Estate of Klieman by and Through Kesner v. Palestinian Auth. ( 2019 )


Menu:
  •  United States Court of Appeals
    FOR THE DISTRICT OF COLUMBIA CIRCUIT
    Argued December 13, 2018              Decided May 14, 2019
    No. 15-7034
    ESTATE OF ESTHER KLIEMAN, BY AND THROUGH ITS
    ADMINISTRATOR, AARON KESNER, ET AL.,
    APPELLANTS
    v.
    PALESTINIAN AUTHORITY, ALSO KNOWN AS PALESTINIAN
    INTERIM SELF-GOVERNMENT AUTHORITY AND PALESTINIAN
    LIBERATION ORGANIZATION, ALSO KNOWN AS PLO,
    APPELLEES
    Appeal from the United States District Court
    for the District of Columbia
    (No. 1:04-cv-01173)
    Edward B. MacAllister argued the cause for appellants.
    With him on the briefs were Richard D. Heideman, Tracy
    Reichman Kalik, and Steven R. Perles.
    Mitchell R. Berger argued the cause for appellees. With
    him on the brief were Gassan A. Baloul and Alexandra E.
    Chopin. Pierre H. Bergeron, John Burlingame, and Laura G.
    Ferguson entered appearances.
    2
    Before: GARLAND, Chief Judge, KATSAS, Circuit Judge,
    and WILLIAMS, Senior Circuit Judge.
    Opinion for the Court filed by Senior Circuit Judge
    WILLIAMS.
    WILLIAMS, Senior Circuit Judge: During the Second
    Intifada, Palestinian terrorists ambushed an Israeli public bus
    traveling in the West Bank and opened fire, killing an American
    schoolteacher, Esther Klieman. Klieman’s estate (along with
    some survivors and heirs) sued numerous defendants—
    including the Palestinian Authority (“PA”) and Palestinian
    Liberation Organization (“PLO”)—under the Anti-Terrorism
    Act (“ATA”), 
    18 U.S.C. §§ 2331
    , et seq., among other laws.
    Having previously dismissed the case against all non-PA/PLO
    defendants for insufficient service of process, Estate of
    Klieman v. Palestinian Auth., 
    547 F. Supp. 2d 8
    , 15 (D.D.C.
    2008), the district court dismissed the case against the PA/PLO
    for want of personal jurisdiction under the constraints of the
    due process clause, Estate of Klieman v. Palestinian Auth., 
    82 F. Supp. 3d 237
     (D.D.C. 2015). Plaintiffs now appeal.
    In Livnat v. Palestinian Authority, 
    851 F.3d 45
    , 48–54
    (D.C. Cir. 2017), cert. denied, 
    139 S. Ct. 373
     (2018), this court
    held that the due process clause of the 5th Amendment barred
    U.S. courts from exercising jurisdiction over non-sovereign
    foreign entities without an adequate nexus to the United States.
    (In contrast, foreign sovereigns sued in the United States do not
    enjoy the benefit of this due process protection.) The district
    court here found that plaintiffs had failed to establish such a
    nexus for the PA/PLO.
    We agree. We conclude that the district court did not abuse
    its discretion in agreeing, in light of the intervening Supreme
    Court case of Daimler AG v. Bauman, 
    571 U.S. 117
     (2014), to
    reconsider its earlier ruling that the court had general personal
    3
    jurisdiction over defendants. As plaintiffs recognize, Daimler
    (and this court’s opinion in Livnat) effectively foreclose a
    ruling that the district court had general jurisdiction over the
    PA/PLO. See Klieman Br. 29. We then consider plaintiffs’
    argument for specific jurisdiction and their request for
    discovery to substantiate that theory, but find both sets of
    arguments inadequate. Finally, we address § 4 of the Anti-
    Terrorism Clarification Act of 2018, Pub. L. No. 115-253, 
    132 Stat. 3183
     (“ATCA”) (codified at 
    18 U.S.C. § 2334
    (e)),
    enacted during the pendency of this appeal and deeming certain
    conduct to qualify as consent to the jurisdiction of U.S. courts
    over terrorism cases. We find that plaintiffs have established
    neither the circumstances rendering § 4 applicable nor facts
    justifying a remand for discovery on the issue. Accordingly,
    we affirm the decision of the district court.
    * * *
    On March 24, 2002, a group of terrorists carried out an
    attack on an Israeli bus in the West Bank, killing Esther
    Klieman. See Estate of Klieman, 82 F. Supp. 3d at 240; see
    also Compl. ¶¶ 23–25 (Jul. 13, 2004), ECF No. 1. 1 Plaintiffs
    brought suit in 2004 against a host of defendants, including the
    PA, PLO, and other Palestinian individuals and entities,
    including the Al Aqsa Martyrs Brigade, a U.S.-designated
    Foreign Terrorist Organization that had “claimed responsibility
    for the attack.” Estate of Klieman, 82 F. Supp. 3d at 240.
    Plaintiffs allege among other things that the PA/PLO,
    acting “by and through their officials, employees and agents,”
    had “provided” other defendants “weapons, instrumentalities,
    1
    Citations to ECF Numbers are to the district court docket in
    Estate of Klieman v. Palestinian Authority, No. 1:04-cv-01173-PLF
    (D.D.C. filed Jul. 13, 2004).
    4
    permission, training, and funding for their terrorist activities,”
    along with “safe haven and a base of operations,” and
    encouraged certain defendants to “plan and execute acts of
    violence, murder and terrorism against innocent civilians in
    Israel, Gaza and the West Bank”—including the attack that
    killed Klieman. Compl. ¶ 40; see also Compl. ¶¶ 41–49.
    Besides asserting various tort claims, plaintiffs alleged
    violations of the ATA, 
    18 U.S.C. §§ 2332
    , 2333, and 2339A.
    See Compl. ¶¶ 50–60. Section 2333 creates a cause of action
    for “[a]ny national of the United States injured in his or her
    person, property, or business by reason of an act of
    international terrorism, or his or her estate, survivors, or heirs.”
    
    18 U.S.C. § 2333
    (a); see 
    id.
     § 2331(1) (defining “international
    terrorism”). And § 2333(d)(2) creates liability for persons who
    have aided or abetted, or conspired with a designated foreign
    terrorist organization (such as the Al Aqsa Martyrs Brigade) in
    the commission of terrorist acts.
    Defendants moved in May 2006 to dismiss the case for
    lack of personal jurisdiction, asserting among other problems
    that they had insufficient “minimum contacts” with the United
    States. See Defs.’ Mot. to Dismiss for Lack of Personal
    Jurisdiction 3 (May 30, 2006), ECF No. 55. As to the PA/PLO,
    the district court initially ruled, in December 2006, that it could
    exercise general jurisdiction over these defendants. Estate of
    Klieman v. Palestinian Auth., 
    467 F. Supp. 2d 107
    , 113 (D.D.C.
    2006). In April 2008, it denied defendants’ motion for
    reconsideration of that decision. Mem. Op. and Order (Apr. 24,
    2008), ECF No. 85. Fact discovery proceeded until 2013.
    In February 2014, defendants filed a motion for
    reconsideration of the 2006 and 2008 rulings, invoking the
    requirements for general personal jurisdiction set forth in
    Daimler, 571 U.S. at 137. See Defs.’ Mot. for Reconsideration
    (Feb. 5, 2014), ECF No. 233. The district court agreed to
    reconsider the matter.      It also embraced defendants’
    5
    jurisdictional argument, finding that the PA/PLO are not “at
    home” in the United States, as required for purposes of general
    jurisdiction under Daimler. It then found unpersuasive
    plaintiffs’ theory of specific jurisdiction and denied their
    request for jurisdictional discovery. As the PA/PLO had been
    the “sole remaining defendants,” the district court dismissed the
    case. Estate of Klieman, 82 F. Supp. 3d at 250.
    Following the roadmap laid out above, we affirm.
    * * *
    The due process limits on judicial exercise of personal
    jurisdiction over non-resident defendants take two forms:
    “general or all-purpose jurisdiction, and specific or conduct-
    linked jurisdiction.” Daimler, 571 U.S. at 122. General
    jurisdiction licenses a court “to hear any and all claims against”
    a defendant, Goodyear Dunlop Tires Operations, S.A. v.
    Brown, 
    564 U.S. 915
    , 919 (2011)—no matter where arising.
    Specific jurisdiction permits a court only to hear disputes that
    “aris[e] out of or relat[e] to the defendant’s contacts with the
    forum.” Daimler, 571 U.S. at 127 (quoting Helicopteros
    Nacionales de Colombia, S.A. v. Hall, 
    466 U.S. 408
    , 414 n.8
    (1984)).
    General jurisdiction entails a relatively demanding
    standard—reflecting its plenary reach over a defendant’s
    affairs. “A court may assert general jurisdiction over foreign
    . . . corporations to hear any and all claims against them when
    their affiliations with the [forum] are so ‘continuous and
    systematic’ as to render them essentially at home in the forum
    . . . .” Daimler, 571 U.S. at 127 (emphasis added) (quoting
    Goodyear, 
    564 U.S. at 919
    ). The upshot is that, absent
    exceptional circumstances, see, e.g., Perkins v. Benguet
    Consolidated Mining Co., 
    342 U.S. 437
     (1952), general
    jurisdiction will lie only where an entity is formally
    6
    incorporated or maintains its principal place of business, see
    BNSF Ry. Co. v. Tyrrell, 
    137 S. Ct. 1549
    , 1558 (2017);
    Daimler, 571 U.S. at 138–39 & n.19.
    Specific jurisdiction’s more limited scope justifies a less
    onerous standard. First, a defendant need not be “at home” in
    the forum. Second, unlike with general jurisdiction, minimum
    contacts must stem from or relate to conduct giving rise to the
    suit. Plaintiffs must establish a relationship among “the
    defendant, the forum, and the litigation.” Walden v. Fiore, 
    571 U.S. 277
    , 291 (2014) (quoting Calder v. Jones, 
    465 U.S. 783
    ,
    788 (1984)). More specifically, for a court “to exercise
    [specific] jurisdiction consistent with due process, the
    defendant’s suit-related conduct must create a substantial
    connection with the forum.” Id. at 284 (emphases added).
    Where, as here, a claim arises under federal law and, as the
    parties agree, a “defendant is not subject to jurisdiction in any
    state’s court of general jurisdiction,” Fed. R. Civ. P. 4(k)(2)(A);
    see Estate of Klieman, 82 F. Supp. 3d at 244, personal
    jurisdiction may be asserted under Rule 4(k)(2), “which
    functions as a federal long-arm statute,” id. Besides proper
    service of process, it requires only that “exercising jurisdiction
    [be] consistent with the United States Constitution and laws.”
    Fed. R. Civ. P. 4(k)(2)(B); see Mwani v. bin Laden, 
    417 F.3d 1
    , 10–11 (D.C. Cir. 2005). With that requirement met, the
    relevant forum is “the United States as a whole.” Mwani, 
    417 F.3d at 11
    ; accord, e.g., Plixer Int’l, Inc. v. Scrutinizer GmbH,
    
    905 F.3d 1
    , 6 (1st Cir. 2018).
    * * *
    In the wake of Daimler, defendants moved for
    reconsideration of the court’s 2006 and 2008 rulings on
    personal jurisdiction. The district court granted the request, and
    plaintiffs now object.
    7
    We review the district court’s decision to reconsider the
    issue for abuse of discretion. See, e.g., Capitol Sprinkler
    Inspection, Inc. v. Guest Servs., Inc., 
    630 F.3d 217
    , 225 (D.C.
    Cir. 2011); accord Connors v. Hallmark & Son Coal Co., 
    935 F.2d 336
    , 341 n.9 (D.C. Cir. 1991) (“[T]he abuse of discretion
    standard ordinarily applies to a district judge’s decision
    whether to consider a new theory raised on motion for
    reconsideration.”). The district court divided the matter into a
    segment on the propriety of reconsideration vel non and the
    plaintiffs’ claim of waiver or forfeiture. We address both
    issues, but in the reverse order.
    Although the PA/PLO raised its personal jurisdiction
    defense in a pre-answer motion under Rule 12(b)(2), thereby
    avoiding forfeiture under Rule 12(h)(1), the plaintiffs argue that
    defendants’ failure to raise the claim promptly after the
    Supreme Court’s decision in Goodyear, 
    564 U.S. at 919
    , the
    precursor of Daimler, waived or forfeited the personal
    jurisdiction defense. See Klieman Br. 17–20; see also Pls.’
    Opp’n to Defs.’ Mot. to Strike 3 (Nov. 2, 2018), Dkt. No.
    1758524.
    Plaintiffs point out that “more than 250 federal court
    cases” have “discussed Goodyear’s ‘at home’ standard,
    including eighteen circuit court cases and three cases in this
    District.” Klieman Br. 24 (quoting Gilmore v. Palestinian
    Interim Self-Gov’t Auth., 
    8 F. Supp. 3d 9
    , 16 (D.D.C. 2014),
    aff’d, 
    843 F.3d 958
     (D.C. Cir. 2016)). They note, too, that
    defense counsel in this litigation at the time of Goodyear had
    invoked the “at-home” language on behalf of the PA/PLO in
    other lawsuits shortly after Goodyear was decided—as well as
    in 2013. 
    Id. at 25
    . Defendants’ wait till 2014 to file the motion,
    plaintiffs conclude, constitutes undue delay. Further, they say
    the delay was prejudicial because the motion wasn’t filed until
    after fact discovery had closed. 
    Id. at 21, 30
    . As plaintiffs see
    it, they were, in effect, precluded from taking discovery to
    8
    support their specific jurisdiction theory, since at the time they
    had (reasonably) relied on the district court’s prior decision
    confirming personal jurisdiction. See 
    id. at 12
    , 20–21, 36.
    Defendants respond that Goodyear, and this circuit’s post-
    Goodyear but pre-Daimler cases, show sufficient room for
    nuance as to the status and reach of Goodyear’s “at-home”
    language that it was not unreasonable to seek reconsideration
    only after Daimler. And they argue that the timing of their
    motion was not prejudicial. See PA/PLO Br. 24–27.
    In finding the motion for reconsideration not barred by
    delay, the district court acknowledged that Goodyear had
    introduced the “at-home” language, but argued that “the reach
    of this language was not immediately clear,” citing the 2013
    supplement of a leading procedure treatise for the view that,
    “[i]f the Goodyear opinion stands for anything . . . it simply
    reaffirms that defendants must have continuous and systematic
    contacts with the forum in order to be subject to general
    jurisdiction.” Estate of Klieman, 82 F. Supp. 3d at 243. The
    court believed that Goodyear’s full import as a departure from
    laxer standards was “appreciated” only after Daimler issued in
    2014. Id. Defendants thus did not proceed with “undue delay.”
    And the court noted that neither plaintiffs nor the court could
    identify a case in which a similar motion was denied on grounds
    of “delay in identifying intervening case law.” Id.
    We see no abuse of discretion in the ruling on forfeiture
    (which the district court styles as a “waiver” analysis). On the
    one hand, in light of in-circuit cases elaborating on the “at-
    home” doctrine pre-Daimler—and defense counsel’s
    arguments on behalf of PA/PLO in other suits—there is some
    force to plaintiffs’ argument that defendants’ delay was
    unjustifiable. But a few points are dispositive in favor of
    defendants’ view. First, as a general matter, a district court has
    leeway “always” to “reconsider[]” interlocutory orders not
    9
    subject to the law of the case doctrine “prior to final judgment.”
    “[S]o long as the court has jurisdiction over an action, it should
    have complete power over interlocutory orders made therein
    and should be able to revise them when it is consonant with
    equity to do so.” Langevine v. Dist. of Columbia, 
    106 F.3d 1018
    , 1023 (D.C. Cir. 1997) (quoting Schoen v. Washington
    Post, 
    246 F.2d 670
    , 673 (D.C. Cir. 1957)); see also Fed. R. Civ.
    P. 54(b). Second, the court properly gave weight to the
    uncertainty in the wake of Goodyear, so clearly reflected in the
    passage quoted above from a leading treatise on procedure.
    Third, the court plausibly concluded that plaintiffs were not
    prejudiced by the timing of the motion.
    To be sure, under some circumstances we would be
    swayed by plaintiffs’ argument that they have been prejudiced
    by the delay in the defendants’ Goodyear-Daimler motion—
    coupled with their reasonable reliance on the district court’s
    finding of general personal jurisdiction and the closure of fact
    discovery. But here, as we’ll develop later in this opinion,
    plaintiffs have been unable to make a showing that discovery
    on their specific jurisdiction theory could have yielded
    evidence to support a finding of specific jurisdiction, and there
    is no sign that the district court relied at all on the “closure” of
    discovery in deciding to deny plaintiffs’ motion for further
    discovery to explore facts relevant to specific jurisdiction.
    Our approach is in keeping with Gilmore v. Palestinian
    Interim Self-Government Authority, 
    843 F.3d 958
    , 963–65
    (D.C. Cir. 2016). There we affirmed the district court’s
    decision under Rule 12(h)(1) that the PA/PLO had waived a
    constitutional personal jurisdiction defense that had been
    “available” because they had altogether failed to raise it in their
    2002 pre-answer motion. The delay argument pressed here is
    quite different from the 12(h)(1) issue in Gilmore; defendants
    here asserted constitutional personal jurisdictional defenses in
    2006 and 2007 on the basis of insufficient “minimum contacts”
    10
    with the forum in advance of filing their answer in May 2008.
    See Defs.’ Mot. to Dismiss for Lack of Personal Jurisdiction 3
    (May 30, 2006), ECF No. 55; see also Answer 2 (May 2, 2008),
    ECF No. 86; cf. Estate of Klieman, 
    467 F. Supp. 2d at 110, 113
    .
    So defendants essentially proceeded as Gilmore’s holding
    would have envisaged—on the basis of defenses “available” at
    the time of their pre-answer filings. In Gilmore we didn’t pass
    on the district court’s alternative theory of forfeiture based on
    acquiescence in the court’s jurisdiction. See Gilmore, 8 F.
    Supp. 3d at 14–16. We need not do so now. Even if we had
    affirmed the district court in reliance on the acquiescence
    theory, finding no abuse of discretion there, and even if the
    district court’s decision were inconsistent with the one we’re
    now reviewing, this outcome would not establish that the latter
    was an abuse of discretion.
    As to the motion for reconsideration viewed separately
    from the delay issue, the district court noted that the Federal
    Rules of Civil Procedure do not state standards governing such
    a motion before judgment, Estate of Klieman, 82 F. Supp. 3d at
    241–42, and in this gap relied on a three-part test from In re
    Vitamins Antitrust Litig., No. 99-1097, 
    2000 WL 34230081
    (D.D.C. Jul. 28, 2000); accord, e.g., McCoy v. FBI, 
    775 F. Supp. 2d 188
    , 190 (D.D.C. 2011) (Wilkins, J.) (adopting the
    Vitamins test). That opinion said that, given the value of
    finality, interlocutory orders may be reconsidered only “when
    the movant demonstrates (1) an intervening change in the law;
    (2) the discovery of new evidence not previously available; or
    (3) a clear error of law in the first order.” Vitamins, 
    2000 WL 34230081
    , at *1; cf. Christianson v. Colt Indus. Operating
    Corp., 
    486 U.S. 800
    , 817 (1988) (“A court has the power to
    revisit prior decisions of its own or of a coordinate court in any
    circumstance, although as a rule courts should be loathe to do
    so in the absence of extraordinary circumstances such as where
    the initial decision was ‘clearly erroneous and would work a
    manifest injustice.’” (citation omitted)). Neither party takes
    11
    issue with the Vitamins test, and we accept it for present
    purposes. (The district court had used the same test in denying
    defendants’ 2008 motion for reconsideration of its 2006 ruling
    on personal jurisdiction. See Mem. Op. and Order 2 (Apr. 24,
    2008), ECF No. 85.)
    We believe the district court acted within the bounds of its
    discretion in finding reconsideration appropriate. Two criteria
    of the Vitamins test seem applicable—“(1) an intervening
    change in the law” and “(3) a clear error of law in the first
    order.” Given that the governing law applicable at the time of
    the district court’s ruling was Daimler, see, e.g., Landgraf v.
    USI Film Prod., 
    511 U.S. 244
    , 273 (1994) (noting that “in many
    situations, a court should ‘apply the law in effect at the time it
    renders its decision’”) (quoting Bradley v. Sch. Bd. of City of
    Richmond, 
    416 U.S. 696
    , 711 (1974)), the prior ruling was
    indeed a clear error of law. Further, it was quite reasonable to
    say that the law had changed since the court’s most recent prior
    ruling on jurisdiction—2008.
    * * *
    We now take up the court’s disposition of the merits of the
    motion, including plaintiffs’ effort to establish specific
    jurisdiction, which we review de novo. See Livnat, 851 F.3d at
    48; FC Inv. Grp. LC v. IFX Markets, Ltd., 
    529 F.3d 1087
    , 1091
    (D.C. Cir. 2008). The district court first concluded that it could
    not properly exercise general jurisdiction over defendants
    because they are not “‘essentially at home’ in the United
    States.” Estate of Klieman, 82 F. Supp. 3d at 245 (quoting
    Daimler, 571 U.S. at 127). We agree. Because the PA’s
    “headquarters, officials, and primary activities are all in the
    West Bank,” it is not subject to general jurisdiction in the
    United States, as we held in Livnat, 851 F.3d at 56; see
    Waldman v. Palestinian Liberation Org., 
    835 F.3d 317
    , 332–
    34 (2d Cir. 2016) (applying similar reasoning to PLO).
    12
    Finding plaintiffs’ effort to salvage the earlier ruling in
    favor of general personal jurisdiction unavailing, we turn to the
    substance of plaintiffs’ theory of specific jurisdiction.
    To advance that theory, plaintiffs sought to develop a link
    between the killing of Esther Klieman and the furthering of
    PA/PLO goals in the United States. They offered a hypothesis
    building on these elements: First, the PA/PLO supported acts
    of terrorism during the Second Intifada in the early 2000s,
    targeting Israelis and areas frequented by Americans. Second,
    they pursued this terrorist program in part with the goal of
    advancing their “campaign in the United States to influence or
    affect United States foreign policy as it related to Israel and the
    Palestinian territories,” Klieman Br. 32; see also 
    id. at 42, 43
    ,
    carrying on the campaign through the use of U.S. offices,
    fundraising, lobbying, speaking engagements, as well as
    commercial dealings, 
    id. at 32
    . Third, as an integral part of this
    blended strategy of terrorism and diplomacy, they facilitated
    the killing of Esther Klieman.
    The first two elements may at first blush seem
    counterintuitive, but their logic is basically that a spate of
    terrorism claiming American (and Israeli) lives could impel
    U.S. policymakers to urge their Israeli counterparts to make
    concessions to defendants in exchange for their exerting their
    influence to halt, or attenuate, the attacks. For example, they
    quote a PA/PLO representative explaining on U.S. national
    television in 2002 that—in order for Palestinian suicide
    bombings to abate—the U.S. Secretary of State should prevail
    on Israel’s prime minister to reduce Israeli troop levels and
    settler presence in contested areas, for, “if the occupation
    continues . . . no one can stop the Palestinians.” 
    Id. at 43
    ; see
    also Reply Br. 22 (same).
    The basic theme here appears reasonable and seems to
    possess historical support. See, e.g., Klieman Br. 35 n.7; cf.
    13
    Peterson v. Islamic Republic of Iran, 
    264 F. Supp. 2d 46
    , 53
    (D.D.C. 2003) (quoting an expert witness on Iran explaining
    that “the foreign policy objective of the October 23rd, 1983,
    attack [on the U.S. Marine barracks in Lebanon] and other like
    attacks by Iran during this period” was “to end the Western,
    especially the American[,] presence in Lebanon”). Rather, our
    focus is on the third element of plaintiffs’ theory—the alleged
    link between the overall strategy and the killing of Esther
    Klieman. Plaintiffs have not alleged tangible facts as to how
    this attack was intended (or even used ex post) to further
    defendants’ political aims in the United States. The assertion
    that “the PA and PLO campaign to influence U.S. policy or
    affect its conduct, by leveraging the carnage of the Second
    Intifada, was expressly directed at . . . the United States,”
    Klieman Br. 43, is a claim that might apply to a welter of
    attacks spanning the years of the Second Intifada. But in a
    “jurisdictional inquiry focuse[d] on the relationship among the
    defendant, the forum, and the litigation,” Walden, 571 U.S. at
    287 (internal quotation marks and citation omitted), the
    “litigation” element requires tangible allegations relating the
    attack that cost Esther Klieman’s life to defendants’ contacts
    with the forum, cf. Waldman, 835 F.3d at 341–42.
    This circuit’s previous decision in Livnat appears
    controlling. The case arose out of a 2011 terrorist attack on
    Jewish worshipers at Joseph’s Tomb, a holy site in the West
    Bank. Livnat, 851 F.3d at 46. Plaintiffs had alleged that the
    attack was “part and parcel of” the PA’s “general practice of
    using terrorism to influence United States public opinion and
    policy” and was “intended, through intimidation and coercion,
    to influence the Israeli and United States government[s’]
    policies.” Id. at 57 (quoting complaint). To reinforce these
    allegations, plaintiffs supplied a declaration by a professor
    attesting that the PA’s support for terrorism was meant to
    “influence U.S. policy in the [PA’s] favor.” Id. But this didn’t
    convince us of an adequate relation between the Joseph’s Tomb
    14
    attack and the United States. Id. We declined even to consider
    the legal sufficiency of plaintiffs’ theory, given their failure to
    “‘make a prima facie showing of the pertinent jurisdictional
    facts’ to survive a motion to dismiss.” Id. at 56–57 (citation
    omitted). Plaintiffs, in essence, had asked us to infer “that
    because some attacks against Jews and Israelis have been aimed
    to influence U.S. policy, the Joseph’s Tomb attack was, too.”
    Id. at 57. “The record before us,” we concluded, “does not
    support that inference.” Id.
    Livnat’s logic governs here. Even if some terrorist acts
    carried out in Israel or the West Bank were used by defendants
    to influence U.S. policy, nothing in the record indicates that this
    attack fills that bill. Plaintiffs would distinguish Livnat by
    noting that whereas the attacks there were against Jews and
    Israelis—the present allegations center on attacks on “areas and
    targets known to be frequented by U.S. citizens.” Klieman Br.
    35. But the distinction doesn’t help plaintiffs on the facts
    presented. After all, they have alleged no facts indicating that
    the attack on an Israeli bus in the West Bank was directed at
    locales with a strong presence of U.S. nationals—either in the
    form of high-level planning or the individual attackers’
    motives. To the extent the attackers had—unbeknownst to
    them—chosen as their target a bus traveling through such a
    locale, the resulting “random, fortuitous, or attenuated
    contacts” with the forum are insufficient under Walden. A
    court’s “exercise of jurisdiction over an out-of-state intentional
    tortfeasor must be based on intentional conduct by the
    defendant that creates the necessary contacts with the forum.”
    Walden, 571 U.S. at 286 (emphasis added).
    In some circumstances allegations of a defendant’s general
    policy might adequately support an inference that the defendant
    aided and abetted a particular attack in furtherance of that
    policy. If two countries are engaged in armed conflict, we
    might be confident in explaining one country’s execution of a
    15
    bombing raid against the other’s territory as part of its general
    policy of inflicting damage on its adversary. But the case here
    plainly differs. Apart from any U.S. nexus there is a wholly
    plausible alternative explanation for defendants’ aiding and
    abetting the attack—dynamics altogether internal to the Israeli-
    Palestinian conflict. Cf. Ashcroft v. Iqbal, 
    556 U.S. 662
    , 682
    (2009) (addressing effect of “obvious alternative explanation”
    (citation omitted)). We think that distinction helps explain
    Livnat’s refusal to draw an inference that “because some
    attacks against Jews and Israelis have been aimed to influence
    U.S. policy, the Joseph’s Tomb attack was, too.” 851 F.3d at
    57.
    Plaintiffs might fill the resulting gap with allegations that
    PA/PLO officials invoked this attack in public or private
    statements in the United States after it took place, or perhaps
    that they took steps in the U.S. to aid and abet this particular
    attack before it occurred with the goal of advancing political
    objectives in the United States. But they offer nothing
    resembling such claims. As to the latter tack, plaintiffs “have
    not alleged [or] provided any prima facie showing . . . that
    either the PA or the PLO engages in fundraising in the United
    States,” let alone fundraising whose proceeds might have
    facilitated the 2002 attack. Estate of Klieman, 82 F. Supp. 3d
    at 247 n.7.
    Nor does Calder’s “effects test” help plaintiffs. See
    Klieman Br. 38–40. That analysis permits courts, in some
    instances, to assert jurisdiction over defendants whose conduct
    outside the forum causes certain “effects” within it. In Calder
    itself the Supreme Court approved a California state court’s
    jurisdiction over two Florida residents—an editor and reporter
    of the National Enquirer, a Florida corporation. Defendants
    penned and published a libelous article about a California
    resident distributed widely in that state. See Calder, 
    465 U.S. at
    784–86. In glossing Calder’s “effects test,” the Walden
    16
    Court stressed defendants’ intentional contacts with the forum.
    The “crux of Calder was that the reputation-based ‘effects’ of
    the alleged libel connected the defendants to California, not just
    to the plaintiff.” Walden, 571 U.S. at 287. “[B]ecause
    publication to third persons is a necessary element of libel . . .
    the defendants’ intentional tort actually occurred in California.”
    Id. at 288. Thus the “effects” of defendants’ libelous article—
    reputational harms arising in California—“connected the
    defendants’ conduct to California, not just to a plaintiff who
    lived there.” Id.
    Unlike the tort in Calder, which had “occurred in” the
    forum, Walden, 571 U.S. at 288, the planning, carrying out, and
    occurrence of Klieman’s killing all took place in the West
    Bank. And the emotional suffering felt by forum residents and
    (perhaps) foreseen by the attackers cannot without more qualify
    as the relevant “effect.” The Walden Court rejected such an
    approach, reasoning that it would “impermissibly allow[] a
    plaintiff’s contacts with the defendant and forum to drive the
    jurisdictional analysis.” Walden, 571 U.S. at 289 (emphasis
    added). Instead, “[t]he proper question is . . . whether the
    defendant’s conduct connects him to the forum in a meaningful
    way.” Id. at 290. Here we lack such allegations.
    Finally, plaintiffs’ invocation of our decision in Mwani v.
    bin Laden is unpersuasive. There defendants’ contacts with the
    United States were manifest in the very act that had precipitated
    the suit—a “devastating truck bomb” outside the U.S. Embassy
    in Nairobi, Kenya, in 1998, which “killed more than 200
    people, including 12 Americans.” Mwani, 
    417 F.3d at 4
    . In
    choosing their target, a U.S.-government building, Osama bin
    Laden and Al Qaeda had manifestly sought “purposefully [to]
    direct their terror at the United States,” 
    id. at 14
    , and “not only
    to kill both American and Kenyan employees inside the
    building, but to cause pain and sow terror in the embassy’s
    home country, the United States,” 
    id. at 13
    . Given conduct “no
    17
    doubt . . . ‘directed at [and] felt in’” the United States, 
    id.
    (alteration in original) (citation omitted), defendants could
    “reasonably anticipate being haled into” court there, 
    id. at 14
    (quoting Burger King Corp. v. Rudzewicz, 
    471 U.S. 462
    , 474
    (1985)), even, as we noted there, if we put aside defendants’
    “ongoing” plots to carry out attacks in the United States, id. at
    13. It was thus of no moment that “the plaintiff group was
    composed of non-U.S. nationals.” Klieman Br. 42; see Mwani,
    
    417 F.3d at 14
    .
    But whereas the Mwani defendants, in attacking a U.S.
    government outpost, indisputably aimed to kill Americans (at
    least in part), here we have no basis for inferring that the
    terrorists who attacked an Israeli bus were instructed, or
    endeavored, to injure American nationals. And absent
    intentional targeting, the fact that an American died in a
    terrorist incident abroad would amount only to a “random,
    fortuitous, or attenuated” contact “ma[de] by interacting with
    . . . persons affiliated with the” United States. Walden, 571
    U.S. at 286. It would thus be inadequate for specific
    jurisdiction absent a firmer link showing “intentional conduct
    by the defendant that creates the necessary contacts with the
    forum.” Id. at 286.
    We note that other circuits have taken a more stringent
    view of the necessary relation between the tort and in-forum
    activities than is manifest in Livnat and this decision. Thus the
    court in Waite v. All Acquisition Corp., 
    901 F.3d 1307
     (11th
    Cir. 2018), ruled that “a tort ‘arise[s] out of or relate[s] to’ the
    defendant’s activity in a [forum] only if the activity is a ‘but-
    for’ cause of the tort.” 
    Id. at 1314
     (first two alterations in
    original); see also O’Connor v. Sandy Lane Hotel Co., 
    496 F.3d 312
    , 318–19 (3d Cir. 2007) (describing the typically stricter
    proximate cause or “legal cause” test). Under a but-for view,
    in-forum activities postdating completion of the wrongful
    conduct—here, for example, any PA/PLO flourishing of the
    18
    killing as part of its U.S. diplomatic efforts—would likely not
    help in establishing minimum contacts for purposes of specific
    jurisdiction. Given that plaintiffs’ theory fails under our Livnat
    decision, we have no need to consider such cases or assess their
    possible application to these facts.
    We conclude that plaintiffs’ prima facie case for specific
    jurisdiction does not meet the Constitution’s requirements.
    Accordingly, we affirm the district court’s determination on
    this score.
    * * *
    The district court also turned down plaintiffs’ request for
    discovery in support of their theory of specific jurisdiction.
    We review the district court’s discovery rulings for abuse
    of discretion. “[A] district court has broad discretion in its
    resolution of discovery problems that arise in cases pending
    before it.” Naartex Consulting Corp. v. Watt, 
    722 F.2d 779
    ,
    788 (D.C. Cir. 1983) (quoting In re Multi-Piece Rim Products
    Liability Litigation, 
    653 F.2d 671
    , 679 (D.C. Cir. 1981)); see
    also Mwani, 
    417 F.3d at 17
    ; Goodman Holdings v. Rafidain
    Bank, 
    26 F.3d 1143
    , 1147 (D.C. Cir. 1994). Just as a plaintiff’s
    personal jurisdiction theory must clear the speculative level, a
    “request for jurisdictional discovery cannot be based on mere
    conjecture or speculation.” FC Inv. Grp. LC, 
    529 F.3d at 1094
    .
    In opposing defendants’ 2014 motion for reconsideration,
    plaintiffs sought discovery intended to disclose facts under two
    headings, both focused on aspects of defendants’ U.S.-centered
    activities:
    1. The extent of Defendants’ activities within the
    United States and this jurisdiction to attempt to
    influence the foreign policy and public opinion in the
    19
    United States to pressure Israel to change its public
    policies vis-à-vis the PA, including, but not limited to,
    information on the consultants, lobbyists and other
    professionals ret[]ained for this purpose.
    2. The financial investment of the Defendants’
    commercial contracts with US companies which allow
    the Defendants to raise revenue in the United States to
    support the operating budgets of the Defendants,
    which funded the joint public relations and terrorism
    campaign. As demonstrated above, funds from the
    Defendants are then used to support terrorism,
    including the very terrorists who murdered Esther
    Klieman.
    Pls.’ Supp. Br. in Opp’n to Defs.’ Mot. for Reconsideration 10
    (Jul. 11, 2014), ECF No. 256, J.A. 94. The district court
    understandably saw the requested discovery as “limited to
    seeking information about defendants’ public advocacy and
    fundraising activities in the United States.” Estate of Klieman,
    82 F. Supp. 3d at 249. It found that “[e]ven if the plaintiffs did
    obtain any such evidence through additional discovery . . . the
    plaintiffs would be unable to meet their burden of showing
    either general or specific personal jurisdiction under Daimler
    and Walden.” Id. Given the failure of these requests to focus
    on what we have identified as the fatal gap in plaintiffs’
    allegations, the purpose of the bus ambush in which the
    terrorists killed Esther Klieman, we can find no abuse of
    discretion in this result. See Livnat, 851 F.3d at 57 (“A district
    court acts well within its discretion to deny discovery when no
    ‘facts additional discovery could produce . . . would affect [the]
    jurisdictional analysis.’”) (alteration in original) (quoting
    Goodman Holdings, 
    26 F.3d at 1147
    )); see also Mwani, 
    417 F.3d at 6, 17
    .
    20
    In their appellate briefs plaintiffs express a new wish to
    seek discovery as to facts far beyond their original request, facts
    which might close the gap that we (and Livnat) have identified:
    They ask for
    jurisdictional discovery on whether the PA and PLO
    directed terrorists to attack Americans, such as in this
    case, or launch their attacks against areas and targets
    frequented by Americans.         Discovery into the
    proximity of PA/PLO-attributed attacks to
    concentrations of U.S. citizens, such as well-known
    tourist areas frequented by U.S. citizens or areas
    where U.S. citizens lived, would be one fruitful area
    of discovery.
    Klieman Br. 33; see also Reply Br. 15. But even if such
    discovery was aimed closely enough at the missing link in
    plaintiffs’ allegations, they failed to make the request to the
    district court, and “issues not raised before judgment in the
    district court are usually considered to have been [forfeited] on
    appeal.” Murthy v. Vilsack, 
    609 F.3d 460
    , 465 (D.C. Cir.
    2010); accord, e.g., Texas v. United States, 
    798 F.3d 1108
    , 1113
    (D.C. Cir. 2015); United States v. Stover, 
    329 F.3d 859
    , 872
    (D.C. Cir. 2003). Accordingly, these late requests provide no
    basis for overturning the district court’s exercise of discretion
    over the requests plaintiffs did make.
    * * *
    Having addressed the case as initially briefed, we now turn
    to the ATCA, enacted during the pendency of this appeal.
    Pursuant to ATCA § 4, certain conduct after January 31, 2019,
    is deemed to qualify as consent to the jurisdiction of U.S. courts
    over terrorism cases.
    21
    The parties spar over the factual predicates for the
    application of ATCA § 4, as well as its constitutionality. We
    conclude that plaintiffs have not made an adequate showing
    that any of § 4’s factual predicates has been triggered between
    February 1, 2019, and the time of the parties’ latest round of
    briefing on the subject on March 13, 2019. Section 4,
    accordingly, does not affect our analysis of personal
    jurisdiction, and we need not reach the defendants’
    constitutional challenges.
    Section 4 identifies five factual predicates grouped under
    two headings to trigger its “deemed to have consented” clause.
    See 
    18 U.S.C. § 2334
    (e). The first heading, § (e)(1)(A), refers
    to “accept[ing]” “any form of assistance, however provided,”
    under the following parts of the Foreign Assistance Act of
    1961, 
    22 U.S.C. §§ 2151
     et seq.:
    (1) chapter 4 of part II, 
    22 U.S.C. §§ 2346
     et seq.;
    (2) section 481, 
    22 U.S.C. § 2291
    ; or
    (3) chapter 9 of part II, 22 U.S.C. §§ 2349bb et seq.
    The second heading, § (e)(1)(B), refers to a defendant
    “benefiting from a waiver or suspension of section 1003” of the
    ATA, 
    22 U.S.C. § 5202
    , and
    (4) “continu[ing] to maintain”— or
    (5) “establish[ing] or procur[ing]”—
    “any office, headquarters, premises, or other facilities or
    establishments within the jurisdiction of the United States.”
    As we noted earlier, once defendants raise personal
    jurisdiction as a defense, “[t]he plaintiffs have the burden of
    establishing the court’s personal jurisdiction over” defendants.
    22
    FC Inv. Grp. LC, 
    529 F.3d at 1091
    . To do so, they must “‘make
    a prima facie showing of the pertinent jurisdictional facts’ to
    survive a motion to dismiss for lack of personal jurisdiction.”
    Livnat, 851 F.3d at 56–57 (quoting First Chicago Int’l v. United
    Exch. Co., 
    836 F.2d 1375
    , 1378 (D.C. Cir. 1988)). We analyze
    the record on the factual predicates as an extension of plaintiffs’
    prima face case for personal jurisdiction, asking whether
    plaintiffs have put forward plausible allegations that meet any
    of the factual predicates for implied consent under § 4. Cf.
    Iqbal, 
    556 U.S. at 679
     (“[O]nly a complaint that states a
    plausible claim for relief survives a motion to dismiss.”).
    The government filed an amicus brief at the invitation of
    the court and agrees with defendants that § 4’s factual
    predicates have not been satisfied. “[A]s of February 1, 2019
    and since that date, defendants have not accepted any of the
    foreign assistance provided under the authorities enumerated in
    Section 4, and they do not currently ‘benefit[]’ from a waiver
    of section 1003 of the Anti-Terrorism Act of 1987, including to
    maintain an office in the United States pursuant to such a
    waiver.” United States’ Response to Feb. 6, 2019, Order 7
    (Feb. 15, 2019) (“U.S. Response”), Dkt. No. 1773566.
    Plaintiffs demur as to both subsections (A) and (B) of
    § (e)(1). We ultimately find, in keeping with the view of the
    United States, that plaintiffs have failed to offer plausible
    allegations that any of the factual predicates of ATCA § 4 has
    been met or to offer credible grounds to support their requested
    remand for discovery.
    Foreign assistance and § 4(e)(1)(A). The PA/PLO offered
    its December 26, 2018, letter to the State Department as
    conclusively rejecting aid covered by ATCA. Plaintiffs say
    that the letter “merely expresses a ‘wish’ to no longer receive”
    relevant forms of assistance. Klieman Supp. Br. 7 (Mar. 13,
    2019), Dkt. No. 1777379. Hardly. The letter is quite emphatic:
    23
    “The Government of Palestine unambiguously makes the
    choice not to accept such assistance.” U.S. Response, Exhibit
    1, Letter at 2. And the State Department and Department of
    Justice readily discerned its meaning. See U.S. Response 7.
    Plaintiffs refer to certain “debt relief grant agreements with
    the PA” dating to 2015 and 2016, Klieman Supp. Br. 7–8,
    which were indeed provided under the Economic Support Fund
    covered by § 4(e)(1)(A)(i), see Foreign Assistance: U.S.
    Assistance for the West Bank and Gaza, Fiscal Years 2015 and
    2016, Gov’t Accountability Office (Aug. 2018),
    https://www.gao.gov/assets/700/693823.pdf. But plaintiffs (1)
    fail to allege that any such forms of debt relief have persisted
    after January 31, 2019; and (2) do not grapple with the
    PA/PLO’s renunciation of all relevant funding sources.
    Because we lack credible allegations that debt relief grants are
    currently being provided to PA/PLO, its instrumentalities, or
    creditors as of February 1, 2019—or that any of these “accept”
    such relief—plaintiffs’ mere allusions to past examples and
    hypothesizing their continuation or renewal is not enough to
    warrant a remand.
    The same goes for plaintiffs’ references to funding for non-
    governmental organizations. See Klieman Supp. Br. 8.
    Plaintiffs rely on a Congressional Research Service report from
    2011, which is unconvincing as to February 2019. Second, a
    gap remains in plaintiffs’ analysis. Section 4(e)(1)(A) requires
    that defendants “accept” the relevant aid, yet plaintiffs allude
    only to payments to non-governmental organizations.
    Although such assistance might constitute a “form of
    assistance, however provided” to PA/PLO, plaintiffs offer
    nothing to establish that link.
    Finally, nothing in the papers before us suggests that if
    granted an opportunity for discovery on remand plaintiffs
    would be able, in spite of the government’s denial, to unearth
    24
    sources of funding that continue to flow to the PA/PLO post-
    January 31, 2019, and come within § 4.
    Benefiting from a waiver or suspension and maintaining
    or establishing an office, headquarters, etc.; § 4(e)(1)(B).
    Subsection (B) sets out two necessary but individually
    insufficient conditions for deeming a defendant to have
    consented to personal jurisdiction. (1) The defendant must
    maintain or establish, etc., “any office, headquarters, premises,
    or other facilities or establishments within the jurisdiction of
    the United States.” (2) The defendant must be “benefiting from
    a waiver or suspension of section 1003.”
    Because the second requirement is dispositive against the
    plaintiffs we address the first requirement only enough to give
    an idea of the context within which the “waiver” is to be
    examined.
    (1) Activities allegedly triggering implied consent if
    defendant is “benefiting from a waiver or suspension of section
    1003.” Plaintiffs’ strongest argument centers on activities
    carried out by defendants under the auspices of the U.N.
    Permanent Observer Mission in New York. They do not
    dispute the Second Circuit’s holding that the ATA—and,
    accordingly, § 1003—do not apply to defendants’ U.N.
    Mission as such. See Klinghoffer v. S.N.C, Achille Lauro Ed
    Altri-Gestione Motonave Achille Lauro in Amministrazione
    Straordinaria, 
    937 F.2d 44
    , 46 (2d Cir. 1991). Rather,
    plaintiffs allege that various activities carried out by personnel
    of the Mission go beyond the legal shield afforded by the
    exclusion of the Mission itself.
    Klinghoffer reasons that “only those activities not
    conducted in furtherance of the PLO’s observer status may
    properly be considered as a basis of jurisdiction,” 
    937 F.2d at 51
    , and offers some examples.            The court mentions
    25
    “proselytizing and fundraising activities,” 
    id. at 52
    , including
    those of the then-Permanent Representative of the PLO Zuhdi
    Labib Terzi, who had “spok[en] in public and to the media in
    New York in support of the PLO’s cause” “[e]very month or
    two,” 
    id.
     (quoting district court opinion). On remand, the
    district court found various activities to exceed the shelter
    accorded the U.N. Mission, including Dr. Terzi’s speeches and
    the Mission’s generation of “informational materials” and
    distribution of them “to those seeking information about the
    PLO.” Klinghoffer v. S.N.C. Achille Lauro Ed Altri-Gestione
    Motonave Achille Lauro in Amministrazione Straordinaria,
    
    795 F. Supp. 112
    , 114 (S.D.N.Y. 1992). Plaintiffs here rely on
    rather similar promotional activities; for example, Dr. Riyad
    Mansour, Permanent Observer for Palestine at the U.N., gave
    speeches well beyond New York itself, to wit, in Orlando,
    Florida. See Klieman Supp. Br. 7; see also 
    id.
     Exhibit 4.
    Even if we were to assume arguendo that the line drawn by
    the Second Circuit in Klinghoffer is correct and that the
    activities of the U.N. Mission in fact ranged beyond that line,
    plaintiffs have not (as discussed below) shown that defendants
    have been “benefiting from a waiver or suspension,” as
    required for an inference of consent to suit triggered by ATCA
    § 4(e)(1)(B).
    (2) “[B]enefiting from a waiver or suspension.” Plaintiffs
    do not and cannot claim an express waiver or suspension. The
    PLO shuttered its D.C. office as of October 10, 2018, after the
    State Department declined to extend its § 1003 waiver. See
    U.S. Response 5–6; see also id. Exhibits 3–5. And the New
    York U.N. Mission operates without a waiver precisely because
    it isn’t subject to the ATA. As the government has stated,
    “[t]here is no waiver of section 1003 currently in effect.” Id. at
    6.
    26
    In fact it appears correct to interpret the phrase “waiver or
    suspension” in (B) as referring solely to an express waiver
    under § 1003(3), as the government assumes.
    For legal authority to issue periodic waivers to the PLO,
    the State Department has relied on annual State Department
    appropriations bills. See U.S. Response, Exhibits 3–4. For
    example, the 2017 letter in Exhibit 3 invokes § 7041(j)(2)(B)(i)
    of the Department of State, Foreign Operations, and Related
    Programs Appropriations Act, 2016, Pub. L. No. 114-113, 
    129 Stat. 2242
    , 2780 (2015), which says:
    The President may waive the provisions of section
    1003 of the Foreign Relations Authorization Act,
    Fiscal Years 1988 and 1989 (Public Law 100-204) if
    the President determines and certifies in writing to the
    Speaker of the House of Representatives, the
    President pro tempore of the Senate, and the
    appropriate congressional committees that the
    Palestinians have not, after the date of enactment of
    this Act [either (1) taken certain steps at the U.N. or
    (2) taken certain actions vis-à-vis the International
    Criminal Court] (emphasis added).
    The natural reading then, of “waiver or suspension” in
    § (e)(1)(B), is the sort of formal exercise of power plainly
    contemplated in this statute setting forth the waiver procedure.
    Plaintiffs point to nothing that could either qualify as or
    substitute for the formal waiver or suspension evidently
    required. They point instead, see Klieman Supp. Br. 3, to: (1)
    an agency’s “constructive” waiver of a deadline by accepting
    payments after that deadline, Morris Commc’ns, Inc. v. FCC,
    
    566 F.3d 184
    , 189 (D.C. Cir. 2009); (2) the unremarkable truth
    that defendants may implicitly consent to personal jurisdiction,
    Ins. Corp. of Ireland v. Compagnie des Bauxites de Guinee, 456
    
    27 U.S. 694
    , 703–04 (1982); (3) the fact that an agency may be
    required to suspend enforcement efforts to collect funds after
    making certain findings, Salazar v. King, 
    822 F.3d 61
    , 78–79
    (2d Cir. 2016); and (4) a statutory provision permitting the
    Secretary of Defense to “expressly waive[], in writing,” a
    certain “limitation,” 10 U.S.C. § 2193b(c)(3)(B).          The
    relevance of items (1)–(3) is remote at best. As to (4),
    plaintiffs’ statement that “Congress certainly knows how to
    specify ‘written waivers’ when it wishes, and did not do so
    here,” Klieman Supp. Br. 3 (emphasis added), appears to
    neglect the actual language of the legal authorization to issue
    waivers under § 1003, namely the one quoted above, which
    creates legal consequences when the President “certifies in
    writing” that a waiver is to be issued.
    Plaintiffs would equate government “failure to prosecute”
    allegedly excessive propaganda activities with provision of a
    waiver or suspension. Klieman Supp. Br. 5. But the statute
    permits no such equation. ATCA § 4 is triggered by a waiver
    of § 1003—not its violation. Thus, the predicate for making
    defendants’ U.N. activities legally material under ATCA § 4
    has not been met.
    * * *
    We affirm the decision of the district court in full.
    So ordered.