You are here

Chapter Two: The Ethics of Emergency

Now in a well-ordered republic it should never be necessary to resort to extra-constitutional measures; for although they may for the time be beneficial, yet the precedent is pernicious, for if the practice is once established of disregarding the laws for good objects, they will in a little while be disregarded under that pretext for evil purposes. Thus no republic will ever be perfect if she has not by law provided for everything, having a remedy for every emergency and fixed rules for applying it.

Niccolò Machiavelli


Terrorist states of emergency raise fundamental questions about the nature of the rule of law.1 If laws can be abridged and liberties suspended in an emergency, what remains of their legitimacy in times of peace? If laws are rules, and emergencies make exceptions to these rules, how can their authority survive once exceptions are made? In this chapter, I consider the impact of emergency suspensions of civil liberties on the idea of the rule of law, and I ask a related question: what remains of the status of human rights if they can be abridged in times of public danger?

Roughly speaking, emergency legislation may take three forms: national, territorial, or selective. In national emergencies, martial law is substituted for civilian rule throughout a country for an indeterminate period of time. A state facing a terrorist insurgency will suspend the normal rule of law in order to give the military full authority to arrest, detain, search, and harass insurgents within a civilian population. In some Latin American countries—Colombia, for example—emergency suspensions of the constitution have been so frequent that they have replaced the rule of law as the norm.2 In territorial emergencies, martial law is confined to special zones of the country, where terrorist or insurgent activity is underway and the state believes it needs the military to have powers of detention, search, and arrest without civilian constraint or review. In active combat zones in Sri Lanka, where the state is combating an insurgency, military, rather than civilian, law prevails. A special subcategory of emergency rule relates to zones of occupation. In the occupied territories under Israeli control, the civilian due process guarantees that obtain in Israel proper are suspended and the territories are governed by military rule, distinct from the Israeli judicial system. A further subvariety of territorial emergency legislation operates in Northern Ireland. The province is not under martial law, but the rule of law functions differently than in other parts of the United Kingdom: for example, in the operation of special courts that try terrorist cases without standard trial by jury. The third form of emergency legislation is piecemeal: no general state of emergency is proclaimed, no part of the country is removed from the normal rule of law, but portions of the law are suspended for terrorist suspects. This third—selective—kind of emergency has been manifested in the United States since September 11: preventive or investigative detention for certain detainees, either aliens or citizens; changes in their right of access to counsel and lawyer-client privilege; widening of police powers of search and seizure; increased wiretap authorizations and other forms of surveillance. Most of what I say here will focus on these piecemeal emergency measures. Some of these measures are limited by sunset clauses providing for their expiration, while others are permanent. Even a selective emergency, therefore, can turn into a constitutional way of life.

Emergencies, in the words of Kathleen Sullivan, raise the specter of a “constitutional black hole,” a break in the continuity of law that contradicts the idea of a constitution as an unvarying arbiter of social and political disputes.3 How can the rule of law be maintained if the law can be suspended as necessity dictates? How can the effectiveness of human rights as a guarantee of dignity be sustained, if rights are suspended in emergency?

Emergencies invariably entail the extraordinary use of executive prerogative. Prerogative, as John Locke defined it, is the “power to act according to discretion for the public good, without the prescription of the law and sometimes even against it.”4 Locke thought such powers necessary to the maintenance of a government in times of crisis, and in thinking this, he was sustaining a line of thought that went back to the Roman republic and to its system of temporary dictatorship. This kind of temporary, delegated dictatorship to save the constitution has always been a feature of republican thought.5 Liberal theories of constitutional government, however, have always feared that an executive might use the pretext of emergency to seize power and abolish constitutional freedom. There is thus a conflict between a republican and a liberal theory of emergency powers.6 A republican account could envisage a democratic rationale for rights abridgments in emergency based on the need for executive decisiveness to protect majority interests, while a liberal view would fear that such a majoritarian rationale would risk permanent damage both to rights and to the system of checks and balances.

One of the most pertinent defenses of this republican use of prerogative power and the necessity of emergency suspensions of liberties was made by Abraham Lincoln in a letter in 1863 justifying his suspension of habeas corpus and indefinite detention of opponents of the draft laws, during the American Civil War. Yes, Lincoln was talking about wartime, and terrorist emergencies are not the same as wartime ones. Still, his words remain relevant to our case, and I will quote at length, since Lincoln speaks directly to the central issue: whether temporary abridgments of freedom do permanent damage to constitutional liberty:

The Constitution is not, in its application, in all respects the same, in case of rebellion or invasion involving the public safety, as it is in time of profound peace and public security. The Constitution itself makes the distinction; and I can no more be persuaded that the Government can take no strong measures in time of rebellion, because it can be shown that the same could not be lawfully taken in time of peace, than I can be persuaded that a particular drug is not good medicine for a sick man, because it can be shown not to be good food for a well one. Nor am I able to appreciate the danger apprehended by the meeting that the American people will, by means of military arrests during the Rebellion, lose the right of Public Discussion, the Liberty of Speech and the Press, the Law of Evidence, Trial by Jury and Habeas Corpus, throughout the indefinite peaceful future, which I trust lies before them, any more than I am able to believe that a man could contract so strong an appetite for emetics during temporary illness as to persist in feeding upon them during the remainder of his healthful life.7

As Lincoln predicted, though he did not live to see the day, habeas corpus returned in robust health after the Civil War. Its temporary suspension in wartime did not erode its legitimacy once peace returned. On the contrary, Lincoln's actions had the effect of hardening judicial opinion against easy suspensions in the future. In Ex parte Milligan in 1866, the U.S. Supreme Court condemned Lincoln's suspension of habeas corpus in uncompromising language:

No doctrine involving more pernicious consequences was ever invented by the wit of man than that any of the Constitution's provisions can be suspended during any of the great exigencies of government. Such a doctrine leads directly to anarchy or despotism, but the theory of necessity on which it is based is false; for the government, within the Constitution, has all the powers granted to it which are necessary to preserve its existence.8

Yet the Milligan decision is not the last word on the long-term impact of Lincoln's actions. Later Supreme Courts, especially those sitting during the Second World War, sided more with Lincoln than with the judges in Milligan.9 When Roosevelt established a military commission in 1942 to try German saboteurs and the case went to the Supreme Court for review, Lincoln's use of military commissions was cited by the government in justification.10 When the Bush administration established military tribunals to try terrorists and illegal combatants for crimes of war and crimes against humanity, rather than putting them through the federal court system, the Lincoln decision again figured as a precedent.11 So if in the short term Lincoln was right—the remedies given a sick man are not those he craves when well—in the longer term, the precedent he set has provided subsequent presidents with the warrant to abridge liberty in emergencies.12

There are two types of question to ask about these emergency circumstances. First, are they actually necessary? And second, even if they are necessary, will they dimmish respect for laws and rights in the future? The two questions need to be distinguished: short-term benefits to order may entail long-term harm to liberty. But there may also be a deeper conflict. A lesser evil approach to a war on terror acknowledges a tension “of tragic dimensions,” as Oren Gross puts it, between what is necessary and what is right.13 Saying the choices are tragic is not meant to excuse indecision—decisions will have to be made—but decisions in favor of necessity should be constrained by awareness of the seriousness of the loss in terms of justice. Weighing loss in this way implies the inadequacy of a solely pragmatic or utilitarian calculation of the balance between order and freedom. A utilitarian calculus may be biased toward security, because it has to value majority interest more than minority rights loss. Such a calculus may also be too short-term, neglecting long-term losses to the framework of rights.

In the previous chapter, I laid out a balanced position in which neither necessity nor liberty should have trumping claims. Terrorist attack can justify abridgments of liberty only if suspensions of liberty do actually enhance security. If they do, it should be lawful to detain suspects and hold them without trial, until the nature of the risk they pose can be determined. At the same time, detainees must retain the right to counsel and judicial review of their detention. A constitutional democracy should keep rights abridgments to a strict minimum and limit their duration with sunset clauses. When organizations are designated as terrorist, they should retain the right to judicial review of their proscription.14 Withdrawing the rule of law from anyone in a democratic state can never be justified. No “constitutional black holes” should be allowed to develop in a war on terror, for they threaten to pull individuals down into realms beyond the control and the ken of other citizens.

Any strategy of preventive or investigative detention that targets a particular social group—say, Arab Americans or aliens of Muslim faith—potentially violates the constitutional norms of equal protection. Rule of law implies both invariance and equality, both unchanging standards of due process and equal ones for all persons. Selectively targeted detentions may be justified in situations of dramatic uncertainty, like the weeks following September 11, but once it became clear that there was no larger plot or further risk of attack, equality considerations alone would mandate the early release of specially targeted suspects, against whom probable cause cannot be demonstrated.

The second question about abridgments of liberty is long-term: what do they do to the status of law once an emergency is over? Lincoln insisted that emergency suspension of civil liberties did not endanger the constitutional fabric in peacetime, but civil libertarians point out that temporary suspensions have a way of becoming permanent.15 Moreover, their ambit has a way of spreading: measures introduced to stop terrorists are then used, once the emergency is over, to catch criminals and other types of offenders. Terrorism becomes a widening wedge, in other words, that weakens due process protections for all. Even if emergency measures are eventually revoked, the very fact that the law is made more severe in a time of emergency, civil libertarians argue, does damage to respect for the law as an abiding set of standards. This is especially the case with national emergencies, which substitute some form of martial law for the rule of law throughout a whole country. There seems little doubt that emergency law, so frequently invoked in Latin America to create temporary military dictatorships, has made it more difficult for committed democrats to anchor the constitutional rule of law throughout their continent.16 But even piecemeal emergencies may do significant damage to the rule of law.

The question that civil libertarians raise is the relation between legitimacy and invariance. Do legal rules have to be invariant in order for them to retain legitimacy? In some constitutions, the rules are written so as to exclude as much variance as possible. U.S. civil libertarians point out, for example, that the U.S. Constitution makes only one reference to emergencies, in the clause allowing the suspension of habeas corpus in “cases of rebellion or invasion” when “the public safety requires it.”17 This, in their view, compares favorably with more recent European constitutions that make elaborate provision for the imposition of emergency regulations in times of war or civil insurrection.18 Clearly, civil libertarians prefer a constitution that allows the fewest possible opportunities for the exercise of emergency power.

The metaphor most often used to describe what is at stake in this discussion is the story of Ulysses and the Sirens in Homer's Odyssey. Ulysses, voyaging home from Troy, is warned by an oracle that death awaits him and his crew if they listen to the haunting voice of the Sirens, who will lure them and their ship onto the rocks. So Ulysses orders the crew to lash him to the mast, and he puts beeswax into the ears of his oarsmen; thus protected, they row safely past the Sirens, while Ulysses himself listens to their alluring cries.

Ulysses' committing, in advance, to the idea of civil liberties' conduct is often cited to help us understand law and rights as strategies of precommitment. Like Ulysses tying himself to the mast so that he will hear but not succumb to the Sirenscommitting, in advance, to the idea of civil liberties' song, democratic states precommit themselves to respecting rights, knowing that they will be sorely tempted to abridge them in times of danger.19 Ulysses commits himself in advance, knowing that when temptation is upon him, it will be too late. Rights, this story tells us, are like Ulysses' beeswax: devices of reason, designed in moments of tranquillity, to master temptation in times of danger.

Terrorism is a supreme test of a liberal society's ability to abide by these precommitments. For they are commitments to invariance and equality alike, and going back on these commitments is easy. Oppression by majorities comes at little direct cost to their own liberties and rights. As Ronald Dworkin has pointed out, the trade-off is not between our liberty and our security in times of terrorist threat, but between our security and their liberty, by which he means the freedoms of small suspect groups, like adult male Muslims and particularly the subset in violation of immigration regulations.20 These abridgments of the rights of a few are easy to justify politically when the threat of terrorism appears to endanger the many.

The idea of rights as precommitments seeks to give special attention to the dignity claims of individuals who may be endangered by majority interest. Special attention here simply means that we cannot accord a trumping claim to the public order interests of the majority. We need to balance competing claims, and we should do so on the assumption that precommitments made in times of safety should be maintained, as far as possible, in times of danger.


Yet this preliminary outline of a lesser evil position is controversial, and to appreciate why, I need to contrast it with other possibilities. To the question “What do exceptions and emergencies do to the rule of law?” there appear to be three answers. A pragmatic answer is that the law's legitimacy has something to do with invariance, but much more to do with effectiveness. So if you have to choose between leaving laws unchanged and changing them to stop terrorists in their tracks, and your concern is to maintain the rule of law, far better to abridge a right than to let a concern for pure consistency stay your hand. Lincoln's justification of his use of emergency powers in 1861—“are all the laws, but one, to go unexecuted, and the government itself go to pieces, lest that one be violated?”—is the great standing defense of sacrificing invariance for the sake of effectiveness.21 A second position, closest to pure civil libertarianism, would be that legitimacy—and, as a result, effectiveness—are tightly tied to invariance, to the idea that law should not bend to necessity, emergency, or political exigency. A third view, closest to the lesser evil position set out in this book, does not want to separate effectiveness, invariance, and legitimacy. According to this view, laws do derive some of their power from being difficult to change, and yet if they are completely unresponsive to emergency situations, they may be ineffective. Emergency alterations of legal protections like habeas corpus may be necessary, but there is a price to pay when you do. Habeas corpus is weakened because once it has been suspended, everyone then understands that an important guarantee of freedom is pliable and susceptible to political pressures.

It is not merely the legitimacy of the law that terrorist emergencies call into question, but the status of the moral standards encapsulated in the idea of human rights. Here the claim would be that the specific civil and political rights to be found in liberal democratic constitutions are not just a revocable privilege of citizenship, but reflect a constitutional commitment to respect the particular and equal moral status of human beings. One way to put this would be that beneath civil and political rights, enshrined in the particular constitutions of liberal democracies, stands an idea of human rights. This is a moral rather than a juridical claim: it does not imply that all liberal democratic states incorporate human rights norms into their constitutions in the same way. Some of these states may have signed on to international human rights conventions, and others, like the United States, may sign on only to some or sign on with reservations, but even if their relation to international human rights is ambivalent or ambiguous, their constitutional doctrine expresses a view of the protected moral status of human beings that is consonant with human rights.22

This commitment to respect the moral status of human beings means treating individuals equally and treating all with at least a minimum respect. Insisting on establishing the guilt of individuals beyond a reasonable doubt is one practical way that the law does this. Another is that the law is supposed to keep these commitments irrespective of the way its officers—policemen, judges, lawyers—may feel toward the prisoner in the dock.23 This commitment—to respect individuals irrespective of conduct—distinguishes human rights from other rights. Some civil rights—like the right to vote—can be forfeited on commission of felony. They are conditional, therefore, on conduct. Others, like the right of judicial review of detention, might be revocable in emergencies if necessity dictates. But human rights are independent of conduct, circumstances, citizenship, desert, or moral worth. Humans have human rights simply because they are humans. Hence even terrorists have rights that they cannot be denied. A liberal society is committed to respect the rights of those who have shown no respect for rights at all, to show mercy to those who are merciless, to treat as human those who have behaved inhumanly. This commitment to observe obligations even when they are not reciprocated is a defining characteristic of any society under the rule of law. Why else do we believe that even the most odious criminal is entitled to a fair trial and proof of guilt beyond a reasonable doubt? We care about these standards, not simply for prudential reasons—to ensure that the correct person is convicted and taken off the streets—but also because we, the democratic we, believe we owe this to everyone. On this reading the rule of law is more than a procedure, treating like cases alike. It also seeks to serve a moral idea: that any human being is entitled to a basic equality of treatment. If this is what is owed common criminals in normal times, there is no convincing reason why it cannot be owed terrorist suspects in times of emergency.

This idea, however, is not shared by all. Of course, every democrat thinks that respect for dignity may be a consequence of a successful regime of laws. But not everyone thinks that this is the law's essential purpose and the test of its legitimacy. Such a view would be a “moral reading” of constitutional rights, to use Ronald Dworkin's phrase, and it is at odds with more pragmatic readings of constitutional rights—readings that would emphasize their procedural utility in arriving at judicial outcomes rather than the idea that they incarnate any particular moral commitments.24 Rights, on a pragmatic reading, are side constraints on the public policy choices of legislatures and governments. As such, the rationale of rights is to protect public interests, not to instantiate dignity commitments. In times of emergency, therefore, they can be properly abridged with no negative consequences since they are, to begin with, a instrument of collective interests and nothing else.

Yet it is hard to see why anyone would care about the rule of law if it is simply a piece of useful procedure or a public policy side constraint. It is hard to understand how the rule of law could generate allegiance unless it expressed a common moral commitment to the dignity of individuals. At least some of the reason why people obey the law has to do with the idea that it is theirs, a product of democratic suffrage, but also that it expresses a commitment to accord them equal respect and equal consideration.

Moreover, without some idea that dignity is at risk when rights are abridged, there are few pragmatic grounds to care when large numbers of people suffer harm from emergency abridgments of rights. Richard Posner, a practicing judge and a leading exponent of the pragmatic view, argues: “In hindsight, we know that interning the Japanese American residents of the West Coast did not shorten World War II. But was this known at the time? If not, should not the government have erred on the side of caution, as it did?”25 This works well as an argument against summary condemnation of Roosevelt's actions from the safety of hindsight, but in saying the government was right “to err on the side of caution,” Posner passes over in silence the irrevocable harm that was done to a hundred thousand Japanese, most of whom were citizens and none of whom turned out to present a danger of subversion or sabotage. This example helps to distinguish a lesser evil position from a pragmatic approach to rights abridgment. Both seek to balance claims of liberty against security, both deny trumping claims to either, but a lesser evil position would put far more emphasis than a pragmatic one on the loss entailed in the abridgment of the rights of the Japanese. This emphasis would be grounded in the idea that rights-based commitments to individual dignity are intrinsic to the definition of what a democracy is. By committing, in advance, to the idea of civil liberties' abridgment as an evil, to be avoided whenever possible, it weighs the internment of the Japanese by a less pragmatic standard than whether it would or would not shorten the war with Japan. A lesser evil approach would reason as follows. Shortening a war is a morally significant goal, but the loss of freedom that internment will entail is such a serious blow to the individuals involved, and the likelihood that internment will shorten the war is so uncertain, that the rights abridgments cannot be justified.


In retrospect, the Japanese internment seems like an obvious abuse. The real difficulty, as Richard Posner correctly says, is to get the balance right when emergencies are underway, and when no one knows what measures will turn out to have been necessary. Getting this balance right means taking a view on whether presidents and prime ministers are entitled to declare emergencies in the first place. Some civil libertarians believe that this exercise of prerogative power is dangerous in itself. But who else can decide when emergencies exist? A constitutional order that does not accord a leader the capacity to declare an emergency and take decisive action to meet it may perish through deliberative irresolution and paralysis.

Yet even if some exercise of prerogative power is inevitable, a civil libertarian might say, this leaves the determination of emergency in the hands of a single leader who is bound to manipulate a crisis to his or her own advantage. This may be true, but where is the remedy? Determination of whether an emergency exists is unavoidably political: there cannot be a science or a law of the matter. But this doesn't mean it has to be capricious or arbitrary. As constitutional scholar John P. Roche wrote more than fifty years ago, “when there is a consensus on the existence of an emergency; i.e., when public and congressional agreement on presidential action are apparent, an emergency exists.”26 While it is presidents and prime ministers who declare emergencies, their use of this power can be subjected to democratic regulation. There must be some consensus, among the public and their elected officials, that the executive is justified in declaring one. Where this consensus does not exist, the proclamation of an emergency will be regarded as an abuse of executive prerogative. If the institutions of adversarial review in a democracy are doing their job, they will resist and set limits to both the proclamation of emergency and the exercise of prerogative.

So the problem with emergency exercises of prerogative power is not, as civil libertarians imply, that they can never be justified, since they obviously can, nor even that their use is undemocratic in itself. For the recurrent rationale of the use of executive prerogative has been a threat to a vital majority interest. In 1861, Lincoln defended the measures he had taken—blockading the southern ports, suspending civil liberties, allocating funds to pay the army and navy without congressional approval—on the grounds that “these measures, whether strictly legal or not, were ventured upon, under what appeared to be a popular demand and a public necessity.”27 Emergencies become a challenge to democracy and to law when proclaimed on grounds that involve bad faith, manipulation of evidence, exaggeration of risk, or prospect of political advantage. Checks and balances cannot stop an executive from declaring an emergency, but Congress and the courts can step in, once emergencies are declared, to subject the exercise of prerogative to adversarial review. When President Truman nationalized the U.S. steel industry in April 1952, arguing that an impending strike threatened the national defense at a time when the nation was at war in Korea, the measure proved unpopular in Congress and the press, and the Supreme Court in June 1952 declared the measure unconstitutional.28

September 11, on the other hand, was an unquestioned emergency. The issue, then, is what measures such emergencies can justify.

A pragmatic view of rights, like the one adopted by Richard Posner, would hold that precommitment is unrealistic because it binds hands that need to be untied in time to cope with danger. Adversarial review, likewise, could prevent a republic from taking swift and robust action in a time of invasion or emergency.

The Roman answer to this problem was temporary dictatorship. The fabled Cincinnatus left his plow and was given dictatorial powers by the Senate in order to lead the republican armies to victory against an invasion that had reached the gates of Rome. Roman thought thus saw dictatorship and republican authority as complementary opposites, not as enemies. This sounds strange to modern democrats, who are more likely to think of dictatorship as a permanent foe or as a potential nemesis. Modern democrats are also likely to think of democracy as being intrinsically self-sustaining, capable of reconciling presidential authority and rule of law at all times. Yet the historical record of the twentieth century suggests otherwise. In times of war, Roosevelt and Churchill were given dictatorial powers by their legislatures, with the approval of the courts, to save their countries.29 The necessity of these powers suggests that democracies are not intrinsically self-sustaining; rather, as the Romans thought, their executives recurrently need dictatorial prerogative to cope with crisis.

Emergencies lay bare an enduring conflict in democracy between adversarial justification, which can make for indecision, and strong authority, which may lead to dictatorial abuse. Yet the wartime experience of democracies shows that these two tendencies can be reconciled. The dictatorial powers exercised by Roosevelt and Churchill remained subject to the supervision of the legislature and the courts and, once the emergency had passed, were quickly dismantled. Churchill was unceremoniously voted out of office within weeks of leading his country to victory. Democracy, therefore, can survive episodes of “constitutional dictatorship,” provided that the dictatorship remains constitutional, and provided that it remains temporary.30 This means that the executive's prerogative ought to be limited by the constitution: presidents should never have the power to prolong themselves in office, to suspend elections, to disband political competition, to alter the constitution itself, or to dissolve legislatures permanently.31

A war on terror sustained indefinitely, against a succession of rogue states or terrorist cells, poses unknown dangers, however, since no one knows how long the emergency will last. In a long twilight war, largely fought by secret means, the key issue is maintaining as much legal and legislative oversight as is compatible with the necessity for decisive action. Sunset clauses—setting time limits to extraordinary powers—seem an essential way to reconcile security and liberty. Thus in Great Britain, the original Prevention of Terrorism Act, which drastically increased police power in Northern Ireland, was not made permanent but was made subject to a periodic process of parliamentary renewal. The new Terrorism Act has no such provision, and this is its chief danger.32

In a war on terrorism, unlike the situation that Lincoln faced in the Civil War or the one that Roosevelt faced after Pearl Harbor, it is not obvious why the president's powers should be increased. He does not have to conscript labor or economic resources, as Roosevelt did, and he does not face a secession, as Lincoln did. There is no good reason, therefore, why his authority should not be kept under constant adversarial review by courts, Congress, and a free media.

The recurrent problem with keeping his authority under review is excessive deference on the part of both legislatures and the judiciary to an executive's power as commander in chief. In the United States, this deference to executive decision during emergencies has evolved into a legal convention, most recently on display in the U.S. federal court's refusal to grant habeas corpus petitions in the case of a U.S. citizen, Yaser Esam Hamdi, held incommunicado on a U.S. Navy brig as an enemy combatant. According to the court, “the federal courts have many strengths, but the conduct of combat operations has been left to others. The executive is best prepared to exercise the military judgment attending the capture of alleged combatants.” The court held, in other words, that the deference accorded to an executive in wartime should also apply in the very different circumstance of a terrorist emergency. Even so, the court sounded troubled by the implication: “we ourselves would be summarily embracing a sweeping proposition—namely that with no judicial review, any American citizen alleged to be an enemy combatant could be detained indefinitely without charges or counsel on the government's say-so.” Yet this is exactly the consequence of its decision, one that has been confirmed by a superior court. The best that can be said is that an extraordinary act of executive power—holding a citizen incommunicado—has been subjected to judicial review.33 The worst that can be said is that courts have deferred to an executive at precisely the place where they should never defer, in a matter relating to the fundamental civil rights of an American citizen. As of this writing, Supreme Court review of this issue is pending. It remains to be seen whether the nation's highest court will accept the view of executive prerogative taken by the president, or whether it will insist that it is up to the judiciary to decide the due process standards that should apply to foreign nationals or U.S. citizens detained in a war on terror.34 On the view of checks and balances taken in this book, the courts rather than the executive must remain control of due process standards for both civilian and military detainees. Failure by the courts to uphold their jurisdiction over this matter would weaken democracy itself.


Emergencies bring into focus two radically different images of the law. In the image personified in Ulysses, law and human rights derive their legitimacy from their status as rules that should not bend to circumstance.35 In the competing image, to use U.S. Supreme Court Justice Robert Jackson's phrase, the constitution cannot be a suicide pact.36 When suicide threatens—that is, when democracy's own deliberative slowness threatens it with disaster, or when rights guarantees extended to the honest many frustrate the pursuit of the dangerous few—rulers must act decisively. But can they, as Locke implied, operate beyond the law, or should their prerogative be constrained by law?

In understanding what is at stake in the use of prerogative power in time of emergency, we need to come to grips with the thought of the Weimar legal theorist Carl Schmitt.37 His jurisprudence was framed in the 1920s by Weimar Germany's struggle to preserve constitutional democracy in the face of terrorist violence from both sides of the political spectrum.38 Schmitt understood our question to be this: does the exception save or destroy the rule? Schmitt was emphatic: without the power to declare exceptions, the rule of law cannot survive. “Sovereign is he,” Schmitt famously wrote, “who decides the exception.”39 A state cannot remain sovereign, cannot reliably maintain an internal monopoly of the means of force together with the rule of law, if the president cannot exempt himself upon necessary occasions from constitutional rules that would prevent him from prevailing in a contest of force with the state's enemies. Schmitt notoriously said that politics was about punishing enemies and rewarding friends. Constitutional regimes, he went on, cannot save themselves from attack unless the friends of constitutionalism withdraw the law's protections from its enemies.

Schmitt also claimed that the question of whether exceptions destroy or save the rule was actually about what place to give political power and executive authority in the enforcement and defense of law. In his day, he was arguing against the legal positivism that dominated German constitutional thinking and sought to view law as a sphere autonomous from politics by virtue of its formal structure as a system of rules. Schmitt thought this image of law failed to represent the conditions of law's own creation, in the realm of politics. Since law was never the codification of an abstract set of absolutes but was, rather, the legal ratification, issue by issue, of political agreements between opposing social forces and interests, those who enforce the law were bound to make exceptions, as these forces and interests exerted their influence on the law's guardians.

In our own day, the defenders of legal invariance are not legal positivists but civil libertarians. They want to keep law as free as possible from political contamination and interference and believe that law's legitimacy derives from its capacity to resist political pressures. This might be a realistic expectation in times of peace. But in liberal democracies under attack, matters might stand differently.

Schmitt believed that this image of autonomous law was not merely unrealistic but foolish, since law was the creature of political power. Its enforcement depended entirely on the viability of a particular constitutional order. This seemed only too evident in Weimar, where law's survival depended on the capacity of a constitutionally elected president to defend an embattled regime by force. Hence, according to Schmitt, rights survive in emergencies to the degree that they enable the particular political regime that defends the constitutional order to survive. Rights that stand in the way of a regime's survival should be suspended in a time of crisis. The Ulysses metaphor would have made no sense to Schmitt. No regime can afford to tie its hands to the mast and stop its ears with beeswax if the ship of state is being boarded by pirates. Ulysses must untie himself, rally the crew, and fight back.

Confidence in Schmitt's judgment on these issues is not enhanced by the knowledge that he went on in the 1930s to become an apologist for Hitler. As Oren Gross has shown, in a close reading of Schmitt's theory, an intellectual project that began in the early 1920s as an attempt to save the Weimar constitution through the use of presidential power became, by the early 1930s, a project to justify extraconstitutional dictatorship.40 From defending the president as guardian of the constitution, he ended up defending dictatorial power at any price. This later career suggests the weakness in his legal realism: yes, law is politics—the codification of often shabby compromises between competing groups in the political realm—but it is not just politics. Law ought to encapsulate right as well as might. Law's commitments to dignity and equal protection are supposed to enable right to prevail over might. The moral content of law that some legal pragmatists and positivists would see as dangerous because it provides a pretext for judicial moralizing and intrusion on legislative authority is, when seen against the catastrophe of Weimar, an essential element to rally moral and political support for constitutional order. Such ethical elements set limits precisely at the point where decisive use of executive authority could shade into extraconstitutional dictatorship. Schmitt's jurisprudence, in its worship of strong authority, lacked any conception of a constitution as a moral order of liberty. Of themselves, commitments to equal protection and dignity cannot save a society from tyranny during an emergency. But these values can operate as a moral depository to remind citizens, judges, and politicians of the limits that ought to guide a democracy in a time of trial.

In the tradition of liberal constitutionalism that descends from Locke, law's ultimate protection lay in morality, in the ability of citizens to rise to the defense of law when morality revealed law's exercise to be unjust. Citizens, judges, and politicians all have moral responsibilities to protect a constitution when it is under attack. Locke's argument for the necessity of executive prerogative was balanced by the people's right to take the government back into their own hands when liberty was usurped. In an argument which provided a key justification for the American Revolution, Locke wrote that when prerogative power threatened to “enslave or destroy” a people, they had the right to “appeal to Heaven,” and by implication to take up arms to defend their freedom.41 Locke's phrase “appeal to Heaven” implied clearly that the armed defense of liberty was a lesser evil, justified only to avert the greater evil of tyranny and enslavement. The Lockean view is more than a defense of revolution: it clearly prioritizes evils, preferring the risks of disorder to despotism. This moral ranking contrasts signally with Schmitt's, for whom the greater evil was disorder and civil war, and for whom dictatorship, in contrast, was the lesser evil.

In 1933, many, though not all, Germans made the same mental choice as Schmitt, believing that dictatorship was a lesser evil than either Communism or civil war. Yet once a constitutional order sacrifices its commitment to liberty, it quickly sacrifices everything else. The racialized jurisprudence of Nazism, by deliberately severing law from any commitment to equal protection and respect, deprived whole categories of German citizens of their rights, creating the nightmarish legal order in which they could ultimately be deprived of their lives. But this racialized jurisprudence found a willing accomplice in a tradition of German legal positivism and legal realism that sought to disinfest law of ethics.

Human rights emerged from the Holocaust as a rejection of legal positivism, as an attempt to provide citizens with an independent moral standard that would enable them, when the law of their country went mad, to say: this may be legal, but it is not right.42 It is this belief in the existence of a higher law, to which statutes and constitutions are ultimately answerable, that was so absent in the fervent apologetics of those like Schmitt whose theory of prerogative power divorced law from ethics. The lesson seems clear: even in emergency, even if some liberties must be suspended, a constitutional state must remain answerable to the higher law, a set of standards that protect foundational commitments to the dignity of every person.


The question is: which higher law? In American constitutional doctrine there is no obvious answer. There is no higher law than the U.S. Constitution, though Supreme Court justices have recurrently resorted to claims of ethics, natural justice, and the higher law in crafting their interpretations of the Bill of Rights.43 A lesser evil argument would make only a modest claim here, that precisely because emergencies appear to place the public interest so overwhelmingly on the side of the public safety, constitutional interpretation should give special weight to the rights claims of those who stand to be arrested, incarcerated, or deported in times of emergency. Human rights conventions, ratified by the U.S. Congress, provide a statutory form of this higher law, and precisely because they are international, they provide a vantage point outside the national jurisprudence of a state, which enables judges and politicians to stand back from the solipsism of threat and victimhood, that state of feeling embattled and attacked that often leads to excessive, arbitrary, and abusive measures.

In Europe, the higher law is more explicitly set out in the terms of the European Human Rights Convention. Yet here, too, the idea of a higher law is obscure because, according to the terms of the convention, many fundamental human rights can be suspended or derogated in times of emergency. Under Article 15 of the Convention, for example, states can suspend human rights guarantees when they are faced with “war or other public emergency threatening the life of the nation.” The loose phrasing allows governments a wide latitude in deciding what constitutes a public emergency, but governments are required to justify to the court why a situation does constitute an emergency requiring the suspension of civil liberties.44 In the United Kingdom, the legality of the government's suspension of civil liberties after September 11 has gone to the House of Lords for review and will probably go on to the European Court. The European human rights system thus attempts to balance the two principles of political necessity and adversarial justification. If a government must suspend civil liberties, it must justify its action to a court, and, moreover, one outside its own national boundaries. This is not an empty requirement. In a judgment against Turkey in 1997, the court found that the government was entitled to suspend certain rights, in order to suppress a Kurdish insurgency in the southeast of the country, but was not entitled to detain suspects beyond fourteen days without their being brought before a court for review of their detention.45 The second restriction that the European Convention imposes is that states are not allowed to derogate from a few “absolute” rights, like the right to be free from torture, extrajudicial killing, slavery, forced labor, and punishment without due process of law.

Any regime to suspend rights is difficult to reconcile with human rights' supposed role as an unconditional, universal, and unchanging set of commitments. The distinction between rights that can be suspended and those that cannot clearly implies a hierarchy of rights. This, too, contradicts a fundamental premise of modern human rights doctrine. The Vienna Declaration of 1993, issued after a meeting of the state signatories to the major human rights conventions, rejected the idea of a hierarchy of rights and reaffirmed that all of them were both indivisible and universal.46 Universality is usually defined as meaning that rights should be universal across cultures. But there is an equally important sense in which human rights should apply universally among all persons and at all times, whether normal times or states of emergency. Terrorist emergencies put these universalist commitments under strain. To say that some rights can be suspended, while others cannot be, is to suggest that some matter more than others, and that the whole is not an indivisible and interconnected package of entitlements.47

We can believe that human rights are indivisible, in the sense that having one right is a precondition for having another. So, to use Amartya Sen's famous example, having a right to free speech and free assembly makes it possible to defend a right to subsistence, since without political rights you will be unable to make your voice heard when food runs short.48 But this sense of causal interdependence is distinct from the idea that all rights are equally important in a time of emergency. We can still argue that the rights are analytically indivisible, in Sen's sense, while admitting that in dangerous times some rights just turn out to be more fundamental than others. Derogation seeks to save what can be saved—an absolute prohibition on torture, cruel and unusual punishment, extrajudicial execution, and penal servitude—from the abridgments of liberty, chiefly habeas corpus, that political necessity may require.

No human rights advocate, let alone liberal democrat, can be happy, however, to discover that the rights which can be derogated in emergencies turn out to be the entire body of political rights essential to the preservation of democracy itself: freedom of thought and opinion, freedom of expression, freedom of assembly. The drafters of the European Convention crafted this derogation structure with the recent catastrophe of Weimar and the rise of fascism in mind, together with the rise of postwar Communist parties in Italy and France.49 Apparently, the risk of a return of fascism or totalitarianism was sufficient, in their minds, to justify suspending democratic rights in order to prevent political parties' seizing power by force. States faced with such a crisis can suspend democratic rights, provided that the means of repression used to put down the challenge do not result in violations of the bodily integrity of detained suspects. Thus, according to the Convention, suspects can be held in indefinite detention as long as they are not tortured, killed, or subjected to penal servitude. This does protect them from the worst, but it sets up a hierarchy of rights, with “absolute” nonderogable protections applied only to infringements of bodily integrity, while all other rights are left in limbo, conditional upon the will of political authorities.50

But there are good reasons to insist on the indivisibility of rights. How are free citizens supposed to prevent their friends and neighbors, their union representatives, and journalists from being tortured, imprisoned without cause, subjected to penal servitude, or executed without trial, unless they can exercise undiminished rights of assembly and expression? Exercising the derogable rights is crucial to protecting the nonderogable ones. The survival of any rights depends on voice, on the capacity of citizens to protest, dissent, organize, and mobilize.51 Once these rights are suspended, citizens would have no recourse other than to seize them back by force, especially since the European Court, which is supposed to protect the Convention, lacks the power to compel states that refuse to abide by its judgments.52

The danger of a radical overthrow by force of democratic governments in Europe seems remote. Much less remote, however, is the possibility that democratic governments might themselves misuse a supposed terrorist threat to justify abrogation of basic democratic rights. The European Convention, as it now stands, offers imperfect protection against that eventuality. To be sure, no court can save a democracy from self-destruction by judicial rulings alone. But courts also have a clear political function. If their rulings protect democratic rights in times of emergency, their clear signal can help rally domestic political support to the defense of civil liberties. The European Court thus has an important political role in signaling to European states that one of its members is tipping from democratic to authoritarian patterns of rule. It is up to strong democratic states to heed these judicial signals and use political leverage to pull states that may be lurching in an undemocratic direction back onto the democratic path. Thus while Turkey is not yet a member of the European Union, it ardently seeks membership, and this gives it a strong incentive to comply with decisions of the European Court, which does have jurisdiction in Turkey. Recently when a right-wing government was elected in Austria, other European governments threatened to suspend the Austrians from membership if they took actions against immigrants that would violate European human rights norms.53 Here external human rights pressure played a role in preventing any recurrence of the 1930s.

But political will is a changeable thing. Austria is a small country. What would happen if Italy or Turkey or Russia were to fall under the sway of a genial demagogue with a popular touch and contempt for both democracy and the judiciary?54 Terrorism would easily provide such a leader with a pretext for declaring a permanent suspension of the rule of law. In such a case, firm political action by other European states would be necessary, but to facilitate such action, European human rights norms—which have statutory authority in each of these countries—must be clear. The standard that defines when emergency powers can be introduced should be tightened up. It should refer only to a repeated and concerted campaign of violence whose purpose is not just political destabilization but the overthrow of the constitution by force. The European Convention should make it harder to suspend political rights, restricting suspensions exclusively to parties and persons explicitly advocating violence. While rights of assembly and free political participation for the population at large should never be suspended, it could be permissible to ban political parties when they engage in incitement, intimidation, or association with terrorist groups. Those who wish to express anticonstitutional views should remain free to vote for parties provided they have no part in violence. In emergencies, authorities should have the power to detain individuals on lower standards of probable cause than would suffice in a criminal matter, simply because, in conditions of uncertainty and possible threat, detaining them may be the only way to determine whether a plot is underway leading to further attacks. But those detained in this manner should always retain right of counsel and access to courts for expeditious judicial review of their detention.55 Thus even if preventive or investigative detention can be justified as a lesser evil, it should not entail suspension of habeas corpus rights.56 Anyone detained for longer than fourteen days, for example, must be brought before a civil magistrate with the power to confirm or terminate detention.

The case for suspension of rights also ought to require proving that you suspend some liberties in order to protect other equally important rights. If these suspensions cannot be shown to enhance the right of the majority to live in security, then they have no justification. You cannot suspend liberties simply to provide a public with a feeling that they are more secure or to gratify their anger at a terrorist outrage or to find some convenient scapegoat. You have to prove—to a legislature, a judiciary, and public opinion—that abridging a particular constitutional liberty will actually enhance the liberty and security of the law-abiding.

What the rule of law requires, as John Finn and other scholars have argued, is not invariance but public justification.57 The question is not whether some restriction of civil liberties can be justified in times of emergency but whether these restrictions are undertaken secretly and arbitrarily or subjected to legislative scrutiny, justified with good reasons to an electorate, and, above all, subjected to full judicial review.

The same principle ought to apply to any derogation from international human rights. The International Covenant on Civil and Political Rights allows states to derogate or suspend rights.58 “In times of public emergency which threaten the life of the nation,” they can suspend rights provided that these measures do not discriminate “on the ground of race, colour, sex, language, religion or social origin.”59 With the Nazi concentration camps clearly in mind, the drafters of the Covenant precluded suspension of legal guarantees against cruel and unusual punishment, penal servitude, and torture, and, with the Nuremberg Laws in mind, they precluded any suspension of the right to remain a person under the law.

Nations that sign the covenant are required to publicly announce and justify their derogations to UN treaty bodies. The United Kingdom formally sought to suspend its obligations in 2001 after enacting legislation that empowered the government to continue detaining terrorist suspects from foreign countries who could not be returned to their home countries except at risk of persecution or torture there.60

This UK action seems a justifiable exercise in the lesser evil. The government did not act in secret. It engaged in public justification. Faced with two ways of jeopardizing the human rights of suspects, it chose the lesser form. Such a policy seeks to save what can be saved from the idea of precommitment, by focusing on preserving terrorist suspects from torture and by insisting on accountability and public review.61

There are two points here. One is that exceptions do not necessarily compromise the status of rules, provided that the exceptions are subjected to adversarial justification, and provided that the exceptions are highly specific, applied to named individuals, rather than blanket provisions applied to large bodies of citizens and noncitizens. The second is that the relevant institutions before which exceptions must be justified are not simply those of the constitutional state under attack. They include the institutions of international law. By going through a formal derogation procedure, the United Kingdom accepted this second point. The United States, by contrast, has not formally derogated from its obligations under the International Covenant on Civil and Political Rights.62 In failing to do so, the United States clearly believes that the duty of justification it owes in a war on terror is restricted exclusively to U.S. institutions. In an international war, however, it is unconvincing to claim that national sovereignty should trump international obligations. The United States is detaining nonnationals, prosecuting its war across national boundaries. If this is so, international law should apply, and the duty of adversarial justification, outlined in this book, extends beyond national courts and legislatures to UN treaty bodies as well.


An emergency is just that: a temporary state, not an indefinite and open-ended revocation of the rule of law. The problem with emergencies is that only the executive has sufficient information to know whether they remain justified. Hence the speedy termination of emergencies remains a recurrent problem. Electorates and legislators are invariably told by their leaders, “If you only knew what we know…,” in justification of the continued suspension of civil liberties. But this is not good enough. It is the very nature of democracy that we should know what they know. It may not always be possible to know immediately: a government can be justified in withholding information on a sensitive operation if disclosure would actually jeopardize lives. But the justification for secrecy can be only temporary, not permanent. Secrecy becomes a greater evil—a danger to democracy itself—when it is used to prevent the process on which constitutional liberty depends, the adversarial justification of lesser evils. That is why sunset clauses, preset limitations on the duration of emergency legislation, remain the legislature's chief weapon in making sure that emergency exceptions do not become the rule.

Yet even when sunset clauses are in place, emergency legislation remains problematic. Legislators invariably respond to attack by giving police additional powers whether they need them or not. A recent scholarly evaluation of Canada's Bill C–36, enacted after September 11 to tighten up Canada's antiterrorist legislation, questioned whether the legislation was actually necessary to meet the terrorist threat. The criminal law already on the books may have been sufficient. The same point is often made about the U.S. Patriot Act.63

The rule of law is not compromised by emergencies per se, but by politicized construal of risk to justify emergency measures that are not actually necessary to meet the threat at hand. It is crucial to distinguish threat assessment from moral repulsion, to separate ethical judgment from the actuarial estimation of danger. The fact that terrorism is an attack on the political character of society does not mean that the society's identity or future is in question. As I shall demonstrate in the next chapter, the historical record shows that democracies are neither less ruthless nor more vulnerable than authoritarian states when faced with what they take to be an ultimate danger. The ruthlessness derives from the great strength of democracy, its capacity to mobilize the allegiance and self-sacrifice of its citizens. Controlling this potential for ruthlessness is partly the business of a good constitution. Checks and balances are there to prevent an executive from pandering to fear. But constitutions are never enough. Good institutions are never enough. Ultimately the protection of these rights devolves upon citizens themselves.

One of the strengths of the liberal tradition is its disabused realism, its belief that abuse of power is inevitable and no constitution can stop it. That is why, in Locke and Jefferson, for example, there remains an articulated right of revolution. Liberal theory places the final defense of constitutional liberty in ordinary citizens, in their willingness, when provoked by greater evil, to rise up and change their government, by peaceful means if possible, by force as a last resort. Civil disobedience has an honored place in the traditions of liberal democracy, precisely because it is the defense of last resort when the constitutional identity of liberal democracy is at risk.64

There is no reason to be complacent about the willingness of citizens to fight for their liberties or prevent their governments from abusing them. Less than half of the American electorate takes the trouble to vote, a sign that they may not care about their own rights, let alone the rights of minorities in their midst. As we have already seen, the number of people who value rights intrinsically will always be small. The capacity of a ruthless government, bent on abridging freedoms, must never be underestimated, especially not in an age in which government has such power to shape public perceptions and manufacture consent through the media. Yet is also dangerous to become cynical and to conclude that the defense of civil liberties is hopeless because of the manipulated apathy of the majority. The civil rights movement in the United States was one long struggle against manipulated apathy—and concerted racist resistance—by a handful of activists. Only in retrospect does their victory seem inevitable. At the time, it often seemed a hopeless cause. But it has always been true that the force which sustains the liberty of the many has been the intransigent courage of the few.

Sticking to the idea of rights as precommitments is hard when a liberal democracy is under threat. Rights express this recognition, this knowledge that we must precommit in times of calm to prevent ourselves, like Ulysses, from succumbing in times of danger. Precommitment is not a commitment to invariance, to never changing the law no matter what, but rather a commitment to adversarial justification, within a framework that maintains equality and dignity standards in times of safety and times of danger alike. To conduct a defense of liberal society in defiance of these precommitments is to betray the order that is being defended, as well as the citizens whose security depends on that order.