NDAA.

The first is the limitation period itself. Codified at § 1605A(b), the FSIA provides that:

An action may be brought or maintained under this section if the action is commenced, or a related action was commenced under section 1605(a)(7)... or

[the Flatow Amendment] not later than the latter of (1) 10 years after April 24, 1996; or (2) 10 years after the date on which the cause of action arose.

The second component is § 1083(c)(3) of the 2008 NDAA, which defines the contours of a "related action" and imposes an additional time limitation on the filing of related actions:

(3) RELATED ACTIONS.—If an action arising out of an act or incident has been timely commenced under section 1605(a)(7)... or [the Flatow Amendment], any other action arising out of the same act or incident may be brought under section 1605A... if the action is commenced not later than the latter of 60 days after—(A) the date of the entry of judgment in the original action; or (B) the date of the enactment of this Act.

Finally, in addition to filing a new action or a "related action," the NDAA offers a second way to avoid the limitation period if the plaintiff had previously brought a claim under § 1605(a)(7). Section 1083(c)(2) of the NDAA provides, in part:

(2) PRIOR ACTIONS.—(A) IN GENERAL.—With respect to any action that—(i) was brought under section 1605(a)(7) of title 28, United States Code, or [the Flatow Amendment] before the date of enactment of this act ... and ... is before the courts in any form ... that action, and any judgment in the action shall, on motion made by plaintiffs ... be given effect as if the action had originally been filed under section 1605A(c).

For these "prior actions" the NDAA removes the "defenses of res judicada, collateral estoppel, and [the] limitations period" if the plaintiff moved to convert his prior action or refiled a new action under § 1605A(c). NDAA § 1083(c)(2)(B). A new claim using § 1083(c)(2) is timely if it complies with the limitation period in § 1605A(b) or was filed within 60 days of enactment of the NDAA. Id. § 1083(c)(2)(C).

Each provision comes into play in Sudan's challenge to the timeliness of the plaintiffs' actions. In this case, the plaintiffs' causes of action arose on August 7, 1998, the date of the embassy bombings. See Vine v. Republic of Iraq , 459 F.Supp.2d 10, 21 (D.D.C. 2006) (holding a claim under the FSIA "arises on the date that the action in question occurred"), rev'd in part on another ground sub nom. Simon v. Republic of Iraq , 529 F.3d 1187, 1194-95 (D.C. Cir. 2008) (describing an argument to the contrary as "rather strained"), rev'd on another ground sub nom. Republic of Iraq v. Beaty , 556 U.S. 848, 129 S.Ct. 2183, 173 L.Ed.2d 1193 (2009). Therefore, unless the plaintiffs can identify a "related action ... commenced under section 1605(a)(7)" or had brought a "prior action" that remained "before the courts in any form," the last day to file a new action under § 1605A was August 7, 2008, ten years after the bombings.

Sudan does not dispute that several of the plaintiffs have filed timely actions under § 1605A. The Owens plaintiffs filed their original action under § 1605(a)(7) in October 2001 and after passage of the NDAA timely moved to convert their prior action pursuant to § 1083(c)(2). Days before the statutory deadline, the Amduso and Wamai plaintiffs filed new actions under § 1605A, and the Osongo and Mwila plaintiffs filed suit on the last possible day. Sudan does not challenge the timeliness of these plaintiffs.

The Khaliq, Opati, and Aliganga plaintiffs are another story. The Khaliq plaintiffs filed a complaint in November 2004 but missed the statutory deadline to convert that prior action under § 1083(c)(2) into a new action under § 1605A. See Khaliq v. Republic of Sudan , No. 1:04-cv-01536, at *3 (D.D.C. Sept. 9, 2009) (denying motion to convert under § 1083(c)(2)). Six months later, they filed a new case under § 1605A, asserting it was "related" both to their earlier suit and to the Owens , Mwila , and Amduso actions. The district court ordered briefing on whether the new suit was a "related action" within the scope of § 1083(c)(3) and ultimately allowed the case to proceed.

After the court held the evidentiary hearing and made its findings on liability and well past August 2008, the Aliganga plaintiffs moved to intervene in the Owens action, which the district court allowed, holding their claims were "related" to the Owens action per § 1083(c)(3). The Opati plaintiffs joined last, filing a suit "related" to the Owens action under § 1083(c)(3) on July 24, 2012. The court allowed both the Aliganga and Opati plaintiffs the benefit of its earlier findings on liability and jurisdiction.

Sudan challenges the timeliness of the Khaliq, Opati, and Aliganga plaintiffs, which raises two issues, only one of which we need to address on appeal. First, Sudan asserts that the limitation period in § 1605A(b) is jurisdictional and therefore bars a court from hearing any untimely action. Unless the limitation period in § 1605A(b) is jurisdictional, Sudan forfeited this affirmative defense by defaulting in the district court. See Practical Concepts , 811 F.2d at 1547. The plaintiffs argue that the time bar, like most statutes of limitation, is not jurisdictional and hence is forfeit. See Day v. McDonough , 547 U.S. 198, 202, 126 S.Ct. 1675, 164 L.Ed.2d 376 (2006) ( "Ordinarily in civil litigation, a statutory time limitation is forfeited if not raised in a defendant's answer or in an amendment thereto").

Assuming the limitation period is jurisdictional, Sudan contends the Khaliq, Opati, and Aliganga claims are barred because they are not "related actions" under § 1605A(b). A "related action," Sudan urges, must be filed by the same plaintiff who had filed an earlier action under § 1605(a)(7), which the Opati and Aliganga plaintiffs did not do. We need not, however, decide what qualifies as a "related action" because we hold the limitation period in § 1605A(b) is not jurisdictional. As a consequence Sudan forfeited its limitation defense by defaulting in the district court. See Harris v. Sec'y, U.S. Dep't of Veterans Affairs , 126 F.3d 339, 343 (D.C. Cir. 1997).

A line of recent Supreme Court cases has defined the circumstances in which a statute of limitation is jurisdictional. These cases uniformly recognize that a limitation period is not jurisdictional "unless it governs a court's adjudicatory capacity, that is, its subject-matter or personal jurisdiction." Henderson ex rel. Henderson v. Shinseki , 562 U.S. 428, 435, 131 S.Ct. 1197, 179 L.Ed.2d 159 (2011). To have a jurisdictional effect, a statute of limitation must "speak in jurisdictional terms," that is, restrict "a court's power" to hear a claim. United States v. Kwai Fun Wong , –––U.S. ––––, 135 S.Ct. 1625, 1633, 191 L.Ed.2d 533 (2015) (quoting Arbaugh v. Y & H Corp ., 546 U.S. 500, 515, 126 S.Ct. 1235, 163 L.Ed.2d 1097 (2006) ). Unless the Congress has "clearly stated" that it "imbued a procedural bar with jurisdictional consequences," the bar does not have them. Id. at 1632 (quoting Sebelius v. Auburn Reg'l Med. Ctr. , 568 U.S. 145, 153, 133 S.Ct. 817, 184 L.Ed.2d 627 (2013) ) (internal quotation marks and alterations omitted). Thus has the Court "made plain that most time bars are nonjurisdictional." Id.

Of course, the Congress need not incant "magic words" in order clearly to demonstrate its intent. Henderson , 562 U.S. at 436, 131 S.Ct. 1197. We look for the Congress's intent in "the text, context, and relevant historical treatment of the provision at issue." Musacchio v. United States , ––– U.S. ––––, 136 S.Ct. 709, 717, 193 L.Ed.2d 639 (2016) (quoting Reed Elsevier, Inc. v. Muchnick , 559 U.S. 154, 166, 130 S.Ct. 1237, 176 L.Ed.2d 18 (2010) (internal quotation marks omitted)). Doing so shows that § 1605A(b) is not a limit on the court's jurisdiction to hear an untimely FSIA claim.

We begin, as we must, with the text of § 1605A(b), which we note does not appear to "speak in jurisdictional terms":

An action may be brought or maintained under this section ... if commenced ... [within] 10 years after April 24, 1996; or 10 years after the date on which the cause of action arose.

Nothing in the section refers to the "court's power" to hear a case. Nothing in § 1605A(a)"conditions its jurisdictional grant on compliance with [the] statute of limitations" in § 1605A(b). Musacchio , 136 S.Ct. at 717 (quoting Reed Elsevier , 559 U.S. at 165, 130 S.Ct. 1237 ). Indeed, § 1605A(b)"is less 'jurisdictional' in tone" than limitation periods held nonjurisdictional in prior cases. See Auburn Reg'l Med. Ctr. , 568 U.S. at 154, 133 S.Ct. 817 (comparing the permissive term "may" in one statute with the mandatory term "shall" in another but holding both were nonjurisdictional). The plain text alone is enough to render the limitation period in § 1605A(b) nonjurisdictional.

Sudan nonetheless contends that the reference to "actions" rather than "claims" imbues the provision with jurisdictional import. For this proposition Sudan cites Spannaus v. U.S. Department of Justice , 824 F.2d 52 (D.C. Cir. 1987), in which we held a statute that similarly barred untimely "actions" was jurisdictional. See 28 U.S.C. § 2401(a). Sudan argues that by using the term "action" in § 1605A(b) the Congress made a clear statement replicating the jurisdictional reach of the similarly phrased statute at issue in Spannaus .

This analogy has several problems. First, as the plaintiffs point out, Spannaus was decided nearly a decade before the Supreme Court erected the presumption against jurisdictional effect, see Carlisle v. United States , 517 U.S. 416, 434, 116 S.Ct. 1460, 134 L.Ed.2d 613 (1996) (Ginsburg, J. concurring) (making the first reference to a presumption against jurisdictional effect), and the Congress enacted § 1605A after that presumption had been fully articulated, see Kontrick v. Ryan , 540 U.S. 443, 455, 124 S.Ct. 906, 157 L.Ed.2d 867 (2004) (criticizing the "less than meticulous" use of the term "jurisdictional" in earlier decisions). Therefore, Spannaus is unpersuasive on the matter. Second, the plaintiffs correctly note we did not rely upon the phrase "every civil action" in Spannaus to hold the limitation period in § 2401(a) jurisdictional. Rather, we relied upon longstanding precedent establishing that " § 2401(a) is a jurisdictional condition attached to the government's waiver of sovereign immunity, and as such must be strictly construed." 824 F.2d at 55 (citing United States v. Mottaz , 476 U.S. 834, 106 S.Ct. 2224, 90 L.Ed.2d 841 (1986) and Soriano v. United States , 352 U.S. 270, 276, 77 S.Ct. 269, 1 L.Ed.2d 306 (1957) ); cf. John R. Sand & Gravel Co. v. United States , 552 U.S. 130, 139, 128 S.Ct. 750, 169 L.Ed.2d 591 (2008) (holding a statute of limitation as jurisdictional when "[b]asic principles of stare decisis" required that outcome). In this case, precedent does not help Sudan because no court has given § 1605A(b)"a definitive earlier interpretation" that could displace the presumption against jurisdictional reach. Id. at 137-38, 128 S.Ct. 750.

Further, Sudan's invocation of the nostrum that identical words in similar statutes demand an identical construction finds little support in the most relevant precedents. See Wong , 135 S.Ct. at 1629 (rejecting the argument that use of the phrase "shall forever be barred" rendered a limitation period jurisdictional despite the inclusion of the identical phrase in a jurisdictional statute of limitation). Therefore, the use of the term "action" in a provision held jurisdictional in Spannaus says little about whether a similarly phrased statute also has jurisdictional reach. Nor have courts attached jurisdictional significance to the word "action" in other statutes. See, e.g. , Reed Elsevier , 559 U.S. at 166, 130 S.Ct. 1237 (holding nonjurisdictional 17 U.S.C. § 411(a), which bars any "civil action" for infringement without prior registration of the copyright); Hardin v. City Title & Escrow Co ., 797 F.2d 1037, 1040 (D.C. Cir. 1986) (stating that 15 U.S.C. § 15b, which bars "[a]ny [untimely] action to enforce any cause of action," is "a good example of a non-jurisdictional time limitation"). Sudan presents no reason we should embrace Spannaus yet ignore these other precedents as well as the Supreme Court's most recent guidance on statutory interpretation. Hence, we find no support for Sudan's textual argument that § 1605A(b) is jurisdictional.

Sudan next argues from the structure of the statute in which § 1605A(b) appears: Because the limitation period follows immediately after the grant of jurisdiction in § 1605A(a), it takes on the jurisdictional nature of the prior provision. Again, precedent suggests otherwise. As the plaintiffs note, the Supreme Court has held the "separation" of a time bar "from jurisdictional provisions" implies the limitation period is not jurisdictional. Gonzalez v. Thaler , 565 U.S. 134, 132 S.Ct. 641, 651, 181 L.Ed.2d 619 (2012) ; cf. Blueport Co., LLC v. United States , 533 F.3d 1374, 1380 (Fed. Cir. 2008) (holding limits on patent infringement suits against the Government are jurisdictional because they appear in the same sentence as a general waiver of sovereign immunity). The limitation period in § 1605A(b) and the grant of jurisdiction in § 1605A(a) appear in two different subsections of the terrorism exception, only one of which speaks in jurisdictional terms. The remaining subsections of § 1605A are plainly nonjurisdictional. See, e.g. , 28 U.S.C. §§ 1605A(c) (private right of action), 1605A(d) (additional damages), 1605A(e) (use of special masters), 1605A(g) (property disposition). That the limitation period follows immediately after the jurisdictional provisions of § 1605A(a) is of little import. See Gonzalez , 565 U.S. at 147, 132 S.Ct. 641 ("Mere proximity will not turn a rule that speaks in nonjurisdictional terms into a jurisdictional hurdle"). If proximity alone were enough, then every subsection in a section containing a jurisdictional provision would, by the transitive property, also abut a jurisdictional subsection and therefore be jurisdictional as well, an absurd proposition. Auburn Reg'l Med. Ctr. , 568 U.S. at 155, 133 S.Ct. 817 ("A requirement we would otherwise classify as nonjurisdictional ... does not become jurisdictional simply because it is placed in a section of a statute that also contains jurisdictional provisions").

Sudan also argues the history of § 1605A supports reading the time bar in § 1605A(b) as jurisdictional. Prior to the enactment of the 2008 NDAA, the FSIA terrorism exception under § 1605(a)(7) contained a similar time bar of ten years. See 28 U.S.C. § 1605(f) (2006). Sudan now contends that § 1605 was "undisputedly a purely jurisdictional statute," rendering both the current and the former limitation periods jurisdictional as well.

This argument mischaracterizes both old § 1605(f) and new § 1605A. The time bar in the former terrorism exception was in a separate subsection of the FSIA, § 1605(f), from the grant of jurisdiction over claims against a state sponsor of terrorism in § 1605(a)(7). Section § 1605 did have several jurisdictional provisions, see §§ 1605(a)(1)-(7), (b), (d), but each one expressly proclaimed its jurisdictional reach. See, e.g. , 28 U.S.C. §§ 1605(a) ("A foreign state shall not be immune from the jurisdiction of courts of the United States or of the States in any case" falling within one of the seven enumerated exceptions). The other four subsections of § 1605 made no mention of jurisdiction. The difference is telling, but understandable as these provisions—much like those in § 1605A —defined terms ( § 1605(e) ), limited discovery ( § 1605(g) ), and governed the choice of law and the calculation of damages ( § 1605(c) ), among other things, none of which could have jurisdictional effect. As in § 1605A, § 1605 demonstrates that when the Congress intends to make a provision jurisdictional, it normally does so expressly. When words of jurisdictional import are absent, so too, we presume, is jurisdictional effect.

Sudan lastly argues that waivers of sovereign immunity must be strictly construed. See Spannaus , 824 F.2d at 55. But see Scarborough v. Principi , 541 U.S. 401, 421, 124 S.Ct. 1856, 158 L.Ed.2d 674 (2004) ("[L]imitations principles should generally apply to the Government 'in the same way that' they apply to private parties") (quoting Franconia Assocs. v. United States , 536 U.S. 129, 145, 122 S.Ct. 1993, 153 L.Ed.2d 132 (2002) ). The Supreme Court has twice addressed this very point and rejected it for time bars that conditioned waivers of the U.S. Government's sovereign immunity. Irwin v. Dep't of Veterans Affairs , 498 U.S. 89, 94-96, 111 S.Ct. 453, 112 L.Ed.2d 435 (1990) ; Wong , 135 S.Ct. at 1636. Treating a time bar as nonjurisdictional, the Court has said, "is likely to be a realistic assessment of legislative intent" and "amounts to little, if any, broadening of the congressional waiver" of sovereign immunity. Irwin , 498 U.S. at 95, 111 S.Ct. 453. Therefore, Sudan's argument that sovereignty gives jurisdictional import to the limitation period in the FSIA terrorism exception is unpersuasive.

In any event, Sudan misses the distinction between a waiver of sovereign immunity and an exception to the statutory grant of foreign sovereign immunity. The Congress "did not waive [a foreign state's] sovereign immunity in enacting [the FSIA terrorism exception]" because "only the sovereign can forswear the sovereign's legal rights." Simon , 529 F.3d at 1196. Rather, "[i]n the terrorism exception the Congress qualified the statutory grant of immunity to [foreign sovereigns]," which is "itself 'a matter of grace and comity.' " Id. (quoting Verlinden , 461 U.S. at 486, 103 S.Ct. 1962 ). Because the FSIA exceptions are not waivers of sovereign immunity, the rule of strict construction does not apply.

Having reviewed the text, structure, or history of the FSIA terrorism exception, we see "no authority suggesting the Congress intended courts to read [ § 1605A(b) ] any more narrowly than its terms suggest." Id. Sudan's arguments to the contrary fail. We therefore hold that the limitation period in § 1605A(b) is not jurisdictional. It follows that Sudan has forfeited its affirmative defense to the Khaliq , Opati , and Aliganga actions by failing to raise it in the district court. See Musacchio , 136 S.Ct. at 717 ; Harris , 126 F.3d at 343. As a consequence, we have no need to consider Sudan's interpretation of a "related action" under NDAA § 1083(c)(3).

V. Jurisdiction and Causes of Action for Claims of Third Parties

Sudan next takes aim at claims brought under state and federal law by the family members of those killed or injured in the embassy bombings. First, Sudan contends § 1605A(a) does not grant the court jurisdiction to hear a claim from a plaintiff (or the legal representative of a plaintiff) who was not physically injured by a terrorist attack. Second, even if jurisdiction is proper, Sudan argues the federal cause of action in § 1605A(c) supplies the exclusive remedy for a FSIA claimant, precluding claims under state law. Finally, Sudan insists a family member who was not present at the scene of the embassy bombings cannot state a claim for intentional infliction of emotional distress (IIED) under District of Columbia law.

A. Jurisdiction

We turn first to Sudan's jurisdictional argument, which we are obliged to address notwithstanding Sudan's default. The plaintiffs in this case have brought two different types of claims under various sources of law. First are the claims of those physically injured by the embassy bombings or by the legal representatives of those now deceased or incapacitated. Second are the claims of family members of those physically injured or killed by the bombings who seek damages for their emotional distress. Sudan contends the FSIA extends jurisdiction only to members of the first group and their legal representatives. The claims of family members for emotional distress, it argues, are outside the jurisdiction conferred upon the court.

Sudan's argument turns upon the meaning of the phrase "the claimant or the victim" in § 1605A(a)(2)(A)(ii). Section 1605A(a) gives the court jurisdiction and withdraws immunity only when "the claimant or the victim" falls within one of four categories: U.S. nationals, members of the armed forces, and employees or contractors of the United States acting within the scope of their employment. A separate subsection of the terrorism exception provides a federal cause of action to the same groups of plaintiffs and their legal representatives. 28 U.S.C. § 1605A(c).

Sudan contends that "the claimant" in § 1605A(a)(2)(A)(ii) refers only to the legal representative of a victim of a terrorist attack. This would effectively align the grant of jurisdiction with the federal cause of action under § 1605A(c). That is, under Sudan's proffered interpretation, a court would have jurisdiction only over claims brought by persons who could invoke the federal cause of action in § 1605A(c). Applied to the case at hand, this might preclude jurisdiction over a claim for emotional distress brought by a relative of someone killed or injured by the embassy bombings because a family member is arguably neither a victim of the attack nor the legal representative of a victim.

Sudan's argument has several problems. First and foremost, Sudan's interpretation is inconsistent with the plain meaning and the structure of the statute, as is clear from the differences between the grant of jurisdiction in § 1605A(a) and the cause of action in § 1605A(c). Section 1605A(a)(2) grants jurisdiction when "the claimant or the victim" is a member of one of the four enumerated groups. In contrast, § 1605A(c) authorizes a cause of action not only for those four groups but also for the legal representative of a member of those groups. If the Congress had intended § 1605A(a)(2) to mirror the scope of § 1605A(c), then it would have used the same term—"legal representative"—in both subsections (i.e., "the legal representative or the victim"), as it did with the verbatim enumeration of the four qualifying groups. That it did not signals its intent to give the term "claimant" in § 1605A(a)(2) a meaning different from and broader than "the legal representative" in § 1605A(c). See Russello v. United States, 464 U.S. 16, 23, 104 S.Ct. 296, 78 L.Ed.2d 17 (1983).

What, then, does the FSIA mean by the terms "claimant" and "legal representative"? The plain meaning of claimant, the plaintiffs correctly note, is simply someone who brings a claim for relief. Who can be a claimant is typically defined by the substantive law under which a plaintiff states a claim. By contrast, the term "legal representative" contemplates a far narrower universe of persons based upon principles of agency or a special relationship, such as marriage. See, e.g. , Fed. Treasury Enter. Sojuzplodoimport v. SPI Spirits Ltd. , 726 F.3d 62, 80 (2d Cir. 2013) ("In its broadest usage, the phrase 'legal representative' may refer simply to '[o]ne who stands for or acts on behalf of another' "). Federal and state procedural law, not the substantive law under which a plaintiff states a claim, typically defines who may serve as a legal representative in a given suit. See FED. R. CIV . P. 17(b)(3) ; Gurley v. Lindsley , 459 F.2d 268, 279 (5th Cir. 1972) (applying Texas law in accord with Rule 17(b) ). Thus, a legal representative is a special type of claimant who proceeds on behalf of an absent party with a substantive legal right.

Sudan nonetheless offers three reasons we should narrowly interpret "claimant" to mean no more than "legal representative." First, Sudan argues that interpreting "claimant" to mean "legal representative" is necessary to "harmonize[ ]" the scope of jurisdiction under § 1605A(a) with the cause of action under § 1605A(c). If the terms had different meanings, Sudan warns, then "certain plaintiffs [could] establish jurisdiction under § 1605A(a)" but anomalously could not "avail[ ] themselves of the private right of action in § 1605A(c)." Here Sudan is assuming a grant of jurisdiction must be no broader than the causes of action that may be brought under it. But that does not follow. Cf. FDIC v. Meyer , 510 U.S. 471, 484, 114 S.Ct. 996, 127 L.Ed.2d 308 (1994) (noting that "whether there has been a waiver of sovereign immunity" and "whether the source of substantive law" "provides an avenue for relief" are "two 'analytically distinct' inquiries"). The other exceptions to sovereign immunity in the FSIA exemplify this distinction because they grant the courts jurisdiction over claims against foreign sovereigns but neither create nor withdraw substantive causes of action for FSIA plaintiffs. See Helmerich & Payne , 137 S.Ct. at 1324 ("Indeed, cases in which the jurisdictional inquiry does not overlap with the elements of a plaintiff's claims have been the norm in cases arising under other exceptions to the FSIA").

Furthermore, even under the prior terrorism exception, the Congress authorized a cause of action—in the Flatow Amendment—with a narrower reach than the grant of jurisdiction in § 1605(a)(7). See Leibovitch v. Islamic Republic of Iran , 697 F.3d 561, 570-71 (7th Cir. 2012). That the Flatow Amendment applied only to state officials, not foreign states, took "nothing away from" the grant of jurisdiction under § 1605(a)(7) because the broader jurisdictional provision operated independently of the narrower cause of action. See Cicippio-Puleo , 353 F.3d at 1035-36. Accordingly, we declined to "harmonize" the broad grant of jurisdiction in the old terrorism exception with the narrower cause of action provided by the Flatow Amendment because doing so would have conflicted with the text of both provisions. Id. at 1032-33. So too here. Again the Congress has authorized a narrower cause of action, § 1605A(c), correlative to a broader jurisdictional grant, § 1605A(a), and as before, we see no reason to distort the plain meaning of either provision in order to make them coextensive.

Second, Sudan contends a broad interpretation of "claimant" would "render [ ] the term 'victim' superfluous." Not so; as the plaintiffs note, the use of both terms affords jurisdiction when "either the claimant or the victim is a national of the United States" or is within one of the other three groups identified in the statute. La Reunion Aerienne v. Socialist People's Libyan Arab Jamahiriya , 533 F.3d 837, 844 (D.C. Cir. 2008).

Third, Sudan argues that reading "claimant" to mean "one who brings a claim" would "greatly expand[ ] the universe of possible plaintiffs, contrary to Congressional intent." The term "claimant," unlike the term "victim," is indeed less bounded by the underlying acts that give the courts jurisdiction: Only a limited set of individuals could properly be considered victims of the 1998 embassy bombings, whereas the term "claimant" may appear to encompass a larger universe of possible plaintiffs. That universe is actually quite limited, however. The FSIA itself limits claimants to those seeking "money damages" "for personal injury or death," 28 U.S.C. § 1605A(a)(1). See La Reunion Aerienne , 533 F.3d at 845 (allowing an insurer to recover payments made to survivors and to estates of those killed in an airline bombing because the insureds' claims were "personal injury claim[s] under traditional common-law principles") (internal quotation marks, emphasis, and citation removed).

Substantive law also limits who is a proper claimant under the FSIA. This is clearly the case with the federal cause of action in the FSIA, which limits claimants to the four enumerated groups and their legal representatives. So too with substantive law outside the FSIA: We have held the common-law tort of IIED limits recovery to the immediate family of a victim who is physically injured or killed. See Bettis v. Islamic Republic of Iran , 315 F.3d 325, 338 (D.C. Cir. 2003) (rejecting claims for IIED brought by nieces and nephews of a U.S. national taken hostage); RESTATEMENT (SECOND) OF TORTS § 46 (1965). Therefore, not every person who experiences emotional distress from a major terrorist attack—a universe that could be large indeed—can state a claim for IIED absent some close relationship to a victim who was injured or killed. Therefore, due to the limitations imposed upon potential claimants both by the FSIA and by substantive law, we are not persuaded by Sudan's argument that the plain meaning of "claimant" produces "absurd results" or is "contrary to Congressional intent."

In sum, by its plain text, the FSIA terrorism exception grants a court jurisdiction to hear a claim brought by a third-party claimant who is not the legal representative of a victim physically injured by a terrorist attack. Who in particular may bring a claim against a foreign sovereign is a question of substantive law, wholly separate from the question of our jurisdiction.

B. Causes of Action

Sudan next contends the foreign family members cannot state a claim under any source of substantive law. Starting from first principles, we reiterate that the question whether a statute withdraws sovereign immunity is "analytically distinct" from whether a plaintiff has a cause of action. See Meyer , 510 U.S. at 484, 114 S.Ct. 996 ; United States v. Mitchell , 463 U.S. 206, 218, 103 S.Ct. 2961, 77 L.Ed.2d 580 (1983). As the district court correctly recognized, we have never required the Congress, in order to effectuate a grant of jurisdiction, expressly to "define the substantive law that applies." Owens V , 174 F.Supp.3d at 286. Indeed, before enactment of the FSIA, the courts—absent objection by the State Department—had jurisdiction to hear suits against a foreign government under state and federal law even though no statute provided rules of decision for such cases. See, e.g. , Victory Transp. Inc. v. Comisaria Gen. de Abastecimientos y Transportes , 336 F.2d 354 (2d Cir. 1964) (enforcing a state-law arbitration agreement against a foreign sovereign via the Federal Arbitration Act). Hence, unless the enactment of the FSIA or of § 1605A somehow changed this situation, a plaintiff proceeding under the FSIA may rely upon alternative sources of substantive law, including state law.

Sudan would have us find an abrogation of a plaintiff's access to state law in § 1606 of the FSIA, which provides in relevant part:

As to any claim for relief with respect to which a foreign state is not entitled to immunity under section 1605 or 1607 of this chapter, the foreign state shall be liable in the same manner and to the same extent as a private individual under like circumstances; but a foreign state except for an agency or instrumentality thereof shall not be liable for punitive damages.

When the original FSIA terrorism exception was in force, § 1606 governed what a claimant could recover from a foreign sovereign. This was because the original exception was codified as a subsection of § 1605, to which § 1606 expressly applied. After we declined in Cicippio-Puleo to infer a federal cause of action against a foreign sovereign arising from § 1605(a)(7) or from the Flatow Amendment, a plaintiff using the old terrorism exception could press a claim under state law, as qualified by § 1606, in the same manner as any other FSIA plaintiff. When the Congress passed the 2008 NDAA, it repealed old § 1605(a)(7) and codified the current terrorism exception in new § 1605A. As a result, § 1606, which references only § 1605 and § 1607, does not apply to the current FSIA terrorism exception. This, Sudan contends, demonstrates the Congress's intent to foreclose a plaintiff from relying upon state law when suing under § 1605A. Essentially, Sudan suggests the Congress struck a deal when it recodified the new terrorism exception in § 1605A : A plaintiff could sue under the new federal cause of action but could no longer press a state-law claim against a foreign sovereign via the pass-through process endorsed by Cicippio-Puleo . Therefore, according to Sudan, plaintiffs who are ineligible for the purportedly exclusive remedy of the federal cause of action—including the foreign family members in this case—were left without a "gateway" to any substantive law under which to state a claim. Contra Leibovitch , 697 F.3d at 572 ("Although § 1605A created a new cause of action, it did not displace a claimant's ability to pursue claims under applicable state or foreign law upon the waiver of sovereign immunity" (quoting Estate of Doe v. Islamic Republic of Iran , 808 F.Supp.2d 1, 20 (D.D.C. 2011) )).

One might wonder, as the plaintiffs do, why we need to reach this nonjurisdictional argument, which Sudan forfeited by failing to appear in the district court. See Practical Concepts , 811 F.2d at 1547. We do so because we have discretion to reach the question, see Acree , 370 F.3d at 58, and this case presents sound reasons for doing so. The question presented is "purely one of law important in the administration of federal justice" because most cases invoking the terrorism exception are filed in this circuit, see 28 U.S.C. § 1391(f)(4), and "resolution of the issue does not depend on any additional facts not considered by the district court." Acree , 370 F.3d at 58 (quoting Roosevelt v. E.I. Du Pont de Nemours & Co. , 958 F.2d 416, 419 n.5 (D.C. Cir. 1992) ). Review is particularly appropriate here because the foreign family member plaintiffs have secured billions in damages against a foreign sovereign. See id. (finding extraordinary circumstances from a "nearly-billion dollar default judgment against a foreign government"). We therefore exercise our discretion to consider Sudan's nonjurisdictional argument that the pass-through approach recognized in Cicippio-Puleo did not survive enactment of § 1605A.

In our view, Sudan assigns undue significance to § 1606. On its face, that section does not authorize a plaintiff to resort to state (or federal or foreign) law in a suit against a foreign sovereign. Nor does it create a substantive body of law for such an action. See First Nat'l City Bank , 462 U.S. at 620-21, 103 S.Ct. 2591. Rather, as the plaintiffs argue and the district court recognized, § 1606 simply limits the liability of a foreign state to "the same manner and to the same extent as a private individual under like circumstances" regardless of what substantive law is being applied. The exclusion of punitive damages from the pass-through approach reinforces our confidence that § 1606 operates only to limit, not to create, the liability of a foreign state. As the Supreme Court has said, the Congress made clear that the FSIA, including § 1606, was not "intended to affect the substantive law of liability" applicable to a foreign sovereign. Id. at 620, 103 S.Ct. 2591 (quoting H.R. REP. NO. 94-1487, at 12 (1976)). In keeping with this straightforward reading, we have recognized that § 1606 does not authorize a court to craft federal common law, but rather requires it to apply state law to suits under the FSIA. See Bettis , 315 F.3d at 333 (noting that § 1606"instructs federal judges to find the relevant law, not to make it").

One might wonder, then, why the Congress moved the FSIA terrorism exception from § 1605, where it was covered by § 1606, to § 1605A, where it is not. Contrary to Sudan's convoluted argument about an implied withdrawal of remedies under state law, the new exception itself provides a ready answer. If the Congress had reenacted the new terrorism exception in the same section as the old one, then it would have created an irreconcilable conflict between the new federal cause of action, which allows the award of punitive damages, and § 1606, which prohibits them. In order to avoid this conflict, a court would have either to disregard a central element of the federal cause of action or to hold the new exception implicitly repealed § 1606 as applied to state sponsors of terror. See Morton v. Mancari , 417 U.S. 535, 549, 94 S.Ct. 2474, 41 L.Ed.2d 290 (1974) (noting the "cardinal rule ... that repeals by implication are not favored") (internal quotation marks removed). Avoiding a conflict between § 1605 and § 1606, rather than Sudan's strained "gateway" argument, more likely explains the Congress's purpose in moving the terrorism exception out of § 1605.

Of course, in most cases brought under the new terrorism exception, the plaintiff need not rely upon state tort law. This does not, however, imply that the Congress intended to foreclose access to state law by those who need it, as do foreign family members. U.S. nationals will continue to sue under § 1605A(c) and benefit from its consistent application. But the pass-through approach remains viable to effectuate the intent of the Congress to secure recoveries for other plaintiffs harmed by a terrorist attack.

C. Intentional Infliction of Emotional Distress

We turn now to Sudan's third and final argument respecting family members who have brought state-law claims for IIED. The district court held that District of Columbia law controls these actions, Owens IV , 826 F.Supp.2d at 157, which Sudan does not contest. Judgments under D.C. law in favor of the foreign family member plaintiffs total more than $7 billion. Sudan contends these awards are invalid because D.C. tort law requires a plaintiff to be present at the scene of a defendant's outrageous and extreme conduct in order to recover for IIED. In particular, Sudan points to Pitt v. District of Columbia , in which this court applied the "presence" requirement to bar a claim for IIED under D.C. law. 491 F.3d 494, 507 (D.C. Cir. 2007).

That case does not extend as far as Sudan contends. In Pitt , we noted "[t]he District of Columbia has adopted the standard for intentional infliction of emotional distress from the Restatement (Second) of Torts." Id. (citing Sere v. Grp. Hospitalization, Inc. , 443 A.2d 33, 37 (D.C. 1982) ). As Sudan points out, the Second Restatement contains a presence requirement:

Where such [extreme and outrageous] conduct is directed at a third person, the actor is subject to liability if he intentionally or recklessly causes severe emotional distress (a) to a member of such person's immediate family who is present at the time, whether or not such distress results in bodily harm, or (b) to any other person who is present at the time, if such distress results in bodily harm."

The Restatement, however, also provides that "there may ... be other circumstances under which the actor may be subject to liability for the intentional or reckless infliction of emotional distress." RESTATEMENT (SECOND) OF TORTS S § 46 (1965) (caveat). A comment to the Restatement expressly applies this caveat to the presence requirement, "leav[ing] open the possibility of situations in which presence at the time may not be required." Id. cmt. l.

Although we did apply the presence requirement in Pitt , the factual situation there was quite different than in the present case. The plaintiff in Pitt alleged emotional distress from the "filing of a false and misleading affidavit and possible evidence tampering." 491 F.3d at 507. Allowing a claim for IIED stemming from a procedural irregularity in law enforcement, we reasoned, would "substantially expand[ ] the scope of the third-party IIED tort under District of Columbia law," id. , without any principled limitation on future actions. In contrast, a massive terrorist attack resulting in widespread casualties and worldwide attention would appear so exceptional that recognizing an appropriate plaintiff's claim for IIED would not broaden the scope of liability to innumerable similar incidents. Therefore, nothing in Pitt suggests D.C. law would apply the presence requirement to an act of international terrorism.

At the same time, we proceed with caution when applying D.C. tort law to this novel situation. The District of Columbia has yet to decide whether it would apply the presence requirement or the exception in the Restatement to an act of international terrorism. Neither has Maryland, the common law of which is authoritative when D.C. law is silent. Clark v. Route , 951 A.2d 757, 763 n.5 (D.C. 2008). Although there are convincing reasons to do so, there are also good reasons to draw back. Some of the first cases applying the caveat in the Restatement dealt with hostage taking. See, e.g. , Stethem , 201 F.Supp.2d at 89-91 ; Sutherland v. Islamic Republic of Iran , 151 F.Supp.2d 27, 50 (D.D.C. 2001). Hostage takers often target the family members of the victim, demanding they pay a ransom for the release of the hostage. The emotional distress of the family member is intended to advance the hostage taker's aims. Therefore, hostage taking seems to be the type of case in which the defendant's extreme and outrageous conduct is "directed at a third person" but is intended also to cause severe emotional distress to the absent plaintiff. See DAN B. DOBBS, THE LAW OF TORTS § 307, at 384 (2000) ("If the defendants' conduct is sufficiently outrageous and intended to inflict severe emotional harm upon a person which [sic] is not present, no essential reason of logic or policy prevents liability"). If so, the plaintiff's contemporaneous physical presence is not required because the plaintiff is the direct target of the tortious conduct, rather than a mere bystander, as the latest version of the Restatement recognizes. See RESTATEMENT (THIRD) OF TORTS: PHYS. & EMOT. HARM § 46 (2012) (cmt. m) ("If an actor harms someone for the purpose of inflicting mental distress on another person, the [presence] limitations ... do not apply").

In contrast, a terrorist bombing is not so precisely targeted at certain absent individuals. Rather than leveraging distress inflicted upon specific third parties to achieve their aims, terrorist bombings typically target the public at large in order to create a general environment of fear and insecurity. Widespread distress, rather than distress "directed at" or confined to particular persons, provides a considerably weaker basis for IIED liability. Indeed, the Second Restatement would preclude an individual's recovery for an event causing widespread emotional distress, absent some unique, foreseeable, and intended harm to the plaintiff. RESTATEMENT (SECOND) OF TORTS § 46 cmt. l. For this reason too, the drafters of the Third Restatement of Torts have criticized several district court decisions for abandoning the presence requirement in FSIA terrorism cases. See RESTATEMENT (THIRD) OF TORTS: PHYS. & EMOT. HARM § 46 (2012) reporter's note cmt. m (criticizing the "questionable determination that the terrorists acts were directed not only to the victims of the attack but also at their family members"). Although we have not decided the matter, we too have expressed skepticism that the sensational nature of a terrorist attack warrants an exception to the limitations of IIED in the Restatement. See Bettis , 315 F.3d at 334 ("If any person that Iran hoped to distress ... could recover under section 46(1) as a direct victim of Iran's conduct, virtually anyone claiming he or she was affected could recover").

We believe a court may reasonably characterize a terrorist bombing as falling either within the caveat in the Second Restatement or beyond the scope of a sovereign's liability to third parties. The plaintiffs once again urge us not to reach this nonjurisdictional question forfeited by Sudan's default, but as with the availability of state law claims, we see sound reasons for exercising our discretion to consider the matter. See Acree , 370 F.3d at 58. Billions of dollars have been awarded to foreign family members as damages for IIED. Furthermore, how to apply the Restatement to terrorist bombings is a question, unfortunately, almost certain to recur in this Circuit. Finally, this is a pure question of law that "does not depend on any additional facts not considered by the district court," Roosevelt , 958 F.2d at 419 & n.5, and potentially may bear upon sensitive matters of international relations. Cf. Acree , 370 F.3d at 58. The situation therefore presents "exceptional circumstances" sufficient to overcome our ordinary reluctance to hear nonjurisdictional arguments not raised before the district court. Id.

That said, the choice is not ours to make. District of Columbia law controls the scope of IIED liability, and the D.C. Court of Appeals has yet to render a decision on the matter. Therefore, we shall certify the question to that court pursuant to D.C. Code Ann. § 11-723. Whether to certify a question "rests in the sound discretion of the federal court." Lehman Bros. v. Schein , 416 U.S. 386, 390-91, 94 S.Ct. 1741, 40 L.Ed.2d 215 (1974). "The most important consideration guiding the exercise of this discretion ... is whether the reviewing court finds itself genuinely uncertain about a question of state law that is vital to a correct disposition of the case before it." Tidler v. Eli Lilly & Co. , 851 F.2d 418, 426 (D.C. Cir. 1988).

This case presents such a question. We are genuinely uncertain whether the D.C. Court of Appeals would apply the presence requirement in the Second Restatement of Torts to preclude recovery for IIED by family members absent from the scene of a terrorist bombing. Other states have reached different conclusions on this question. See Peterson , 515 F.Supp.2d at 43-44 & n.19 (identifying Florida, California, and Vermont as states that apply the presence requirement and Louisiana, and Pennsylvania as states that do not).

Furthermore, the question is one of significant public interest in the District of Columbia. See Eli Lilly & Co. v. Home Ins. Co. , 764 F.2d 876, 884 (D.C. Cir. 1985). Because the great majority of claims under the FSIA terrorist exception are brought in the federal district court in D.C. pursuant to the FSIA venue provision in 28 U.S.C. 1391(f)(4), this question of D.C. tort law will likely arise in future cases before our district court. And the District, as the home of thousands of government employees, military service members, and contractors, and as itself a potential target of terrorist attacks, has a substantial interest in determining who may recover for the emotional distress caused by a terrorist attack.

We therefore certify the following question to the D.C. Court of Appeals:

Must a claimant alleging emotional distress arising from a terrorist attack that killed or injured a family member have been present at the scene of the attack in order to state a claim for intentional infliction of emotional distress?

VI. Punitive Damages

Having affirmed that the district court properly asserted jurisdiction over the plaintiffs' claims and held Sudan liable for their injuries, we now review the amount in damages it awarded to the plaintiffs. The court awarded $10.2 billion in damages, including more than $4.3 billion in punitive damages under both state and federal law. See, e.g. , Opati , 60 F.Supp.3d at 81-82. In post-judgment motions under Rule 60(b)(6), Sudan asked the district court to vacate the awards of punitive damages. The court declined, reasoning that any nonjurisdictional legal error in assessing punitive damages against Sudan did not present an "extraordinary circumstance" that would justify vacatur. Owens V , 174 F.Supp.3d at 288 ; see Gonzalez v. Crosby , 545 U.S. 524, 536, 125 S.Ct. 2641, 162 L.Ed.2d 480 (2005) ("[R]elief under Rule 60(b)(6)... requires a showing of 'extraordinary circumstances' ").

Sudan's renewed request to vacate these awards is now before us both on appeal from the denial of Sudan's Rule 60(b) motions and on direct appeal from the final judgments. Sudan principally contends the FSIA terrorism exception does not retroactively authorize the imposition of punitive damages against a sovereign for conduct occurring before the passage of § 1605A. As explained below, we agree. But before reaching the merits, we first explain why we are addressing the matter despite Sudan's default in the district court.

A. Whether to Review the Awards of Punitive Damages

The plaintiffs contend, and the district court agreed, we need not consider Sudan's argument against the awards of punitive damages because it forfeited this nonjurisdictional challenge by failing to appear in the district court. While this is true, see Practical Concepts , 811 F.2d at 1547, there are sound reasons to exercise our discretion to hear Sudan's argument, whether under Rule 60(b) or on direct appeal.

First, Supreme Court precedent generally favors more searching appellate review of punitive damages than of other nonjurisdictional matters. See Pac. Mut. Life Ins. v. Haslip , 499 U.S. 1, 18, 111 S.Ct. 1032, 113 L.Ed.2d 1 (1991) (warning against "unlimited judicial discretion" in fixing punitive damages). Heightened scrutiny is appropriate because punitive damages are in the nature of criminal punishment. Id. at 19, 111 S.Ct. 1032. Accordingly, the Court has closely reviewed the size of punitive damage awards relative to compensatory damages, State Farm Mut. Auto. Ins. v. Campbell , 538 U.S. 408, 426, 123 S.Ct. 1513, 155 L.Ed.2d 585 (2003), the availability of punitive damages for conduct occurring outside a court's territorial jurisdiction, BMW of N. Am., Inc. v. Gore , 517 U.S. 559, 572, 116 S.Ct. 1589, 134 L.Ed.2d 809 (1996), and the factors a court may consider in imposing punitive damages, Haslip , 499 U.S. at 21-22, 111 S.Ct. 1032. In particular, the Court has emphasized the importance of judicial review to ensure awards of punitive damages comport with the Constitution. Honda Motor Co. v. Oberg , 512 U.S. 415, 432, 114 S.Ct. 2331, 129 L.Ed.2d 336 (1994). Consistent with these concerns, the scope of appellate review for a timely challenge to an award of punitive damages is broad. See Cooper Indus., Inc. v. Leatherman Tool Grp., Inc. , 532 U.S. 424, 436, 121 S.Ct. 1678, 149 L.Ed.2d 674 (2001) (reviewing de novo constitutional challenges to punitive damages). We think the same concerns call for a similarly exacting standard for review of an untimely challenge to an award of punitive damages. Our view is reinforced by the Court's warning that the "[r]etroactive imposition of punitive damages would raise a serious constitutional question." Landgraf v. USI Film Prods ., 511 U.S. 244, 281, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994).

In order to avoid possible constitutional infirmities, other Circuits too have reviewed denials of Rule 60(b)(6) motions to vacate punitive damages awarded in default judgments. See Watkins v. Lundell , 169 F.3d 540, 545 (8th Cir. 1999) ; Merrill Lynch Mortg. Corp. v. Narayan , 908 F.2d 246, 253 (7th Cir. 1990). Although review of punitive damages entered upon default is not always warranted, we think the circumstances of this case merit appellate review. Of particular note are the size of the awards (totaling $4.3 billion), the presentation of a novel question of constitutional law (retroactivity), and the potential effect on U.S. diplomacy and foreign relations. We believe these factors present the "extraordinary circumstances" needed for review under Rule 60(b)(6). This issue also comes before the court on direct appeal from the default judgments. As previously mentioned, we may consider nonjurisdictional questions not raised by the parties on direct appeal in "exceptional circumstances." Acree , 370 F.3d at 58. Our discretion is properly exercised over pure questions of law—such as the retroactivity of punitive damages—that need no further factual development. Roosevelt , 958 F.2d at 419 & n. 5. Direct review of forfeited arguments is also warranted for questions that bear upon sensitive matters of international relations. Acree , 370 F.3d at 58 (finding exceptional circumstances from a "nearly-billion dollar default judgment against a foreign government"). Furthermore, because most cases invoking the FSIA exception for terrorism are brought in this district, our decision on retroactivity will provide useful guidance to the district court. Compare Owens V , 174 F.Supp.3d at 291 (doubting whether punitive damages apply retroactively but declining to vacate award) with Flanagan v. Islamic Republic of Iran , 190 F.Supp.3d 138, 182 (D.D.C. 2016) (vacating punitive damages despite the defendant's default) and Kumar v. Republic of Sudan , No. 2:10-cv-171, at 39 n.17 (E.D. Va. Oct. 25, 2016) (approving retroactive assessment of punitive damages); see also Leatherman , 532 U.S. at 436, 121 S.Ct. 1678 (noting that "[i]ndependent review [of punitive damages] is ... necessary if appellate courts are to maintain control of, and to clarify, the legal principles"). Given the size of the awards, the strength of Sudan's contentions, and the likelihood of this question recurring, we believe reviewing the award of punitive damages both promotes "the interests of justice" and "advance[s] efficient judicial administration." City of Newport , 453 U.S. at 257, 101 S.Ct. 2748. We therefore exercise our discretion to consider Sudan's belated objections.

B. Retroactivity of Punitive Damages Under § 1605A(c)

In challenging the punitive damage awards, Sudan raises the "presumption against retroactive legislation" explicated in Landgraf v. USI Film Products , 511 U.S. 244, 265, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994). Courts "have declined to give retroactive effect to statutes burdening private rights unless Congress had made clear its intent." Id. at 270, 114 S.Ct. 1483. This presumption avoids "the unfairness of imposing new burdens on persons after the fact," absent a clear signal of congressional intent to do so. Id. The Court in Landgraf noted the retroactive authorization of punitive damages, in particular, "would raise a serious constitutional question." Id. at 281, 114 S.Ct. 1483.

An analysis of retroactivity entails two steps. First, the court must determine "whether Congress has expressly prescribed the statute's proper reach." Id. at 280, 114 S.Ct. 1483. If the Congress has clearly spoken, then "there is no need to resort to judicial default rules," and the court must apply the statute as written. Id. When "the statute contains no such express command," the court must then evaluate whether the legislation "operate[s] retroactively," as it does if 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 operates retroactively but lacks a clear statement of congressional intent to give it retroactive effect, then the Landgraf presumption controls and the court will not apply the statute to pre-enactment conduct. Sudan argues both that the new FSIA terrorism exception does not contain a clear statement of retroactive effect and that it operates retroactively.

1. Section 1605A operates retroactively

As for the latter point, it is obvious that the imposition of punitive damages under the new federal cause of action in § 1605A(c) operates retroactively because it increases Sudan's liability for past conduct. Under § 1605(a)(7), the predecessor to the current terrorism exception, and the pass-through approach recognized in Cicippio-Puleo , § 1606 expressly barred courts from awarding punitive damages against a foreign sovereign. The 2008 NDAA plainly applies the new cause of action in § 1605A(c) to the pre-enactment conduct of a foreign sovereign. Further, recall that, pursuant to NDAA § 1083(c), a plaintiff may convert a pending, prior action under § 1605(a)(7) into a new action under § 1605A(c) or file a new suit arising from the same act or incident as an action "related" to an original suit timely filed under § 1605(a)(7). In both cases, the new actions under § 1605A(c) necessarily are based upon the sovereign defendant's conduct before enactment of § 1605A.

The plaintiffs dispute this, relying upon Republic of Austria v. Altmann , 541 U.S. 677, 124 S.Ct. 2240, 159 L.Ed.2d 1 (2004), in which the Supreme Court held the jurisdictional provisions of the FSIA apply to conduct occurring prior to its enactment notwithstanding the absence of a clear statement to that effect in the statute. Id. at 692-96, 700, 124 S.Ct. 2240. That jurisdiction under the FSIA applies retroactively, however, has no bearing upon the question whether the authorization of punitive damages does as well.

Unlike the grant of jurisdiction held retroactive in Altmann , the authorization of punitive damages "adheres to the cause of action" under § 1605A(c), making it "essentially substantive" and thereby triggering retroactive operation. Id. at 695 n.15, 124 S.Ct. 2240 ; cf. Landgraf , 511 U.S. at 274, 114 S.Ct. 1483 ("Application of a new jurisdictional rule usually takes away no substantive right," causing it not to operate retroactively) (internal quotation marks omitted). Furthermore, while the original FSIA codified only the preexisting "restrictive theory" of foreign sovereign immunity, leaving the scope of a sovereign's potential liability unchanged, see Altmann , 541 U.S. at 694, 124 S.Ct. 2240, the new terrorism exception authorizes a quantum of liability—punitive damages—to which foreign sovereigns were previously immune.

Having failed to distinguish the FSIA terrorism exception from the Supreme Court's core concerns in Landgraf , the plaintiffs advance a policy argument transplanted from Altmann . There the Court explained the "aim of the presumption [against retroactivity] is to avoid unnecessary post hoc changes to legal rules on which parties relied in shaping their primary conduct." 541 U.S. at 696, 124 S.Ct. 2240. In contrast, the plaintiffs urge "the principal purpose of foreign sovereign immunity ... reflects current political realities and relationships, and aims to give foreign states and their instrumentalities some present 'protection from the inconvenience of suit as a gesture of comity.' " Id. (quoting Dole Food Co. v. Patrickson , 538 U.S. 468, 479, 123 S.Ct. 1655, 155 L.Ed.2d 643 (2003) ). Because the Congress was motivated by these "sui generis " concerns of comity in initially passing the FSIA, id. , the plaintiffs contend the presumption in Landgraf should not apply to a subsequent FSIA amendment, even if it appears to operate retroactively.That argument misses the central point of authorizing punitive damages against a state sponsor of terrorism, viz ., to deter terrorism. By its nature, deterrence attempts to influence foreign sovereigns in "shaping their primary conduct." Id. And when the law affects a defendant's past actions, "[e]lementary considerations of fairness dictate that individuals should have an opportunity to know what the law is and to conform their conduct accordingly." Landgraf , 511 U.S. at 265, 114 S.Ct. 1483.

This principle applies equally to state sponsors of terrorism. As the Supreme Court has said, "[e]ven when the conduct in question is morally reprehensible or illegal, a degree of unfairness is inherent whenever the law imposes additional burdens based on conduct that occurred in the past." Id. at 282 n.35, 114 S.Ct. 1483. Therefore, without a clear statement of retroactivity, courts have properly declined to apply statutes authorizing an award of punitive damages, even for outrageous conduct. See, e.g. , Ditullio v. Boehm , 662 F.3d 1091, 1100 (9th Cir. 2011) (holding that punitive damages under the Trafficking Victims Protection Act are unavailable to punish child sex trafficking that occurred before enactment); Gross v. Weber , 186 F.3d 1089, 1091 (8th Cir. 1999) (holding the same for the Violence Against Women Act as applied to pre-enactment sexual abuse). Hence, unlike the grant of jurisdiction in Altmann , the authorization of punitive damages in § 1605A(c) cannot be dismissed as a reflection of "current political realities and relationships" but rather goes to the heart of the concern in Landgraf about retroactively penalizing past conduct.

2. Clear statement of retroactive effect

Having concluded that § 1605A(c) operates retroactively, the next question is whether the Congress has made a clear statement authorizing punitive damages for past conduct. We will find that authorization only if the statute is "so clear that it could sustain only one interpretation." See Lindh v. Murphy , 521 U.S. 320, 328 n.4, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997). With this in mind, we agree with the district court that the FSIA contains no such statement. Owens V , 174 F.Supp.3d at 289.

As a starting point, we look for a clear statement in § 1605A(c), which provides that a designated state sponsor of terrorism:

shall be liable ... for personal injury or death caused by acts described in subsection (a) (1) of that foreign state, or of an official, employee, or agent of that foreign state, for which the courts of the United States may maintain jurisdiction under this section for money damages. In any such action, damages may include economic damages, solatium, pain and suffering, and punitive damages. In any such action, a foreign state shall be vicariously liable for the acts of its officials, employees, or agents.

On its face, nothing in the text of § 1605A(c) speaks to whether punitive damages are available under the federal cause of action for pre-enactment conduct. Nor does precedent provide support for retroactivity. Although Altmann held the grant of jurisdiction in § 1605(a) applies retroactively (despite lack of a clear statement to that effect), the authorization of punitive damages under the current terrorism exception lies in the cause of action under § 1605A(c), not in the grant of jurisdiction under § 1605A(a).

The plaintiffs contend that § 1083(c) of the 2008 NDAA, when combined with the authorization of punitive damages in § 1605A(c), provides a clear statement of retroactive effect. As we have seen, supra part IV, both a converted prior action under § 1083(c)(2) and a related action under § 1083(c)(3) necessarily arise out of conduct that occurred before the enactment of the 2008 NDAA, and both provisions allow a plaintiff to proceed under the federal cause of action in § 1605A(c), which authorizes punitive damages. Accordingly, the plaintiffs contend, both § 1083(c)(2) and (c)(3), when read in conjunction with § 1605A(c), clearly allow a court to award punitive damages under the federal cause of action for pre-enactment conduct.

This argument takes one too many a logical leap. Yes, by allowing a plaintiff to convert an action brought under § 1605(a)(7), § 1083(c)(2) clearly authorizes the federal cause of action to apply retroactively. This, however, does not mean that § 1083(c) authorizes the punitive damages in § 1605A(c) to apply retroactively as well. Cf. Roeder v. Islamic Republic of Iran , 646 F.3d 56, 61-62 (D.C. Cir. 2011) (finding no clear statement that § 1083(c)(3) abrogated the Algiers Accords simply by allowing plaintiffs to bring actions under § 1605A related to those formerly dismissed by reason of the Accords). Instead, § 1083(c) operates as a conduit for a plaintiff to access the cause of action under § 1605A(c). If punitive damages under § 1605A(c) were not available retroactively to any plaintiff (including those who did not make use of § 1083(c)), then nothing in § 1083(c) would change that. Inversely, if § 1083(c) did not exist, then one plaintiff's inability to convert his pending case or to bring a related action under § 1083(c) would not detract from the retroactive availability of punitive damages for another plaintiff if such relief were clearly authorized by the Congress. At most, Sudan has identified § 1083(c) as a plausible mechanism through which the Congress could have authorized punitive damages for past conduct. But Landgraf demands more, and no clear statement emerges from the union of § 1083(c) and § 1605A(c).

There being no clear textual command, the plaintiffs urge that the purpose of § 1083(c) supplies the necessary clear statement of congressional intent. An argument based solely upon the purpose of a statute can hardly supply a "clear statement" of any sort. See Bowen v. Georgetown Univ. Hosp ., 488 U.S. 204, 208, 109 S.Ct. 468, 102 L.Ed.2d 493 (1988) ("congressional enactments and administrative rules will not be construed to have retroactive effect unless their language requires this result"). Because an expansion of punitive damages would operate retroactively by "increas[ing] [Sudan's] liability for past conduct," the presumption in Landgraf applies and bars an award of punitive damages for the embassy bombings, which occurred before the enactment of the 2008 NDAA. Therefore, we vacate the award of punitive damages to plaintiffs proceeding under the federal cause of action.

C. Retroactivity of Punitive Damages Under State Law

The same principle applies to the awards of punitive damages to plaintiffs proceeding under state law. Sudan makes two arguments against the availability of punitive damages for them. Sudan first contends that § 1605A(c) provides the sole source for seeking punitive damages against a foreign sovereign. Sudan rests this view upon § 1606 of the FSIA, which precludes punitive damages against a sovereign defendant. As we have recognized, supra p. 808, § 1606, by its terms, applies only to claims brought under § 1605 and § 1607 of the FSIA. Owens V , 174 F.Supp.3d at 290. Section 1606 therefore has no bearing upon state law claims brought under the jurisdictional grant in § 1605A.

If this were the end of the analysis, however, a puzzling outcome would arise from our holding that punitive damages are not available retroactively to plaintiffs proceeding under the federal cause of action in § 1605A(c). As we have said, in creating a federal cause of action, the Congress sought to end the inconsistencies in the "patchwork" pass-through approach of Cicippio-Puleo . See Leibovitch , 697 F.3d at 567. Allowing punitive damages for pre-enactment conduct under state but not federal law would frustrate this intent: Plaintiffs otherwise eligible for the federal cause of action, for which punitive damages are unavailable, would instead press state law claims for punitive damages, which would effectively perpetuate the inconsistent outcomes based upon differences in state law that the Congress sought to end by passing § 1605A.

As it happens, the retroactive authorization of punitive damages under state law fails for the same reason it does under the federal cause of action: The authorization of § 1605A, read together with § 1606, lacks a clear statement of retroactive effect. Without the Landgraf presumption, the enactment of § 1605A would have lifted the restriction on punitive damages in § 1606 from state law claims. If the express authorization of punitive damages under § 1605A(c) lacks a clear statement of retroactive effect, then the implicit, backdoor lifting of the prohibition against punitive damages in § 1606 for state law claims fares no better. Cf. Landgraf , 511 U.S. at 259-60, 114 S.Ct. 1483 (finding that cross-references between several sections of the Civil Rights Act did not impliedly make a clear statement of retroactive effect). As a result, a plaintiff proceeding under either state or federal law cannot recover punitive damages for conduct occurring prior to the enactment of § 1605A. Accordingly we vacate all the awards of punitive damages.

VII. Vacatur Under Rule 60(b)

Finally, Sudan argues the district court abused its discretion in denying its motions to vacate the default judgments, invoking three sections of the Rule 60(b) : the judgments are void for lack of subject matter jurisdiction per § (b)(4); default was due to "excusable neglect" per § (b)(1); and relief may be justified for "any other reason" per § (b)(6). The first jurisdictional ground is nondiscretionary, Bell Helicopter , 734 F.3d at 1179, and has been rejected already in the sections on extrajudicial killing, jurisdictional causation, and the ability of family members of a victim physically injured by the bombings to press a claim under § 1605A.

We review the district court's decision to deny vacatur on the other two grounds for abuse of discretion. Gonzalez , 545 U.S. at 535, 125 S.Ct. 2641 (" Rule 60(b) proceedings are subject to only limited and deferential appellate review"). In doing so, we recognize "the district judge, who is in the best position to discern and assess all the facts, is vested with a large measure of discretion in deciding whether to grant a Rule 60(b) motion." Twelve John Does v. District of Columbia , 841 F.2d 1133, 1138 (D.C. Cir. 1988). Deferential review preserves the "delicate balance between the sanctity of final judgments ... and the incessant command of a court's conscience that justice be done in light of all the facts." Good Luck Nursing Home, Inc. v. Harris , 636 F.2d 572, 577 (D.C. Cir. 1980) (emphasis and internal quotation marks removed). With respect to Rule 60(b)(1), relief for excusable neglect "is rare" as "such motions allow district courts to correct only limited types of substantive errors," Hall v. CIA , 437 F.3d 94, 99 (D.C. Cir. 2006), and relief for "any other reason" under Rule 60(b)(6) is even more rare, being available only in "extraordinary circumstances," Ackermann v. United States , 340 U.S. 193, 199, 71 S.Ct. 209, 95 L.Ed. 207 (1950). Factual determinations supporting the district court's decision are, of course, reviewed only for clear error. Gates v. Syrian Arab Republic , 646 F.3d 1, 4 (D.C. Cir. 2011).

Sudan, as "the party seeking to invoke Rule 60(b)," bears "the burden of establishing that its prerequisites are satisfied." Id. at 5 (internal alterations and quotation marks removed). As we have said before, "no principle of sovereign immunity law upsets the parties' respective burdens under Rule 60(b) ; nor do oft cited ephemeral principles of fairness" demand a different result for a foreign sovereign than for a private litigant. Id. In order to secure vacatur, therefore, Sudan must show the district court, in denying its motion for relief, relied upon an incorrect understanding of the law or a clearly erroneous fact. Sudan has not met this burden.

A. Excusable Neglect Under Rule 60(b)(1)

We begin with Sudan's claim of excusable neglect, which the district court addressed in detail. In evaluating a claim of excusable neglect, a court makes an equitable determination based upon "the danger of prejudice to the [non-moving party], the length of the delay and its potential impact on judicial proceedings, the reason for the delay, including whether it was within the reasonable control of the movant, and whether the movant acted in good faith." Pioneer Inv. Servs. Co. v. Brunswick Assocs. , 507 U.S. 380, 395, 113 S.Ct. 1489, 123 L.Ed.2d 74 (1993). Additionally, a party seeking vacatur must "assert a potentially meritorious defense." FG Hemisphere Assocs., LLC v. Democratic Republic of Congo , 447 F.3d 835, 842 (D.C. Cir. 2006).

In its motion, Sudan submitted a three-page declaration from Maowia Khalid, the Ambassador of Sudan to the United States, explaining its failure to participate in much of the litigation. First, the Ambassador asserted Sudan's ongoing domestic problems, including natural disasters and civil war, rendered it unable to appear. Khalid Decl. ¶ 4. Second, the Ambassador said a "fundamental lack of understanding in Sudan about the litigation process in the United States" accounted its prolonged absence from the litigation. Id. ¶ 5. The district court soundly rejected both reasons. On Sudan's domestic troubles, the district court noted that "[s]ome of that turmoil ... has been of the Sudanese government's own making," but, regardless of blame, Sudan could not excuse at least six years of nonparticipation without sending a single communication to the court. Owens V , 174 F.Supp.3d at 255. The court further doubted the credibility of Sudan's alleged ignorance of U.S. legal procedure. After all, Sudan had used this excuse to escape an earlier default in the same litigation, and the "fundamental-ignorance card cannot convincingly be played a second time." Id. at 256.

Although the district court, in denying Sudan's Rule 60(b) motion, addressed all the elements of "excusable neglect" mentioned in Pioneer , on appeal Sudan challenges only the "reason for the delay" and the "length of the delay." The district court's unchallenged finding that "vacatur would pose a real risk of prejudice to the plaintiffs," Owens V , 174 F.Supp.3d at 257, makes it difficult to imagine Sudan could prevail even if it were to succeed on the two elements it does raise, Pioneer , 507 U.S. at 397, 113 S.Ct. 1489 (affirming a holding of excusable neglect when the "petitioner does not challenge the findings made below concerning ... the absence of any danger of prejudice" to him), but we consider its arguments nonetheless.

Preliminarily, Sudan also contends the district court "ignored" the "policy favoring vacatur under Rule 60(b)" as it applies to a foreign sovereign. Sudan then claims error in the district court purportedly blaming Sudan for the circumstances that prompted its default. Finally, Sudan faults the district court's comparison of the instant case to FG Hemisphere , in which this court vacated a default judgment against the Democratic Republic of Congo (DRC).

On the first point, Sudan correctly notes that precedent in this Circuit supports a liberal application of Rule 60(b)(1) to default judgments. See Jackson v. Beech , 636 F.2d 831, 836 (D.C. Cir. 1980). This stems from the general policy favoring adjudication on the merits. Id. ; Foman v. Davis , 371 U.S. 178, 181-82, 83 S.Ct. 227, 9 L.Ed.2d 222 (1962). The policy has particular force with respect to a defaulting sovereign because "[i]ntolerant adherence to default judgments against foreign states could adversely affect this nation's relations with other nations and undermine the State Department's continuing efforts to encourage foreign sovereigns generally to resolve disputes within the United States' legal framework." FG Hemisphere , 447 F.3d at 838-39 (quoting Practical Concepts , 811 F.2d at 1551 n.19 ). Further, we have noted, "[w]hen a defendant foreign state has appeared and asserts legal defenses, albeit after a default judgment has been entered, it is important ..., if possible, that the dispute be resolved on the basis of [ ] all relevant legal arguments." Practical Concepts , 811 F.2d at 1552.

For these reasons, the U.S. Government on many occasions has submitted an amicus brief urging vacatur of a default judgment against a foreign sovereign. See, e.g ., id. ; FG Hemisphere , 447 F.3d at 838 ; Gregorian v. Izvestia , 871 F.2d 1515, 1518 (9th Cir. 1989) ; Jackson v. People's Republic of China , 794 F.2d 1490, 1495 (11th Cir. 1986). In this case, however, we think it significant that the Government has not taken a position on Sudan's motion to vacate. Indeed, with only two factually unique exceptions, see Beaty , 556 U.S. at 855, 129 S.Ct. 2183 and Roeder , 646 F.3d at 56, the Government has not weighed in on behalf of a defendant state sponsor of terrorism. Cf. Doe v. Exxon Mobil Corp. , 473 F.3d 345, 360 (D.C. Cir. 2007) (noting that "courts give deference ... when the Executive reasonably explains that adjudication of a particular civil lawsuit would adversely affect the foreign policy interests of the United States").

Absent an expressed governmental concern with the liability of a foreign sovereign, the general policy favoring vacatur, by itself, cannot control the resolution of Sudan's Rule 60(b) motion. After all, the FSIA expressly authorizes default judgments against absent sovereigns. See 28 U.S.C. § 1608(e). If policy considerations alone made vacatur of judgments against foreign sovereigns under Rule 60(b) near-automatic, then the general policy favoring vacatur would render the specific authorization of default judgments in the FSIA a nullity. A district court would abuse its discretion if it were simply to apply the general policy, as Sudan asks us to do now, without considering the specific facts at hand. See FG Hemisphere , 447 F.3d at 838-42 (noting the general policy opposing vacatur but considering the Pioneer factors). Considering those facts, we see why the district court said that "shouldering [Sudan's] burden is a Herculean task." Owens V , 174 F.Supp.3d at 254. Indeed, if we were to vacate the default judgment in this case, then we could not expect any sovereign to participate in litigation rather than wait for a default judgment, move to vacate it under Rule 60(b), appeal if necessary, and then reenter the litigation to contest the merits, having long delayed its day of reckoning. Cf. H. F. Livermore Corp. v. Aktiengesellschaft Gebruder Loepfe , 432 F.2d 689, 691 (D.C. Cir. 1970) (approving of default judgments "when the adversary process has been halted because of an essentially unresponsive party" in which case "the diligent party must be protected lest he be faced with interminable delay and continued uncertainty as to his rights").

Sudan's own actions place it well outside the general policy favoring vacatur. In the cases it cites, relief was justified because the defendant had no notice of the default and promptly responded once made aware of the judgment. See Bridoux v. E. Air Lines , 214 F.2d 207, 209 (D.C. Cir. 1954) ; FG Hemisphere , 447 F.3d at 839. In contrast, Sudan knew of the Owens action, twice obtained sophisticated legal counsel in 2004, and fully participated in the litigation before absenting itself in 2005. In another case involving a foreign sovereign, there was no abuse of discretion in denying vacatur because the defendant had "received actual or constructive notice of the filing of the action and failed to answer" or to provide a good-faith reason for its unresponsiveness. See Meadows v. Dominican Republic , 817 F.2d 517, 521 (9th Cir. 1987). Moreover, unlike the foreign sovereigns in some cases vacating default judgments, see, e.g. , Gregorian , 871 F.2d at 1525 ; Jackson , 794 F.2d at 1495-96, Sudan cannot claim to have defaulted in the reasonable belief that it enjoyed sovereign immunity. Several decisions of the district court and this court served on Sudan suggested the evidence proffered by the Owens plaintiffs could meet or met their burden of production to establish the jurisdiction of the court. Even when served with the district court's 2011 opinion on liability, which definitively established Sudan's lack of immunity, Sudan let three years pass before filing its motion to vacate. For these reasons, Sudan's lack of diligence in pursuing its Rule 60(b) motion weighs heavily against vacatur. Cf. Reinsurance Co. of Am. v. Administratia Asigurarilor de Stat , 902 F.2d 1275, 1276, 1278 (7th Cir. 1990) (affirming denial of Rule 60(b) motion made by a state-owned insurance company for failure to "demonstrate the diligence necessary" to vacate a default judgment).

Furthermore, this is not the first time Sudan has sought to vacate its default or default judgment. In May 2003 the district court entered a default against Sudan for failure to appear. Ten months later, Sudan secured counsel and moved for vacatur under Rule 55(c), which the court granted based upon the very "presumption against an entry of default judgment against a foreign state" that Sudan claims the court ignored in 2016. Owens I , 374 F.Supp.2d at 9, 10 n.5. But the presumption against a default judgment is just that—a presumption. The rationale for leniency is necessarily weaker when a defendant seeks to excuse its second default. See Flanagan , 190 F.Supp.3d at 158 (noting, as well, Sudan's prior default in Rux v. Republic of Sudan , No. 2:04-cv-0428, 2005 WL 2086202, at *2-3, *12-13 (E.D. Va. Aug. 26, 2005) ). A double-defaulting sovereign also loses the ability to assert certain "reasons for the delay," including ignorance of the law and a reasonable belief in its own immunity. It is still more difficult to show "good faith" by a defendant that has walked away a second time without so much as a fare thee well. Hence, the general policy favoring relief from default judgments is not enough to overcome Sudan's double default in this case. Finally, it bears mentioning that the district court and now this court have afforded Sudan, as a foreign sovereign, substantial protection against the harsh consequences of a default judgment. Notwithstanding Sudan's failure to participate, the district court assessed whether the plaintiffs' evidence was satisfactory, once to prevail on the merits and twice to establish jurisdiction. See Owens IV , 826 F.Supp.2d at 139-46 (applying 28 U.S.C. § 1608(e) ); Owens V , 174 F.Supp.3d at 275-80. Furthermore, the district court (and now this court de novo) reviewed Sudan's jurisdictional arguments pursuant to its Rule 60(b)(4) motion. We have also exercised our discretion to consider several of Sudan's nonjurisdictional objections, even though Sudan forfeited these arguments by defaulting. We even granted Sudan relief from punitive damages despite its failure timely to object to these awards in the district court. Therefore, Sudan cannot complain "the dispute [has not been] resolved on the basis of ... all relevant legal arguments." See Practical Concepts , 811 F.2d at 1552.

Beyond relying upon the general policy in favor of vacatur, Sudan challenges the reasoning behind the district court's decision. In particular, Sudan faults the district court for holding it responsible for its domestic troubles, contending a court may not consider "the question of blame" in analyzing excusable neglect. Sudan is twice wrong. Not only have courts consistently recognized that a defendant's "culpable conduct" may justify denying it relief under Rule 60(b)(1), see Mfrs.' Indus. Relations Ass'n v. E. Akron Casting Co. , 58 F.3d 204, 206 (6th Cir. 1995) (inquiring "[w]hether culpable conduct of the defendant led to the default"); Gregorian , 871 F.2d at 1523 ; Info. Sys. & Networks Corp. v. United States , 994 F.2d 792, 795 (Fed. Cir. 1993), but the district court expressly based its decision upon Sudan's unresponsiveness, not its blameworthiness; "setting aside the question of blame," it said:

Domestic turmoil would surely have justified requests by Sudan for extensions of time in which to respond to the plaintiffs' filings. It would have also probably led the Court to forgive late filings. And perhaps it would have even justified a blanket stay of these cases. But Sudan was not merely a haphazard, inconsistent, or sluggish litigant during the years in question—it was a complete and utter nonlitigant. Sudan never sought additional time or to pause any of these cases in light of troubles at home. Sudan never even advised the Court of those troubles at the time they were allegedly preventing Sudan's participation—not through formal filings, and not through any letters or other mode of communication with the Court. The idea that the relevant Sudanese officials could not find the opportunity over a period of years to send so much as a single letter or email communicating Sudan's desire but inability to participate in these cases is, quite literally, incredible.

Owens V , 174 F.Supp.3d at 256. Therefore, we find no abuse of discretion in the district court's brief reference to the Sudan's possible responsibility for its domestic turmoil.

Sudan also objects to the district court's discussion of its unresponsiveness, arguing the court demonstrated "a lack of appreciation of the operational realities of a least developed nation in turmoil." But the one conclusory paragraph in the three-page declaration of its Ambassador to the United States that Sudan cites as evidence for this proposition does not show it was incapable of maintaining any communication with the district court. Indeed, Sudan participated in the litigation during its civil war and while negotiating a peace treaty bringing that war to a close. See UNMIS Background , UNITED NATIONS MISSION IN THE SUDAN , http://www.un.org/en/peacekeeping/missions/past/unmis/background.shtml (last visited July 19, 2017). This shows Sudan could participate in legal proceedings despite difficult domestic circumstances. Without record evidence supporting Sudan's complete inability to participate, the district court did not abuse its discretion in holding Sudan failed to carry its burden of proving excusable neglect.

As a final argument under Rule 60(b)(1), Sudan faults the district court's comparison of this case to FG Hemisphere . In FG Hemisphere we vacated a default judgment against the Democratic Republic of Congo (DRC) rendered under the FSIA exception for commercial activity, § 1605(a)(2). 447 F.3d at 843. Sudan's reliance upon FG Hemisphere is unsurprising as there we noted the DRC "was plainly hampered by its devastating civil war" which justified, in part, its delayed response. Id. at 841. But the outcome in FG Hemisphere did not turn solely, or even primarily, upon the domestic turmoil in the DRC. Problems with notice and service, not internal strife, principally excused the DRC's default. In that case, the defendant sovereign was first notified that its diplomatic properties were in jeopardy when it was served with a motion to execute a default judgment a mere six days before a response was due. Id. at 839-40. The plaintiffs' failure to translate the motion from English into French, the official language of the DRC, "virtually guaranteed the DRC's inability to file a timely response." Id. That the DRC was then engaged in a "devastating civil war" merely diminished its "capacity ... for [the] swift and efficient handling of ... English-language materials"; it did not ultimately prevent the DRC from responding to the motion, which it did shortly after receipt. Id. at 840-41.

Unlike the DRC in FG Hemisphere , Sudan had notice of the litigation from the time it was first sued. The district court's 2011 opinion on liability was translated into Arabic, Sudan's national language, and delivered through diplomatic channels. Sudan cannot, and does not, complain about defects in notice or service of process. See Owens V , 174 F.Supp.3d at 255 (noting that "Sudan's council conceded, 'there's no dispute about service being proper' ").

Nor can Sudan claim to be surprised by the suits, as was the defendant in FG Hemisphere . Sudan actively participated in the litigation from February 2004 until January 2005. Even after disengaging from the case, Sudan contacted its counsel for a status update in September 2008. If Sudan indeed needed to divert "all [its] meager legal and diplomatic personnel" to the "cession of south Sudan," as its Ambassador now suggests, then it could have communicated this affirmative decision to the court, along with a request to stay the proceedings. In light of this history, it was not unreasonable for the district court to demand something more than a conclusory assertion without virtually any record evidence of Sudan's inability to participate in the litigation.Also, as the district court noted, the length of delay in FG Hemisphere pales in comparison to Sudan's absence in this case. The DRC initiated efforts to secure counsel within one day of receiving notice of the motion to execute. 447 F.3d at 838. Within two months, its counsel filed motions to vacate the default judgment and to stay its execution. Id. In contrast, Sudan filed its motions to vacate the judgments 17 months after service of the complaint in Opati , the last of the consolidated cases, 40 months after the district court's 2011 opinion on liability, and 53 months after the evidentiary hearing that Sudan did not attend. Indeed, Sudan ceased regular communication with counsel in the Owens action nearly eight years before filing its present motions. Cf. Smith v. District of Columbia , 430 F.3d 450, 456 n.5 (D.C. Cir. 2005) (noting that delay of "well over a year" militated against excusable neglect). By defaulting, then appearing, then defaulting again, Sudan delayed this case for years beyond its likely end had it simply failed to appear at all. These affirmative actions extended the delay and make Sudan's second default even less excusable than its first. We therefore find no error in the district court's unfavorable comparison of Sudan's default to that of the DRC in FG Hemisphere . In sum, none of Sudan's arguments shows the district court abused its discretion in failing to vacate the default judgments for "excusable neglect."

B. Extraordinary Circumstances Under Rule 60(b)(6)

Sudan also challenges the district court's denial of its motion under Rule 60(b)(6), claiming its failure to appear was justified by "extraordinary circumstances." Because Rule 60(b)(1) contains a one-year filing deadline for claims of "excusable neglect," which Sudan missed with respect to the Mwila and Khaliq judgments, Sudan's Rule 60(b)(6) motions are the only way it may obtain vacatur of those default judgments.

Perhaps recognizing this, Sudan rephrased its earlier arguments asserting "excusable neglect" as requests for relief from those default judgments under Rule 60(b)(6). As with the other cases, the declaration of Ambassador Khalid figures prominently in Sudan's Mwila and Khaliq motions. This gets Sudan nowhere. In order to receive relief under Rule 60(b)(6), a party must show "extraordinary circumstances" justifying vacatur. Gonzalez , 545 U.S. at 534, 125 S.Ct. 2641. As the Supreme Court has explained, the grounds for vacatur under Rule 60(b)(1) and(b)(6) are "mutually exclusive." Pioneer , 507 U.S. at 393, 113 S.Ct. 1489. Therefore, "a party who failed to take timely action due to 'excusable neglect' may not seek relief more than a year after the judgment by resorting to subsection (6)." Id.

The district court acknowledged this distinction and denied Sudan's motion under Rule 60(b)(6) as merely a "rehash of Sudan's Rule 60(b)(1) argument for excusable neglect." Owens V , 174 F.Supp.3d at 258. Instead of grappling with the district court's actual decision, Sudan takes issue with the court's reference to Ungar v. Palestine Liberation Organization , 599 F.3d 79 (1st Cir. 2010), in which the First Circuit held that a sovereign's willful default did not per se preclude vacatur. Id. at 86-87. The district court was understandably puzzled by Sudan's fleeting reference to Ungar in light of its assertions that its default was involuntary. If Sudan's default was intentional, as in Ungar , the court noted, then relief under Rule 60(b)(1) would be unavailable. Owens V , 174 F.Supp.3d at 258. But these musings were not the basis of the district court's decision and therefore cannot be an abuse of discretion.

Undeterred, Sudan now argues Ungar demands vacatur when there would be "political ramifications[ ] and [a] potential effect on international relations" from a default judgment, as Sudan claims there would be in this case. Ungar , 599 F.3d at 86-87. In its view, these political considerations supply the "extraordinary circumstances" needed to vacate a default judgment under Rule 60(b)(6). Sudan failed to raise this argument before the district court, and it is therefore forfeit on appeal. Accordingly, we affirm the district court's denial of vacatur under Rule 60(b).

* * * * *

To conclude, we (1) affirm the district court's findings of jurisdiction with respect to all plaintiffs and all claims; (2) affirm the district court's denial of vacatur; (3) vacate all awards of punitive damages; and (4) certify a question of state law—whether a plaintiff must be present at the scene of a terrorist bombing in order to recover for IIED—to the District of Columbia Court of Appeals.

So ordered .