Anthony Kennedy
Anthony Kennedy; drawing by David Levine


Boumediene v. Bush is one of the most important Supreme Court decisions in recent years.1 The Court held by a 5–4 vote that aliens detained as enemy combatants in Guantánamo have a constitutional right to challenge their detention in American courts. The decision frees none of them, some of whom have been held without trial for six years, but it makes it possible for them to argue to a federal district court judge that the administration has no factual or legal ground for imprisoning them. If that judge is persuaded, he must order their release. American law has never before recognized that aliens imprisoned by the United States abroad have such rights. The disgrace of Guantánamo has produced a landmark change in our constitutional practice.

The case raised complex constitutional issues that I must describe, but the principle the Court vindicated is simple and clear. Since before Magna Carta, Anglo-American law has insisted that anyone imprisoned has the right to require his jailor to show a justification in a court of law. (The technical device through which this right is exercised is called a writ of habeas corpus. Addressed to the jailor, it announces that he has custody of a certain person’s body and demands that he justify that custody.)

The Bush administration, as part of its so-called “war on terror,” created a unique category of prisoners that it claims have no such right because they are aliens, not citizens, and because they are held not in an American prison but in foreign territory. The administration labels them enemy combatants but refuses to treat them as prisoners of war with the protection that status gives. It calls them outlaws but refuses them the rights of anyone else accused of a crime. It keeps them locked up behind barbed wire and interrogates them under torture. The Supreme Court has now declared that this shameful episode in our history must end. By implication, moreover, the decision goes even further. It undermines the assumption, widespread among lawyers and scholars for decades, that the Constitution as a whole offers substantially less protection against American tyranny to foreigners than it does to America’s own citizens.

Boumediene was decided by the slimmest of margins. The Court now often divides, in cases of high importance, into a conservative phalanx of Chief Justice John Roberts and Justices Antonin Scalia, Clarence Thomas, and Samuel Alito, and a more liberal group of Justices John Paul Stevens, David Souter, Ruth Ginsberg, and Stephen Breyer.2 The ninth justice, Anthony Kennedy, holds the balance of power; in this case he rejected the phalanx, joined the more liberal group, and wrote the Court’s opinion on their behalf.

The conservatives were outraged but self-contradictory in dissent. Roberts declared that the Court’s decision would have at best a “modest” impact and would be of no use to the detainees because it left them, as a practical matter, with no more opportunity for freedom than they had before. If anything, he said, the decision made it less likely that any of them would be released soon. Scalia insisted, with his usual splenetic flamboyance, that the decision would free dangerous terrorists and “almost certainly cause more Americans to be killed.” Each of the four conservatives signed both dissenting opinions, apparently unconcerned by the contradiction.

Senator John McCain called the decision “one of the worst” in the country’s history. The conservative press was horrified: The Wall Street Journal said that Kennedy had turned the Constitution into a “suicide pact.” No one explained why it would destroy America to allow people who claim innocence of any crime, or threat, a chance to defend that claim before an American judge who is presumably just as worried about his family’s security as the president is. Why would it be suicidal to allow them the same opportunity for defense that we allow people indicted as serial killers?

Senator Barack Obama, on the other hand, welcomed the decision, so the Court’s action may well become an important issue in the coming presidential election. McCain has already promised that if elected he will appoint more justices like Roberts and Alito. It would take only one such appointment to make further decisions like Boumediene impossible, and probably reverse that decision itself.


The immediate background to the Boumediene case is a four-year turf war between the Supreme Court, on one side, and the Bush administration and a Republican Congress on the other. In Hamdi v. Rumsfeld, in 2004, five Supreme Court justices accepted that the government had authority to detain people captured in the fighting in Afghanistan; Congress, they said, had implicitly authorized detention of enemy combatants “for the duration of the particular conflict in which they were captured” when it adopted its Authorization for Use of Military Force (AUMF) after September 11. But a plurality of four justices3 said that an American citizen designated an enemy combatant nevertheless has a right, under the Constitution’s due process clause, to “notice of the factual basis for his classification, and a fair opportunity to rebut the Government’s factual assertions before a neutral decision-maker.” Justice Sandra Day O’Connor, writing for the plurality, said that the administration might satisfy this requirement through appropriately constituted military tribunals that do not provide the procedures and protections of ordinary criminal courts.


In response, the Defense Department established what it called Combatant Status Review Tribunals (CSRTs) to determine whether detainees, including aliens held at Guantánamo, were correctly designated as enemy combatants.4 These tribunals complied with only a minimal interpretation of O’Connor’s description of what due process required. Detainees were provided with special legal “representatives” appointed by the administration rather than lawyers of their choice. They were not allowed to confront government witnesses, and could call only those witnesses the government decided could be produced “reasonably.” Hearsay evidence was allowed against them, and the government’s factual claims were to be presumed correct unless rebutted. On the same day in 2004 as it decided the Hamdi case, however, the Court also decided, in Rasul v. Bush,5 that the Guantánamo detainees were entitled to bring a habeas corpus challenge to their detention in the federal district court for the District of Columbia.

Lawyers in the Justice Department, relying on their interpretation of earlier Supreme Court decisions, had assured the President that prisoners held there, outside the sovereign territory of the United States, could not bring a habeas petition in an American court. In Rasul, the Court declared that opinion wrong: it held that the statutory scheme in which Congress had stipulated the details of habeas corpus review should be interpreted to grant rights to prisoners located not just within the United States itself but also on territory, like Guantánamo Bay, that is subject to the longstanding, exclusive, and permanent control of the United States.6

The decision in Rasul opened the way for Lakhdar Boumediene and thirty-six others held at Guantánamo to file habeas corpus petitions in federal courts challenging their detention. Boumediene is one of six men born in Algeria who were arrested in Bosnia in October 2001, on suspicion of planning an attack on the US embassy there. They were all ordered released by the Bosnian Supreme Court for lack of any evidence against them; but they were taken into custody by American troops in Bosnia and transported to Camp X-Ray at Guantánamo where they were classified as enemy combatants but neither charged nor tried for any crime.

After the joint habeas corpus petition on behalf of Boumediene and thirty-six other detainees was filed, however, Congress quickly overturned the Rasul decision. In 2005, it passed the Detainee Treatment Act (DTA), providing that

no court, justice, or judge shall have jurisdiction to hear or consider…an application for a writ of habeas corpus filed by or on behalf of an alien detained by the Department of Defense at Guantánamo Bay, Cuba….

The act gave the Court of Appeals for the District of Columbia Circuit Court the “exclusive” jurisdiction to review decisions of the CSRTs; it specified that this review was limited to determining whether these decisions were “consistent with the standards and procedures specified by the Secretary of Defense” and “to the extent the Constitution and laws of the United States are applicable, whether the use of such standards and procedures to make the determination is consistent with the Constitution and laws of the United States.”7

In the next round of the battle, however, in 2006, the Supreme Court held, in Hamdan v. Rumsfeld,8 that this provision was not intended to apply to petitioners who had already filed habeas corpus applications when the DTA was enacted. Congress returned fire: in the Military Commissions Act of 2006 (MCA), it ruled that the DTA was indeed to apply retroactively, so that even those petitioners had no remedy beyond an appeal from their CSRTs to the D.C. Circuit Court. (McCain voted for this act; Obama voted against it.)

The Court could no longer interpret the habeas corpus statute to provide a remedy for Guantánamo detainees, since Congress had changed the statute. But Congress cannot overrule the Constitution, which provides—in what is called the “suspension clause” —that Congress may suspend the writ of habeas corpus only during an invasion or a rebellion. No one has argued that terrorist attacks constitute either, so Congress had forced the justices to face two new questions. Does the constitutional guarantee of habeas corpus apply to aliens imprisoned at Guantánamo outside the formal territory of the United States? If so, is the scheme provided by Congress in the Detainee Treatment Act and the Military Commissions Act—military tribunals that could be followed by a limited appeal to the D.C. Circuit Court—an adequate substitute for the traditional writ? In Boumediene, Kennedy, for the five-justice majority, answered the first question yes and the second no, and he therefore declared the congressional scheme unconstitutional.



Scalia’s dissent mainly challenged Kennedy’s disposition of the first of these two issues. How should a contemporary judge decide whether the Constitution’s guarantee of habeas corpus applies to aliens held by United States forces outside the United States? Scalia insisted on an historical test. “The writ as preserved in the Constitution, he said, “could not possibly extend farther than the common law provided when that Clause was written.” Kennedy argued that it is impossible to know what the common law was in 1789. In the absence of definitive historical evidence, he said, the issue should be decided on principle. Scalia disagreed on both points. The history was clear beyond doubt, he said, that the writ was not available in an English court for prisoners held outside the formal realm of England— it was not available to prisoners held in Scotland before the Act of Union, for example, even when the crowns of England and Scotland were united in the same king. But even if the history were ambiguous, he added, the DTA/ MCA scheme should not be held unconstitutional because when constitutional issues are doubtful, the Court should defer to Congress as representative of the people.

Was Scalia right to make the history of the Constitution decisive? We can read the suspension clause in two ways. We can take it to declare, as Scalia did, that the United States should never deny any prisoner the rights he would have had if he had lived in England or America in 1789, except in rebellions or invasions. Or we can read it to state a constitutional principle: that except in those cases government must allow anyone it imprisons the right to challenge his imprisonment in court. Like other constitutional principles, this requirement could not be read as absolute. It would not require habeas rights when it would be impossible or particularly burdensome to grant them—for example, if a prisoner would have to be flown to a US court from a battlefield abroad.

That qualification would not, however, allow government to escape habeas responsibility by building its prison camps in a foreign territory that is as much under its control as any base in this country. As Kennedy said, “The test for determining the scope of this provision must not be subject to manipulation by those whose power it is designed to restrain.” Scalia’s historical reading demeans the Constitution and insults those who made it. It is absurd to translate their clear declaration of principle into a rule pointlessly limiting prisoners’ rights to those enjoyed at some fixed and essentially arbitrary date.9 Kennedy adopted the second, principled, reading. The scope of the constitutional right to habeas corpus, he said, should be determined by what he called a “functional” test: the right should be available unless it would be, in his words, “impracticable and anomalous” to grant it—as it would be in the midst of military operations.

Scalia had a further argument. In a 1950 decision, Johnson v. Eisenträger, the Court refused habeas rights to German soldiers who were captured in China, where they had gone to continue the fight against America after Germany surrendered in 1945. They were convicted of war crimes by a military commission there and then imprisoned in an American military base in occupied Germany. Justice Robert Jackson said, for the majority, that since the Germans were aliens who had never been in the United States they had no constitutional right to contest their imprisonment in an American court. Scalia insisted that Jackson had decided that aliens held abroad never have such rights; he said that the majority in Boumediene was overruling an established precedent.10

But though some of the language in Jackson’s 1950 opinion supports that reading, the facts of that case were sufficiently different from those of Boumediene to allow Kennedy to distinguish the two cases. Jackson’s opinion emphasized, for instance, that the German prisoners conceded that they were citizens of a defeated enemy and were continuing to wage war against the United States. He said that if prisoners like those were entitled to the writ, it

would be equally available to enemies during active hostilities…. It would be difficult to devise more effective fettering of a field commander than to allow the very enemies he is ordered to reduce to submission to call him to account in his own civil courts and divert his efforts and attention from the military offensive abroad to the legal defensive at home.

The decision in Eisenträger is therefore consistent with the principled reading of the suspension clause if we assume, as Kennedy did, that it would have been anomalous and impractical to grant the right to the German soldiers in that case, but not to Boumediene and the other petitioners held at Guantánamo.


Kennedy next held that the review scheme Congress created in the DTA and the MCA was not an adequate substitute for the constitutional right to habeas corpus he had declared the detainees to have. A federal court with habeas jurisdiction has extensive fact-finding power. When someone convicted of a crime in a state criminal court trial petitions a federal court for review of the conviction through a writ of habeas corpus, the federal court normally accepts the state court’s determinations of fact. But a federal court may be much less deferential when it is reviewing an executive tribunal, like a CSRT, particularly when the tribunal’s procedures fall short of the normal trial court standards. Kennedy listed some of the tribunal’s flaws: a detainee, he said,

does not have the assistance of counsel and may not be aware of the most critical allegations that the Government relied upon to order his detention…. [He] can confront witnesses that testify… but given that there are in effect no limits on the admission of hearsay evidence—the only requirement is that the tribunal deem the evidence “relevant and helpful”—the detainee’s opportunity to question witnesses is likely to be more theoretical than real.

In these circumstances, he continued, a habeas review could consider exculpatory evidence not submitted to the tribunal. It could also allow witnesses to testify whom the government had claimed not to be reasonably available, and consider new evidence made available only after the tribunal hearing was concluded.11

Roberts devoted his lengthy dissent to rebutting this part of Kennedy’s argument. He claimed, first, that the CSRT procedures that Kennedy found inadequate in fact satisfy the due process standard O’Connor laid out for citizen detainees in her Hamdi opinion. A detainee who had a right to go to federal court on a habeas writ would, he said, have nothing to complain about when he got there. That did not respond to Kennedy’s point that a federal court would have extensive powers to make new findings of fact. It is worth noticing, moreover, that O’Connor’s standards of due process did not command a majority of the justices in Hamdi and do not appear, at least in retrospect, to satisfy the test she herself proposed: any procedure for classifying citizens as enemy combatants, she said, must balance the grave harm done to someone unjustly detained against the risk of releasing someone who would rejoin a terrorist group.

The balance O’Connor struck in 2004, which allowed a military commission to appeal to hearsay evidence and to presume that the military’s determination is correct unless shown otherwise, might well be deemed to leave prisoners without adequate protection.12 In any case, she assumed that those detained after a review of the kind she described would not be subject to harsh interrogation, which is routine at Guantánamo.

Roberts also argued that the D.C. Circuit Court, which the congressional scheme authorized to conduct “exclusive” review of CSRT determinations, could remedy any defects in a CSRT’s fact-finding as well as a federal district court exercising habeas jurisdiction could. The DTA does not explicitly give the D.C. Circuit power to remedy such defects, but Roberts said that the Court should assume that Congress meant it to have that power in order to save the scheme from being declared unconstitutional. The assumption seems implausible since, as Kennedy pointed out, the evident purpose of the congressional scheme enacted after Rasul was to give the detainees less protection than habeas rights would give them. But even if we assume that the D.C. Circuit could rectify any factual errors, the procedures it would have to use would be much clumsier and take much longer than those that a federal district court, which is constituted to try matters of fact, has available. D.C. Circuit review is therefore not the equivalent of a habeas corpus review, particularly when we remember, as Souter emphasized in his brief concurring opinion, that the detainees have been imprisoned without trial and subjected to coercive interrogation, in some cases for six years.


The rationale of the majority in Boumediene challenges two distinctions that many lawyers have accepted as obvious but that must now be reexamined. The first is the distinction between the constitutional rights of American and alien prisoners; the second is the distinction between the rights of those we imprison on American soil and those we imprison everywhere else in the world.

Neither distinction makes sense in principle. Of course America owes special concern to its own citizens and it must of course grant rights to participate in its political processes only to them. Aliens have no right to enter the country, moreover, and they have only limited rights when the government wishes to expel them and detains them for that purpose. But America owes its duty to respect fundamental human rights, including the right not to be imprisoned unjustly, to all people who come under its authority; and there can be no moral justification for discriminating against foreigners either in the definition or the enforcement of those rights. The Constitution’s text suggests that this moral principle is a constitutional principle as well, because it declares that all “persons” are owed due process. But Jackson, in his Eisenträger opinion, said that foreigners are owed less by way of due process and other rights against unjust imprisonment than citizens, particularly when they are not in America.

It is easy to understand the appeal of that idea in the circumstances of that case: it seemed alarming to suppose that German soldiers who confessed to continuing to war against the United States had the same rights against imprisonment under our Constitution as any American has. It is a different question, however, whether foreigners who deny that they have made war on the US or pose any threat to it have those rights.

The Boumediene decision should now be recognized as ending the discrimination that Jackson claimed. The Constitution’s structure of protections against unjust imprisonment is an interlinked system. It makes little sense to think that aliens have full constitutional rights to habeas corpus without also assuming that they have the rest of the due process rights the Constitution has been understood to grant. The United States should not imprison anyone in circumstances in which its fundamental law prohibits imprisoning its own citizens. We do treat our soldiers accused of war crimes differently from ordinary criminals: they are subject to the Uniform Code of Military Justice. Foreigners accused of waging illegal war against us must be tried under standards no less rigorous than that code.

The Boumediene decision also undercuts the geographical distinction that Jackson took to be crucial and on which the Bush administration relied. The Court, it is true, emphasized the special history and circumstances of Guantánamo Bay. But the functional test Kennedy described as setting the boundaries to those rights cannot justify an automatic distinction between that base and other bases, like Bagram in Afghanistan and our bases in Iraq, where our control is more temporary but is nevertheless sufficiently complete to permit courts to treat the executive branch as fully responsible for what happens there.13 It follows that constitutional rights should extend to prisoners of the United States wherever and in whatever circumstances enforcing those rights would be effective, not offensive to another nation, not inimical to military operations, and not unduly expensive or otherwise “anomalous.”

That functional standard would exclude Jackson’s nightmare; it would not require our military to fly anyone they capture in battle to Washington, together with the soldiers who captured him as witnesses, for a trial to justify his detention. Kennedy’s test makes that an easy case. But the standard does mean that the president cannot escape his constitutional responsibilities by finding some spot on a map to hold those he wants to torture that is fully controlled by but not leased to the US.

Courts applying that functional standard could develop a new legal scheme to replace the flawed detention system that the Bush administration and Congress constructed and the Supreme Court has now dismantled— unless new right-wing appointments to the Supreme Court make that impossible. But that would in any case take years of litigation and probably several fresh appeals to that court, and it would be better if Congress, in the next administration, created a new statutory system. Any new scheme must have three goals. It must enable the government efficiently to try those prisoners it can show guilty of war crimes. It must permit the government, at least for some specified period, to prevent truly dangerous terrorists who cannot be prosecuted from carrying out more attacks. And it must try to insure that those who have been wrongly imprisoned are promptly freed.

Several schemes have been suggested by constitutional lawyers and other commentators. One fairly modest proposal would require the government, within some months after capture or arrest, to designate any prisoner it wishes to continue to detain either as (1) in effect a prisoner of war who can be tried for war crimes but is otherwise entitled to the protection of such prisoners under international law or (2) as a criminal suspect who must then be indicted and tried for crimes under our domestic law.14 Under the Geneva Conventions, prisoners of war cannot be subjected to coercive interrogation; they must be given accommodation as comfortable as that of the soldiers who guard them; and they have other important rights. It is unclear how long international law would permit those classified as prisoners of war to be detained: traditionally prisoners can be held until the cessation of the hostilities in which they were engaged and it is unclear what, if anything, would count as the end of the terrorist threat. It would therefore be necessary for any statute enacting this scheme to provide some time limit for this detention that Congress could, of course, later extend if it deemed necessary. The Boumediene decision guarantees that anyone doubtfully detained as a POW could eventually challenge the government’s classification in court.

Members of terrorist groups do not, however, fit the classical profile of POWs. Even those the US might capture in battle do not fight as soldiers of some enemy power; they belong to extremist groups rather than the forces of sovereign powers, they do not wear uniforms, and they operate as guerrillas. Some of those the government wishes to detain, moreover, were not captured in battle at all. Lakhdar Boumediene was arrested because the American embassy in Bosnia suspected he was in touch with members of al-Qaeda.

It would therefore be necessary to expand the definition of a prisoner of war, and some commentators have suggested that it would be better to avoid the traditional POW regime altogether. They favor designing a new category of persons who can be detained on a showing, perhaps before a new and specially constituted security court, that their release would be very dangerous.15

That suggestion has obvious advantages; it would permit classifications and discriminations that make more sense than the traditional POW structure in a world of continuing terrorist threat. But it has disadvantages as well. Preventive detention—imprisoning someone because he is deemed to be dangerous and not because he has broken a law—is always regrettable and it might be better to limit its use to the well-established POW model than to add a further precedent. It would probably be politically easier to adopt the traditional POW system of rights, moreover, than to enact equivalent protections for a new category. I therefore prefer adapting to our new circumstances the present structure of international law.

The alarmist claims of Scalia, McCain, and others that the Boumediene decision imperils national security are groundless. It remains to be seen what standards the federal district courts adopt on habeas review, how the different lawsuits are consolidated in a small number of courts, what procedures judges develop for allowing classified information to be studied only by the judge and defense lawyers with security clearances and how appellate courts, probably including the Supreme Court on other occasions, react to these district court decisions. There is no reason to think that new procedures will release a greater number of genuinely dangerous terrorists than would be released under the Bush administration’s procedures, which are inefficient as well as unfair. We will, however, gain back some of the national honor that the cowardly decision to imprison without charges anyone who might conceivably threaten us or might conceivably have information useful to us has cost. If McCain makes the Boumediene decision a political issue, we must hope that Obama will make recapturing that national honor a goal in which Americans of both parties can take pride.

——July 16, 2008