Before 2004, few Americans had ever heard of the Office of Legal Counsel or knew the power it wields. A small, elite law office within the Justice Department, the OLC is inconspicuous, and many of its opinions remain secret. Its two dozen lawyers provide legal advice to the entire executive branch of government and by custom, if not law, its opinions are binding on federal agencies. That gives the office substantial power. The OLC attracts talented lawyers, and its alumni include Supreme Court justices, solicitors general, attorneys general, and distinguished academics.

The OLC has gained notoriety in recent years almost entirely because of John Yoo, a law professor it employed from 2001 to 2003. In April 2004, five weeks after the Abu Ghraib revelations transfixed the country, newspapers broke the story of a secret “torture memo,” written by OLC lawyers in August 2002. Torture is a serious federal crime, but the torture memo provided interrogators with the authority to act with maximum impunity. It concluded that inflicting physical pain does not count as torture unless the interrogator specifically intends the pain to reach the level associated with organ failure or death; that inflicting mental suffering is lawful unless the interrogator intends it to last months or years beyond the interrogation; that enforcing criminal laws against presidentially authorized torturers would be unconstitutional; that lawful self-defense can include torturing helpless detainees in the name of national defense; and that interrogating detainees under torture may be justifiable as the lesser evil, through the legal defense of necessity.1

The memo’s legal arguments were widely regarded as preposterous. It defined torture by using language from a Medicare statute on medical emergencies, ignored inconvenient Supreme Court precedents, misrepresented sources, and at one point argued that while torture might be justified as a lesser evil, the same would not necessarily be true of lifesaving abortions.

The torture memo could hardly have been leaked at a worse time. The Bush administration was scrambling to reassure the world that Abu Ghraib was an aberration, not a result of official US policy. The OLC’s opinion on torture had been signed by its head, Jay S. Bybee (now a federal judge), and is also known as the Bybee memo. But news reports soon made clear that John Yoo, not Bybee, was the principal author. Attorney General John Ashcroft hastily disowned the memo, and the OLC withdrew it and replaced it with a second memo on interrogations and treatment of detainees a week before Alberto Gonzales’s confirmation hearing for attorney general began on January 6, 2005. Lawyers who worked on the torture memo were subjected to an internal investigation by the Justice Department. (No independent investigation has yet been carried out.) As Yoo makes clear in War by Other Means, he remains bitter about his repudiation by the Justice Department and his disdain for Ashcroft is palpable. He correctly observes that the OLC’s substitute memo made mostly cosmetic changes, and approved the same treatment of detainees as the original torture memo.2

John Yoo is a young, energetic, mild-mannered law professor at Berkeley’s Boalt Hall law school. He took a leave of absence to join the Office of Legal Counsel the summer before September 11. Yoo’s academic publications had already established his reputation as an aggressively conservative voice on issues of executive power and international law. He had defended nearly limitless war powers for the president, and he fended off unwelcome constraints of international law on US policy by arguing that treaties, which the Constitution declares to be “supreme law of the land,” actually have no legal force unless Congress implements them.3 Neither view was widely accepted by lawyers, but they fit perfectly with the pro-executive, unilateralist outlook of Vice President Dick Cheney and others in the Bush administration.

Then came September 11. In Yoo’s words, “In the months following 9/11, OLC went into overdrive.” According to The New York Times, only ten days after September 11 Yoo prepared a memo on how to overcome constitutional objections to the use of military force within the US, for example “to raid or attack dwellings where terrorists were thought to be, despite risks that third parties could be killed or injured by exchanges of fire.”4 Five days later, Yoo produced another memo, sketching out the president’s war powers in the broadest terms, and concluding with the startling proposition that no act of Congress can

place any limits on the President’s determinations as to any terrorist threat, the amount of military force to be used in response, or the method, timing, and nature of the response.5

The torture memo, written ten months later, restated this theory of the “commander-in-chief override”: that the president’s constitutional authority as commander of the armed forces in war simply trumps other laws, even the criminal law against torture.

Yoo wrote other important opinions concerning the “war on terror.” In January 2002 he worked on two memos arguing that the Geneva Conventions do not protect detainees suspected of belonging to either al-Qaeda or the Taliban. In an opinion of March 14, 2003, which was never leaked, and which the Pentagon has refused to release despite repeated requests from members of Congress, Yoo discusses specific interrogation techniques. (We know of its existence because a letter of February 2005 from the OLC to the Defense Department formally withdraws Yoo’s opinion, and notes that more than a year earlier the OLC had warned the Department of Defense not to rely on it. The letter states that twenty-four techniques are still approved; the obvious inference is that Yoo approved more and tougher techniques than that.6 ) And Senator Carl Levin (D-Mich.) has complained that the Senate was denied access to another memo issued under the name of Jay Bybee, written around the same time as the torture memo, which also discusses specific interrogation techniques; it seems logical to surmise that Yoo wrote this memo as well.7 Yoo mentions reports that he wrote the secret 2002 opinion authorizing the President’s warrantless wiretapping program—but neither confirms nor denies that he did so.


Yoo gives a revealing account of his own ambitions. He identifies strongly with Alexander Hamilton, the most pro-executive and militarist of the constitutional founders, and also the most prolific and polemical. Yoo quotes Hamilton’s adversary Thomas Jefferson: “Hamilton is really a colossus…. Without numbers, he is an host within himself.” Yoo comments: “I decided to take Hamilton as my role model.” He has written a staggering number of speeches, Op-Ed pieces, and articles defending hard-line policies in the war on terror. War by Other Means, his second book since leaving the OLC, offers an uncompromising defense of every one of these controversial policies, including the torture memo, from which Yoo does not back down. He claims that because the OLC’s work is confidential he cannot reveal any inside information; he cannot even tell us directly whether he wrote the torture memo. But he leaves no doubt that he was involved in virtually all the most controversial decisions, and that on sensitive matters he did not collaborate with others. Since he is unable to tell an insider’s story, Yoo’s book is mostly a series of arguments that deserve scrutiny.


A single argument lies at the core of Yoo’s book. The struggle against al-Qaeda, he insists, is a war rather than a matter of law enforcement. Therefore, in his view, the president’s powers as commander in chief apply full force in the fight against al-Qaeda. These powers include the authority to gather intelligence, whether by wiretapping or by harsh interrogation of captives; to kill the enemy (including targeted killings); to capture enemy troops and imprison them until the war is over; and to try them for war crimes before military commissions. It would, Yoo writes, encroach on the president’s constitutionally mandated authority for Congress to regulate these activities. In his view, Congress’s sole constitutional check over the commander in chief is the authority to withhold funds for the war on terror, an act that seems politically impossible (although some members of Congress wish to hold back funds for the Iraq war). Furthermore, he believes, it is unprecedented hubris for courts to intervene by second-guessing what are, in essence, battlefield decisions. As Hamilton argued in The Federalist, the executive branch alone possesses the “energy” and “dispatch” to deal with military emergencies, and so the president’s exclusive war powers are a matter of necessity as well as constitutional design.

Furthermore, Yoo writes, the unconventional nature of al-Qaeda and its tactics “erases the traditional boundaries between the battlefield and the home front.” Al-Qaeda can be anywhere, in or out of the United States. Therefore the battlefield can be anywhere; and on the battlefield, the commander in chief calls the shots, both figuratively and literally. So too, the commander in chief’s time-honored power to gather intelligence becomes more important than ever in an unconventional war, and this war power prevails over constitutional protections of privacy in peacetime. To say that there should be a “wall” between domestic law enforcement and foreign intelligence-gathering makes no sense when no wall exists between the battlefront and the home front. Requirements that the government apply for warrants are senseless red tape in an emergency. Libraries should not be sacrosanct from spying when computer terminals are the enemy’s command and control system.

In Yoo’s view, civil libertarians and human rights advocates make one central mistake, and they make it again and again, on every issue in the war on terror: they regard the struggle against al-Qaeda as a matter of criminal justice, in which all the protections rightly built into our criminal justice system should apply. But according to Yoo, protections that are right for criminal law are dangerously wrong for confronting terrorist threats. As a legal matter, it follows that civil libertarians, by citing peacetime rather than wartime liberties cases, are simply looking at the wrong body of law. The wartime precedents invariably deny judicial review to captured enemies. Judges are not generals and should not pull military personnel from the battlefield to testify in court.


These are not frivolous arguments. Some of Bush’s critics deny that the struggle against al-Qaeda is a war, but I think this is wrongheaded. In Yoo’s words, “Necessity creates war, not a hovering zeitgeist called ‘law.'” September 11 was an aerial attack against US targets using, in effect, a stolen air force, and was one of many attacks by Islamist extremists who, to greater and lesser degrees, are determined to kill for ideological purposes. It is a campaign, not a crime wave. What al-Qaeda shares with traditional war-makers, and what differentiates it from an equally violent and powerful narcotics cartel, is that it uses violence in service of politics. “Politics by other means” was Clausewitz’s definition of war, and it would be foolish to deny that al-Qaeda’s ends are geopolitical. It is equally obvious that al-Qaeda uses unconventional means that demand nontraditional strategies and tactics in response. Here, too, Yoo’s initial premise seems right.

The problem lies not in the label, but in the consequences that supposedly follow from it. For Yoo, labeling the struggle “war” activates every war power formerly associated with battle commanders. The central contradiction, which Yoo never overcomes, is that while he insists that the US is fighting a new kind of war, he also insists that it should be fought with the full panoply of traditional presidential war powers. But these war powers were designed for conflicts in which the enemy is in uniform and belongs to an identifiable foreign government, and whose duration and conclusion are defined by victories, surrenders, and peace treaties.

Thus, in a traditional war, captured enemy troops do not get federal court hearings on whether they can rightly be held captive; but (as Richard Epstein has observed in a recent Op-Ed article8 ), in a traditional war there is seldom any question whether uniformed captives are actually enemy combatants and not cases of mistaken identity. Likewise, in a traditional war, captives can be held until hostilities end. Yet in the new kind of war, where no enemy commander has the authority to surrender, the war ends only when we say it is over. Yoo reassures us that “there is no reason to believe it will go on for a generation.” There is equally no reason to believe it will go on for less, and the potential internment of enemy combatants for decades on end differs only in name from what Yoo denies it is: indefinite imprisonment.9

Most importantly, Yoo never confronts the defining property of the kind of war now being fought: the difficulty of distinguishing it from peace. It is, for the US and other countries, mostly a dormant conflict, with brief, intermittent flare-ups of horror. Most of the time, life goes on without any attack by terrorists or other determined enemies of the US. To protect against such attacks is obviously necessary, but the measures used must be workable in everyday life. Yoo is misleading when he characterizes the world as the “battlefield,” which he tells us is anywhere or everywhere. For all you know, he tells us, you are in the middle of the battlefield right now; therefore the commander in chief’s powers override the rest of the statute book, and courts have no business second-guessing military decisions. But that is a circular argument: to call the decisions “military” is already to place them under the president’s command authority. What we really need to know is how far that authority extends. To the question “when and where are the president’s war powers not in play?” Yoo has no answer beyond “that’s for the president to determine.”

The result is a system with two sets of rules superimposed on each other: the rule of law as we have come to expect it in civilian life, and a regime of war powers with vastly diminished civil liberties. Both regimes can apply everywhere, with the president deciding at his discretion which applies in any given case. Yoo regards as a deep violation of the separation of powers the idea that some civilian protections might apply to the new kind of “military” decision (for example, the decision to arrest a US citizen in a Chicago airport). But there is absolutely no reason to believe that either the framers of the Constitution or its early interpreters would have given broad war powers to a commander in chief if they thought that those powers could displace civilian law anywhere, perhaps for decades on end, just on the president’s say-so.


Yoo might reply that whatever we think the framers would have intended, they made the president commander in chief. But this brings us to a disabling weakness in Yoo’s constitutional theory: the relevant clause of the Constitution says nothing about the breadth of the president’s war powers. The clause designates the president as commander in chief of the army, navy, and militias “when called into the actual Service of the United States”—period. It never explains what powers the commander in chief possesses, and the Philadelphia debates were equally silent on this issue. Is the authority of the commander in chief a narrow power of military command or a vast set of “war powers”? Yoo assumes the latter, but his assumption has no textual support in the Constitution, and he falls back again and again on Hamilton’s call for executive “energy” and “dispatch.” He does not mention that many of the founders had deep suspicions of Hamilton’s pro-executive views.

All the commander-in-chief clause makes clear is that the framers wanted civilian control of the military, presumably to protect against coups and adventurism. The framers certainly remembered the near mutiny of the Continental Army at Newburgh, and the humiliating flight of Congress from Philadelphia to Princeton three months later to escape angry, unpaid Continental soldiers. But they were equally concerned that a civilian leader commanding an army might seize power. Many had in mind the historical examples of Julius Caesar, who brought down the Roman Republic, and Oliver Cromwell, whose New Model Army staged five military coups in fifteen years. Anti-Federalists repeatedly invoked their names, and even Hamilton warned against the danger of a Caesar or a Cromwell.

That is why, during the founding-era debates, the very idea of a standing federal army proved controversial. Many feared that its commander would use it to suppress liberties and launch military adventures. Nor was the fear of military adventures far-fetched: when Hamilton finally saw his dream of a standing army fulfilled, with himself as its inspector-general, he immediately devised what his biographer Ron Chernow calls a “harebrained plot” to attack Latin America.10 Contemplating the Iraq quagmire, we might sympathize today with eighteenth-century worries about messes that a president’s “energy” and “dispatch” can land us in.

To weaken the president’s war powers, the Constitution granted some to Congress and others to the states. Congress got the power to declare war and to regulate the army and navy. More subtly, at a time when far more troops would come from state militias than the federal army, the Constitution specified that militia officers would be appointed by governors, not the president, and only Congress could call out the state militias although, once called out, they are under the president’s command. Even the Second Amendment, providing for the right to “keep and bear arms,” arose partly out of fear that the national government might disarm the militias and seize power. The proposition that, in writing the commander-in-chief clause, the framers had in mind a civil and military super-executive is farfetched. So is the idea that the commander in chief can override the laws; the Constitution instructs the president to “take care that the laws are faithfully executed.”

News reports have suggested that the Vice President and officials around him viewed September 11 as an opportunity for Bush to extend executive power. It is of interest that Yoo’s first memo on the overriding power of the commander in chief appeared two weeks after September 11—and seven days after Congress had authorized the President to use military force against al-Qaeda and its sponsors. The memo declared that the President has “inherent” power to respond militarily to September 11, and that Congress’s authorization merely “acknowledges” this inherent power. In other words, even after Congress gave the President the military authority he asked for, Yoo was composing an unnecessary opinion declaring that the President did not need Congress’s authorization.

Only once has Yoo complained that a president “exercised the powers of the imperial presidency to the utmost…in our dealings with foreign nations.” He added, “Unfortunately, the record of the administration has not been a happy one, in light of its costs to the Constitution and the American legal system,” and “the administration has played fast and loose with the law.” He added that “when it comes to using the American military, no president in recent times has had a quicker trigger finger.” Yoo wrote those words about President Clinton in 2000.11


Yoo argues forcefully and intelligently, but not always honestly. Half-truths, straw men, double standards, selective quotations, significant omissions, and caricatures of his opponents’ positions—all are characteristic of War by Other Means. He writes, for example, that the Bush administration “never challenged the courts’ jurisdiction to review writs of habeas corpus or any other claims involving American citizens,” and that “no one has questioned the role that the judiciary plays.” The first assertion is true but disingenuous; the second is simply wrong. In the case of the US citizen Yaser Hamdi, the administration insisted that “courts may not second-guess the military’s determination that an individual is an enemy combatant and should be detained as such”12 (my emphasis). While technically this argument was not framed as a challenge to the jurisdiction of the courts, the government was claiming that there was no judicial recourse against detention of an enemy combatant. The government offered the same argument in the case of the US citizen Jose Padilla. Had the Bush administration succeeded, these arguments would have required courts—on the basis of nothing more than a presidential assertion—to automatically endorse executive determinations that an American citizen is an enemy combatant and therefore not entitled to even basic access to the legal system. The Bush administration did in fact “question the role that the judiciary plays”: it argued that unless courts defer to the president, they would be “interfering with military decision-making” and the president’s “war-making powers” (in the words of the government’s brief in Padilla).

Equally disingenuous is Yoo’s assertion that

intelligence obtained in military detention usually can’t be used in any kind of criminal prosecution, since it would have been obtained without Miranda rights. We will obtain information that may prevent a future al Qaeda attack, but that information cannot be used to convict the detainee of a crime.

As Yoo knows very well, the Bush administration planned to use evidence obtained during coercive interrogations in presenting cases before military commissions. The new Military Commissions Act excludes evidence obtained through torture but permits other evidence obtained coercively, if the government disputes that the coercive methods were “torture.” Because of Yoo and his successors, the administration interprets torture very narrowly, and denies that any of its techniques are torture. Under the new law, this evidence can be admitted.13 It is entirely possible that Guantánamo defendants could receive the death penalty based on evidence obtained from witnesses who were subjected to waterboarding.14 The Bush administration, like John Yoo, has never conceded that waterboarding is torture.

Yoo is given to the kind of overstatement that one would think a professor of law would correct in students. Human rights groups believe, he writes, that “all the lawyers in the Department of Justice, the White House, and the Defense Department are engaged in a conspiracy to twist the law….” Any limits on warrantless wiretapping, he claims, would “disable” the president from gathering intelligence—a proposition the administration itself has now contradicted by saying it will submit all wiretapping requests to the secret FISA court. Critics would not even allow the president to “search for” terrorists. “According to the critics,” he writes, the president must “check back with Congress on every strategy and tactic in the war on terrorism.” Zacarias Moussaoui’s lawyers, according to Yoo, argued that “the US was really to blame for 9/11.” He cites no convincing evidence for any of these gross overstatements.

Yoo’s descriptions of the Bush administration’s interrogation tactics and treatment of prisoners are equally unreliable. He tells us that the President ordered that “al Qaeda and Taliban operatives…be treated humanely.” In fact, the President ordered only that the armed forces treat them humanely, leaving a conspicuous loophole for the CIA.15 In describing the interrogation of Mohamed al-Kahtani, Yoo refers to

unverified media reports…that some interrogators went beyond their orders and made Kahtani wear women’s underwear, put him on a leash and made him bark like a dog, and put him on an IV when he went on a hunger strike.”

The first two tactics were described not in “unverified media reports” but in General Randall Schmidt’s official Pentagon report on Guantánamo abuses—and the report labels the tactics “authorized.”16 Yoo implies that sleep-deprivation tactics mean giving detainees only six hours’ sleep a night; but Schmidt reveals that “SECDEF [Rumsfeld] approved 20-hour interrogations for every 24-hour cycle,” and that Kahtani was interrogated eighteen to twenty hours a day for forty-eight days out of fifty-four (a two-day break was granted when his heart rate plunged dangerously).

Yoo insists that working dogs could confront Guantánamo inmates only “in a nonthreatening posture.” General Schmidt found that “in November 2002 a military working dog was brought into the interrogation room and directed to growl, bark, and show his teeth at the subject”—and, again, the technique was “authorized,” indeed authorized by the secretary of defense.17 According to Yoo’s false statement, however, “Rumsfeld did not approve use of dogs.” Yoo does not even mention other “authorized” techniques, such as blasting detainees with “futility music” (General Schmidt’s term), having a female interrogator straddle a detainee and describe to him the deaths of fellow al-Qaeda members, and threatening to have the detainee’s mother arrested and sent to Guantánamo.

At least four of the Guantánamo tactics described by General Schmidt—intimidation with dogs, forced nudity in front of American servicewomen, placing women’s underwear on the detainee’s head, and leading the detainee around on a leash—were reenacted at Abu Ghraib. Yet Yoo dismisses those who wonder if this is more than coincidence as “conspiracy theorists.” Yoo neglects to mention that the Abu Ghraib abuses occurred soon after the Guantánamo commander, Major General Geoffrey Miller, was reassigned there, where, in his own words, his “team used JTF-GTMO operational procedures and interrogation authorities as baselines.”18 Nor does Yoo mention that the Army’s Jones/Fay report on Abu Ghraib says that some interrogators at Abu Ghraib had previously served at Guantánamo and raises the possibility that “‘word of mouth’ techniques…were passed to the interrogators in Abu Ghraib by assistance teams from Guantánamo….”19 To suspect that the techniques migrated there from Guantánamo is hardly what Yoo calls it—“an exercise in hyperbole and partisan smear”—unless the Army itself was engaged in hyperbole and partisan smear.

If it is true, the hypothesis of a migration of interrogation techniques illustrates how the consequences of official legal opinions can overflow whatever boundaries their authors imagined would hold them in check. The Bybee memo justifying coercive interrogation was carefully studied by the Defense Department, and large sections of it were incorporated verbatim into a DOD working group report on interrogations issued on April 4, 2003, months before General Miller went to Abu Ghraib. When the heads of the Judge Advocate Generals, the departments responsible for military law in the Army, Navy, Air Force, and Marines, learned of the memo, they wrote outraged letters complaining that “OLC does not represent the services; thus, understandably, concern for service members is not reflected in their opinion”; three of them protested that the OLC’s approach to interrogation might corrupt the uniformed services.20 Yoo never mentions these letters or acknowledges their concerns; nor does he mention that the torture memo was written without consulting military lawyers. Instead, he reassures his readers that the memo went through proper review channels within the Justice Department.


Since Congress passed the Military Commissions Act last September, John Yoo’s views have largely prevailed, at least for the time being. (It is hard to imagine the courts upholding all of it.) The Supreme Court found that the courts have habeas jurisdiction over Guantánamo; the new act removes it. The act bans courts from using international law to interpret the US war crimes statute, and grants the president authority to interpret the Geneva Conventions. In Hamdan the Court held that Common Article 3 of the Geneva Conventions protects detainees, and thus that torture, humiliation, and unfair trials are war crimes that violate their Geneva rights. The act retroactively states that humiliation and unfair trials are not crimes, and forbids detainees from invoking their Geneva rights in military commissions. Furthermore, it draws fine-grained distinctions between torture (which requires “severe pain or suffering”) and cruel and inhuman treatment (which requires only “serious pain or suffering”—an effect that is bewilderingly defined in another clause as “extreme pain” as if there is some comprehensible difference between “extreme pain” and “severe pain”). The act exempts bruises, cuts, and abrasions from the kind of bodily injury that counts as cruelty. The specific standards may differ from those in the Bybee memo, but that memo launched the entire misguided project of defining torture through arcane legalisms carefully tailored to let the administration do what it wishes.

I have seen no evidence that Yoo’s writing after he left the Office of Legal Counsel played any part in the debate over the Military Commissions Act. But it may well be that the memos he wrote while in government helped shift us to a new standard by which “energy” and “dispatch” can prevail over justice and law, and the job of lawyers is to bury that fact beneath a mass of technicalities. It remains now for a Democratic Congress to hold hearings to clarify Yoo’s role in changing government policy and go on to propose changes in repressive laws—as the chair of the Senate Judiciary Committee, Patrick Leahy, has urged. For unless it is properly addressed by the government, Yoo’s troubling legacy could affect the history of American liberties for decades to come.

—February 14, 2007

This Issue

March 15, 2007