John Doe v. James Mattis [REISSUED OPINION]

• Court: Court of Appeals for the D.C. Circuit
• Decided: June 28, 2019
• Docket: 18-5032
• Status: Published

Opinion

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued April 5, 2018 Decided May 7, 2018 Reargued April 27, 2018 Reissued June 28, 2019

No. 18-5032

JOHN DOE, APPELLEE

v.

JAMES MATTIS, IN HIS OFFICIAL CAPACITY AS SECRETARY OF DEFENSE, APPELLANT

Consolidated with 18-5110

Appeal from the United States District Court for the District of Columbia (No. 1:17-cv-02069)

James M. Burnham, Attorney, U.S. Department of Justice, argued the cause for appellant. With him on the briefs were Jessie K. Liu, U.S. Attorney, Matthew M. Collette and Sonia M. Carson, Attorneys. Catherine H. Dorsey, Attorney, entered an appearance.

Jonathan L. Hafetz argued the cause for appellee. With him on the briefs were Arthur B. Spitzer and Hina Shamsi. 2 Before: HENDERSON, SRINIVASAN, and WILKINS, Circuit Judges.

Opinion for the Court filed by Circuit Judge SRINIVASAN, with whom Circuit Judge WILKINS joins.

Dissenting opinion filed by Circuit Judge HENDERSON.

SRINIVASAN, Circuit Judge: This case involves a United States citizen who has been detained by the United States military in Iraq for several months. He seeks release from military custody in a habeas corpus action brought under the pseudonym John Doe. Doe is a citizen not only of the United States but also of Saudi Arabia.

Doe was initially captured in Syrian territory controlled by the Islamic State of Iraq and the Levant (ISIL). The Department of Defense determined that he is an enemy combatant for ISIL, and the Department has been detaining him at a military facility in Iraq. Doe’s habeas petition contends that he must be released because, he claims, ISIL combatants do not come within any existing authorization for use of military force. He also contends that he is not in fact an ISIL combatant. At this stage of the proceedings, no court has addressed the merits of those claims.

This appeal instead concerns a separate claim by Doe: that the government, while his habeas petition remains pending, cannot forcibly—and irrevocably—transfer him to the custody of another country. Transfer of Doe to another country’s custody would, naturally, obviate any occasion to seek release from physical custody at the hands of the United States. In connection with the possibility of Doe’s forcible transfer to the custody of another country, the district court has entered two orders we now review. 3 In the first order, the court required the government to give 72 hours’ notice before transferring Doe to the custody of any other country. The notice period was meant to afford the court an opportunity to review the circumstances of a planned transfer before it takes place. The government seeks to set aside any obligation to give advance notice with regard to two specific countries. We will refer to those countries as Country A and Country B because of the government’s desire to withhold public release of their identities due to apparent sensitivities associated with ongoing or future diplomatic discussions.

The district court’s second order came about after the government reached an agreement with Country B to transfer Doe to its custody. The government gave the district court the requisite notice of its intent to transfer Doe to that country. The court then enjoined the government from effecting the transfer. In the court’s view, the government had failed to demonstrate the necessary legal authority (specifically, a statute or treaty) for the transfer.

We sustain both of the district court’s orders. In claiming the authority to forcibly transfer an American citizen held abroad to the custody of another country, the government ultimately relies on two species of argument. Neither, in our view, gives the government the power to effect its desired handover of Doe to Country B, at least as things currently stand.

The first rationale advanced by the government has no necessary grounding in Doe’s designation as an enemy combatant or in the military’s authority under the law of war. Rather, the government relies on Supreme Court decisions recognizing that, when a foreign country wants to prosecute an American citizen already present in its territory for a crime 4 committed within its borders, the Executive can relinquish her to that country’s custody for purposes of criminal proceedings. See Munaf v. Geren, 553 U.S. 674 (2008); Wilson v. Girard, 354 U.S. 524 (1957). Those decisions, on the government’s reading, extend past their facts in two ways: (i) they enable a forcible transfer of a U.S. citizen to a different foreign country than the one in which she is already present, and (ii) they enable a forcible transfer as long as the receiving country has some legitimate sovereign interest in her (whether or not related to criminal prosecution). No. 18-5032, Gov’t Opening Br. 23-25.

We cannot accept the government’s argument. We know of no instance—in the history of the United States—in which the government has taken an American citizen found in one foreign country and forcibly transferred her to the custody of another foreign country. Under the logic of the government’s position, it could pick up an American traveling in Europe and involuntarily relinquish her to, say, the custody of Afghanistan, as long as Afghanistan is thought to have some cognizable sovereign interest in her. We cannot conclude that the government possesses that kind of authority over a U.S. citizen, at least without a statute or treaty specifically providing for it.

The government’s second line of argument differs from its first in an important respect: the second argument turns on Doe’s status as an alleged enemy combatant and on the military’s attendant authority in a time of war. We agree with the government that, if Doe is an enemy combatant, the military can transfer him to the custody of Country B, a partner in the campaign against ISIL. But under the precedents of the Supreme Court and our court, two conditions must exist for an American citizen to be subject to military transfer or detention as an enemy combatant: (i) there must be legal authority for the Executive to wage war against the enemy, and (ii) there must be an opportunity for the citizen to contest the factual 5 determination that he is an enemy combatant fighting on behalf of that enemy. See Hamdi v. Rumsfeld, 542 U.S. 507, 517, 533 (2004) (plurality opinion); Omar v. McHugh, 646 F.3d 13, 24 (D.C. Cir. 2011). Neither the legal inquiry nor the factual inquiry has taken place in this case. In the absence of those inquiries, we see no basis to set aside the district court’s injunction barring the forcible transfer of Doe to Country B.

What about the district court’s order requiring the government to give 72 hours’ notice before transferring Doe to either Country A or Country B? Because the government gave notice of the proposed transfer to Country B, the government’s appeal of the notice order as it applies to Country B is now moot. With regard to Country A, the government has yet to come forward with any information about the circumstances of a prospective transfer to that country, including the specific purpose or interest that will give rise to the transfer. The government instead seeks ex-ante, carte-blanche authorization to transfer Doe to Country A, regardless of the particular circumstances or reasons, and without any opportunity for judicial review. We conclude that the district court did not err in denying the government that sort of blanket preapproval.

While we sustain the district court’s orders, we do so respectful of—and with appreciation for—the considerable deference owed to the Executive’s judgments in the prosecution of a war. That latitude of course extends to military decisions about what to do with enemy combatants captured overseas in a zone of active hostilities. Virtually all such decisions will be unaffected by our decision today.

But when an alleged enemy combatant—even one seized on a foreign battlefield—is an American citizen, things are different. See Hamdi, 542 U.S. at 532-33, 535-37 (plurality); id. at 558-59 (Scalia, J., dissenting). In that “surely . . . rare” 6 circumstance, id. at 571 n.3 (Scalia, J., dissenting), the Executive’s authority to wage war as it sees fit is cabined by the Supreme Court’s decision in Hamdi, which requires that a citizen be afforded certain guarantees before the military detains or transfers him under the law of war. Id. at 517, 533 (plurality); Omar, 646 F.3d at 24. That precedent, in our view, governs the disposition of this appeal.

I.

A.

In September 2017, Syrian Democratic Forces encountered Doe at a screening point on an active battlefield in Syrian territory controlled by ISIL. Doe surrendered, informed the Syrian Democratic Forces that he was an American citizen, and asked to speak to U.S. officials. The Syrian Democratic Forces transferred Doe to the custody of U.S. military forces in the region. The military reached a preliminary determination that Doe is an enemy combatant, and has detained him at a U.S. facility in Iraq for the past seven months.

The military’s preliminary determination that Doe is an enemy combatant is based on evidence that he is a member or substantial supporter of ISIL. The evidence against Doe includes the following: the circumstances of his surrender, his statements upon surrender and during detention, and records of his ISIL membership.

ISIL, also known as the Islamic State of Iraq and Syria (ISIS), has been designated as a terrorist group. It controls territory in Iraq and Syria, and has perpetrated and aided terrorism there and around the world, killing several thousand civilians, including American aid workers and journalists. See U.S. Dep’t of State, Bureau of Counterterrorism, Country 7 Reports on Terrorism 2016: Chapter 6, Terrorist Organizations (July 2017). Since September 2014, the United States has pursued a counterterrorism strategy against ISIL, and is an active member of a 75-country coalition working to defeat ISIL in Iraq and Syria.

B.

In October 2017, the American Civil Liberties Union Foundation, acting on Doe’s behalf, petitioned the district court for a writ of habeas corpus. The petition asserts that the military’s existing authority to engage in armed conflict does not extend to ISIL, that the military thus lacks legal authority to detain an alleged member of ISIL, and that, as a result, the government must either prosecute Doe in an Article III court or release him. In addition to those legal arguments, Doe contends as a factual matter that he is not an ISIL combatant.

The district court determined that the ACLU had standing to bring the action on Doe’s behalf. The court ordered the government to give the ACLU access to Doe to ascertain whether he wanted to continue the action. Am. Civil Liberties Union Found. v. Mattis, 286 F. Supp. 3d 53, 60-61 (D.D.C. Dec. 23, 2017). On January 5, 2018, the ACLU informed the court that Doe wanted to continue pursuing the habeas petition with the ACLU representing him. The ACLU then asked for an order barring the government from transferring Doe to another country until the court decided the merits of his petition.

On January 23, the district court granted Doe’s request in part. The court entered a preliminary injunction requiring the government to provide 72 hours’ notice before transferring Doe to any other country. 8 The court determined that Doe had proven a likelihood of success because the government had failed to demonstrate that it had the requisite legal authority to transfer him to another country. The court further concluded that Doe had shown irreparable injury, reasoning that transfer out of U.S. custody would render him “unable to pursue his habeas petition.” Doe v. Mattis, 288 F. Supp. 3d 195, 200 (D.D.C. Jan. 23, 2018). Finally, the court weighed the government’s interest in maintaining productive diplomatic relations with potential transferee countries against a U.S. citizen’s right to contest the lawfulness of his detention, concluding that both the balance of equities and the public interest favored Doe. Finding the requirements for a preliminary injunction to have been met, the court entered its order requiring 72 hours’ notice so that Doe would have an opportunity to challenge a proposed transfer before it happened.

The government appealed. It initially asked this court to vacate the preliminary injunction so that it could transfer Doe to any country without providing advance notice. No. 18-5032, Gov’t Opening Br. 27-28. In the alternative, the government asked for vacatur of the notice requirement as applied to one specified country “or any other country that the Executive Branch determines has a legitimate interest in petitioner.” Id. at 38. Later, in its reply brief, the government narrowed the scope of its appeal still further, such that it now seeks vacatur of the notice requirement only as applied to Countries A or B. No. 18-5032, Gov’t Reply Br. 2 n.1.

On April 16, 2018, while the government’s appeal of the notice injunction was pending, the government filed a notice in the district court in compliance with that injunction. The notice communicated the government’s intent to transfer Doe to the custody of Country B in 72 hours. Attached to the notice was a sworn declaration from a Deputy Assistant Secretary of State, 9 who averred that Country B had expressed a “strong interest” in taking custody of Doe and continuing to detain him in some form. Doe v. Mattis, No. 17-cv-2069, Notice attach. 1 at 4-5 (D.D.C. Apr. 17, 2018), ECF No. 80. Doe moved for a preliminary injunction or temporary restraining order to block the proposed transfer.

On April 19, 2018, the district court granted the preliminary injunction, barring the government “from transferring [Doe] from U.S. custody.” Doe v. Mattis, No. 17- cv-2069, Prelim. Inj. (D.D.C. Apr. 19, 2018), ECF No. 88. While the order could be read to bar transfer to any foreign country, we understand it to grant only the relief Doe requested (and thus only the relief the government had notice might be imposed)—that is, a bar on transfer to Country B specifically. Cf. Capital City Gas Co. v. Phillips Petrol. Co., 373 F.2d 128, 131 (2d Cir. 1967). In support of the order, the court again concluded that Doe had demonstrated a likelihood of success on the merits because the government had failed to identify the requisite legal authority for a forcible transfer of Doe to Country B. And for the same reasons it gave when it entered the notice injunction, the court concluded that Doe would be irreparably injured absent an injunction and that the balance of equities and public interest weighed in his favor.

The government appealed the second injunction to this court. It then moved for consolidation of the two appeals and expedited treatment, both of which we granted. This opinion thus resolves both of the government’s appeals. In view of the presumption of public access to judicial proceedings, we have endeavored to fashion the opinion so as to manage redactions while still not revealing the identities of Countries A and B. 10 II.

The government appeals two orders granting injunctive relief to Doe: the order requiring the government to give 72 hours’ notice before transferring Doe to Country A or B (the only countries as to which the government appeals the notice obligation); and the order prohibiting the government from transferring Doe to Country B. While both orders are denominated preliminary injunctions, the latter appears to function as a permanent injunction.

A district court facing a request for a preliminary injunction must balance four factors: (i) whether the party seeking the injunction is likely to succeed on the merits of the action, (ii) whether the party is likely to suffer irreparable harm without an injunction, (iii) whether the balance of equities tips in the party’s favor, and (iv) whether an injunction would serve the public interest. Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). The same factors apply when a party seeks a permanent injunction, except the party must show “actual success” on the merits rather than just a likelihood. Id. at 32. We review the district court’s balancing of those considerations for an abuse of discretion, but review any underlying legal conclusions de novo. Abdullah v. Obama, 753 F.3d 193, 197-98 (D.C. Cir. 2014).

We first consider the order enjoining the Secretary from transferring Doe to Country B. We address each of the injunction factors in order. 11 1.

In assessing whether Doe has succeeded on the merits, the relevant question is whether, in the circumstances of this case, involuntarily transferring Doe to Country B would be unlawful. We hold that it would be.

The government makes two species of arguments as to why the Executive has the power to transfer Doe to Country B without his consent. The first rationale has no necessary connection to Doe’s designation as an enemy combatant, or even to the wartime context of this case. It instead relies on a general understanding that, when a foreign country wants to prosecute an American citizen already in its territory for a crime committed within its borders, the Executive can relinquish him to that country’s custody for criminal proceedings. The government’s second rationale, unlike the first, hinges on Doe’s status as an enemy combatant. That second strand of the argument relies on the military’s asserted authority under the law of war to transfer an enemy combatant (including an American citizen) to an allied country in the conflict.

Neither of the government’s rationales, we conclude, supports the involuntary transfer of Doe to Country B, at least as things currently stand. In reaching that conclusion, we rely on the same undisputed facts as our dissenting colleague: that Doe is an American citizen, that he is in U.S. custody in Iraq, that the government believes he is an ISIL combatant, and that he objects to the government’s forcible transfer of him to the custody of Country B. Dissent, at 3-4, 27. While our colleague would conclude that the Executive can forcibly transfer Doe to Country B in those circumstances, we respectfully disagree for the reasons explained in this opinion. 12 a.

A fundamental attribute of United States citizenship is a “right to . . . remain in this country” and “to return” after leaving. Mandoli v. Acheson, 344 U.S. 133, 139 (1952). That right is implicated when the government seeks to forcibly transfer an American citizen from the United States to a foreign country. To effect such a transfer, the government must both (i) demonstrate that a treaty or statute authorizes the transfer, and (ii) give the citizen an opportunity to challenge the factual basis for the transfer. Valentine v. United States ex rel. Neidecker, 299 U.S. 5, 9 (1936); Collins v. Loisel, 259 U.S. 309, 316-17 (1922).

The government’s first argument in this case, though, is that a citizen loses both of those protections the instant he leaves U.S. territory. When a citizen sets foot outside the United States, the government says, the Executive can forcibly transfer him to the custody of any country having a “legitimate sovereign interest” in him. The transfer, the government emphasizes, would be “total.” No. 18-5110, Gov’t Second Supp. Br. 8. Following the citizen’s transfer, then, he would be fully—and irrevocably—subject to the power of the foreign sovereign now holding him.

i. The government’s contention that it possesses that kind of transfer authority over an American citizen is centrally predicated on Munaf v. Geren, 553 U.S. 674, which is itself predicated on Wilson v. Girard, 354 U.S. 524. We disagree with the government’s understanding of those decisions.

In Wilson, William Girard, a U.S. soldier stationed in Japan, was accused by Japan of committing a homicide in its territory. 354 U.S. at 525-26. The Army agreed to relinquish Girard to Japanese custody for pretrial detention. Id. at 526. 13 Girard filed a habeas petition, and the district court issued a preliminary injunction prohibiting the transfer. Id. The Supreme Court vacated the order and allowed the handover of Girard to Japanese custody.

The Court began by recognizing that, as a general matter, a “sovereign nation has exclusive jurisdiction to punish offenses against its laws committed within its borders.” Id. at 529. Japan had voluntarily surrendered that prerogative in a security agreement with the United States that governed the treatment of U.S. soldiers stationed in Japan. But the agreement permitted the United States to cede back to Japan the authority to prosecute a service member in a given instance. Id. at 527-29. In Girard’s case, the United States had done just that. Id. at 529. So the question, the Court said, was whether there was any “constitutional or statutory barrier” to the Executive (i) waiving the United States’s jurisdiction and (ii) transferring Girard to Japan to face criminal prosecution. Id. at 530. Finding no such barrier, the Court sanctioned Girard’s transfer to Japanese custody. Id.

In Munaf, the Court again applied the principle recognized in Wilson—i.e., that, when a foreign country wishes to prosecute an American citizen who is within its borders for a crime he committed while there, the Executive can relinquish him to the country’s custody. Munaf involved two American citizens who voluntarily traveled to Iraq and allegedly committed crimes while there. 553 U.S. at 679. A multinational military coalition identified the two citizens as security risks, and they were held by U.S. military forces in Iraq “[p]ending their criminal prosecution for those offenses” in Iraqi courts. Id. at 705; see id. at 681, 683. Both of the citizens filed habeas petitions, asserting (i) that the Executive lacked the power to transfer them to Iraq’s custody for criminal proceedings, and (ii) that transferring them thus would violate 14 the Due Process Clause. Id. at 692. The Court rejected their arguments and allowed the military to relinquish them to Iraqi custody. Id. at 705.

Relying on Wilson, the Court emphasized that a country has a “sovereign right to ‘punish offenses against its laws committed within its borders.’” Id. at 692 (quoting Wilson, 354 U.S. at 529). That sovereign entitlement, the Court observed, was one that the Court had long and repeatedly recognized. Id. at 694-95 (citing, e.g., Schooner Exchange v, McFaddon, 11 U.S. (7 Cranch) 116 (1812); Neely v. Henkel, 180 U.S. 109 (1901); Kinsella v. Krueger, 351 U.S. 470 (1956)). An order prohibiting the Executive from transferring the two petitioners to Iraqi authorities would infringe that time-honored right. 553 U.S. at 697-98. The Executive thus could transfer the petitioners to Iraqi custody without violating the Due Process Clause. Id. at 699-70.

In both Munaf and Wilson, the authority of the Executive to transfer U.S. citizens had no roots in any military authority over enemy combatants under the law of war. Wilson, after all, concerned “the peacetime actions of a [U.S.] serviceman,” not the wartime actions of an enemy combatant. Id. at 699. In Munaf, meanwhile, it is true that the alleged crimes involved insurgent acts committed in a time of war, for which both suspects had been designated “security internees” and one had been deemed an enemy combatant. See id. at 681-84, 705. But the Court’s recognition of the Executive’s power to transfer the two men did not depend on those designations or on the nature of the alleged crimes. That is evident from the Court’s heavy reliance on Wilson, a case having nothing to do with military authority in wartime.

In accordance with that understanding, the Court in Munaf observed that “[t]hose who commit crimes within a sovereign’s 15 territory may be transferred to that sovereign’s government for prosecution” even if the “crime at issue” is an inherently non- war offense like “embezzlement.” Id. at 699-700 (discussing Neely v. Henkel, 180 U.S. 109 (1901)). To be sure, “there is hardly an exception to that rule when the crime” is “unlawful insurgency directed against an ally during ongoing hostilities.” Id. at 700. So while the war-related context in which the crimes arose in Munaf was not a necessary condition for the Executive to possess the transfer authority recognized in Wilson, that context of course did not diminish the Executive’s authority.

ii. In holding that the Executive had the power to transfer the Munaf petitioners, the Court distinguished its previous decision in Valentine v. United States ex rel. Neidecker, 299 U.S. 5. Because Doe chiefly relies on Valentine in arguing that the military lacks authority to transfer him to Country B, whereas the government centrally relies on Munaf in arguing the opposite, the Munaf Court’s treatment of Valentine warrants our careful examination.

In Valentine, three American citizens fled to New York City after being accused by France of committing crimes within its territory. Id. at 6. France requested the citizens’ extradition, and U.S. officials arrested the three men. Id. The men then filed habeas petitions, arguing that, because the extradition treaty between the United States and France contained no obligation for either country to hand over its own citizens, the Executive lacked the power to extradite them. Id. The Court agreed, holding that the power to extradite “is not confided to the Executive in the absence of treaty or legislative provision.” Id. at 8. Valentine thus establishes that the Executive’s power to extradite a citizen from the United States to another country must come from a treaty or statute. Id. at 9; see Munaf, 553 U.S. at 704. 16 Relying on Valentine, Doe contends that the Executive cannot transfer him from U.S. custody to another country’s custody unless the transfer is authorized by a treaty or statute. The petitioners in Munaf made the same argument in resisting their transfer to Iraqi custody. Munaf, 553 U.S. at 704. The Court, though, found Valentine “readily distinguishable.” Id. It explained that Valentine “involved the extradition of an individual from the United States.” Id. The Munaf petitioners, by contrast, had “voluntarily traveled to Iraq and [were] being held there.” Id. They were therefore “subject to the territorial jurisdiction of that sovereign, not of the United States.” Id.

The Court, for that reason, denied the contention that the Executive invariably “lacks the discretion to transfer a citizen absent a treaty or statute.” Id. at 705. Wilson, the Court said, “forecloses” that contention. Id. That is because the only conceivable authority in Wilson was the security agreement governing the treatment of U.S. service-members in Japan— which, while authorized by a treaty, was not itself a treaty or statute. Id. “Nevertheless,” the Munaf Court observed, “in light of the background principle that Japan had a sovereign interest in prosecuting crimes committed within its borders,” the Wilson Court had “found no ‘constitutional or statutory’ impediment to the United States’s waiver of its jurisdiction” over Girard and its ensuing transfer of him to Japanese custody. Id.

iii. Because Munaf and Wilson recognized the Executive’s authority to transfer American citizens to foreign custody without having to satisfy Valentine’s treaty-or-statute rule, it is apparent that the Executive need not invariably meet the Valentine test to effect a forcible transfer. So some transfers of American citizens to foreign custody are governed by Valentine; others are not. Into which of those camps does the proposed transfer of Doe to Country B fall? 17 In arguing that it can forcibly transfer Doe, the government reads Valentine, Munaf, and Wilson to yield the following set of rules. Under Valentine, an American citizen in the United States cannot be forcibly transferred to a foreign country absent a statute or treaty (such as an extradition treaty) authorizing the transfer. But under Munaf and Wilson, the government says, once a citizen voluntarily leaves the United States, the Executive can pick her up and deliver her to any foreign country that has a “legitimate sovereign interest” in her. No. 18-5032, Gov’t Opening Br. 27; No. 18-5032, Gov’t Reply Br. 15; No. 18-5110, Gov’t Supp. Br. 5; No. 18-5110, Gov’t Second Supp. Br. 3. And a country’s interest in a person qualifies as “legitimate,” the government submits, if, under international law, the country would have “prescriptive jurisdiction” over her—that is, the power to prescribe legal rules regulating her pertinent conduct. No. 18-5032, Gov’t Opening Br. 23 (citing Restatement (Fourth) of the Foreign Relations Law of the United States § 211 (Am. Law Inst. Draft No. 2, 2016)); see also No. 18-5032, Gov’t Reply Br. 15; No. 18-5110, Gov’t Supp. Br. 4-5; No. 18-5110, Gov’t Second Supp. Br. 4.

We cannot accept the government’s submission. Munaf and Wilson do not suggest a general prerogative on the part of the Executive to seize any American citizen voluntarily traveling abroad for forcible transfer to any country with some legitimate sovereign interest in her. Consider again the facts of Valentine. There was no doubt of the legitimacy of France’s interest in the U.S.-citizen petitioners in that case: they had allegedly committed crimes in France. The Executive nonetheless lacked unilateral authority to “dispose of the[ir] liberty” by extraditing them. 299 U.S. at 9. That is because, the Court said, there is generally “no executive discretion to surrender [a person] to a foreign government, unless . . . [a] statute or treaty confers the power.” Id. 18 Under the government’s theory, though, everything would have changed the moment one of the Valentine petitioners voluntarily ventured outside the United States—say, on a family vacation to the Canadian side of Niagara Falls. At that moment, the unilateral “executive discretion” found lacking in Valentine ostensibly would have sprung to life, such that the person—though an American citizen—could have been seized by the Executive and forcibly transferred to France. Cf. United States v. Alvarez-Machain, 504 U.S. 655, 669-70 (1992) (involving the seizure in Mexico (of a non-U.S. citizen) for transfer to the United States).

That expansive vision of unilateral Executive power over a U.S. citizen who ventures abroad does not follow from Munaf and Wilson. Those cases did not involve a citizen forcibly transferred from one foreign country they voluntarily visited to the custody of another foreign country. The cases instead involved “the transfer to a sovereign’s authority of an individual . . . already . . . in that sovereign’s territory.” Munaf, 553 U.S. at 704. The petitioners in Munaf had “voluntarily traveled” to Iraq, id. at 681, 683, and the petitioner in Wilson, an Army specialist, was stationed in Japan, 354 U.S. at 525-26. They were “therefore subject to the territorial jurisdiction of [those] sovereign[s], not of the United States.” Munaf, 553 U.S. at 704. The petitioners in those cases, already present in the sovereign’s territory, could be relinquished by the Executive to that sovereign for prosecution of offenses allegedly committed while there.

That transfer power, the Munaf Court explained, is grounded in the receiving country’s “territorial jurisdiction” over a person who has “voluntarily traveled” to its territory and is “being held there.” Id. The government, though, reads Munaf and Wilson to embrace a transfer power extending to a receiving country’s “prescriptive jurisdiction,” not just its 19 territorial jurisdiction. E.g., No. 18-5032, Gov’t Opening Br. 23. And a country’s prescriptive jurisdiction under customary international law, the government emphasizes, extends to any “individual with a ‘genuine connection’ to the state, even when the individual is located outside the state’s territory.” Id. (emphasis added); see also Restatement (Fourth) of the Foreign Relations Law of the United States § 211 (Draft No. 2, 2016).

The government is surely correct that a sovereign’s prescriptive jurisdiction—its power to regulate conduct— extends to persons located beyond its borders. The practice of extraditing individuals from abroad, and the existence of laws with extraterritorial reach, both illustrate the point. But the fact that a foreign country may have prescriptive jurisdiction over an American citizen who is outside its territory hardly means that, as long as the citizen is somewhere else abroad, the Executive has power to seize her and deliver her to that foreign country.

Indeed, we know of no instance—in the history of the United States—in which the government has forcibly transferred an American citizen from one foreign country to another. (That includes the case of Amir Meshal, in which the government ardently denied a citizen’s allegations that foreign officials, who had moved him from Kenya, to Somalia, to Ethiopia, were acting at the United States’s behest. See Meshal v. Higgenbotham, 47 F. Supp. 3d 115, 119 (D.D.C. 2014), aff’d, 804 F.3d 417 (D.C. Cir. 2015)). Especially in habeas cases like this one, “history matters.” Omar, 646 F.3d at 19.

To that end, the absence of even a single known example of the unilateral power the Executive claims here is illuminating. Indeed, we are unaware of any involuntary transfer of a U.S. citizen from one foreign country to another even pursuant to a treaty or statute. There is all the more 20 reason, then, to proceed with considerable caution before recognizing such a power as a unilateral (although apparently never-before-exercised) prerogative of the Executive.

The implications of the government’s reading of Munaf and Wilson amplify the reasons to reject it. Consider, for example, a U.S. citizen who becomes a journalist, travels to Thailand for a multi-year assignment, and, on returning to the United States, writes articles critical of the Thai King that are alleged to play some role in sparking demonstrations in Thailand. Thailand might well argue that she falls within its prescriptive jurisdiction. And its arguments would have force if, for instance, she underpaid her Thai taxes while there, or her articles were deemed to have had a “substantial effect” within Thailand. See Restatement (Fourth) of the Foreign Relations Law of the United States §§ 211 & cmt. f, 213 (Draft No. 2, 2016).

If the government were right about Munaf and Wilson, then the moment the journalist stepped outside the United States, the Executive would have unilateral power to apprehend her and forcibly transfer her to Thailand if she were accused of violating Thai law. (Incidentally, there is a good reason to think the U.S.-Thai extradition treaty would not apply in that instance, given that it covers only “persons found in the territory of one of the Contracting Parties.” Extradition Treaty, U.S.-Thai., art. 1, Dec. 14, 1983, S. Treaty Doc. No. 98-16.) By the government’s logic, then, alleged breaches of the Thai tax code would authorize a forcible transfer. So too would alleged violations of Thailand’s lèse-majesté statute—under which anyone who “defames, insults, or threatens the [Thai] King . . . shall be punished with imprisonment of three to fifteen years.” Crim. Code B.E. 2499 § 112 (1956), amended by Crim. Code (No. 17), B.E. 2547 (2003) (Thai.); see Lese- Majeste Explained: How Thailand Forbids Insult of its 21 Royalty, BBC.com (Oct. 6, 2017) (discussing recent lèse- majesté prosecutions).

We cannot accept that, if Thailand were to accuse the American journalist of underpaying taxes or penning articles critical of the King, the Executive would have unilateral power to apprehend and forcibly (and irrevocably) transfer her to Thai custody whenever she ventures outside the United States. Indeed, the implications of the government’s argument are more far reaching still. Imagine that the journalist is a dual citizen of the United States and Thailand. If so, Thailand would have prescriptive jurisdiction over her regardless of any violation of Thai law, because, like all sovereigns, it has an “interest in retaining control over its nationals and residents, wherever they may be.” Restatement (Fourth) of the Foreign Relations Law of the United States § 214 cmt. a (Draft No. 2, 2016). Under the government’s theory, then, the Executive could forcibly transfer the journalist to Thai custody for any reason Thailand saw fit, including, say, that she would be a useful witness in a Thai trial. Cf. Blackmer v. United States, 284 U.S. 421, 436-37 (1932).

Thailand’s mere desire to have one of its citizens back cannot give the Executive the unilateral authority to forcibly transfer an American there, just because she steps outside the United States. After all, a dual citizen “is entitled to all the rights and privileges of [U.S.] citizenship.” Perkins v. Elg, 307 U.S. 325, 349 (1939). That includes the “right to return to and remain” in the United States after having left. Mandoli, 344 U.S. at 139.

To be sure, if Thailand asked the United States for help in delivering the journalist to its custody (Thailand presumably would be reluctant to seize a U.S. citizen on its own), the Executive could (and presumably would) decline to do so as a 22 matter of discretion. But the question for us is an antecedent one: whether, in the first place, the Executive would have the unilateral power to forcibly transfer an American citizen to another country merely because she travels abroad. We think the answer is no.

The government emphasizes that, on the facts of this case, Doe is not just any citizen who traveled someplace abroad and is suspected of conduct like tax evasion. Rather, he went to an active battlefield; and Country B, a “coalition partner[] in an ongoing armed conflict” against ISIL, has, the government says, “an obvious and legitimate interest in taking custody of” him. No. 18-5032, Gov’t Reply Br. 6.

Those circumstances, however, do not give the Executive transfer power under Munaf and Wilson that it would otherwise lack. Munaf and Wilson, as explained, do not rest on the military’s authority under the law of war. And we have declined to read those decisions to manifest a principle of prescriptive jurisdiction under which the Executive can forcibly transfer a U.S. citizen who has traveled abroad to any other country with a legitimate sovereign interest in her. That a country may have an especially important interest in a citizen—including by reason of her allegedly hostile actions against the country’s interests in a time of war—does not affect that conclusion.

Does this mean that the military necessarily is without power in a time of war to transfer an enemy combatant who is a U.S. citizen to an allied country’s custody? No, it does not. It means that the authority to effect such a transfer does not come from the general transfer power recognized in Munaf and Wilson. The authority instead would come from the Executive’s wartime powers under the law of war, a subject we turn to next. 23 b.

The government, as noted, has said in this case that its “determination that [Doe] is an enemy combatant . . . is not the basis for the U.S. military’s authority to transfer” him to Country B. No. 18-5032, Gov’t Reply Br. 8. At the same time, though, the government has also said that “battlefield detainees” like Doe are “lawfully transferrable under the laws of war.” Id. at 11; see also id. at 13 (“[P]etitioner’s status as a U.S. citizen imposes no special constraints on the U.S. military’s ability to transfer him consistent with the laws of war.”); No. 18-5110, Gov’t Second Supp. Br. 3 (arguing that transfer is permissible, in part because of “the Department of Defense’s good-faith determination . . . that [Doe] is an enemy combatant”).

We now take up the latter facet of the government’s claim of authority to transfer Doe: that it can do so pursuant to the Executive’s wartime powers under the law of war. We conclude that the Executive does generally possess authority under the law of war to transfer an enemy combatant to the custody of an ally in the conflict. But that authority, we hold, could potentially support a transfer of Doe only if the government (i) demonstrates that it is legally authorized to use military force against ISIL, and (ii) affords Doe an adequate opportunity to challenge the Executive’s factual determination that he is an ISIL combatant.

i. The starting point for our analysis is the Supreme Court’s decision in Hamdi v. Rumsfeld, 542 U.S. 507 (2004). (Because the plurality in Hamdi issued the controlling opinion, which our court has treated as binding, see Al-Bihani v. Obama, 590 F.3d 866, 872 (D.C. Cir. 2010), we will treat the plurality opinion as that of the Court for purposes of this opinion.) There, the Court spoke directly to the military’s authority over 24 an American citizen under the law of war. The case involved Yaser Esam Hamdi, who, like Doe, was captured on a foreign battlefield, where the government alleged he had fought with the Taliban against the United States. Id. at 510, 512-13. Hamdi, again like Doe, was a dual citizen of the United States and Saudi Arabia. See Man Held as Enemy Combatant to Be Freed Soon, CNN.com (Sept. 22, 2004.)

The military initially detained Hamdi in Afghanistan and at Guantanamo Bay, and then, upon learning he was an American citizen, brought him to the United States for continued detention. 542 U.S. at 510. Hamdi then filed a habeas petition seeking release from his military custody, alleging that his detention without criminal charge violated his rights under the Due Process Clause. Id. at 511.

The Court first held that the military had legal authority to detain Hamdi for the duration of the conflict in which he was captured. That power flowed from the 2001 Authorization for Use of Military Force (AUMF), Pub. L. No. 107-40, 115 Stat. 224. 542 U.S. at 517. The 2001 AUMF authorized the President to “use all necessary and appropriate force against those nations, organizations, or persons [that] he determines planned, authorized, committed, or aided the terrorist attacks” of September 11, 2001. Id. at 510 (quoting 115 Stat. 224, § 2(a)). The Court found “no doubt” that Taliban combatants (like Hamdi was alleged to be) fit within that description. Id. at 518. And the Court explained that detention of enemy combatants “for the duration of the particular conflict in which they were captured” is “so fundamental and accepted an incident to war as to be an exercise of the ‘necessary and appropriate force’ Congress ha[d] authorized the President to use.” Id. 25 The Court next addressed whether Hamdi’s U.S. citizenship affected the Executive’s power to detain him. On that issue, the Court found “no bar to this Nation’s holding one of its own citizens as an enemy combatant.” Id. at 519. After all, “[a] citizen, no less than an alien, can be part of or supporting forces hostile to the United States or coalition partners and engaged in an armed conflict against the United States.” Id. (internal citation and quotation marks omitted).

Finally, the Court turned to “the question of what process is constitutionally due to a citizen who disputes his enemy- combatant status.” Id. at 524. The government argued that its determination to that effect should be subject to highly deferential review, solely to confirm the existence of some evidence supporting it. Id. at 527. The government emphasized the “limited institutional capabilities of courts in matters of military decision-making in connection with an ongoing conflict.” Id. The Court disagreed with the government.

Because “due process demands some system for a citizen- detainee to refute his classification,” the Court explained, “the proposed ‘some evidence’ standard [was] inadequate.” Id. at 537. Rather, “a citizen-detainee seeking to challenge his classification as an enemy combatant must receive notice of the factual basis for his classification, and a fair opportunity to rebut the Government’s factual assertions before a neutral decisionmaker.” Id. at 533. That process, the Court observed, could potentially be afforded in a military proceeding. Id. at 538. The Court also clarified, however, that “initial captures on the battlefield need not receive the process” the Court had outlined. Id. at 534. Rather, that “process is due only when the determination is made to continue to hold” a combatant. Id. 26 After Hamdi, we know that if there is legal authority to exercise military force against an enemy, that authority encompasses detention of an enemy combatant for the duration of the conflict. And we further know that the detention authority more generally extends to an enemy combatant who is an American citizen. But a citizen, Hamdi instructs, must have a meaningful opportunity to challenge the factual basis for his designation as an enemy combatant in accordance with the procedures set forth by the Court.

ii. Whereas Hamdi addressed whether the Executive can detain an alleged enemy combatant who is a citizen, this case (at least at this stage) instead involves whether the Executive can transfer him to the custody of another country. That naturally raises two sets of questions. First, is the Executive’s transfer authority (this case) on par with its detention authority (Hamdi) as a fundamental incident of waging war? Second, if so, is the Executive’s exercise of transfer authority against a U.S. citizen subject to the same conditions attending the exercise of detention authority against a U.S. citizen? In other words, do transfer authority over citizens and detention authority over citizens essentially rise or fall together? We conclude they do.

First, the military possesses settled wartime authority under the law of war to transfer enemy combatants to allied countries. That power, in the words of Hamdi, is “a fundamental incident of waging war,” such that the Executive generally has the authority to transfer when it has legal authorization to engage in hostilities. Id. at 519.

Congress confirmed as much in the National Defense Authorization Act (NDAA) for Fiscal Year 2012, Pub. L. No. 112-81, 125 Stat. 1298 (Dec. 31, 2011). There, Congress elaborated on the authority conferred by the 2001 AUMF. It 27 affirmed that the AUMF grants detention authority pending decision of an enemy combatant’s “disposition under the law of war”; and it enumerated the available “dispositions” to include “[t]ransfer to the custody or control of the person’s country of origin, any other foreign country, or any other foreign entity.” Id. § 1021(a), (c). Congress thus expressly considers transfer of an enemy combatant to be one option available to the military under the law of war. The Department of Defense’s directives are to the same effect. U.S. Dep’t of Def., Directive No. 2310.01E, § 3.m (May 24, 2017).

That understanding is firmly rooted in historical practice. “Throughout the 20th Century, the United States transferred or released hundreds of thousands of wartime alien detainees— some of whom had been held in America—back to their home countries, or in some cases, to other nations.” Kiyemba v. Obama, 561 F.3d 509, 519-20 (D.C. Cir. 2009) (Kavanaugh, J., concurring). In World War I, for instance, the United States regularly transferred captured combatants to France, an ally. See George G. Lewis & John Mewha, History of Prisoner of War Utilization by the United States Army 1776-1945, Dep’t of the Army Pamphlet No. 20-213, at 59 (1955), available at https://cgsc.cdmhost.com. And in World War II, the United States transferred hundreds of thousands of Axis soldiers to allies like Belgium, France, and Luxembourg, where the soldiers were used as agricultural workers and underwent rehabilitation. Id. at 240-41. Transfers to allies were also commonplace during the Vietnam and Gulf Wars. See George S. Prugh, Law at War: Vietnam 1964-1973, at 62 (1975); U.S. Dep’t of Def., Office of Gen. Counsel, Law of War Manual at 633 n.742 (Dec. 2016). “Transfers,” in short, “are a traditional and lawful aspect of U.S. war efforts.” Kiyemba, 561 F.3d at 519 (Kavanaugh, J., concurring). 28 Even if transfers of alien combatants have been a regular feature of warfare, does the traditional authority to transfer enemy combatants extend to a U.S. citizen? On this score, the historical evidence is sparse. As noted, we know of no instance in which the Executive has forcibly transferred a citizen from one foreign country to another; and that includes wartime transfers of enemy combatants.

Hamdi, however, instructs that a traditional military power over enemy combatants in wartime should generally be assumed to encompass American citizens. The Court reasoned that a citizen, “no less than an alien,” can be a part of an enemy force. 542 U.S. at 519. For that proposition, the Court relied on its decision in Ex parte Quirin, 317 U.S. 1 (1942), in which it had upheld the military trial of a U.S. citizen for his unlawful belligerency in support of the enemy in World War II, id. at 30- 31.

To be sure, Justice Scalia, dissenting in Hamdi, discounted Quirin as “not [the] Court’s finest hour.” 542 U.S. at 569 (Scalia, J., dissenting). He would have held that the military’s wartime authority over enemy combatants—including, presumably, transfer authority—does not extend to a U.S. citizen (at least absent a suspension of the writ by Congress). See id. at 554. The Court, though, adhered to Quirin notwithstanding Justice Scalia’s critique. Id. at 522-23. It thus found no reason to exclude U.S. citizens from the Executive’s fundamental authority under the law of war to detain enemy combatants for the duration of a conflict. Id. at 519. Following the approach set out in Hamdi, we similarly see no basis for excluding a citizen—at least as a categorical matter—from the Executive’s wartime authority to transfer enemy combatants.

Hamdi referenced a Ninth Circuit decision upholding the Executive’s power to detain, as a prisoner of war, a dual U.S.- 29 Italian citizen who was a member of the Italian forces in World War II. Id. at 524 (discussing In re Territo, 156 F.2d 142 (9th Cir. 1946)); see also Ronald D. Rotunda, The Detainee Cases of 2004 and 2006 and Their Aftermath, 57 Syracuse L. Rev. 1, 13 n.73 (discussing Territo’s dual citizenship). That decision also contemplated that he would be sent from the United States back to Italy at the war’s end. See 156 F.2d at 144. True, that contemplated transfer would have been a “repatriation” to the enemy state, which, under the law of war, is distinct from a transfer to an ally (and which, presumably, would result in release rather than continued detention). Compare Geneva Convention (III) Relative to the Treatment of Prisoners of War, art. 12, Aug. 12, 1949, 6 U.S.T. 3316, with id. at art. 118. And Territo’s repatriation might well have been voluntary, especially given his family and other connections to Italy (he sought release from his detention in the U.S, and the opinion gives no indication that he wanted to stay here if released). See 156 F.2d at 143. Still, Territo offers modest support for the conclusion that the Executive’s power to transfer under the law of war applies to both aliens and citizens. And Hamdi, again, teaches that both aliens and citizens may be subject to the Executive’s wartime authority.

Second, having determined that the Executive has authority to transfer enemy combatants under the law of war, and that there is no blanket exemption from that power for U.S. citizens, we now assess whether Hamdi’s conditions on the exercise of detention authority equally govern any exercise of transfer authority. Those conditions, again, are that the Executive have legal authority to use military force against the relevant enemy (here, ISIL), and that the citizen be afforded the process laid out in Hamdi for challenging the factual determination that he is an enemy combatant. 30 In considering whether transfer should be subject to those conditions, an initial point bears noting: the transfer of a citizen to another country’s custody, unlike continued detention of that citizen, is irrevocable. Once the Executive relinquishes custody of an American citizen to another country, our government, and our laws—including our law’s habeas guarantee, which a detainee can use to seek relief from detention over time—would be unavailable to her, perhaps in perpetuity. Decisions about the duration and conditions of her custody, and about the availability to her of a means of challenging her confinement, would be entirely up to the detaining sovereign.

The government asserts that, when we assess a potential transferee’s liberty interests, we cannot factor in her continued detention in the receiving country. That, the government says, follows from our holding in Kiyemba. 561 F.3d at 515-16. Here, though, the central issue is not the prospect of continued detention in Country B, but rather the forcible transfer itself, which would involuntarily send an American citizen from U.S. custody to the custody of another country.

In that regard, Kiyemba is starkly different; there, it was undisputed that the detainees had no cognizable interest against being moved from Guantanamo to a foreign country. (Indeed, because transfer was the only relief available to the petitioners—who, as aliens, had no right to be released into the United States—they affirmatively sought to be moved to a foreign country. Id. at 519 n.5 (Kavanaugh, J., concurring)). Here, by contrast, the transfer centrally implicates Doe’s interest in not being forcibly moved into Country B’s custody. Indeed, involuntary transfer of a citizen to the custody of another sovereign—including via extradition—undoubtedly involves fundamental liberty interests that can be vindicated in habeas corpus. E.g., Valentine, 299 U.S. at 9 (“no executive 31 prerogative to dispose of the liberty of the individual” by way of extradition); Landon v. Plasencia, 459 U.S. 21, 36 (1982). Cf. Sessions v. Dimaya, 138 S. Ct. 1204, 1213 (2018) (deportation from the United States can be viewed a more “severe penalty” for criminal misconduct than imprisonment in the United States).

Given that transfers involve fundamental liberty interests, we see no basis for concluding that, for the transfer of a citizen (as opposed to the detention of a citizen), the Executive need not satisfy the Hamdi conditions. The 2012 NDAA is instructive in this regard. There, Congress set out four types of “disposition[s] under the law of war” that the Executive could choose for an enemy combatant, including “[d]etention under the law of war without trial until the end of the hostilities,” and “[t]ransfer to the custody or control of the person’s country of origin [or] any other foreign country.” Pub. L. No. 112-81 § 1021(c)(1), (4). The statutory structure indicates that Congress saw transfer and detention as two options falling on largely the same plane—not as one option (transfer) broadly available in circumstances in which the other (detention) would not be.

Significantly, our decisions draw an equivalence between transfer of citizens and detention of citizens. We have rejected the notion “that the Executive Branch may detain or transfer Americans or individuals in U.S. territory at will, without any judicial review of the positive legal authority for the detention or transfer.” Omar, 646 F.3d at 24 (emphases added). And we have said that “Congress cannot deny an American citizen or detainee in U.S. territory the ability to contest the positive legal authority (and in some situations, also the factual basis) for his detention or transfer unless Congress suspends the writ.” Id. (emphasis added). For either “detention or transfer,” then, an 32 “American citizen” is entitled to challenge both “legal authority” and “factual basis,” as Hamdi envisions.

The government reads the just-quoted language from our decision in Omar to say that an American citizen can bring a “legal authority” or “factual basis” challenge to her “detention or transfer” only if she is in the United States. See No. 18-5032, Gov’t Reply Br. 14. That is an unsustainable reading. Hamdi itself rejects the notion that it could “make a determinative constitutional difference” if an American citizen were detained overseas rather than in the United States. 542 U.S. at 524. The Court understood that any such conclusion would “create[] a perverse incentive” to hold American citizens abroad. Id.

The Omar court’s reference to a challenge brought by “an American citizen or detainee in U.S. territory” thus plainly speaks to a challenge brought by a citizen anywhere or by an alien detained in U.S. territory (such as Guantanamo Bay). Omar, 646 F.3d at 24 (citing Boumediene v. Bush, 553 U.S. 723, 785-86 (2008)); see also Al Bahlul v. United States, 767 F.3d 1, 65 n.3 (D.C. Cir. 2014) (Kavanaugh, J., concurring in the judgment in part and dissenting in part) (“As a general matter, the U.S. Constitution applies to U.S. citizens worldwide and to non-U.S. citizens within the 50 states and the District of Columbia[.]”). There is no basis for thinking that a citizen relinquishes her right to bring a legal challenge to her detention—or, equivalently, to her transfer—if she is detained in (or transferred from) a foreign country. That is why the court in Omar went on to explain that Omar (one of the two Munaf petitioners), who was still being held in Iraq, had the requisite opportunity to contest the legal authority for his transfer. Id. That discussion would have been entirely unnecessary if he had no right to bring that challenge in the first place since he was held overseas. 33 Consider the implications if there were, in fact, an asymmetry between transfer and detention, such that the Executive could transfer a U.S. citizen to another country without meeting the Hamdi conditions. With regard to legal authority, the military could irrevocably transfer a citizen thought to be an enemy combatant even if judicial review would have revealed that the Executive lacked lawful authority to use military force against the particular enemy. In that event, detainees in U.S. custody—and thus protected by U.S. law— would need to be released or criminally charged. But for those who had already been transferred to another country, an American court could not order their return or grant them comparable relief.

With regard to a factual-basis challenge, the Hamdi Court sought to “meet the goal of ensuring that the errant tourist, embedded journalist, or local aid worker has a chance to prove military error.” 542 U.S. at 534. The procedural guarantees prescribed by the Court were intended to guard against an undue risk of an erroneous military determination. See id. But if the transfer of a citizen could be accomplished without affording her those protections, a risk of error thought unacceptable for continued detention would be present for an irrevocable transfer to another country. An “errant tourist” might then be protected against detention but unable to avoid an irrevocable transfer to another country’s custody. Compare 31A Am. Jur. 2d Extradition § 120 (2d ed. 2018) (describing process granted to persons subject to extradition); 18 U.S.C. § 3191.

The government, in that respect, relies on its having made a “good-faith determination, supported by extensive record evidence, that [Doe] is an enemy combatant.” No. 18-5110, Gov’t Second Supp. Br. 3. We do not doubt the government’s good faith. Nor do we discount the importance of the need to 34 avoid unduly burdening the Executive’s prosecution of a war, which concerned the Hamdi Court as well. See 542 U.S. at 531-35. But in Hamdi, one point on which eight Justices agreed was that, in the case of an American citizen, the government’s good-faith determination that he is an enemy combatant is not enough to justify his detention for the duration of a conflict. Id. at 537; id. at 553 (Souter, J., concurring in part, dissenting in part, and concurring in the judgment); id. at 564-65 (Scalia, J., dissenting). We find the same to be true of an irrevocable transfer to another country’s custody.

In that regard, it is instructive to consider the implications of the government’s argument here for the facts of Hamdi itself. Upon holding that the government’s continued detention of Hamdi was contingent on his having a meaningful opportunity to challenge the factual basis for his detention, the Court remanded the matter so that the government could conduct the factfinding process the Court had outlined. See 542 U.S. at 538-39. That process would result in a determination of whether Hamdi was a person against whom military force could be applied.

Under the government’s argument here, though, the Executive, rather than grant Hamdi that process following remand, could have simply avoided it by choosing instead to forcibly and irrevocably transfer him to the custody of another country (pursuant to its authority under the 2001 AUMF). True, the government eventually did in fact transfer Hamdi to Saudi Arabia—but with his consent, not over his objection (and after he renounced his American citizenship). Jerry Markon, Hamdi Returned to Saudi Arabia, Washington Post (Oct. 12, 2004). There is, of course, a vast difference between a voluntary transfer and an involuntary one. As to the latter, we do not believe the Hamdi Court would have countenanced Hamdi’s forcible transfer to another country unless he were 35 first afforded the process the Court held he was constitutionally due.

The government’s final argument on this score is that transfer without process is permissible if effected in conjunction with “initial capture[] on the battlefield.” No. 18- 5110, Gov’t Supp. Br. 8-9 (quoting Hamdi, 542 U.S. at 534). But while Hamdi allows for temporary detention without process attending “initial capture,” a citizen can be released if there ends up being an insufficient factual basis to continue detention. Transfer may be different because it, by nature, is not temporary.

In addition, there would be no citizenship-based limit on transfer unless there were reason to know that a person is a citizen. Cf. Asbury Aff. at 4, United States v. Lindh, No. Crim. 02-MJ-51 (E.D. Va. Jan. 15, 2002) (“[Harakat ul-Mujahideen] officials told [John Walker Lindh] not to admit to anyone that he was American but to say, if asked, that he was from Ireland.”) Here, at any rate, the Executive decided to transfer Doe—and reached an agreement to do so—several months after his capture. Doe v. Mattis, No. 17-cv-2069, Notice at 1 (D.D.C. Apr. 16, 2018), ECF No. 77; Status Hr’g Tr. at 8 (D.D.C. Jan. 22, 2018), ECF No. 55 (stating that no final decision had been made on whether to transfer Doe). This transfer decision, then, was not a battlefield judgment. For those reasons, the Executive cannot transfer Doe at this stage unless he receives the process required by Hamdi.

c.

In light of the above analysis, can the Executive involuntarily transfer Doe to Country B? We conclude it cannot, at least as things stand now. We take up the two strands of the government’s argument in order. 36 i. We first address whether the Executive can forcibly transfer Doe to Country B based on the general transfer authority recognized in Munaf and Wilson. That authority, as we have explained, does not encompass the forcible transfer of a citizen from one foreign country to the custody of another foreign country. Insofar as the transfer of Doe to Country B would be an inter-country transfer, it falls outside of Munaf and Wilson.

The government contends that the transfer nonetheless should be allowed because Doe is a dual citizen of the United States and Country B (Saudi Arabia). As a result, the government emphasizes, Country B has an especially strong interest in accepting custody over Doe: the interest in repatriating one of its own nationals in order to attempt to rehabilitate him consistent with its own laws and practices.

To that end, the government notes that a country always has prescriptive jurisdiction over its own nationals, including when they are abroad. See No. 18-5032, Sealed Gov’t Opening Br. 23 (citing Restatement (Fourth) of the Foreign Relations Law of the United States § 214 & cmt. a (Draft No. 2, 2016)); No. 18-5032, Sealed Gov’t Reply Br. 15; No. 18-5110, Sealed Gov’t Supp. Br. 4; No. 18-5110, Sealed Gov’t Second Supp. Br. 4. And by carrying out the transfer, the government urges, the United States also would further its own interest in maintaining constructive relations with an ally in the military efforts against ISIL (which would, among other benefits, allow for productive discussions with Country B about the transfer of additional combatants in the future). See No. 18-5032, Gov’t Opening Br. 24.

We do not doubt the weight of Country B’s sovereign interests in (and prescriptive jurisdiction over) Doe based on all of those considerations, including, in particular, his Saudi 37 citizenship. Nor do we question the Executive’s assessment of Country B’s interests. See Kiyemba, 561 F.3d at 515. But the strength of Country B’s interests in Doe as a Saudi citizen does not diminish the force of Doe’s rights as a U.S. citizen: here, the right to resist the Executive’s forcible seizure and transfer of him to the custody of another country. After all, “dual citizenship presupposes rights of citizenship in each country.” Kawakita v. United States, 343 U.S. 717, 725 (1952); see Elg, 307 U.S. at 345. And the limits on unilateral Executive authority ultimately “protect the individual.” Bond v. United States, 564 U.S. 211, 222 (2011).

Recall, for instance, the example introduced earlier based on the facts of Valentine: while Valentine held that the Executive lacked unilateral authority to extradite the petitioners to France, the Executive, under the government’s theory, would have gained that authority the moment one of the petitioners stepped across the border into Canada.

Now imagine that the same petitioner had been a dual U.S.-French citizen. True, his French citizenship would have fortified France’s sovereign interests in him (which were already substantial given his alleged commission of crimes in France). But his U.S. citizenship still would have meant that the Executive could not extradite him to France from the United States. See Valentine, 299 U.S. at 10-13. And it still would be anomalous to suppose that the Executive gained the ability to transfer him merely because he set foot in Canada.

Doe’s dual citizenship, in short, does not affect our conclusion that the transfer authority recognized in Munaf and Wilson is inapplicable in this case.

ii. We now turn to whether the forcible transfer of Doe to Country B can be supported by the Executive’s wartime 38 authority over enemy combatants under the law of war. That authority, as we have explained, encompasses transfers of enemy combatants to an allied country. But before the Executive could exercise that transfer power against Doe, the two Hamdi conditions would need to be met.

The first condition is a determination that the Executive has legal authority to wage war against ISIL. “For wartime military transfers,” we have said, “Article II and the relevant Authorization to Use Military Force generally give the Executive legal authority to transfer.” Omar, 646 F.3d at 24. Second, Doe would need to be afforded a meaningful chance to rebut the government’s factual assertion that he is an ISIL combatant, per the requirements set out in Hamdi.

Neither condition has been met at this point. Until those conditions are satisfied, the Executive lacks power under the law of war to transfer Doe to Country B on the basis of his status as an alleged ISIL combatant.

2.

Having addressed Doe’s success on the merits of his claim that a forcible transfer to Country B would be unlawful, we now consider whether he has shown he would be irreparably injured absent the injunction. See Winter, 555 U.S. at 20. We conclude he has made that showing.

A forcible transfer of Doe to the custody of Country B, the government explains, would be “bona fide and total,” in that “[o]nce transfer is effectuated,” he “would be entirely in [Country B’s] custody,” without any continuing oversight by— or recourse to—the United States. No. 18-5032, Gov’t Reply Br. 15. Doe, wishing to avoid that irrevocable change in his station, objects to his proposed transfer to the custody of 39 Country B. No more is required to demonstrate that he would face irreparable injury if he were involuntarily (and irreversibly) handed over to Country B in violation of his constitutional rights.

In contending that Doe fails to establish irreparable injury, the government observes that the point of a habeas petition is to obtain release from U.S. custody. And if the planned transfer of Doe to Country B goes forward, the government observes, he would no longer be in U.S. custody. So transfer, the government says, is thus tantamount to release, and there can be no “irreparable harm from obtaining the very relief his habeas action seeks to obtain.” No. 18-5110, Gov’t Supp. Br. 10.

The government’s position cannot be correct. It would mean that any habeas petitioner objecting to a planned extradition of him would be unable to demonstrate irreparable injury if he were extradited. We know that is not the case. See Belbacha v. Bush, 520 F.3d 452, 456 (D.C. Cir. 2008) (collecting cases granting stays of extradition); Demjanjuk v. Meese, 784 F.2d 1114, 1118 (D.C. Cir. 1986) (“extradition of petitioner to Israel may qualify as a threat of irreparable harm”); see also Nken v. Holder, 556 U.S. 418, 434-35 (2009) (noting “irreparable nature of harm from removal before decision on a petition for review”). Of course, a transfer to a foreign country’s custody necessarily ends U.S. custody; but the transfer itself is a harm that cannot be remedied.

The government similarly observes that, if Doe were released from his U.S. custody in Iraq, he would likely be detained by Iraq. Or, the government says, he might be seized in Iraq by another country. As a result, the government contends, there would be a limited practical difference between 40 the release sought by Doe and a transfer to another country. No. 18-5110, Gov’t Supp. Br. 10-11.

That argument again proves too much. As to the first point, if detention in Iraq were equivalent to detention in another country for purposes of irreparable injury, then a citizen who could lawfully be transferred to one country could never secure an injunction prohibiting his ensuing transfer to any other country. By the government’s logic, once the Munaf Court blessed the petitioners’ transfer to Iraq, they would have been categorically precluded from getting an injunction barring their transfer to Albania, Zambia, or any country in between. As to the second point, once a petitioner is released, he could conceivably be seized by any country. So if the mere possibility of apprehension by a country meant that a petitioner would not be harmed by transfer there, then courts could never enjoin a transfer to any country on the globe. That is not the law.

3.

When a private party seeks injunctive relief against the government, the final two injunction factors—the balance of equities and the public interest—generally call for weighing the benefits to the private party from obtaining an injunction against the harms to the government and the public from being enjoined. See Pursuing America’s Greatness v. FEC, 831 F.3d 500, 511 (D.C. Cir. 2016). We find the balance to tip in Doe’s favor.

The equities at stake on both sides are manifestly weighty ones. The government seeks to avoid undue interference with its military judgments in connection with ongoing hostilities and with its conduct of foreign relations with a coalition partner in that campaign. Doe, meanwhile, seeks to vindicate his rights 41 as an American citizen to avoid a forcible and irrevocable transfer to (potentially indefinite) custody at the hands of a foreign sovereign.

As the Supreme Court observed in Hamdi, a citizen’s “interest in being free from physical detention” is the “most elemental of liberty interests.” 542 U.S. at 529. The Court therefore denied the Executive the ability to continue detaining an alleged enemy combatant in wartime unless it afforded him procedural protections the Court thought he was constitutionally owed. And the Court did so despite the government’s belief that affording additional process would be unnecessary and unworkable. See id. at 525. Here, we conclude an injunction barring Doe’s forcible transfer to Country B’s custody is warranted for substantially similar reasons and in substantially similar circumstances.

The government also appeals the district court’s order requiring it to give 72 hours’ notice before transferring Doe to either Country A or Country B. With regard to Country B, the government gave the district court the requisite notice before attempting to effect an agreed-upon transfer. When a defendant complies with an injunction in that fashion, its appeal of the injunction becomes moot. See People for the Ethical Treatment of Animals, Inc. v. Gittens, 396 F.3d 416, 421 (D.C. Cir. 2005). At any rate, now that we have sustained the injunction barring Doe’s transfer to Country B, any requirement to give advance notice of such a transfer is beside the point.

The notice requirement still presents an ongoing controversy with regard to Country A, however. An order requiring the government to give advance notice before 42 transferring a detainee to another country cannot be sustained if there could be no grounds for enjoining the transfer. See Kiyemba, 561 F.3d at 514. The government relies on that principle here, contending that any transfer of Doe to Country A invariably would be lawful. We are unpersuaded.

As an initial matter, we note that, because of the way this case developed, Doe did not have a meaningful opportunity to address a potential transfer to Country A. In the government’s opening brief, it made three alternative requests for relief: (i) vacatur of the injunction in its entirety, (ii) vacatur of the injunction as applied to any “country that the Executive Branch determines has a legitimate interest” in Doe, or (iii) vacatur as applied only to one specified country. See No. 18-5032, Gov’t Opening Br. 38. Indeed, the government’s opening brief noted the possibility of transferring Doe to Country A only in passing in a footnote. Id. at 31 n.5. Such a reference is ordinarily inadequate to preserve an argument. See CTS Corp. v. EPA, 759 F.3d 52, 64 (D.C. Cir. 2014). And while the government specifically included Country A as a possible transferee country in its reply brief, that was too late. See Abdullah, 753 F.3d at 199-200.

The lateness of the government’s suggestion that it might wish to transfer Doe to Country A is magnified, because, on the existing record, we know very little about what such a transfer would entail. Unlike with Country B, with whom the government has reached an agreement to transfer Doe, we are aware of no concrete plans in the works (or on the horizon) to transfer Doe to Country A. Indeed, the government has not submitted a single affidavit or declaration discussing a transfer of Doe to Country A, the reasons that might give rise to an agreement to transfer Doe there, the terms or expectations surrounding such a transfer, or the anticipated conditions of his custody after that transfer. The government has listed at a high 43 level of generality some possible interests Country A could have in mind if it were to accept custody of Doe. See No. 18- 5032, Gov’t Reply Br. 8-9. But even with regard to that array of potential interests, we do not know whether a transfer of Doe would occur only for those reasons.

The government thus essentially seeks blanket preapproval to transfer Doe to Country A, regardless of the reasons or circumstances. We decline to recognize that sort of carte-blanche license in the present circumstances. In Munaf, the Supreme Court upheld the transfer of the two habeas petitioners to Iraq’s custody, but only after examining the reasons for the proposed transfers and the governing law. See Omar, 646 F.3d at 24. Here, the government asks for an all- purpose preapproval without any opportunity to assess a particular transfer before it takes place. Particular transfers to Country A may or may not be unlawful depending on the circumstances. The notice requirement secures the ability to make that assessment at a suitable time.

In these circumstances, we cannot set aside the notice requirement as to Country A. In terms of likelihood of success on the merits, with notice of the possibility of a transfer to Country A and at least some factual information about what such a transfer might entail, Doe would have had an opportunity to show that a particular transfer to Country A would be unlawful. With regard to irreparable injury, a particular transfer arrangement, depending on the circumstances, could irrevocably injure his interests, and Doe did not have an opportunity to address in his briefing the potential harm he would suffer if transferred to Country A. And the remaining injunction factors could favor Doe in the context of a concrete transfer proposal. 44 None of this is to say that, in the end, Doe necessarily will be able to show that any agreed-upon transfer to Country A (Iraq) is unlawful. He may or may not be able to do so, depending on considerations such as: (i) whether, given the way Doe came to be in Iraq, his presence there qualifies as “voluntar[y],” Munaf, 553 U.S. at 704, and (ii) whether the reasons for an agreed-upon transfer should bring the case within Munaf’s rule allowing transfers of citizens already within a country’s borders. At this point, without any information about an agreed-about transfer, we decline to set aside the notice requirement with regard to Country A.

* * * * *

We affirm the district court’s injunction barring the government from transferring Doe to Country B, and we also affirm the district court’s injunction requiring the government to give 72 hours’ notice before transferring him to Country A.

Our disposition will constrain the government’s ability to transfer an American citizen believed to be an enemy combatant more than the government would like. That is an important consideration in this case in light of the deference owed to military judgments in wartime. But “such cases,”— i.e., those in which “a United States citizen [is] captured in a foreign combat zone”—“must surely be rare.” Hamdi, 542 U.S. at 571 n.3 (Scalia, J., dissenting) (formatting altered).

In those rare cases, the constraints on the Executive could, in theory, discourage the Executive from taking custody of a suspected enemy combatant known to be an American citizen. That was equally true, though, of the Supreme Court’s decision in Hamdi, which established constraints on the Executive’s treatment of U.S. citizens captured on a foreign battlefield. We 45 adhere to that decision and apply it to military transfers, consistent with our precedent. See Omar, 646 F.3d at 24.

The Hamdi Court believed it “unlikely” that its decision would have a “dire impact on the central functions of warmaking.” 542 U.S. at 534. At the same time, the Court thought it “vital” that it “not give short shrift to the values that this country holds dear or to the privilege that is American citizenship.” Id. at 532; see id. at 558-59 (Scalia, J., dissenting). We follow the Court’s guidance today.

It is so ordered. KAREN LECRAFT HENDERSON, Circuit Judge, dissenting: A reader, having just reviewed the majority opinion, might well be thinking it declares a lead-pipe result. Caveat lector. The opinion treats all but silently the judiciary’s dispositively downsized role in the theater of war. See Al Bahlul v. United States, 792 F.3d 1, 28 (D.C. Cir. 2015) (Henderson, J., dissenting) (in the “thicket” of international politics and “waging war,” “our lack of competence is marked,” “our democratic unaccountability glaring” and “the ramifications of our actions unpredictable” (internal quotation, citations and ellipses omitted)), judgment vacated upon grant of reh’g en banc (Sept. 25, 2015). The majority affirms a preliminary injunction (Order) that ventures well beyond the district court’s limited authority. The Order blocks our military from transferring a battlefield captive, petitioner John Doe, to a country that has a sovereign interest in him based on his citizenship there. 1 The district court does not find—because there is no evidence—that Doe will be mistreated if transferred. Instead, the point of the Order is to ensure that Doe can challenge his custody in the hope of winning release therefrom on his own terms. The Order is without precedent: in Munaf v. Geren, 553 U.S. 674 (2008), the United States Supreme Court vacated the only comparable injunction. And worse than the Order’s novelty is its effect: it disrupts military operations and sovereign-to-sovereign relations half a world away.

1 This case involves materials that have been sealed to protect sensitive diplomatic interests. Consistent with the “presumption of openness in judicial proceedings,” United States v. Microsoft Corp., 56 F.3d 1448, 1464 (D.C. Cir. 1995) (per curiam) (internal quotation omitted), I rely on the public portion of the briefs and record where possible. Where not possible, I rely on sealed information—mainly in footnotes—and redact it from the public version hereof. 2 Affirmance portends a hazardous expansion of the judiciary’s role in matters of war and diplomacy. In defending the Order, Doe relies on Hamdi v. Rumsfeld, 542 U.S. 507 (2004), by which a habeas court reviews the lawfulness of a U.S. citizen’s extended military detention. But Hamdi does not empower a court to enjoin our military from transferring a battlefield captive not facing extended detention. Much less does it authorize injunctive relief where, as here, the receiving country has a facially strong interest in the captive and the Executive Branch has determined in good faith that he is an enemy combatant. Habeas is concerned with Executive Branch “custody,” 28 U.S.C. § 2241(c), not relinquishment of it. Doe erroneously blurs the distinction. He claims that the Executive cannot relinquish custody absent authority to maintain it or, alternatively, absent the “positive legal authority” of an extradition treaty. I discern no such requirement in Hamdi or any other precedent Doe cites.

Further, I believe the Order is at odds with Munaf. There, the Supreme Court vacated a preliminary injunction that blocked our military from transferring to Iraqi custody an American citizen determined by military officers—without Hamdi’s judicial review—to be an enemy combatant. I see no reason for a different result here: the facts are closely analogous and the comity and separation of powers considerations that animated Munaf apply with similar force. If that were not sufficient to align this case with Munaf, our own decision in Kiyemba v. Obama, 561 F.3d 509 (D.C. Cir. 2009) (Kiyemba II), bridges any gap.

I would vacate the Order. It is valid only if Doe shows that all of the preliminary injunction factors support the district court’s intrusion into Executive Branch affairs. In my view, Doe has not carried his burden. Because my colleagues conclude otherwise, I respectfully dissent. 3 I. BACKGROUND

The majority opinion recounts many of the relevant facts and much of the procedural history. I include my own recitation to complete the picture and to amplify points that I think distance this case from Hamdi and bring it within Munaf’s ambit. I draw the bulk of the recitation from the government’s factual return. Public Appendix (App.) 155- 309; see 28 U.S.C. § 2243 (habeas court may require custodian to “make a return certifying the true cause of the detention”); Order, Dkt. No. 41 at 1 (Jan. 12, 2018) (court required return).

Doe claims in passing that “the government’s allegations are riddled with inaccuracies” and “are fundamentally misleading.” Appellee’s Br. 24 n.3. Yet he does not give us his own factual account, except to say that terrorists “kidnapped and imprisoned” him while he was in Syria seeking to “understand” and “report about” the conflict there. Id. Otherwise, he accepts the government’s allegations for the purpose of litigating its authority to detain him. And in his view, the government’s authority to detain him is all but coextensive with its discretion to transfer him. Make no mistake, he is wrong about that. When read together, Hamdi and Munaf make plain that a putative transferee like Doe is not on the same legal footing as a detainee the military has decided to “continue to hold” indefinitely. Hamdi, 542 U.S. at 534 (emphasis altered). Still, given Doe’s view of the purported overlap—and because he reserves his right to challenge the government’s allegations only at “a later stage,” Pet’r’s Resp. to Factual Return, Dkt. No. 59 at 1 (Feb. 9, 2018)—I would hold him to his stance regarding the government’s authority to detain him and its discretion to transfer him.

To me, then, it does not matter that Doe baldly “contends that he is not in fact an [ISIS] combatant.” Maj. Op. 2. His 4 pro forma assertion is contrary to all evidence of record. For our purpose today, he was found in a foreign war zone during active hostilities and he admitted training with and working for a terrorist organization. Accordingly, for our purpose today, he is on far different ground from a tourist, tax evader or political dissident. Maj. Op. 4, 18, 20-22.

A. DOE’S BACKGROUND, ISIS MEMBERSHIP AND CAPTURE

Doe is a citizen of Saudi Arabia. He is also a citizen of the United States but has not lived here since 2006 and has not visited since 2014. 2

In July 2014, Doe voluntarily traveled to Syria to join the Islamic State of Iraq and the Levant, a terrorist organization better known as ISIS. ISIS has committed

systematic abuses of human rights and violations of international law, including indiscriminate killing and deliberate targeting of civilians, mass executions and extrajudicial killings, persecution of individuals and entire communities on the basis of their identity, kidnapping of civilians, forced displacement of Shia communities and minority groups, killing and maiming of children, rape and other forms

2 Doe was born in the United States. When he was ten years old, he moved to Saudi Arabia and became a citizen there. He returned to the United States for college but moved back to Saudi Arabia two years later. Before taking up with ISIS, Doe owned businesses in Saudi Arabia, got married there and fathered a daughter there. Members of his extended family still live there. 5 of sexual violence, along with numerous other atrocities.

Dep’t of State, The Global Coalition to Defeat ISIS (Sept. 10, 2014), perma.cc/W9ZV-Y4DV. 3 The United States and 74 other countries have committed to defeating ISIS through military force and other means. Id.; see Dep’t of State, The Global Coalition to Defeat ISIS: Partners (Sept. 10, 2014), perma.cc/SQ57-GQ7R. 4

Starting in or about March 2015, Doe attended ISIS training in Syria with fellow recruits. At the training site, he swore allegiance to Abu Hafs al-Maghrebi, who acted on behalf of ISIS’s leader, Abu Bakr al-Baghdadi. ISIS assigned Doe to be a fighter in the Zarqawi Brigade, a military unit that “guard[ed] the front lines” in Syria. App. 195. There, Doe procured fuel for ISIS vehicles, handled funds for ISIS expenses and performed other administrative tasks. He was later assigned to guard the gate of an ISIS oil field and then to monitor personnel who worked on ISIS’s heavy equipment.

Doe worked for ISIS for about two and one-half years “until air strikes and other military offensives against [ISIS] forced him to flee.” App. 162. On or about September 11,

3 ISIS has committed many such acts in Saudi Arabia. See, e.g., Dep’t of State, Country Reports on Terrorism 2016, 219-20 (July 2017) (Country Reports), perma.cc/Q2J9-LPCX. 4 Saudi Arabia has been “a key member and active participant” in this coalition. Country Reports, supra, at 219. It has, for example, joined the United States in launching military strikes against ISIS targets. Dep’t of Defense, U.S., Saudi Arabia Conduct Airstrikes Against ISIL in Syria (Oct. 13, 2014), perma.cc/3YEF- RET3. 6 2017, Syrian Democratic Forces captured him on an active battlefield as he tried to escape Syria into Turkey. He was carrying thumb drives that contained ISIS personnel spreadsheets as well as “military style handbooks” about techniques for interrogation, handling weapons and building bombs. App. 199-200. Doe told his captors he had been walking for two days. ISIS controlled all of the territory within a two-day walk. Doe’s physical appearance was “typical of an [ISIS] devotee.” App. 245. And, indeed, he expressly identified himself as “daesh,” another name for ISIS. Id. Claiming American citizenship, he said he “wanted to speak to the Americans” and “turn himself in.” Id.

Because Doe claimed American citizenship, the Syrian Democratic Forces transferred him “to U.S. forces stationed in Iraq,” App. 161, within the same theater of combat as his capture, Public Oral Arg. Tr. 18-19, 31 (Apr. 5, 2018). According to the factual return, “[t]he Government had not set out to capture” him but has since “worked diligently to investigate [him] . . . and determine an appropriate disposition of him.” App. 161. During custodial interrogation, Doe admitted that he attended ISIS training and “became an active member of ISIS.” App. 262-63. Based on those admissions and other facts, the Executive Branch has concluded that he is an enemy combatant.

B. DISTRICT COURT PROCEEDINGS

Doe was in United States custody by September 12, 2017. On September 14, the Defense Department confirmed a news report that it had a citizen in custody abroad. Betsy Woodruff & Spencer Ackerman, U.S. Military: American Fighting for ISIS “Surrenders,” DAILY BEAST (Sept. 14, 2017), thebea.st/2x1RfeZ. 7 On October 5, 2017—i.e., 23 days after our Armed Forces took custody of Doe—the American Civil Liberties Union Foundation (ACLUF) filed a petition for writ of habeas corpus as his next friend. The petition, which remains pending in district court, claims that Doe’s custody at the hands of the United States military is “[u]nauthorized and [u]nlawful.” App. 19. It asks the court to order the government to “charge [Doe] with a federal criminal offense in an Article III court or release him.” App. 23.

On December 23, 2017, the district court ordered the government to give the ACLUF “immediate and unmonitored access” to Doe. App. 39. In the same order, the court prohibited the government from transferring Doe until the ACLUF informed the court whether Doe wanted the ACLUF “to continue this action on his behalf.” Id. The government complied with the order and the ACLUF spoke with Doe by videoconference. Doe confirmed that he wanted to pursue the habeas case with the ACLUF as his counsel.

On January 5, 2018, Doe sought “interim relief” prohibiting the government from transferring him “until the Court issues a final judgment on his habeas petition.” Pet’r’s Mot. for Continued Interim Relief, Dkt. No. 32 at 2. He argued that such relief was necessary “to prevent the United States from pretermitting this habeas action while the Court considers the lawfulness of his detention.” Id. at 3 n.4. He disclaimed then—and has not alleged since—that he will be mistreated if transferred. Less than two weeks later, the district court prohibited any transfer pending its ruling on Doe’s motion for interim relief. Then, on January 23, the court ordered the government to provide the court and counsel 72 hours’ notice before transferring Doe, “at which time [he] may file an emergency motion contesting his transfer.” App. 50. 8 On April 16, 2018—pending an expedited appeal of the notice requirement and after “extensive diplomatic discussions” with the receiving country—the government notified the district court and counsel of its intent to transfer Doe. 5 Resp’t’s Notice, Dkt. No. 80-1 at 7 (redacted version). 6 On April 18, Doe sought a preliminary injunction blocking the transfer. He renewed his contention that the government “should not be allowed to pretermit [his] habeas action seeking his release from unlawful detention by forcibly transferring him.” Mot. for Prelim. Inj., Dkt. No. 82-1 at 2 (redacted version).

On April 19, 2018, in the Order sub judice, the district court granted Doe’s motion for a preliminary injunction. The Order prohibits the government “from transferring [Doe] from U.S. custody” absent “further order” of the district court. Prelim. Inj., Dkt. No. 88. Explaining its Order, the court recognized that Doe had to show “[1] he is likely to succeed on the merits, [2] he is likely to suffer irreparable harm in the

5 The government intends to transfer Doe to Saudi Arabia, which has “an interest in returning its citizens to Saudi custody.” Sealed Appendix 166. In past negotiations, Saudi Arabia has pursued that interest and the United States has honored it, to the benefit of both countries. Id. (returning Saudi citizens to Saudi Arabia “has been a significant benefit to the United States in combatting terrorism”); see, e.g., Dep’t of Defense, Detainee Transfers Announced (Jan. 5, 2017) (noting transfer of four detainees from Guantanamo Bay to Saudi Arabia), perma.cc/5CRP-74K8. 6 I agree that, because the government satisfied the notice requirement as to the proposed transfer to Saudi Arabia, the validity of the requirement as to that transfer is moot. Maj. Op. 41. And because I would permit the government to effectuate the transfer to Saudi Arabia, I do not address whether the notice requirement is valid as to any other country. 9 absence of preliminary relief, [3] the balance of equities tips in his favor, and [4] an injunction is in the public interest.” Mem. Op., Dkt. No. 91-1 at 2 (unsealed Apr. 23, 2018) (quoting Winter v. NRDC, 555 U.S. 7, 20 (2008)) (ellipses omitted). The court concluded that Doe meets all four requirements. In the court’s view:

• Doe is likely to succeed on the merits because the government is required to, and has failed to, “present positive legal authority for his transfer.” Mem. Op. 3 (internal quotation omitted).

• Doe will suffer irreparable harm absent the Order because, upon transfer to another country, he “will lose his constitutional right to contest his detention in a U.S. court.” Id. at 5.

• The equities favor blocking the transfer because “the potential harm to bilateral relations between the United States and its strategic ally does not outweigh [Doe’s] constitutional right to seek habeas relief.” Id. at 6.

• Similarly, the public interest favors blocking the transfer because the government’s military and diplomatic interests do not override “citizens’ rights to contest the lawfulness of their detentions and transfers.” Id.

II. ANALYSIS

A preliminary injunction is a “drastic remedy” to be granted only if the movant makes a “clear showing” that he is entitled to it. Mazurek v. Armstrong, 520 U.S. 968, 972 (1997) (per curiam) (internal quotation omitted). To my mind, Doe does not come close. 10 A. LIKELIHOOD OF SUCCESS

As a threshold matter, Doe misunderstands his burden. He says “the government . . . must show” his transfer will be “lawful.” Appellee’s Br. 14 (emphasis added). He adds that, in deciding whether the government has made that showing, we cannot “accept[] as true” the allegations in the factual return. Id. And he suggests the government must possess foursquare “precedent for the proposition that it may transfer a U.S. citizen to the custody of a foreign sovereign without positive legal authority.” Appellee’s Suppl. Br. 5. Doe’s contentions erroneously treat a preliminary injunction as the baseline. Such relief is “the exception,” not “the rule.” Munaf, 553 U.S. at 690. Because it is “extraordinary”—especially where, as here, it disrupts core functions of a coequal branch of government—“it is never awarded as of right.” Id. at 689-90 (internal quotation omitted). It is Doe who must justify the Order, relying on precedent that can bear its weight.

We also have every reason at this stage to accept the government’s factual allegations. To the extent Doe’s likelihood of success depends on his being a wayward bystander kidnapped in ISIS territory, see Appellee’s Br. 23-24 n.3, it is incumbent on him as the moving party to support that story with evidence and to explain why the government’s contrary account is inaccurate, see 13 JAMES WM. MOORE ET AL., MOORE’S FEDERAL PRACTICE § 65.23[2] (3d ed. 2012) (“Submission of affidavits in support of a motion for a preliminary injunction is customary.”). Granted, Doe has reserved the right to challenge the government’s account at “a later stage.” Pet’r’s Resp. to Factual Return 1. But for now he provides no “clear evidence”—indeed, no evidence at all— to rebut the “presumption of regularity” we accord military assertions like the ones contained in the return. Latif v. 11 Obama, 677 F.3d 1175, 1178 (D.C. Cir. 2012) (internal quotation omitted).

With Doe’s burden in mind, I turn to the leading cases and their application vel non here.

1. Law of detention and transfer

Relying heavily on Hamdi, Doe argues that the Executive Branch cannot transfer him absent “positive legal authority” or ex ante judicial review of the military’s determination that he is an enemy combatant. The government argues that, under Munaf and Kiyemba II, principles of comity and separation of powers prevent the district court from blocking Doe’s transfer. I agree with the government.

a. Extended detention in Hamdi

Yaser Hamdi, an American citizen, allegedly took up arms with the Taliban before September 11, 2001 and remained with his unit afterward. Hamdi, 542 U.S. at 512-13 (plurality opinion). Later in 2001, a coalition of our allies captured him in an active combat zone in Afghanistan. Id. at 510, 514, 516. They transferred him to the United States military, which in turn sent him to Guantanamo Bay and later to stateside naval brigs. Id. at 510. More than six months after Hamdi’s capture on the battlefield, his father filed a habeas petition as his next friend. Id. at 511. With no apparent intention of transferring him to another country, the government claimed the authority to detain him indefinitely as an enemy combatant. Id. at 510.

Faced with that claim of authority—to detain Hamdi without charge “for the duration of the particular conflict in which [he was] captured,” 542 U.S. at 518—the Supreme Court agreed that a 2001 congressional enactment supplied the 12 authority if Hamdi was in fact an enemy combatant, id. at 516- 24. The Court turned, then, to “the question of what process is constitutionally due to a citizen who disputes his enemy- combatant status.” Id. at 524. Balancing the competing interests under Mathews v. Eldridge, 424 U.S. 319 (1976), the Court concluded “that a citizen-detainee seeking to challenge his classification as an enemy combatant must receive notice of the factual basis for his classification, and a fair opportunity to rebut the Government’s factual assertions before a neutral decisionmaker,” Hamdi, 542 U.S. at 533; see id. at 527-35.

Importantly, however, the Court emphasized “that initial captures on the battlefield need not receive the process we have discussed” and that such “process is due only when the determination is made to continue to hold those who have been seized.” 542 U.S. at 534 (emphasis altered). Moreover, the Court repeatedly made plain that its due process analysis applies only to detention. See, e.g., id. at 509 (concluding that “citizen held in the United States” must “be given a meaningful opportunity to contest . . . that detention”); id. at 524 (focusing on procedures attendant to “detention of enemy combatants”); id. at 525 (examining relief “available to [an] individual detained within the United States”); id. at 529 (weighing “interest in being free from physical detention by one’s own government”); id. at 530 (considering “interest of the erroneously detained individual” (emphasis omitted)); id. at 535 (referring to “protections that accompany challenges to detentions”); cf. id. at 519 (discussing authority for “detention”); id. at 523 (finding “authority to detain” enemy combatant). The Court’s analysis mentioned the concept of sovereign-to-sovereign transfer only once and only in passing. Id. at 518-19. Even then, it equated transfer with repatriation or release, not continued detention. Id. (noting that “object of capture is to prevent the captured individual from serving the enemy” until he is “exchanged, repatriated or otherwise 13 released” (quoting In re Territo, 156 F.2d 142, 145 (9th Cir. 1946))).

b. Transfer in Munaf

Acting under a United Nations resolution, a coalition force of 26 countries took Shawqi Omar and Mohammad Munaf into military custody for their “serious hostile acts” in Iraq. Munaf, 553 U.S. at 699; see id. at 679, 681, 684. I focus here on Omar. He was a citizen of the United States and Jordan. Id. at 681. He was “believed to have provided aid to” al Qaeda. Id. He was held in Iraq “in the immediate physical custody of American soldiers.” Id. at 685 (internal quotation omitted). A tribunal of three American military officers concluded that he was an enemy combatant. Id. at 681. A coalition review board reached the same conclusion. Id. at 682. The coalition later “decided to refer” Omar to Iraqi criminal court “for criminal proceedings.” Id.

Members of Omar’s family filed a habeas petition on his behalf. Omar v. Harvey, 479 F.3d 1, 4 (D.C. Cir. 2007). They asserted that Omar was an “innocent civilian[] . . . unlawfully detained by the United States in violation of the Due Process Clause.” Munaf, 553 U.S. at 692. As here, the United States decided to relinquish custody to another country. Omar, 479 F.3d at 3. As here, the district court “issued a preliminary injunction barring transfer in order to preserve its jurisdiction to entertain the habeas petition.” Id. As here, no criminal charges were pending in the receiving country when the court issued the preliminary injunction. Id. at 8. As here, this Court upheld the preliminary injunction on the theory that it “properly preserve[d]” the district court’s jurisdiction “to test the lawfulness of . . . extrajudicial detention.” Id. at 8, 15.

The Supreme Court vacated this Court’s decision and the preliminary injunction itself. Munaf, 553 U.S. at 705. I 14 recognize that the Supreme Court’s holding was narrow: the Court concluded that district courts cannot “exercise their habeas jurisdiction to enjoin our Armed Forces from transferring individuals detained within another sovereign’s territory to that sovereign’s government for criminal prosecution.” Id. at 689. But the Court’s reasoning swept more broadly. Because it weighs heavily against the Order here, I discuss it in detail.

The Court observed that, “at its core,” habeas is directed at “unlawful executive detention,” the “typical remedy” for which is “release.” 553 U.S. at 693. In the Court’s view, the atypical remedy of blocking Omar’s transfer to Iraq was “not appropriate.” Id. The Court emphasized that habeas “is governed by equitable principles,” which means that “prudential concerns, such as comity and the orderly administration of criminal justice, may require a federal court to forgo the exercise of its habeas corpus power.” Id. (internal citation and quotations omitted). And the Court concluded that comity—specifically, respect for Iraq’s sovereign interest in prosecuting crimes committed within Iraq’s borders, even by citizens of the United States—prevented the district court from enjoining Omar’s transfer. Id. at 694 (“Iraq has a sovereign right to prosecute Omar and Munaf for crimes committed on its soil.”); see id. at 705 (invoking Wilson v. Girard, 354 U.S. 524 (1957) (per curiam), for “background principle” that foreign country has “sovereign interest in prosecuting crimes committed within its borders”); id. at 692, 694-99 (same).

In a passage my colleagues downplay, the Court found further support for its conclusion in separation of powers principles. Even in peacetime, the Court noted, “the Constitution allows the Executive to transfer American citizens to foreign authorities for criminal prosecution.” 553 U.S. at 699. The Court remarked on how “strange” it would be “to 15 hold that the Executive lacks that same authority where, as here, the detainees were captured by our Armed Forces for engaging in serious hostile acts against an ally in what the Government refers to as ‘an active theater of combat.’” Id. at 699-700. “Such a conclusion,” the Court cautioned, “would implicate . . . concerns about unwarranted judicial intrusion into the Executive’s ability to conduct military operations abroad.” Id. at 700.

Finally, the Court rejected Omar’s contention that “the Government may not transfer a citizen” to another country “without legal authority” in the form of “a treaty or statute.” 553 U.S. at 704 (internal quotations and brackets omitted). Omar had relied on Valentine v. United States ex rel. Neidecker, 299 U.S. 5 (1936), which the Court found “readily distinguishable” because “[i]t involved the extradition of an individual from the United States.” 553 U.S. at 704. The Court acknowledged that, in the context of extradition from the territorial United States, the government cannot “‘seize [a fugitive criminal] and surrender him to a foreign power’” absent authority conferred by “a pertinent constitutional or legislative provision.” Id. (quoting Valentine, 299 U.S. at 9). “But Omar . . . voluntarily traveled to Iraq” and was “captured and already detained” there. Id. Because he was not within the territorial jurisdiction of the United States, Valentine was inapposite. Id.

c. Transfer in Kiymeba II

The district court in the Kiyemba litigation required the government to provide 30 days’ notice to the court and counsel before transferring nine Uighurs from Guantanamo Bay to any “country where they might be tortured or further detained.” Kiyemba II, 561 F.3d at 511. In this Court, the Uighurs defended the district court’s order as essential “to protect[ing] 16 the court’s jurisdiction over their underlying claims of unlawful detention.” Id. at 513 n.3. Treating the order as a preliminary injunction, this Court vacated it because the Uighurs did not “make the required showing of a likelihood of success on the merits.” Id. at 516.

Even “assum[ing] arguendo these alien detainees have the same constitutional rights with respect to their proposed transfer as did the U.S. citizens facing transfer in Munaf,” 561 F.3d at 514 n.4, this Court held that “Munaf precludes the district court from barring the transfer of a Guantanamo detainee on the ground that he is likely to be tortured or subject to further prosecution or detention in the recipient country,” id. at 516. The Court accepted the government’s representation that “any prosecution or detention the petitioners might face would be effected ‘by the foreign government pursuant to its own laws and not on behalf of the United States.’” Id. at 515 (quoting declaration of Defense Department official). And the Court reasoned that, under Munaf, “comity and respect for foreign sovereigns . . . bar[] a court from issuing a writ of habeas corpus to shield a detainee from prosecution and detention by another sovereign according to its laws.” Id. (internal quotation omitted).

Taking a further cue from Munaf, the Court added that “separation of powers principles” “preclude the courts from second-guessing” the Executive Branch with respect to transfer. 561 F.3d at 515. The Court concluded that the district court’s notice requirement alone—even without regard to potentially blocking the transfer itself—unduly “interfere[d] with the Executive’s ability to conduct the sensitive diplomatic negotiations required to arrange safe transfers for detainees.” Id. 17 2. Application to Doe’s transfer

Under the foregoing framework, Doe has not shown—in fact, cannot show—that he will likely succeed on the merits.

a. As Judge Brown recognized in Omar, “we must first [ask] in what sense” a putative transferee “must be likely to succeed.” 479 F.3d at 18 (Brown, J., dissenting in part). The Supreme Court answered that question in Munaf: we look to whether he will likely succeed on “the merits of [his] habeas petition.” 553 U.S. at 690.

Here, Doe’s habeas petition challenges his detention at the hands of the Executive Branch. App. 11 (alleging that Doe is “being unlawfully detained by the United States military”); App. 13 (stating that Secretary of Defense “is detaining [Doe] under or by color of the authority of the United States”); App. 17 (claiming that detention violates 18 U.S.C. § 4001(a), which applies to “citizen . . . imprisoned or otherwise detained by the United States”); App. 20 (challenging “detention . . . by Respondent”). Doe therefore cannot succeed on the merits of his habeas petition unless he remains “detained by the United States.” App. 17 (quoting section 4001(a)). And he will not remain detained by the United States if the district court has improperly blocked the government from relinquishing custody to Saudi Arabia.

b. To repeat, habeas “at its core” is aimed at “unlawful executive detention,” not at a transfer that ends it. Munaf, 553 U.S. at 693. Accordingly, if it is ever “appropriate,” as a matter of “equitable principles,” to enjoin a captive’s transfer from Executive Branch custody simply to allow him to challenge that soon-to-be-erstwhile custody, such relief ought to be reserved for the most “extreme case” of Executive Branch malfeasance. Id. at 693, 702 (internal quotation omitted); see, e.g., id. at 702 (suggesting relief might be warranted if “the 18 Executive has determined that a detainee is likely to be tortured but decides to transfer him anyway”). Doe’s case is by no means extreme in that sense. Indeed, it tracks Munaf in two crucial respects.

First, as in Munaf, the receiving country here has a facially strong—for that matter, all but undisputed—interest in the transfer.7 Granted, the particular interest here is slightly different from that in Munaf. There, the Court relied on Iraq’s “sovereign right to prosecute Omar and Munaf for crimes committed on its soil.” Id. at 694. Here, by contrast, Doe did not (as far as the record discloses) commit crimes within the receiving country’s territory and he has not (to date) been charged with any offense there. But the difference in the two cases is not as stark as Doe would have it: recall that Omar had not been charged with a crime in Iraq before the district court issued the preliminary injunction, or even before this Court issued a decision. Compare Appellee’s Br. 31 (attempting to distinguish Omar’s case on basis that Iraq “was actively prosecuting” him), with Omar, 479 F.3d at 8 (noting that “Omar has not been charged with a crime related to the allegations now lodged against him”). More to the point, focusing on a receiving country’s interest in prosecuting territorial offenses misses the ocean for the boat: in the habeas context, comity is why the prosecutorial interest matters. Munaf, 553 U.S. at 693 (“Habeas corpus is governed by

7 Doe does not dispute that he is a citizen of Saudi Arabia and has spent much of his life there. See supra note 2. Nor does he dispute that the Saudi government has an interest in repatriating its citizens, including radicalized ISIS-affiliated citizens who are detained abroad and are candidates for rehabilitation. See supra notes 3-5; see also Decl. ¶¶ 2-3 (Apr. 16, 2018) (detailing Saudi interest in repatriation and rehabilitation); Country Reports, supra, at 222 (same). 19 equitable principles,” including “prudential concerns . . . such as comity.” (internal quotations omitted)); see id. at 698-99 (relying on “principles of comity and respect for foreign sovereigns” (quoting Omar, 479 F.3d at 17 (Brown, J., dissenting in part))).

Comity is “[c]ourtesy” towards “the laws and usages” of another nation. III OXFORD ENGLISH DICTIONARY 539 (2d ed. 1989). By definition, it counsels “mutual recognition of legislative, executive, and judicial acts” that go well beyond prosecutorial prerogatives. BLACK’S LAW DICTIONARY 324 (10th ed. 2014). In some cases, then, comity weighs against blocking a captive’s transfer even if the receiving country claims no immediate interest in prosecuting him for a territorial offense.

Perhaps the most obvious case is one like Doe’s, in which the captive is a citizen of the receiving country. Customary international law recognizes a state’s “sovereign[] interest in retaining control over its nationals and residents, wherever they may be.” RESTATEMENT (FOURTH) OF THE FOREIGN RELATIONS LAW OF THE UNITED STATES § 214 cmt. a (2016 draft); see id. § 211 cmt. c (state has jurisdiction over “conduct occurring on the state’s territory or being committed by or against its nationals” (emphasis added)); see also Blackmer v. United States, 284 U.S. 421, 437 n.2 (1932) (under international law, state has “jurisdiction over its subjects travelling or residing abroad, since they remain under its personal supremacy” (internal quotation omitted)).

In any event, Kiyemba II demonstrates that the availability of habeas relief does not depend on whether the putative transferee will be prosecuted by the receiving state. This Court held that “Munaf precludes the district court from barring the transfer of a Guantanamo detainee on the ground that he is 20 likely to be tortured or subject to further prosecution or detention in the recipient country.” 8 Kiyemba II, 561 F.3d at 516 (emphasis added).

Second, the separation of powers considerations highlighted in Munaf also apply here. When “‘adjudicating issues inevitably entangled in the conduct of our international relations,’” a court is “to proceed ‘with . . . circumspection.’” Munaf, 553 U.S. at 689 (quoting Romero v. Int’l Terminal Operating Co., 358 U.S. 354, 383 (1959)). Far from circumspect, the Order upends the Executive Branch’s decision to relinquish Doe to a country the district court acknowledges is a “strategic ally.” Mem. Op. 6. Much as in Munaf, the Executive’s decision was informed by the ally’s sovereign interest in Doe and by our military’s good-faith determination that he committed “serious hostile acts” in “an active theater of combat” where he was captured and remains detained. 9 Munaf, 553 U.S. at 699-700; see generally App. 155-309 (factual return). Thus, the Order is every bit the “judicial intrusion” into “military operations” and “sensitive diplomatic negotiations” that the preliminary injunctions in Munaf and

8 Doe tries to distinguish Kiyemba II on the ground that the Uighurs were non-citizens. It is a fair point but goes only so far; the Court “assume[d] arguendo” the Uighurs had “the same constitutional rights with respect to their proposed transfer as did the U.S. citizens facing transfer in Munaf.” 561 F.3d at 514 n.4. 9 At oral argument, Doe contended that the record contains evidence only that the determination was made, not that it was made in good faith. Because there is no evidence that the determination was made in bad faith, however, I see no reason to question the government’s motives. Cf. Latif, 677 F.3d at 1178-85. 21 Kiyemba II were. Munaf, 553 U.S. at 700; Kiyemba II, 561 F.3d at 515.

c. Doe argues that Hamdi justifies the intrusion. He contends that, absent an applicable extradition treaty, Appellee’s Suppl. Br. 10; see infra pp. 23-27, the government cannot transfer him unless it “can lawfully detain [him] as an enemy combatant in the first place,” Appellee’s Br. 23. It follows, in his view, that the district court can appropriately block his transfer in order to review the government’s “unilateral and untested assertion” that he is a detainable enemy combatant. Id. To hold otherwise, he says, would wrongly deprive him of the due process protections to which Hamdi entitles him.

I disagree. For starters, the Supreme Court in Munaf did not read Hamdi the way Doe does. Omar was merely “alleged to have committed hostile or warlike acts in Iraq.” 553 U.S. at 679 (emphasis added); see id. at 694 (he was “alleged to have committed serious crimes in Iraq” (emphasis added)). Based on those alleged hostile acts, military authorities decided that Omar was an enemy combatant. Id. at 681-82. The Court did not hold that a federal judge had to review that determination as a prerequisite to transfer, whether as a matter of “positive legal authority” or due process. To the contrary, the Court concluded that it was “not appropriate” to block Omar’s transfer for the sake of ensuring he could litigate, via habeas, his claim that he was an “innocent civilian[] . . . unlawfully detained by the United States in violation of the Due Process Clause.” Id. at 692-93.

Notably, the Court in Munaf cited Hamdi only once, for the proposition that “[h]abeas is at its core a remedy for unlawful executive detention.” 553 U.S. at 693 (emphasis added). Conversely, as mentioned above, the Court in Hamdi 22 invoked the concept of sovereign-to-sovereign transfer only once, equating it with repatriation or release rather than detention. 542 U.S. at 518-19. Reading the cases together, I can only conclude that detention and transfer are not flipsides of the same coin but two entirely different currencies. Hamdi, in short, does not apply to Doe’s transfer. It is a case about detention potentially “for the duration of the relevant hostilities.” Id. at 519. To reiterate, the Court excepted “initial captures on the battlefield” from “the process we have discussed,” emphasizing that such “process is due only when the determination is made to continue to hold those who have been seized.” Id. at 534 (emphasis altered); see id. at 529 (Hamdi’s “liberty interest[]” was “in being free from physical detention by [his] own government”).

Nevertheless, according to Doe, wherever one draws the line between battlefield captive and long-term detainee, he falls on the latter side. In his telling, this case involves an Executive Branch decision to detain him without charge for an extended period, now exceeding six months. Appellee’s Second Suppl. Br. 6 (asserting “government decide[d] not to release him . . . six months ago” when it moved to dismiss his habeas petition). I reject that characterization.

Rewind to September 12, 2017, when our military took custody of Doe. In an active combat zone, it faced the real- time decision of what to do with a battlefield captive who admitted affiliation with ISIS. Should it detain him indefinitely as an enemy combatant? Transport him to the United States and charge him with a crime? Transfer him to a country with a sovereign interest in him? When the ACLUF filed the habeas petition on October 5, “the Government was still engaged in this decisional process” and had yet to choose a course of action. App. 161. No surprise there: the government had had a mere 23 days to investigate Doe. Since 23 then, this litigation has left the government in a poor position to consider, negotiate and effectuate Doe’s transfer. App. 153 (according to State Department, litigation contingencies have “hinder[ed] the Department’s ability to engage constructively with” receiving country); see Kiyemba II, 561 F.3d at 515 (“[T]he requirement that the Government provide pre-transfer notice interferes with the Executive’s ability to conduct the sensitive diplomatic negotiations required to arrange safe transfers[.]”). 10

The end result is the judicial equivalent of mission creep. After today, a habeas court is authorized to review not only a decision to “continue” Executive Branch custody of a citizen captured abroad on an active battlefield, Hamdi, 542 U.S. at 534, but also—extraordinarily—a decision to discontinue it. Indeed, if the captive’s next friend gets to the courthouse quickly enough, nearly any Executive decision about the captive will be subject to judicial review. Doe makes no showing—much less a clear showing—that Hamdi reserves so little breathing room for the military’s on-the-ground judgment.

d. Doe likewise makes no clear showing that Valentine v. United States ex rel. Neidecker, 299 U.S. 5 (1936), or Wilson v. Girard, 354 U.S. 524 (1957) (per curiam), supports the Order. He claims that Valentine forbids the government to

10 Doe himself argues that the government could not appropriately transfer him once the habeas petition was filed. Public Oral Arg. Tr. 59-61 (Apr. 5, 2018). And at Doe’s urging, the district court has issued a series of orders restricting his transfer. I do not suggest there was anything improper about his litigation choices. But they inevitably delayed the decision to transfer him. In my view, affirmance of the Order mistakenly lays the consequences of Doe’s choices at the Executive Branch’s feet. 24 relinquish him to another country absent “positive legal authority” set forth in a statute or extradition treaty. But Valentine involved “fugitive criminal[s]” apprehended in the United States. 299 U.S. at 9, 11 (internal quotation omitted); see id. at 6 (they were United States citizens arrested in New York based on criminal charges in France). In that “very narrow” context, United States ex rel. Neidecker v. Valentine, 81 F.2d 32, 33 (2d Cir. 1936), the Supreme Court required a “statute or treaty confer[ring] the power” to extradite, 299 U.S. at 9.

Doe bears no resemblance to the fugitives in Valentine. He voluntarily traveled abroad to an active war zone. He was captured on a foreign battlefield by foreign military forces. He admitted affiliation with a terrorist organization the United States is combatting militarily. And he was taken at his own request into United States military custody within the same theater of combat. Nothing in Valentine indicates that extradition rules apply to such a person any more than the laws of war apply to a fugitive criminal apprehended in the United States. The Supreme Court in Munaf drew a line between the two types of cases in rejecting Omar’s argument that the Executive Branch could not transfer him to Iraq “without legal authority” in the form of “a treaty or statute.” 553 U.S. at 704 (internal quotations omitted). “Valentine,” the Court observed, was “readily distinguishable” because “[i]t involved the extradition of an individual from the United States.” Id. (emphasis added). Valentine is distinguishable here for the same reason. Moreover, because Doe has the burden of persuasion, I think it significant that—despite numerous armed conflicts since 1936—he cites no case that has ever applied 25 Valentine to the wartime transfer of a battlefield captive abroad. 11

Doe is similarly mistaken in suggesting that Wilson v. Girard requires “positive legal authority” for his transfer. At issue in Wilson was a bilateral Status of Forces Agreement between the United States and Japan. 354 U.S. at 527-28. The Agreement provided that the American military had jurisdiction over acts committed in Japan by American servicemen in performance of their duties. Id. The Agreement also required the United States to “notify” Japanese authorities “as soon as practicable” if it “decide[d] not to exercise jurisdiction.” Id. at 528. Finally, the Agreement provided that the United States, when making that decision, was to give Japan’s interests “sympathetic consideration.” Id.

Against this backdrop, Girard, an American serviceman in Japan, was alleged to have killed a Japanese national there.

11 The phrase “positive legal authority” does not appear in Valentine, Munaf or any other Supreme Court precedent. Doe draws it from Omar v. McHugh, 646 F.3d 13 (D.C. Cir. 2011), a slim reed on which to base such a requirement. There, on remand from Munaf, this Court held that Omar lacked any “right to judicial review of conditions in Iraq before he is transferred,” id. at 18, but emphasized it was not holding that “the Executive Branch may detain or transfer Americans or individuals in U.S. territory at will, without any judicial review of the positive legal authority for the detention or transfer,” id. at 24. The caveat was as obiter as dictum can be: the Court acknowledged that Munaf had already settled the government’s “authority” to transfer Omar and that, on remand, the Court was “addressing Omar’s separate argument . . . about conditions in the receiving country.” Id. At all events, the Court did not purport to eliminate the sharp distinction between fugitives “in U.S. territory” on the one hand and “wartime military transfers” on the other. Id. 26 354 U.S. at 525-26. Because he did so, arguably in performance of his duties, id. at 529, the Agreement “seemed to give [him] a right to be tried by an American military tribunal, not a Japanese court,” Munaf, 553 U.S. at 705 (discussing Wilson). But the Executive Branch “decided not to exercise . . . jurisdiction.” Wilson, 354 U.S. at 529. Per the Agreement, it notified the Japanese government that it intended to transfer Girard to Japanese custody for trial in a Japanese court. Id. at 526, 529. In turn, the Japanese government indicted him. Id. at 526. Girard petitioned for habeas relief and a district court here in the United States enjoined his transfer. Id. Far from requiring affirmative authority for the transfer, the Supreme Court vacated the injunction because the Court discerned “no constitutional or statutory barrier” to the transfer. Id. at 530 (emphasis added). “In the absence of such encroachments,” the Court deferred to the “wisdom” of the political branches. Id. The Court apparently saw nothing of relevance in Valentine, which it nowhere mentioned.

Doe nevertheless reads Wilson to hold that “a treaty satisfied the requirement of positive legal authority for the transfer.” Appellee’s Suppl. Br. 6. He misunderstands the Status of Forces Agreement. The Agreement—which was “[t]he only ‘authority’ at issue in Wilson”—permitted the United States to refuse a transfer and to exercise jurisdiction itself notwithstanding the “background principle” that Japan, absent the Agreement, “had exclusive jurisdiction ‘to punish offenses . . . committed within its borders.’” Munaf, 553 U.S. at 696, 705 (quoting Wilson, 354 U.S. at 529). The Agreement’s mere procedural requirements—to give Japan’s interests “consideration” and to promptly “notify” Japan when the United States “decide[d] not to exercise jurisdiction,” Wilson, 354 U.S. at 528 (emphasis added)—were hardly “authority” for a transfer, let alone the sort of “positive legal authority” that Doe demands here. The Court in Munaf 27 recognized as much. It held that Wilson outright “forecloses” the “argument that the Executive lacks the discretion to transfer a citizen absent a treaty or statute.” 553 U.S. at 705. That holding makes sense only if the Munaf Court rejected the notion that the Agreement served as “authority” for the transfer in Wilson.

e. Doe suggests the foregoing analysis cannot possibly be correct because, as he sees it, it gives the Executive Branch license to run roughshod over the rights of American citizens with no judicial check. See, e.g., Appellee’s Br. 52 (it “make[s] a mockery of the Great Writ”); Appellee’s Suppl. Br. 6 (it means citizens “surrender [their] constitutional rights when abroad” (internal quotation omitted)); Appellee’s Second Suppl. Br. 10 (it means “government’s power to dispose of citizens” is not “constrained by law”). None of this is so.

When someone in Executive Branch custody files a habeas petition, the federal courts ensure that the Executive handles him “in accordance with law,” including due process. Hamdi, 542 U.S. at 525 (citing INS v. St. Cyr, 533 U.S. 289, 301 (2001)). But there are limits to a habeas court’s equitable power, even if the petitioner is a citizen. Under Munaf, for example, “prudential concerns . . . may require [the] court to forgo the exercise of its habeas corpus power.” 553 U.S. at 693 (internal quotations omitted). I have explained why, in my view, considerations of comity and separation of powers preclude the Order here. Especially important to me are Doe’s voluntary travel abroad to a war zone during active hostilities; his capture on a foreign battlefield by foreign military forces; his admitted affiliation with a terrorist organization the United States is combatting militarily; the Executive Branch’s resulting good-faith determination that Doe is an enemy combatant; Doe’s continued presence in the same active theater 28 of combat as his capture; and the receiving country’s facially compelling interest in his transfer.

If these facts differed, the prudential considerations might differ and the district court might have equitable authority to block a transfer. For instance, Munaf reserves the possibility of judicial intervention if the Executive Branch “determine[s] that a detainee is likely to be tortured but decides to transfer him anyway.” 553 U.S. at 702. Similarly, the government appears to concede “that the courts have a role to play” in ensuring that the Executive Branch does not transfer a battlefield captive to a country that lacks a “legitimate basis” in law to receive him. Public Oral Arg. Tr. 10, 17, 34 (Apr. 5, 2018).

Here, however, we have no record-based reason to assume Executive Branch bad faith or negligence. Rather, as the Supreme Court admonished in Munaf, “we need not assume the political branches are oblivious” to a transferee’s well-being. 553 U.S. at 702 (quoting Omar, 479 F.3d at 20 n.6 (Brown, J., dissenting in part)). Nor should we be distracted by any “farfetched hypothetical[],” Gutierrez v. Waterman Steamship Corp., 373 U.S. 206, 210 (1963), that “veers far from the case before us,” Digital Realty Trust, Inc. v. Somers, 138 S. Ct. 767, 781 (2018); see, e.g., Public Oral Arg. Recording 34:20-34:48 (Apr. 27, 2018) (Doe hypothesizes transfer “to Bolivia or Madagascar” or some other country with no sovereign interest in him); see also, e.g., Maj. Op. 20-22 (majority hypothesizes transfer to Thailand based on political criticism).

The long and short of it is that Doe does not dispositively differ from the petitioners in Munaf. Necessarily, I do not read that opinion the same way my colleagues do. On their view, “the war-related context” of Munaf “did not diminish” the military’s discretion to transfer Omar to Iraqi authorities, at 29 least as compared to the military’s discretion to transfer Girard to Japanese authorities during peacetime. Maj. Op. 15. If my colleagues imply that the war context of Munaf made no difference, I disagree: the Supreme Court was explicit that “more [was] at issue” in Munaf than in Wilson, which did not involve a petitioner captured on a battlefield in “‘an active theater of combat’” “during ongoing hostilities.” Munaf, 553 U.S. at 699-700 (accepting government’s characterization to that effect). Although my colleagues do not mention it, those are the very circumstances that gave rise to the Court’s “concerns about unwarranted judicial intrusion into the Executive’s ability to conduct military operations abroad.” Id. at 700. Those same circumstances—and, thus, those same separation of powers concerns—are equally in play here. Doe’s battlefield capture during ongoing hostilities and his admitted affiliation with ISIS align him with the Munaf petitioners and readily distinguish him from the civilians in my colleagues’ counterfactual detours. Maj. Op. 4, 18, 20-22.

B. OTHER FACTORS

Because I believe Doe has not demonstrated a likelihood of success, I do not think it strictly necessary to consider the other preliminary injunction factors. See Kiyemba II, 561 F.3d at 516 (vacating injunction without consideration of other factors because Uighurs did not “make the required showing of a likelihood of success on the merits”); see also, e.g., Greater New Orleans Fair Hous. Action Ctr. v. HUD, 639 F.3d 1078, 1088 (D.C. Cir. 2011) (“When a plaintiff has not shown a likelihood of success on the merits, there is no need to consider the remaining factors.”). Nevertheless, I briefly address the remaining factors because in my view the Order badly misjudges them. 30 Irreparable harm. This Court “has set a high standard” for irreparable harm: “the injury must be both certain and great” and “must be actual and not theoretical.” Chaplaincy of Full Gospel Churches v. England, 454 F.3d 290, 297 (D.C. Cir. 2006) (internal quotation omitted). On this requirement, too, Doe falls short.

The district court finds that Doe will suffer irreparable harm absent the Order because, once transferred, he “will lose his constitutional right to contest his detention in a U.S. court.” Mem. Op. 5. That is half right: because Doe’s petition challenges his detention by the Executive Branch, he will no longer have a viable habeas case once it divests itself of custody. See Qassim v. Bush, 466 F.3d 1073, 1074-77 (D.C. Cir. 2006) (per curiam) (transfer of Uighurs from Guantanamo Bay to Albania effected release and mooted habeas claims); see also supra pp. 17-18. Contrary to the district court’s view, however, Doe has no cognizable interest in pursuing his petition once he is released from the very custody he challenges. See Munaf, 553 U.S. at 692-93 (it was “not appropriate” to enjoin transfer to ensure petitioners could litigate their claim that “they are innocent civilians”). And because Doe has no cognizable interest in litigating a moot petition, he will suffer no “certain,” “great” and “actual” harm from being denied the opportunity to pursue it. England, 454 F.3d at 297 (internal quotation omitted); see Ralph v. Lucent Techs., Inc., 135 F.3d 166, 170 (1st Cir. 1998) (“A federal court must find a cognizable threat of irreparable harm as an essential prerequisite to the issuance of a preliminary injunction.” (emphasis added)).

Even apart from the mootness problem, Doe’s litigating position precludes him from showing irreparable harm. He says he seeks nothing more than “release simplicit[er]”—that is, “for the United States simply to open the jailhouse doors and 31 let him go” in Iraq. Public Oral Arg. Tr. 80 (Apr. 5, 2018). He does not ask to be transported to the United States. He concedes that the Executive Branch is free to notify Iraqi authorities upon his release and that, immediately thereafter, the Iraqi government or other foreign authorities are free to apprehend him.

These are major concessions, and necessary ones. See Munaf, 553 U.S. at 689 (district court could not forbid Executive from “sharing” with Iraqi government “details concerning any decision to release Omar”); id. at 694 (it could not require Executive to “shelter” Omar from prosecution in Iraq); id. at 697 (it could not order Executive to “smuggle” Omar “out of Iraq”). As the government aptly observes, the concessions mean there is “little practical difference . . . between the ‘release’ that [Doe] seeks and the ‘transfer’ that the Government proposes to undertake.” Appellant’s Suppl. Br. 11.

Doe resists this logic because it is “speculat[ive].” Appellee’s Suppl. Br. 11. For all we know, he says, no one will seek to detain him if our military lets him go. This is classic wishful thinking. Because of his admitted affiliation with ISIS—and because Iraq and Saudi Arabia are both members of the coalition against ISIS, see Partners, supra—I believe it is all but certain he will again be held abroad if the United States releases him. 12 And any uncertainty on that

12 I recognize that the country and circumstances of such further detention might differ if the United States were to relinquish Doe to the Saudi government instead of “simply . . . open[ing] the jailhouse doors” and subjecting him to recapture. Public Oral Arg. Tr. 80 (Apr. 5, 2018). But Doe does not allege, let alone show, that the conditions of detention in the latter scenario would be cognizably preferable to the conditions in the former. In any event, judges are ill positioned to compare conditions of detention. Cf. Munaf, 553 32 score operates against Doe, not for him. After all, he must prove that, absent the Order, he will suffer “certain,” “great” and “actual” harm. England, 454 F.3d at 297 (internal quotation omitted). He has failed that task.

Balance of equities. The district court finds that “the potential harm to bilateral relations between the United States and its strategic ally does not outweigh [Doe’s] constitutional right to seek habeas relief.” Mem. Op. 6. But as just discussed, Doe seeks an end to Executive Branch custody; the Executive Branch will in fact end that custody by relinquishing him to the Saudi government; and Doe does not demonstrate that he will suffer cognizable harm if the Executive Branch so relinquishes him instead of “simply . . . open[ing] the jailhouse doors” in Iraq. Public Oral Arg. Tr. 80 (Apr. 5, 2018); see supra note 12. On the other side of the balance, I take the Executive Branch’s word that blocking the transfer complicates our diplomatic relations with the receiving country. Decl. ¶¶ 8-10 (Apr. 16, 2018); see Latif, 677 F.3d at 1178-85 (applying presumption of regularity in analogous setting). Common sense and circuit precedent support its assertion. Kiyemba II, 561 F.3d at 515 (even requiring “pre-transfer notice interferes with the Executive’s ability to conduct the sensitive diplomatic negotiations required to arrange safe transfers”).

In these circumstances, the Executive Branch’s interest in completing the transfer is at least as strong as Doe’s interest in blocking it. See De Arellano v. Weinberger, 788 F.2d 762, 764 (D.C. Cir. 1986) (en banc) (per curiam) (where injunction will “intrude[] into the conduct of foreign and military affairs” and “‘embarrass the accomplishment of important

U.S. at 702 (“The Judiciary is not suited . . . to pass judgment on foreign justice systems[.]”). 33 governmental ends, a court of equity acts with caution and only upon clear showing that its intervention is necessary in order to prevent an irreparable injury’” (quoting Hurley v. Kincaid, 285 U.S. 95, 104 n.3 (1932))).

The public interest. Most of what has already been said also goes to the question of where the public interest lies. But some final observations are in order. The district court concluded that a citizen’s right to contest his military transfer outweighs the government’s military and diplomatic priorities. Mem. Op. 6. That conclusion is shortsighted for at least two reasons.

First, judicial intrusions like the Order cost the Executive Branch valuable diplomatic capital. App. 152-54 (declaration of State Department official); Decl. ¶¶ 8-10 (Apr. 16, 2018) (subsequent declaration of same official); see Kiyemba II, 561 F.3d at 515. Within bounds that have nowise been exceeded in Doe’s case, Executive Branch officials have wide discretion to spend that limited capital as they see fit. Judges ought not lightly cause them to waste it, especially if it might better be spent on ensuring that the United States, in future negotiations, obtains custody of persons in whom it has a compelling sovereign interest.

Second, contrary to Doe’s hyperbole, the Order and its affirmance will not necessarily favor “the errant tourist, embedded journalist, or local aid worker [who seeks] to prove military error.” Appellee’s Br. 24 (quoting Hamdi, 542 U.S. at 534). What if our military had known before taking custody of Doe that it would not be permitted to relinquish him to an ally with a facially strong interest in him unless it first litigated—in distant courts, for months, if not years, on end— the ability to do so? Would our commanders in the field have declined custody, leaving a citizen to the actions of other 34 countries or, even worse, to the chaos of the battlefield? It seems to me that today’s result gives the military an incentive to avoid custody when possible, especially if it is not immediately clear in the heat of combat that the captive is a U.S. citizen. And I doubt that the innocent American citizen who finds himself on a foreign battlefield could fare better than in the custody of our military.

*****

To borrow an understatement, the Order is “not appropriate.” Munaf, 553 U.S. at 693. I would vacate it. Accordingly, I respectfully dissent.