Appeals dismissed by published opinion. Judge KING wrote the opinion, in which Chief Judge TRAXLER and Judges MOTZ, GREGORY, DUNCAN, AGEE, DAVIS, KEENAN, WYNN, DIAZ, and FLOYD joined. Judge DUNCAN wrote a concurring opinion, in which Judge AGEE joined. Judge WYNN wrote a concurring opinion. Judge WILKINSON wrote a dissenting opinion, in which Judge NIEMEYER and Judge SHEDD joined. [209]*209Judge NIEMEYER wrote a dissenting opinion, in which Judge WILKINSON and Judge SHEDD joined.
OPINION
KING, Circuit Judge:
Following the 2003 invasion of Iraq, the United States military took control of Abu Ghraib prison near Baghdad, using it to detain criminals, enemies of the provisional government, and other persons thought to possess information regarding the anti-Coalition insurgency. The United States contracted with CACI International, Incorporated (with CACI Premier Technology, Incorporated, together referred to herein as “CACI”), and Titan Corporation, now L-3 Services, Incorporated (“L-3”), to provide civilian employees to assist the military in communicating with and interrogating this latter group of detainees.
On June 30, 2008, a number of Iraqis who had been detained at Abu Ghraib and elsewhere filed lawsuits against CACI and L-3 in the Southern District of Ohio and the District of Maryland, alleging that the contractors and certain of their employees were liable in common law tort and under the Alien Tort Statute (“ATS”), 28 U.S.C. § 1350, for torturing and abusing them during their incarceration. Following the unopposed transfer of the Ohio action to the Eastern District of Virginia, where CACI is headquartered, Suhail Najim Abdullah A1 Shimari and three co-plaintiffs submitted an Amended Complaint asserting that CACI, through its employees, agents, and government coconspirators, deprived them of basic human necessities, beat them and ran electric current through their bodies, subjected them to sexual abuse and humiliation, and traumatized them with mock executions and other sadistic acts. In the operative Second Amended Complaint filed in the companion litigation, seventy-two plaintiffs, headed by Wissam Abdullateff Sa’eed AI-Quraishi, detailed similar allegations against L-3 and Adel Nakhla, an L-3 employee residing in Maryland.1
I.
A.
On September 15, 2008, CACI moved to dismiss the Amended Complaint filed in the Eastern District of Virginia, maintaining generally that, among other things: (1) the dispute presented a nonjusticiable political question; (2) the inevitable application of the law of occupied Iraq rendered CACI, as part of the occupying power, immune from suit under Coleman v. Tennessee, 97 U.S. 509, 24 L.Ed. 1118 (1878), and Dow v. Johnson, 100 U.S. 158, 25 L.Ed. 632 (1879); (3) the plaintiffs’ claims were preempted by the “combatant activities” exception to the Federal Tort Claims Act (the “FTCA”), see 28 U.S.C. § 2680(j), discussed in Ibrahim v. Titan Corp., 556 F.Supp.2d 1 (D.D.C.2007), and subsequently adopted on appeal, see Saleh v. Titan Corp., 580 F.3d 1 (D.C.Cir.2009) (citing Boyle v. United Tech. Corp., 487 U.S. 500, 108 S.Ct. 2510, 101 L.Ed.2d 442 (1988)); and (4) the company was entitled to absolute official immunity in accordance with Mangold v. Analytic Services, Inc., 77 F.3d 1442 (4th Cir.1996), because its employees had performed delegated governmental functions. With respect to the [210]*210ATS claims, CACI proffered, several additional arguments, none of them relevant here in light of the claims’ eventual dismissal. See infra at 210.
L-3’s motion to dismiss the Second Amended Complaint in the Maryland action, filed on November 26, 2008, and in which Nakhla joined, was predicated essentially along the same lines as CACI’s, though it characterized Mangold as involving the application of derivative sovereign immunity instead of absolute official immunity. As CACI had previously done, L-3 invoked the political question doctrine, cited the Supreme Court’s decisions in Coleman and Dow (the “law-of-war defense”), and requested (through supplemental briefing) that the court adopt the combatant activities exception ultimately applied in Saleh {“Saleh preemption”). L-3 similarly advocated for dismissal of the ATS claims on substantially the same grounds identified by CACI.2
1.
On March 19, 2009, the district court in Virginia entered a Memorandum Order dismissing the ATS claims against CACI, but permitting the common-law tort claims to proceed. See Al Shimari v. CACI Premier Tech., Inc., 657 F.Supp.2d 700 (E.D.Va.2009). In so ruling, the court acknowledged its considerable reservations that the action implicated a political question, in that CACI, a private entity, was not the United States, and only low-level military and governmental personnel appeared to have been involved in the alleged mistreatment. See id. at 708-14. The court was similarly doubtful that the foreseeable application of Iraqi law required dismissal in light of CACI’s apparent status as an arms-length contractor, “because even if the law of a foreign jurisdiction were to govern any of the Plaintiffs’ claims, it would not regulate the conduct of the United States, a non-party to this suit between private parties.” Id. at 725.
The dividing line between the bona fide military and its civilian support personnel also fueled the district court’s uncertainty that the latter could have engaged in wartime activities as a “combatant” for purposes of adopting the D.C. Circuit’s theory of FTCA preemption. See Al Shimari, 657 F.Supp.2d at 720-21. The court concluded that, in any event, the plaintiffs’ allegations of torture at the hands of CACI failed to implicate the uniquely federal interests or irreconcilable conflict with state law that animated the Supreme Court’s decision in Boyle, on which Saleh relied. See id. at 722-25.
Regarding CACI’s claim of derivative immunity under Mangold, the district court set forth its view that the validity of such a claim depends on whether its proponent, in committing the act complained of, was “ ‘exercising discretion while acting within the scope of their employment.’ ” Al Shimari, 657 F.Supp.2d at 715 (emphasis omitted) (quoting Mangold, 77 F.3d at 1446). Citing “a very limited factual record,” the court expressed its skepticism that CACI had established at the dismissal stage that its treatment of the plaintiffs at Abu Ghraib involved the exercise of discretion. Id. The court stated further that it was “completely bewildered” by the suggestion that it could accept CACI’s representations that the company had performed within the scope of its agreement with the government “when the contract is not before the Court on this motion.” Id. at 717. On March 23, 2009, CACI noted [211]*211its appeal (No. 09-1335) from the district court’s ruling.
2-
The assertion of Mangold immunity was viewed much the same way by the district court in Maryland, which, in its Opinion of July 29, 2010, concluded that, “relying on the information in the [Second Amended] Complaint, it is clearly too early to dismiss Defendants.” Al-Quraishi v. Nakhla, 728 F.Supp.2d 702, 735 (D.Md.2010).3 The dis[212]*212trict court perceived no such record deficiencies concerning L-3’s and Nakhla’s alternative bases for dismissal, however, deeming the facts as pleaded sufficient to reject outright both defendants’ arguments. The court thus denied the motion to dismiss with respect to all claims, including those premised on the ATS. See id. at 724-33, 736-60. From the court’s accompanying Order, L-3 noted its appeal (No. 10-1891) on August 4, 2010, followed two days later by another appeal (No. 10-1921) noted on behalf of Nakhla.
B.
The appeals in Alr-Quraishi were consolidated and argued in seriatim with the Al Shimari appeal before a panel of this Court on October 26, 2010. Apart from urging our affirmance on the merits, the plaintiffs in each matter alternatively maintained that we lacked appellate jurisdiction over the district courts’ non-final orders denying the contractors’ respective motions to dismiss. On September 21, 2011, we issued opinions in both cases, in which a majority of the panel concluded that jurisdiction was proper in this Court, and that the district courts had erred in permitting the claims against the contractors to proceed. See Al Shimari v. CACI Int’l, Inc., 658 F.3d 413 (4th Cir.2011); Al-Quraishi v. L-3 Servs., Inc., 657 F.3d 201 (4th Cir.2011).4 Consistently therewith, we entered separate judgments reversing the orders on appeal and remanding with instructions to dismiss both proceedings.
On November 8, 2011, upon the timely petitions of the plaintiffs, see Fed. R.App. P. 35(b)-(c), we entered an Order granting en banc rehearing of all three appeals, thereby vacating our prior judgments. The appeals were thereafter consolidated for purposes of oral argument, which was conducted before the en banc Court on January 27, 2012.5 Having fully considered the briefs and arguments of the parties, together with the written and oral submissions of the amici curiae permitted leave to participate, we conclude that we lack jurisdiction over these interlocutory appeals, and we therefore dismiss them.6
II.
Except for the limited categories of interlocutory orders set forth at 28 U.S.C. § 1292, federal appellate jurisdiction is reserved for “final decisions of the district courts of the United States.” 28 U.S.C. § 1291. It is undisputed that the decisions underlying these putative appeals are interlocutory, at least in the procedural sense, in that no final order or judgment has been entered by either district court. [213]*213It is also without contest that neither order has been certified appealable by the issuing court pursuant to 28 U.S.C. § 1292(b), and that none of that statute’s provisions otherwise apply to confer jurisdiction on this Court.
Consequently, the only way we may be entitled to review the orders on appeal is if they are among “that small class [of decisions] which finally determine claims of right separable from, and collateral to, rights asserted in the action, too important to be denied review and too independent of the cause itself to require that appellate consideration be deferred until the whole case is adjudicated.” Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541, 546, 69 S.Ct. 1221, 93 L.Ed. 1528 (1949). Expounding on the topic, the Supreme Court has emphasized that an appealable Cohen order must “[1] conclusively determine the disputed question, [2] resolve an important issue completely separate from the merits of the action, and [3] be effectively unreviewable on appeal from a final judgment.” Will v. Hallock, 546 U.S. 345, 349, 126 S.Ct. 952, 163 L.Ed.2d 836 (2006) (alterations in original) (internal quotation marks omitted).
Cohen involved a stockholder’s derivative action for mismanagement and fraud, in which the Supreme Court reviewed the district court’s threshold decision declining to enforce a state law requiring plaintiffs in such cases to post security ensuring payment of attorney fees in the event the defendant corporation prevailed. Deeming the appeal properly taken, the Court declared no exception to the jurisdictional prerequisites of 28 U.S.C. § 1291, but instead described what would subsequently be coined the “collateral order doctrine,” MacAlister v. Guterma, 263 F.2d 65, 67 (2d Cir.1958), as a “practical, rather than a technical construction” of the statute. Cohen, 337 U.S. at 546, 69 S.Ct. 1221.
The federal courts of appeals have consistently been charged with keeping a tight rein on the types of orders suitable for appeal consistent with Cohen. We are therefore bound to maintain “a healthy respect for the virtues of the final-judgment rule.” Mohawk Indus., Inc. v. Carpenter, — U.S. —, 130 S.Ct. 599, 605, 175 L.Ed.2d 458 (2009); see also Will, 546 U.S. at 350, 126 S.Ct. 952 (“[W]e have not mentioned applying the collateral order doctrine recently without emphasizing its modest scope.”).7
The Supreme Court’s concern, as expressed through its repeated admonitions, is amply justified. The appellate courts are, by design, of limited jurisdiction; thus, accepting prejudgment appeals as a matter of course would “undermine[ ] efficient judicial administration and encroach[ ] upon the prerogatives of district court judges, who play a special role in managing ongoing litigation.” Mohawk, 130 S.Ct. at 605 (internal quotation marks omitted). In addition, routine interlocutory review would unacceptably subject meritorious lawsuits to “the harassment and cost [214]*214of a succession of separate appeals from the various rulings to which a litigation may give rise, from its initiation to entry of judgment.” Firestone Tire & Rubber Co. v. Risjord, 449 U.S. 368, 374, 101 S.Ct. 669, 66 L.Ed.2d 571 (1981) (internal quotation marks omitted).
Moreover, there is no need to construe Cohen broadly given the existence of a suitable alternative. The “safety valve” of discretionary interlocutory review under 28 U.S.C. § 1292(b) is frequently a “better vehicle for vindicating [certain] serious ... claims than the blunt, categorical instrument of [a] § 1291 collateral order appeal.” Digital Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 883, 114 S.Ct. 1992, 128 L.Ed.2d 842 (1994). Accordingly, the collateral order doctrine should “never be allowed to swallow the general rule that a party is entitled to a single appeal, to be deferred until final judgment has been entered.” Id. at 868, 114 S.Ct. 1992 (citation omitted).
Although a properly appealable collateral order under Cohen must of course satisfy all of the Will requirements, its hallmark is the encapsulation of a right whose abridgement is “effectively unreviewable” should appellate review await final judgment. See Henry v. Lake Charles Am. Press LLC, 566 F.3d 164, 177 (5th Cir.2009) (describing unreviewability as “the fundamental characteristic of the collateral order doctrine” (citation omitted)). The “critical question” in determining whether the right at issue is effectively unreviewable in the normal course “is whether the essence of the claimed right is a right not to stand trial” — that is, whether it constitutes an immunity from suit. Van Cauwenberghe v. Biard, 486 U.S. 517, 524, 108 S.Ct. 1945, 100 L.Ed.2d 517 (1988) (internal quotation marks omitted). Absent an immediate appellate review of the denial of an immunity claim, the right not to stand trial “would be irretrievably lost.” Id. (internal quotation marks omitted). By contrast, if the right at issue is one “not to be subject to a binding judgment of the court” — that is, a defense to liability— then the right can be vindicated just as readily on appeal from the final judgment, and the collateral order doctrine does not apply. Id. at 527, 108 S.Ct. 1945.
In assessing whether the right sought to be protected constitutes a true immunity and not merely a defense, “§ 1291 requires [the court] of appeals to view claims of a ‘right not to be tried’ with skepticism, if not a jaundiced eye.” Digital Equip., 511 U.S. at 873, 114 S.Ct. 1992. As the Supreme Court has cautioned, “[o]ne must be careful ... not to play word games with the concept of a ‘right not to be tried,’ ” Midland Asphalt Corp. v. United States, 489 U.S. 794, 801, 109 S.Ct. 1494, 103 L.Ed.2d 879 (1989), as “virtually every right that could be enforced appropriately by pretrial dismissal might loosely be described as conferring a right not to stand trial,” Digital Equip., 511 U.S. at 873, 114 S.Ct. 1992. It is within the foregoing framework that we review de novo the appealability of the district courts’ denial orders. See Mitchell v. Forsyth, 472 U.S. 511, 528-30, 105 S.Ct. 2806, 86 L.Ed.2d 411 (1985) (equating denials of qualified immunity to collateral denials of other asserted immunities or of double jeopardy invocations, and deeming de novo standard proper based on non-deferential review of latter claims).
III.
In Doe v. Exxon Mobil Corp., 473 F.3d 345 (D.C.Cir.2007), the District of Columbia Circuit confronted an attempted appeal from the district court’s interlocutory order refusing to dismiss an action brought [215]*215by Indonesian villagers alleging serious injuries visited upon them by members of that nation’s military in the defendants’ private employ. According to the defendants, the dispute presented a nonjustieiable political question. The court of appeals declined to address the merits of the issue, noting the absence of “a single case in which a federal appeals court held that denial of a motion to dismiss on political question grounds is an immediately appealable collateral order.” Id. at 352.8
That case yet appears to be lacking, and the appellants do not contend to the contrary. L-3, however, ventures that an appellate court may determine whether an action is a political question or otherwise nonjusticiable when it has proper jurisdiction over a different issue pursuant to Cohen or § 1292(b), if consideration of the former is “necessary to ensure meaningful review.” Swint v. Chambers Cnty. Comm’n, 514 U.S. 35, 51, 115 S.Ct. 1203, 131 L.Ed.2d 60 (1995). We may also exercise so-called “pendent” appellate jurisdiction in circumstances where the question is “inextricably intertwined” with another that may be immediately reviewed. Id.; see Rux v. Republic of Sudan, 461 F.3d 461, 476 (4th Cir.2006).
L-3’s argument necessarily supposes the existence of an otherwise valid jurisdictional basis for its appeal. Absent an independently reviewable issue with which the political question doctrine may be inexorably bound, or one that cannot be reviewed in a meaningful fashion without addressing the justiciability of the underlying dispute, we are without authority to make any pronouncement on that aspect of the appellants’ defense. We therefore withhold for the moment substantive comment on the political question doctrine, at least until we evaluate whether the law-of-war defense, Saleh preemption, or Mangold immunity provides the jurisdictional green light for us to proceed.
The appellants characterize their former presence in Iraq as “occupying forces” (L-3) or “occupying personnel” (CACI) that are answerable “only to their country’s criminal laws,” Opening Br. of CACI at 25, and thus “not subject to civil suits by the occupied,” Opening Br. of L-3 at 22-23. In that regard, the appellants equate their situation with those of the Civil War soldiers in Coleman v. Tennessee, 97 U.S. 509, 24 L.Ed. 1118 (1878), and Dow v. Johnson, 100 U.S. 158, 166, 25 L.Ed. 632 (1879), who sought relief from judgments entered against them for their wartime acts. The defendant in Coleman had been convicted and sentenced to death by a Tennessee state court for murdering a civilian, though the same judgment and sentence had been previously imposed as the result of a United States Army court-martial. Dow, by contrast, involved a challenge to a civil judgment entered in Louisiana against a Union general after forces under his command had seized the plaintiffs private property in furtherance of the war effort.
[216]*216Neither judgment was permitted to stand. In both cases, the Supreme Court considered the states of the Confederacy to have been “the enemy’s country,” to whose tribunals the “[ojfficers and soldiers of the armies of the Union were not subject.” Coleman, 97 U.S. at 515. The Court expressed its bewilderment that a contrary result could obtain “from the very nature of war,” concluding that “the tribunals of the enemy must be without jurisdiction to sit in judgment upon the military conduct of the officers and soldiers of the invading army. It is difficult to reason upon a proposition so manifest; its correctness is evident upon its bare announcement.” Dow, 100 U.S. at 165.
Some differences between the disputes at bar and those underlying Coleman and Dow are readily evident. Most salient is that the civilian employees of CACI and L-3 assigned to Abu Ghraib were not soldiers. The idea that those employees should nonetheless be treated like full-fledged members of the military pervades this litigation, though the concept resonates with more force as to some of the appellants’ other defenses, particularly Saleh preemption and Mangold immunity. But cf. Ford v. Surget, 97 U.S. 594, 601-02, 24 L.Ed. 1018 (1878) (relieving Mississippi civilian from liability for burning landowner’s cotton where destruction- ordered by Confederate army in face of Union advance and those “commands would have been undoubtedly enforced by the same means of coercion as if he had been an enlisted soldier”). The potential liability of government contractors was front and center in both Saleh and Mangold, and if the legal principles in either case (or both) are deemed apposite to the dispute at bar, there is little question that the appellants, as contractors themselves, may avail themselves of them.
Another distinction is that the appellants attempt to invoke the law-of-war defense exclusively on the assertion that them alleged wrongs will be evaluated under Iraqi law, and not the laws of Virginia, Maryland, or another state. If true, that may or may not be enough to bring Coleman and Dow into play, inasmuch as the overriding concern in those cases appears to have been less about the application of the criminal law of Tennessee or of Louisiana tort law (there being no suggestion that either differed significantly from the analogous law applied by the defendants’ states of citizenry), and more about the jurisdiction of the “foreign” courts. See Coleman, 97 U.S. at 516 (musing that “there would be something incongruous and absurd in permitting an officer or soldier of an invading army to be tried by his enemy”); Dow, 100 U.S. at 163 (identifying “[tjhe important question” for resolution as whether nation’s military could be held liable “in the local tribunals”). Here, of course, the appellants are being sued on their home turf, in courts that are indisputably domestic.
Even assuming that the facts before us can be viewed in such a fashion to permit Coleman and Dow to apply, there is no indication from the opinions in those cases that the Supreme Court intended to construe the law-of-war defense as an immunity from suit, rather than merely an insulation from liability. See Dow, 100 U.S. at 165 (characterizing dispute as concerning personal jurisdiction); Lauro Lines s.r.l. v. Chasser, 490 U.S. 495, 500, 109 S.Ct. 1976, 104 L.Ed.2d 548 (1989) (“[W]e have declined to hold the collateral order doctrine applicable where a district court has denied a claim ... that the suit against the defendant is not properly before the ... court because it lacks jurisdiction.”). In its subsequent Ford opinion, with judgment having been entered against the defendant on a jury verdict, the Court in no [217]*217way indicated that trial should not have been had.
Indeed, it seems a bit curious to imagine the nineteenth century Court regarding its decisions in the Civil War cases as having durable precedential effect; the appeals afforded an unusual opportunity for substantive domestic review of what were, in effect, foreign pronouncements of judgment. But to the extent that Coleman and Dow possess continued relevance beyond their immediate context, it is nonetheless clear that the issues presented in those cases were effectively reviewed and disposed of on appeal, and, as such, the manner in which the Supreme Court chose to resolve them fails to compel the conclusion that immunity must be accorded all prospective defendants who insist they are similarly situated. The law-of-war defense thus provides no basis for an interlocutory appeal in this case.
In a like fashion, Saleh preemption falls squarely on the side of being a defense to liability and not an immunity from suit. Immunity, according to the Supreme Court, derives from “an explicit statutory or constitutional guarantee that trial will not occur.” Midland Asphalt Corp. v. United States, 489 U.S. 794, 801, 109 S.Ct. 1494, 103 L.Ed.2d 879 (1989) (emphasis added).9 There is no contention that the Supreme Court in Boyle v. United Technologies Corp., 487 U.S. 500, 108 S.Ct. 2510, 101 L.Ed.2d 442 (1988), from which Saleh preemption is derived, relied on any such explicit guarantee embodied in statute or in the Constitution. Boyle preemption (and, thus, Saleh preemption) is, ipso facto, not immunity.
We are not the first court to arrive at this ineluctable conclusion. In Martin v. Halliburton, 618 F.3d 476, 487 (5th Cir.2010), the Fifth Circuit similarly reckoned that “the combatant activities exception is not subject to a sui generis exemption from the ordinary jurisdictional requirements for denials of preemption claims.”10 Indeed, the Boyle Court itself repeatedly framed the preemption it recognized as creating a mere defense to liability. See, e.g., 487 U.S. at 507, 108 S.Ct. 2510 (“The [218]*218imposition of liability on Government contractors [in the military procurement context] will directly affect the terms of Government contracts.”); id. at 511-12, 108 S.Ct. 2510 (“The financial burden of judgments against the contractors would ultimately be passed through ... to the United States itself.”); id. at 512, 108 S.Ct. 2510 (“[S]tate law which holds Government contractors liable for design defects in military equipment does in some circumstances present a ‘significant conflict’ with federal policy and must be displaced.”).
It is tempting, we suppose, to blur the line between an eventual frustration of liability and the more immediate right to avoid suit altogether. One might be persuaded to consider the words “preemption” and “immunity” as mere labels that are more or less synonymous with each other, or to presume that the former can effectively operate as the latter. But merely repackaging for the sake of convenience the preemption defense derived from Boyle as “combatant activities immunity,” as our good colleague Judge Niemeyer does in speaking for the dissenters, post at 259, is patently incorrect.
Though Boyle preemption, like sovereign immunity, may be invoked to bar state law claims, the encapsulated rights serve distinct purposes. State law claims are preempted under Boyle simply because the imposition of liability in such situations is irreconcilable with uniquely federal interests. The right conferred through federal preemption, in other words, is the right not to be bound by a judgment stemming from state law duties.
In stark contrast, immunity has consistently been administered as a protection against the burden of litigation altogether. See Mitchell v. Forsyth, 472 U.S. 511, 525-27, 105 S.Ct. 2806, 86 L.Ed.2d 411 (1985). Further, as the court of appeals explained in Rodriguez v. Lockheed Martin Corp., 627 F.3d 1259, 1265 (9th Cir.2010), “Although the source of the government contractor defense [recognized in Boyle ] is the United States’ sovereign immunity,” the preemption defense is not itself a species thereof. To the contrary, entitlement to preemption “is only a corollary financial benefit flowing from the government’s sovereign immunity.” Id. Accordingly, Boyle’s “government contractor defense does not confer sovereign immunity on contractors,” and as such, the denial of the defense is not immediately appealable. Id. (internal quotation marks omitted).
Importantly, the law requires that we assess the appealability of a potentially qualifying collateral order in a categorical sense, and not on a case-by-case basis.11 Conducting that assessment here leads to the conclusion that the denial of a preemp[219]*219tion claim stemming from the combatant activities exception would not necessarily entail significant scrutiny of sensitive military issues. Fundamentally, there is little intrusion because the court’s inquiry focuses on whether the contractor complied with the government’s specifications and instructions, and not the wisdom or correctness thereof. The Boyle and Saleh decisions themselves well illustrate the lack of intrusion that would result from deferring review until after entry of a final judgment. Boyle, for example, involved an appeal from a jury verdict for the plaintiff, while “the two appeals in Saleh reached the D.C. Circuit using the normal machinery of §§ 1291 and 1292(b).” Martin, 618 F.3d at 488.12
Moreover, the district court in Saleh had conducted extensive discovery “regarding the military’s supervision of the contract employees as well as the degree to which such employees were integrated into the military chain of command,” 580 F.3d at 4, with no ill effects. The Fifth Circuit, while acknowledging that Boyle preemption is underpinned by “a respect for the interests of the Government in military matters,” has nonetheless reasoned that those interests can be safeguarded without resort to interlocutory review. Martin, 618 F.3d at 488. For example, a district court “should take care to develop and resolve such defenses at an early stage while avoiding, to the extent possible, any interference with military prerogatives.” Id. Additionally, a trial court should consider “limiting discovery initially to such defenses” and “certifying orders denying [the] defense[ ] where the law is unsettled but, after refinement on appeal, might warrant dismissing plaintiffs’ claims.” Id.13
When properly conducted, suits against private contractors pose minimal risk that military personnel will be improperly haled into court or their depositions taken, because “[w]here discovery would hamper the military’s mission, district courts can and must delay it.” Saleh, 580 F.3d at 29 (Garland, J., dissenting) (citing, inter alia, Watts v. SEC, 482 F.3d 501, 508-09 (D.C.Cir.2007)). Other procedural and substantive rules, such as Rule 45 of the Federal Rules of Civil Procedure and the state secrets doctrine, also adequately safeguard military interests. See id. at 29 n. 18 (Garland, J., dissenting). Accordingly, we decline to recognize denials of Saleh preemption as a new class of collateral order.14 Insofar as it would be founded on [220]*220the false premise that immediate appeals are necessary in preemption cases to protect the government’s legitimate military interests, such recognition would reflect an impermissibly indulgent view of appellate jurisdiction.
C.
Before jurisdiction can be invoked under the collateral order doctrine, a district court must issue a “fully consummated decision” that constitutes “a complete, formal, and ... final” resolution of the issue. Abney v. United States, 431 U.S. 651, 659, 97 S.Ct. 2034, 52 L.Ed.2d 651 (1977). In other words, the court’s ruling must be “the final word on the subject addressed.” Digital Equip., 511 U.S. at 867, 114 S.Ct. 1992. If a ruling lacks finality, the threshold requirement for collateral order review—that the question in dispute be definitively resolved—is likewise left wanting. See Will v. Hallock, 546 U.S. 345, 349, 126 S.Ct. 952, 163 L.Ed.2d 836 (2006) (confining review of non-final orders to disputed questions conclusively determined, which raise important non-merits issues that are effectively unreviewable if not immediately appealed).
A question in dispute cannot be said to have been conclusively resolved if a district court “ma[kes] clear that its decision [is] a tentative one, ... and that it might well change its mind” after further proceedings. Jamison v. Wiley, 14 F.3d 222, 230 (4th Cir.1994). Disputed questions that arise with respect to claims of immunity are not the exception to that ironclad rule. Fundamentally, a court is entitled to have before it a proper record, sufficiently developed through discovery proceedings, to accurately assess any claim, including one of immunity. And even a party whose assertion of immunity ultimately proves worthy must submit to the burdens of litigation until a court becomes sufficiently informed to rule.
Manifestly, with respect to the appellants’ attempts to invoke Mangold immunity in their respective actions, sufficient information was lacking. The Maryland and Virginia district courts each perceived that the validity of such invocations depended in significant part on whether the contractor involved was acting within the scope of its agreement with the United States. One could hardly begin to answer that question without resort to any and all contracts between the appellants and the government pertinent to the claims, defenses, and related matters below. See, e.g., Al-Quraishi v. Nakhla, 728 F.Supp.2d 702, 741 n. 11 (D.Md.2010) (reasoning that contract could show, for example, that “ ‘federal wartime policy-making’ was not behind Defendants’ alleged actions,” in which case plaintiffs’ “state law claims [would] not intrude upon the preempted field”). While other evidence and testimony could also be relevant to ascertain the appellants’ business relationship with the government in general, and the parties’ agreed duties and responsibilities in Iraq and at Abu Ghraib in particular, the analysis must necessarily begin with the written contract or contracts. Cf. Harris v. Kellogg Brown & Root Servs., Inc., 618 F.3d 398, 402 (3d Cir.2010) (rejecting appellate jurisdiction for failure of Will’s “conclusively determined” requirement, where only limited discovery had been conducted on combatant activities and political question defenses).15
[221]*221In dissent, Judge Niemeyer contends that Behrens v. Pelletier, 516 U.S. 299, 116 S.Ct. 834, 133 L.Ed.2d 773 (1996), and Ashcroft v. Iqbal, 556 U.S. 662, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009), each a qualified immunity proceeding, provide for collateral order jurisdiction of the district courts’ orders denying Mangold immunity, as illustrated by other of our qualified immunity cases. See post at 254, 255 (citing McVey v. Stacy, 157 F.3d 271 (4th Cir.1998); Jenkins v. Medford, 119 F.3d 1156 (4th Cir.1997) (en banc); Winfield v. Bass, 106 F.3d 525 (4th Cir.1997) (en banc)). According to Judge Niemeyer, Behrens and Iqbal counsel that Rule 12 denials of immunity invariably constitute final decisions appealable under § 1291, and those authorities “clearly establish that these appeals fit comfortably with the Cohen collateral order doctrine.” Post at 249.
It is more accurate to say that orders denying dismissal motions, insofar as those motions are based on immunities that are not absolute but conditioned on context, such as qualified immunity in a § 1983 action or the derivative immunities at issue here, are, in accordance with Behrens and Iqbal, sometimes immediately appealable. Winfield makes the point:
[W]e possess no jurisdiction over a claim that a plaintiff has not presented enough evidence to prove that the plaintiffs version of the events actually occurred, but we have jurisdiction over a claim that there was no violation of clearly established law accepting the facts as the district court viewed them.
106 F.3d at 530. More generally, we would have jurisdiction over an appeal like the ones attempted here “if it challenged] the materiality of factual issues.” Bazan ex rel. Bazan v. Hidalgo Cnty., 246 F.3d 481, 490 (5th Cir.2001). By contrast, we lack jurisdiction if such an appeal “challenges the district court’s genuineness ruling — that genuine issues exist concerning material facts.” Id. Of course, “[w]e always have jurisdiction to determine whether the facts relevant to our jurisdiction exist.” Wireko v. Reno, 211 F.3d 833, 835 (4th Cir.2000) (citation omitted).
In Iqbal, the Supreme Court framed the genuineness-materiality distinction as one between “fact-based” or “abstract” issues of law, with only the latter supplying a proper foundation for immediate appeal. 556 U.S. at 674, 129 S.Ct. 1937 (quoting Johnson v. Jones, 515 U.S. 304, 317, 115 S.Ct. 2151, 132 L.Ed.2d 238 (1995)). The Iqbal Court concluded that whether a particular constitutional right was clearly established for qualified immunity purposes presents an abstract issue of law that permits an appeal at the' dismissal stage. See id. at 674-75, 129 S.Ct. 1937. Here, as in Iqbal, there is no “vast pretrial record” to encumber our decisionmaking, id. at 674, 129 S.Ct. 1937, but the issues before us are more factually entrenched and far less amenable to meaningful analysis by resort merely to the plaintiffs’ pleadings. Thus, unlike Iqbal, these appeals encompass fact-based issues of law, with the need for additional development of the record being among those “matters more within a district court’s ken.” Id.
Hence, insofar as an interlocutory appeal of a denial of immunity requires resolution of a purely legal question (such as whether an alleged constitutional violation was of clearly established law), or an ostensibly fact-bound issue that may be re[222]*222solved as a matter of law (such as whether facts that are undisputed or viewed in a particular light are material to the immunity calculus), we may consider and rule upon it. See Behrens, 516 U.S. at 313, 116 S.Ct. 834 (deeming appellate jurisdiction to have been properly asserted over denial of summary judgment in § 1983 action where adverse ruling was premised on defendant’s alleged conduct having violated clearly established law); McVey, 157 F.3d at 276 (approving jurisdiction over similar legal issue at dismissal stage, where appeal did not “raise factual questions concerning the defendants’ involvement, which would not be appealable”).16
Behrens, then, confers jurisdiction of these appeals only if the record at the dismissal stage can be construed to present a pure issue of law. We might discern such an issue if we were of the opinion, as the dissenters evidently are, that persons similarly situated to the appellants are inevitably and invariably immune from suit premised on any and all conduct occurring (1) when they are in a war zone, by virtue of (2) a contract with the government. But not even Saleh, which receives a ringing endorsement in both dissents, went that far.
The court in Saleh adopted the following rule; “During wartime, where a private service contractor is integrated into combatant activities over which the military retains command authority, a tort claim arising out of the contractor’s engagement in such activities shall be preempted.” 580 F.3d at 9. The D.C. Circuit therefore conditions preemption on the presence of a certain level of public/private integration, the conduct of activities that may be classified as combat, and the military’s retained prerogative concerning the decisionmaking process. Though the Saleh court had the luxury of a complete record developed through discovery to assist it in pondering those issues, there has been no discovery in the cases at bar, and the pleadings provide nothing approaching definitive answers.17
[223]*223Indeed, the questions that will require proper answers in order to gauge the appellants’ entitlement to immunity have yet to be fully ascertained. In Mangold v. Analytic Services, Inc., supra note 3, the relevant issues on appeal from summary judgment included whether government personnel were conducting an “official investigation,” and whether the contractors’ statements giving rise to potential liability were responsive to the investigators’ queries, as opposed to being extraneous thereto. See Mangold v. Analytic Services, Inc., 77 F.3d at 1449-50. Subsequently, in Butters v. Vance International, Inc., supra note 3, also a summary judgment appeal, we were constrained to decide whether withholding a job promotion from the plaintiff was a “commercial activity,” and whether that employment decision was made by the defendant or the foreign government with which it had contracted. See Butters v. Vance International, Inc., 225 F.3d at 465-67. As with Mangold and Butters, this case too requires careful analysis of intrinsically fact-bound issues, which may resemble any or all of the Saleh considerations, and will almost certainly entail an exploration of the appellants’ duties under their contracts with the government and whether they exceeded the legitimate scope thereof.
The appellants are requesting immunity in a context that has been heretofore unexplored. These are not disputes in which facts that might be material to the ultimate issue have been conclusively identified. Moreover, those facts that may have been tentatively designated as outcome-determinative are yet subject to genuine dispute, that is, a reasonable fact-finder could conclude in favor of either the plaintiffs dr the defendants. See Metric/Kvaerner Fayetteville v. Fed. Ins. Co., 403 F.3d 188, 197 (4th Cir.2005). Because the courts’ immunity rulings below turn on genuineness, we lack jurisdiction to consider them on an interlocutory appeal. See Winfield, 106 F.3d at 530; Bazan, 246 F.3d at 490.18
Thus, although Mangold immunity confers upon those within its aegis the right not to stand trial, the appellants have yet to establish their entitlement to it. See Martin, 618 F.3d at 483 (concluding that claims of immunity must be “substantial,” and not “merely colorable”). Because these appeals were taken before the district courts could reasonably render a decision on the applicability of Mangold and, perhaps, Butters, there is no collateral order fulfilling the Will requirements for appealability pursuant to Cohen, and therefore no jurisdiction in this Court to review any related aspect of the proceedings below.19
[224]*224D.
There being no independent basis for appellate jurisdiction premised on the law-of-war defense, Saleh preemption, or Man-gold immunity, we are without pendent jurisdiction to further consider the appellants’ contentions that the plaintiffs’ claims present nonjusticiable political questions. Our rejection of each of the three proffered bases also precludes the exercise of jurisdiction regardless of whether the appellants’ political question defense is inextricably intertwined with any of them, or whether those bases are similarly interdependent with one another.
IV.
Pursuant to the foregoing, these consolidated appeals must be dismissed.
APPEALS DISMISSED