In the wake of 9/11, many political scientists and theorists in the United States of America turned their attention to the topic of emergencies. That required them to confront a fundamental question: Are emergencies to be studied as important in their own right, as altogether exceptional events that threaten the very existence of a society in unforeseeable ways? Or are they important, not because they are radically distinct from the normal situation of politics, but because they bring to the surface otherwise implicit aspects of normal politics?

If they are altogether exceptional events, our ordinary tools of analysis of normal politics as well as the norms we adopt for evaluating such politics will not apply in states of emergency. For example, it might be thought that we may disregard the legal norms of due process that we think are otherwise appropriate when someone is detained on a criminal charge in a situation when the government supposes that survival depends on the large-scale detention of individuals whom it considers to be risks to national security. We might also want the criminal trials of such individuals to be conducted by military tribunals, not civilian courts. Or we might consider it justifiable to torture an individual when there is reason to believe that this is the only way to extract information in time to save thousands of lives threatened by a “ticking time bomb.” Since detention and criminal trials in liberal democratic societies are carefully regulated by due process norms, and since torture is prohibited, such responses will in some sense (I will come back to this qualification below) take place “outside the law”—the title of one of the books discussed in this essay.

However, if there is no radical distinction between normality and emergency, then we have to decide whether normal politics is more like emergency politics, or vice versa. We either have to revise our understanding of the tools and norms we have used for normal politics up until now, or show that these tools and norms are perfectly appropriate for emergencies. We might conclude either that the lack of legal constraint on the executive during emergencies simply highlights the fact that the executive is hardly constrained by law even in ordinary times, or that the fact that the executive can be meaningfully constrained by law in ordinary times shows that legal constraints can be effective during emergencies. But whatever we decide, we presuppose that inquiry into emergencies will assist our understanding of the public institutions and the separation of powers in the legal order of the liberal-democratic state. We will then better understand in particular the liberal commitment to the rule of law and its requirement that all public acts of power be legally authorized.

It is not, of course, obvious that liberalism and the rule of law are the same, and one benefit of the study of emergencies is that it challenges that equation. Indeed, two of the authors discussed here, Clement Fatovic and Nomi Lazar, take their inspiration from John Locke, who did not make this equation. For while Locke extolled the advantages to liberty of life under “settled, standing” legislated rules common to all, in contrast to “the inconstant, uncertain, unknown, arbitrary will of another man,” he also insisted that in emergencies the government had to have a prerogative or legally unconstrained power to “act according to discretion, for the publick good, without the prescription of the Law, and sometimes even against it.”Footnote ¹ Fatovic and Lazar claim to find in Locke a much more pragmatic form of liberalism than the rigid, principled and legalistic model they associate with Kant, since the latter is, they allege, incapable of responding to the radically contingent nature of a state of emergency. Among other things, Lockean liberalism makes explicit allowance for what we can think of as an “extra-legal measures model”Footnote ² of response to states of emergency, one that acknowledges the necessity for the state to respond to states of emergency without prior legal authorization, or even in ways that are legally prohibited, but that also attempts to show that such a response can take place on liberal terms.

For Fatovic and Lazar as for other Lockeans, an emergency is by definition ungovernable by the rules that govern normal life. Thus, an effective response to an emergency will require that some state institutions respond quickly and effectively to threats without legal authority, or even against the law. As Fatovic and Lazar see it, since neither the legislature nor the judiciary is capable of the swift, energetic action required to deal with an emergency, the executive must be the authoritative body when one occurs.

In contrast, Bonnie Honig argues that liberalism's inability to respond to emergency is symptomatic of its inability to respond to or even to understand politics in general, and that we should appreciate the fecundity of a democratic approach to politics inspired by Hannah Arendt. On this approach, politics is essentially agonistic or conflictual in a way that is not controllable in advance by norms, for the norms of politics emerge in politics. An emergency is then simply a context in which such emergence is particularly stark. But its unruliness is not to be feared, since even ordinary politics is unruly; indeed, we should celebrate unruliness, because once the political is seen as open to contest, the prospects for democratic engagement in politics are enhanced.

All three authors share the view that what we can think of as liberal legalism is incapable of responding to states of emergency. But while Honig equates liberalism with legalism—the bid to subject politics to the control of the rule of law—and thinks that this incapability means an alternative to liberalism is required, Fatovic and Lazar think that this incapacity shows why we should prefer Lockean liberalism, with its readiness to consider state action outside the law.

My own position is that a liberal democratic state must adopt a liberal-legalist approach to emergencies, one which requires not only that all executive action be authorized by law, but also that all executive action is subject to the control of the rule of law—the control of norms such as those of due process. This position, I will argue, is supported by the fact that Honig does not succeed in providing any concrete alternative to liberalism, and that Fatovic and Lazar cannot resist the pull of legalism. I will also argue that if one mark of a sound political theory is its pragmatism—its sensitivity to experience—then liberal legalism should be preferred to Lockean liberalism, both because it is better attuned to the constitutive ideas of the experience of emergencies and because it seeks to ensure that the state's response to emergency has the normative substance to which liberalism aspires.Footnote ³

Of the three books considered here, Fatovic's makes the most valuable contribution to our understanding of liberalism and emergencies, as it traces clearly the attempt in modern political thought to show the necessity and the legitimacy of the extra-legal prerogative power of the executive for preserving a liberal state. Fatovic starts with Locke, but also provides excellent, chapter-length discussions of the place of the prerogative in Hume, in Blackstone, and in the American constitutional debates of the late eighteenth century. In addition, Fatovic wishes to demonstrate the relevance of the historical treatment of executive prerogative to the present. But, as I will explain below, this demonstration is unsuccessful for reasons he acknowledges but to which he cannot effectively respond, since his sympathy for Locke leads him to reproduce Locke's own ambivalence about extra-legal measures, and about whether the authority of the executive is personal or official.

Both Honig and Lazar also deal with the history of political thought, but their principal aim is to argue that a focus on emergencies illuminates contemporary debates about liberalism, democracy, and the rule of law. But neither provides a compelling account of how it does so, in Lazar's case because her Lockean liberalism creates problems for her argument similar to the ones Fatovic encounters, and in Honig's case because the substance of her positive argument is unclear. In Honig's case, moreover, were she to clarify her argument in a way that escapes banality, she would come perilously close to adopting a position that she claims to reject.

Before I proceed, we need to keep three things in mind. First, liberal legalism is still vigorously defended—for example, by Ronald Dworkin, the most eminent philosopher of law of the last fifty years and a leading exponent of liberal political philosophy, and by David Cole, a leading scholar and practitioner of human rights law, whose essays in the New York Review of Books since 9/11 have made a consistent argument for a legal-liberal response to the threat of terror attacks.

Second, in the twentieth and twenty-first centuries, the responses of liberal democratic states to emergency have been through law, and on the evidence of state practice since 9/11 it is highly unlikely that any liberal democratic state would claim an authority openly to act outside of the law. Consider that even Berkeley law professor John Yoo, and members of the Bush Administration Yoo advised when he worked in the Department of Justice, have always claimed that the measures taken in response to 9/11 by the United States were authorized by an inherent legal authority in the President as the chief executive officer, an authority that descends from the prerogative authority of which Locke spoke, and from the Constitution.Footnote ⁴ Moreover, the Authorization for Use of Military Force (AUMF) passed by Congress in 2001, which empowered the President to “use all necessary and appropriate force against those … he determines planned, authorized, committed, or aided the terrorist attacks,” was argued by the Administration's lawyers to give the President legislative authority to act as he saw fit in the “war on terror.”Footnote ⁵ Consider also that Harvard law professor Alan Dershowitz argued that, since torture would inevitably be deployed to combat terror, it should be brought within the law and regulated.Footnote ⁶ Finally, consider that even though Yale political theorist and lawyer Bruce Ackerman has argued that legal controls are impractical in a time of emergency (in part because judges always defer to the executive in an emergency), so that one should resort to political safeguards to control the executive rather than legal ones, he too succumbs to the pull of legalism by making it an essential component of his model for dealing with emergencies; these safeguards must, according to Ackerman, be put into a written constitution or a statute, so that judges may review their observance.Footnote ⁷

But this does not render Lockean liberalism and the extra-legal measures model entirely irrelevant, for the reason that such responses may still in some specially-qualified sense be said to take place outside the law. This point is best illustrated by the claim that the executive has the inherent power to proclaim martial law in order to deal with an emergency. This is a puzzling claim, since it is an allegedly valid use of law by the executive to do away with law's control over the executive.Footnote ⁸ Of course, those who regard martial law or something like it as inevitable in times of severe political stress want to justify it as only a temporary killing-off of law—a suspension. They also say that the acts done under martial law are both lawful and in the long-term interest of legal order. On their view, martial law is not a complete absence of law, nor is it a special kind of law—a scheme of legal regulation. Rather, it is an absence of law prescribed by law under the concept of necessity—a legal black hole, but one created, perhaps even in some sense bounded by, law. However, even on these terms, the use of law to do away with law for the sake of preserving legal order presents a conceptual puzzle—“Is martial law really law?”—which has obvious echoes in post-9/11 debates. Moreover, as we know from these debates, the puzzle has important political implications.

For example, if officials may not be legally authorized to act outside the rule of law, those subject to such acts are entitled to a judicial declaration that the officials acted illegally, and so, if the illegal act is a detention, they are entitled to be set free, unless there has been a constitutionally valid suspension of habeas corpus. But when Congress does authorize the President to suspend habeas corpus, what we get is a valid statutory installation within the law of the conceptual puzzle created by the claim that there is an inherent power in the executive to proclaim martial law. For the executive would then be legally authorized to act without being subject to legally enforceable controls. And even if Presidents do not have the inherent authority to do as they please in response to a perceived emergency, and even if there is no written constitutional document that gives them such authority, they can be given legislative authority to act as they see fit, as the example of the AUMF might show. Such authority might seem circumscribed by other legal documents—for example, a constitutional bill of rights like the American one might preclude a presidential authorization to subordinate officials to torture individuals in ticking-time-bomb situations. But as Alan Dershowitz demonstrates, lawyers can find arguments to support surprising interpretations of constitutional prohibitions—or, failing that, arguments that such prohibitions are subject to exceptions.

In sum, while governments might not openly claim the authority to act outside of the law, they do seem to be able to procure, or at least to claim, a legal authority to exercise the powers that past governments claimed under the rubric of prerogative. In doing so, they exhibit what I have called elsewhere the “compulsion of legality”—the perceived necessity to have a legal authorization for state action based on the belief that legally-unauthorized action is widely considered illegitimate.Footnote ⁹ Put differently, compliance with legality is seen as a necessary if not sufficient condition for legitimate state action. But, again as I have argued elsewhere, the compulsion of legality can set in motion two very different cycles of legality. In one, virtuous cycle, the institutions of legal order cooperate in devising controls on public actors that ensure that their decisions comply with the principle of legality, understood as a substantive conception of the rule of law. In the other, less virtuous cycle, the content of legality is understood in an increasingly formal or vacuous manner, resulting in the mere appearance or pretence of legality. In this second cycle, the compulsion of legality results in the subversion of constitutionalism, of government according to the rule of law. Arbitrariness is thus covered by what an English judge referred to as a “thin veneer of legality.”Footnote ¹⁰

This brings me to the third thing we need to keep in mind regarding the study of and debate about emergencies: the omnipresence of Carl Schmitt, the conservative (even fascist) lawyer who during the years of Germany's Weimar Republic wrote extensively on the nature of the state of emergency and what it revealed about the limits of liberalism and liberalism's commitment to the rule of law. Each of the authors discussed here acknowledges Schmitt as the preeminent thinker on the state of emergency, and rightly so, for Schmitt identified the main problem liberalism faces in trying to craft a response to emergencies. However, none of them fully grasps Schmitt's diagnosis of liberalism's inability to deal with emergencies, and that failure is behind the problems their projects and arguments encounter.

Schmitt argued that even when liberals recognize that an emergency is a state of exception to regular norms and principles, they do not have the theoretical resources to cope with such an exception. As he put it, the state of emergency or exception is “something incommensurable to John Locke's doctrine of the constitutional state.”Footnote ¹¹ According to Schmitt, this is not just a theoretical problem, but a practical one as well. On his account, liberalism, to remain consistent, must refuse to countenance an extra-legal measures model. As a result, when a liberal state responds to existential threats, protections for individual liberty associated with the rule of law will become ever more attenuated, until the point where the rule of law is reduced to the executive's claim that it has a valid or purely formal authorization for its actions. At that point, liberal legalism ceases to be liberal, and becomes a purely formal mode of legitimizing state conduct. As long, for example, as due process for detainees, including judicial review of their detentions, is precluded by law, or as torture warrants are explicitly authorized by law, the state is by definition complying with the rule of law and so acting legitimately. This vacuous understanding of legality does more than legitimate the use of law for illiberal ends; it also, in Schmitt's view, permits political factions hostile to liberalism to capture politics from within.

Schmitt thus sees the choice for liberalism as between liberal legalism and vacuous legalism. However, if vacuous legalism permits everything that an extra-legal measures model permits, albeit with the caveat that there is some form of legal authorization, then nothing substantive might seem to turn on the fact that governments feel the compulsion of legality. Indeed, such a claim is entailed by Schmitt's argument. In order for liberalism to deal with emergencies, it has to evacuate the idea of legality of liberal content, with the result that the politics of emergencies are determined by highly personalized decision-making under a thin veneer of legality.

It is thus significant that Fatovic, despite his support for an extra-legal measures model, notes that there has in practice been a “dramatic shift away from reliance on extralegal exercises of prerogative towards the use of formal grants of authority when it comes to dealing with emergencies.”Footnote ¹² He also concedes that such developments might “seem to suggest that prerogative has become nothing more than the outmoded relic of an empirically questionable and normatively dangerous theory.”Footnote ¹³ Finally, he acknowledges that Locke himself was equivocal about extra-legalism, for it is not clear whether Locke thought that the prerogative has its source in the legal constitution, that is, whether the authority to act extra-legally is somewhat paradoxically a legal authority that inheres in the executive.Footnote ¹⁴ That Locke might have held the latter view is indicated by his insistence that the principle that should guide the executive in using this power is a legal principle that he regarded as fundamental or constitutional—that the safety of the people is the supreme law.

This equivocation tracks another: whether the executive, personified in Locke in the figure of the prince, exercises the prerogative in his natural capacity, or in his artificial role as the public official at the apex of the hierarchy of state.Footnote ¹⁵ If it is the former, then it seems that the authority would come not from his official position, but from whether he handles the emergency appropriately, which would in turn depend on his moral character or virtue. If it is the latter, then the authority would be constitutional, bestowed on him by a legal (albeit unwritten) constitution.

Fatovic's sympathy with Locke's position has the result that he reproduces these ambivalences. For example, he says that in Locke “the person and the office are clearly separable,” but also that both “the person and the office determine the legitimacy of certain exercises of executive power.”Footnote ¹⁶ He consistently calls the exercise of prerogative extra-legal, and titles one section “Extraconstitutional Criteria of Legality.”Footnote ¹⁷ But he also insists that the exercises of prerogative power have, according to Locke, to meet the criterion of “constitutional permissibility” contained in the maxim that the safety of the people is the supreme law, so that they are “in strict compliance with the moral and legal order that matters most.”Footnote ¹⁸

Lazar encounters similar problems. Her stated aim is to “undermine the norm/exception perspective” by developing an “alternative conceptual framework,” one that understands emergency powers “in terms of their continuities with everyday institutions or values.”Footnote ¹⁹ Her core argument is that liberal norms of justice can be served in doing wrong, where “doing wrong” means “breaking the law in order to sustain liberal democratic institutions.”Footnote ²⁰ She insists that the rule of law is of purely instrumental value—it is an instrument that serves underlying values of uniformity and predictability.Footnote ²¹ It seems to follow for her that when these values are better secured in an emergency by breaking the law, then the moral disvalue in breaking the law might be outweighed by the moral value gained by successful illegal action. Thus, at times, it seems that Lockean liberalism for her means going outside of the law—but on liberal terms, since in emergencies liberal values can sometimes be served, and even better served, outside of the law.

But at other times it is not clear that she is actually advocating extra-legal measures. For example, she suggests that scarce resources in the ordinary criminal justice system have the unfortunate result that due process protections are “only roughly enforced,” and she seems to think that it follows that it is legitimate to detain those suspected of terror without appropriate due process.Footnote ²² She does not say that security detentions are extra-legal, but justified, because they serve the same liberal values to which the rule of law is instrumental. Nor does she argue that such detentions comply with the rule of law. Rather, her argument seems to be that even though they do not comply with the rule of law, since we live with such lack of compliance in the legal regime of criminal law, there is no problem. Something can be lawful, even if not in accordance with the rule of law.

However, even a criminal justice system that detains people for far too long prior to trial concludes in a trial, and it is that end that justifies the detention. In contrast, the kind of detention used during emergencies usually does not conclude in a trial, and the lack of due process in detainment decisions are not justified mainly by claims about scarce resources, but by claims about considerations of national security, information about which those who have it refuse to reveal. The fact that we live with flaws in the operation of a system of detention that serves a legitimate purpose cannot be used to justify a system of security detention that serves very different purposes, purposes themselves in need of justification.

In Lazar's discussion of Roman emergency regimes, she explores the possibility that informal constraints on officeholders, as well as their personalities, can supplement or substitute for formal (that is, legal) constraints on an office.Footnote ²³ But in the next chapter, she seems to find an analogy between the Roman instrument of emergency rule in the senatus consultum ultimum and martial law,Footnote ²⁴ which forces her to withdraw rather rapidly from a willingness to rely on informal constraints and personality. As she notes, the incentives in an extra-legal measures response to emergency “naturally run in the direction of” irresponsible rather than good behavior, which is why formal constraints are considered mandatory.Footnote ²⁵ And while she states at the outset of her final chapter that we must choose between liberal legalism and an extra-legal measures model, in fact the main argument of her book does not explore extra-legal measures, but develops instead the sensible claim that political and informal constraints on executive power can matter as much as, if not more than, formal legal constraints.

In sum, the kind of ambivalence one finds in Fatovic and Lazar between liberal legalism and Lockean liberalism shows that one cannot make Locke's position coherent. Schmitt gives us the explanation for this, as his anti-liberal position permits him to make consistently the following three claims: first, that the political sovereign is he who decides, unconstrained by law, when there is an emergency and how to respond to it; second, that no liberal sovereign can admit to this openly, given that liberals must adopt the view that a condition of the legitimacy of a political decision is that it is authorized by law; and third, that sovereignty so understood hollows out legality, with the result that the fact of political sovereignty prevails under the guise of legality.Footnote ²⁶

The main virtue of Honig's book is that she hones in on what should be the central issue in the debate about emergencies, for she recognizes that in the twentieth and twenty-first centuries, states have responded and will continue to respond to emergencies through law, and that the questions this raises are the daily diet of administrative law, the law that governs the activity of the officials who exercise the authority delegated to them by the state. Honig thus suggests that she will reconceptualize Schmitt's idea of the state of emergency by demonstrating that the moment of decision in a legal void—a space uncontrolled by legal norms—does not occur only in the drama of a state of emergency, but also is part of the normal business of politics and law.Footnote ²⁷ In other words, she takes Schmitt to accept the distinction described by Locke between the regular law, whose rules control normal times, and the exception, in which no law governs, and argues against Schmitt that this is a false contrast, since even in the most normal times, officials have to decide when and how to apply the law in a way that involves an extra-legal exercise of discretion.

However, Schmitt did not accept this distinction. Rather, from his very first work, he argued that judicial interpretation of the law necessarily involves an extra-legal exercise of discretion, which can be controlled only by the fact that a typical judge will rely on established social values.Footnote ²⁸ An emergency is for him in one sense simply judicial discretion writ large. But it is also qualitatively different, in that the decision that there is an emergency, and decisions about how to respond to it, are constitutive of rather than reliant on social values, because the very fact that there is an emergency exposes the lack of the political homogeneity Schmitt thinks is required to sustain normal order.

Honig, then, far from moving away from Schmitt's understanding of the ordinary situation of law, in fact reproduces it. For it is precisely Schmitt's point that the exception is the norm because it is a feature of every legal decision. In his view, the liberal dream of a legislative state in which public coercive judgments would be made by a centralized legislature, and put into laws of general application, deteriorates inevitably into the nightmare of the administrative state, in which such judgments are made by the decision of particular officials at the point of application of the laws. But what is applied is neither the law nor something authentically public; rather, it is an exercise of arbitrary power by a particular official.Footnote ²⁹

In Honig's view, that the positive law does not provide determinate answers to many questions about its application provides the occasion for what she calls “agonistic cosmopolitics.” She only gestures at what this phrase might mean, referring to “movements” that seek to diminish global inequality or to sustain local economies, or individuals who marry citizens of less privileged nations in order to bring them into privilege. In describing the steps democratic citizens might take to revitalize democracy, she suggests that they should “[e]lect different representatives, protest institutional injustices, educate members and demand accountability, and strive for better legislation and demand better court appointments.”Footnote ³⁰

These suggestions are, of course, ones that many liberals might enthusiastically endorse. They are also at odds with Honig's reliance on the language of “interpellation,” a term coined by Louis Althusser—a Marxist philosopher of a decidedly deterministic tendency—to convey the relationship between the legal subject in a liberal democracy and the “ideological state apparatuses” that constitute this subject. Honig seeks to defend against an earlier critique of her use of this term by saying that it “continues to capture something no newer term does: the ways in which our entire being is swept up in the address of the state and its agents, even in anticipation of such address.”Footnote ³¹ But this is a capitulation, not a defense. The image of the liberal state and its law that it presupposes is of an all-powerful monolith that constructs its subjects in so totalizing a fashion that the only possible mode of resistance to the apparatuses of state is their overthrow, and yet her concrete suggestions do not take her beyond the banal politics of a well-functioning liberal democracy.Footnote ³²

This displacement from the political drama suggested by “agonism” to the banality of liberal democracy creates a severe tension between Honig's wish to understand the state and law as totalizing, and thus to be resisted only by resorting to emergency politics, and her sense that liberal democracy differs from many other political orders in that it makes possible resistance from within, indeed resistance which makes use of the resources that law (or, more accurately, the rule of law) provides. She is thus unable to give a satisfactory account of the fascinating story she relates of Louis Freeland Post, an Assistant Secretary of Labor in Woodrow Wilson's administration, who sought in 1920—in the middle of the post-World War I Red Scare—to subject decisions to deport anarchist aliens under the 1918 Anarchist Exclusion Act to constitutional principles, against pressures from Attorney General A. Mitchell Palmer, and from the Director of the Justice Department's Bureau of Investigation J. Edgar Hoover.

Honig says that the “question that hovered” over Post was whether “his actions were contained in the congressional act” that Attorney General Mitchell understood to authorize the deportations, or whether Post was distorting the legislation by interpreting its authority as bounded by constitutional constraints. On her view, “Post in effect split the difference between the two options by resorting to technicality,” by which Honig appears to mean an imaginative or creative exercise in legal interpretation. And second, “he solicited through his publicly addressed testimony, defending his actions before a Congressional committee in 1920, “a public that could rise from the constraints of survival politics as mere life to champion a bivalent democratic commitment to mere and more life.”Footnote ³³ She also says that for Post the “mere life of democracy was not a fair trade for the mere and more life of equality and justice that he valued and on which democracy's mere life was deeply dependent.”Footnote ³⁴

It is unclear what Honig means here, and her lack of clarity is puzzling, because Post offered a clear and simple justification of his actions, which Honig quotes at length.Footnote ³⁵ Post explained that while there was no judicial review possible of the exercise of his discretion, nevertheless, as an official who wielded legal authority, he was bound to exercise that authority in accordance with the fundamental legal principles to which he took his legal order to be committed, principles that were amply expressed in criminal law and that, while not directly applicable to an exercise of administrative discretion, should guide executive decisions given the importance of the interests at stake in deportation on the basis of an individual's political beliefs.

Honig regards this as an ambivalent posture, and claims this ambivalence masks the fact that Post was simply legislating his will. She makes this claim because she is committed to the Schmittian position that such an exercise of official authority is a kind of “highly discretionary,” and for the most part legally uncontrolled, “decisionism,” “something like the decisionism that Schmitt approvingly identifies with a dictator” that he identified “with administrators and with administrative government.”Footnote ³⁶ As a result, she explains Post's actions in terms that would have been incomprehensible to him.

A more plausible account, which bears some resemblance to Honig's claims about democracy, is that Post believed that there is at the basis of any democratic legal order a political commitment to a constitutional principle of equality, such that all officials of the order, whether executive or judicial, are obliged to interpret delegations of legal authority as requiring the delegates to treat those subject to their authority with equal concern and respect. But it would follow from that description that Post was right not to regard himself as legislating. He was simply exercising his judgment on a liberal-legalist interpretation of his duties as a public official, and thereby adopting the kind of stance advocated by Ronald Dworkin for legal officials, one that requires an official to show the legal materials relevant to his decision in their best moral light.Footnote ³⁷

Honig is prevented from appreciating this because of her aversion to liberal legalism, which she thinks is committed in its Dworkinian version to a “demonological or purifying logic.”Footnote ³⁸ This is a curious charge. As I have indicated, Dworkin has consistently argued that the legal response to 9/11 must avoid constructing or demonizing an “other” who is somehow less worthy of the constitutional protections to which the United States is officially committed. Moreover, Honig's charge presupposes a view of law that is contested within legal theory and has implications for the practice of law and politics. Since liberal legalism has to reject this view in order to fill out its conception of legal control, it is important briefly to explore this issue.

Honig, along with Locke, Schmitt, Fatovic, and Lazar, is committed to a kind of naïve legal positivism in which “controlled by law” means that general legal rules determine an outcome so that no human agency or judgment is required to decide what the law says. It follows that when the law does not explicitly dictate to officials what they should do to resolve particular situations, they have the legally-uncontrolled discretion to resolve them as they see fit. Locke's claim that the executive has to act sometimes “without the prescription of the Law” thus seems to correspond with “uncontrolled by law” in this sense, whereas his claim that the executive will sometimes have to act “even against” the law seems to correspond with a different sense of the term: illegal action.Footnote ³⁹

The combination of this view of what it means to be controlled by law with the inevitable insight that contingencies—most dramatically, emergencies—cannot be controlled in this way does not, as we have seen, lead to a Lockean extra-legal measures model. Rather, naïve legal positivism combines with legalism—the view that legal authorization is a precondition of valid official action—and ceases to be naïve. It becomes sophisticated positivism, of the sort elaborated in the twentieth century. On this account of law, it is sufficient for an act to be legally authorized that an official is given the authority to decide, even if the ultimate basis for the official's decision is extra-legal.

As we can see, Schmitt shares with Honig, Locke, Fatovic and Lazar the view that when the law that delegates authority to the executive does not dictate the executive's decision, the discretion that the executive has is legally uncontrolled. But he sees far more clearly than they do that liberal states will not openly say that the executive may act in a legally uncontrolled fashion, which is why he argues that such states will engage in the paradoxical exercise of using law to free the executive of legal control. In other words, liberal states “solve” the problem of the state of emergency (and of contingency in general) at a conceptual level, by bringing it within the legal order. The executive during an emergency is exercising the same kind of discretion, albeit a much larger and more politically significant version, that one finds in ordinary decisions by judges and other public officials. However, this is not a real solution, for, as Schmitt claimed, the upshot of bringing the emergency inside the legal order is to make the idea of legal control vacuous, and entails that the rule of law becomes detached from the liberal principles it was meant to serve. The blessing of legal authorization is bestowed on the legally uncontrolled exercise of discretion, and so the conceptual solution comes with an unacceptable normative cost: It permits the state to use the law as an instrument of illiberal ends.

Here it is important to distinguish between two strategies that can be exploited within this structure. First, the state can explicitly empower officials to pursue illiberal ends, for example, by giving them clear statutory authority to detain security suspects indefinitely without effective judicial review of the detentions, as well as the clear authority to torture the detainees should the officials think this is necessary. Second, the state can delegate to its officials a very general authority—as, for example, Congress did with the AUMF. That general authority might then in turn be interpreted by judges as giving the President authority to detain without legislative authorization, as the majority of the Supreme Court did in Hamdi.Footnote ⁴⁰

I say “interpret” because it is unclear what sustains the claim that, when officials are given a general legal authority to act, they may act as they please—the view Honig takes of Post's discretion to deport. Recall that Post did not regard his authority as legally uncontrolled. Rather, he thought his discretion was controlled by the moral and legal principles of the rule of law developed in criminal law. As I remarked, his theory of what it means to be controlled by law is in this way similar to Dworkin's. On this understanding, judges and other legal officials do not run out of law in the situation where they find that their decision is not dictated by law, since they have available to them other legal resources: the moral and legal principles of the rule of law that are inherent in their legal order.

From the perspective of an official committed to liberal legalism, a general statutory delegation of authority to respond to a crisis, like the AUMF, is not a delegation to officials of unconstrained power to detain, because the authority to deprive individuals of their liberty requires very explicit authorization. Moreover, if that authority is given, it is subject to the principles developed in other legal regimes to protect the individual against arbitrary power, unless the legislature explicitly specifies the application of these principles, or explicitly excludes their application. An official who endorses Schmitt's view of law, by contrast, would come to a very different interpretation of what is permitted by the AUMF, or of some more specific delegation of authority.

But that leaves liberal legalism with the problem of the situation in which a valid law—one that is properly promulgated and not vulnerable to a constitutional challenge—explicitly excludes due process norms and judicial review, or inappropriately specifies the content of due process norms, permitting (for example) conviction on the basis of secret evidence, or torture warrants. As I have indicated, such situations represent the installation of the equivalent of the extra-legal measures approach in law. It is this problem that makes Schmitt's work on emergencies of continuing relevance. However, the problem shows at most that a legislature may use the law as an instrument toward illiberal ends, not that those ends are justified.

In addition, there is a distinct political virtue in requiring a government openly to declare through legislation that it wishes to govern illiberally, rather than permitting it, as sophisticated positivism does, to hide those wishes behind the cover of a general legal authority. Liberal legalism has political salience because an official who adopts it will require the legislature to declare its intentions, and that has the advantage of bringing the matter into the public gaze, and thus has the potential to provoke exactly the kind of democratic contestation and debate that Honig values.

However, an insistence on a “clear statement” rule, a requirement that there be completely explicit legislative authorization of departures from the rule of law and from fundamental rights, makes sense only if it is accompanied by meaningful review of the executive decisions thus authorized, for one needs institutional checks on executive compliance with the terms of the delegated authority. More precisely, the authorization makes such a review possible, because it forces the executive to bring its activity within the scope of a deliberately and democratically designed statutory regime, one which has at least the potential of providing rule-of-law teeth to legal responses to emergencies.

Of course, the demos might decide that it wishes to be governed illiberally, and liberal legalism provides no guarantee against that. All that it guarantees is that the demos will have the chance to learn about and react to the government's desire to govern without the constraints of the rule of law. But that guarantee is an important one, and we should note that American experience after 9/11 shows that, over time, the judiciary can bring even a recalcitrant executive back under legal control and, more importantly, can do a lot to restore a virtuous cycle of legality—as long, of course, as the judges are committed to liberal legalism.Footnote ⁴¹

In such a process, judges and other like-minded officials are making creative, interpretive decisions about how best to understand their mandate. These decisions are not determined by law, if what one means by that phrase is that the officials simply transmit the determinate content of settled law to those subject to it. But in a looser sense of determination—one that requires all the reasons for a decision to be legal ones, and these reasons to be marshaled so as to display the best account available of the decision's legality—the decisions are as determined as any normative decision can be.Footnote ⁴² Such decisions are contested, for example, by those who have positivistic inclinations, and by those who reject liberal legalism. But that contest takes place within a political and legal space that Locke and Schmitt did not acknowledge.

It is precisely within that space that what Honig calls the “paradox of politics” gets played out—as Rousseau put it, that one needs good men to make good law, but one needs good law to make good men. As she correctly sees, the way to break this “vicious circle” is to “recast the problem from a problem thought to pertain only to origins … into a feature of daily democratic life.”Footnote ⁴³ But, as I have suggested, she does not fully realize that this requires us to reject the language of interpellation, and engage in the critique of practices from the inside. Moreover, in making that critique from inside the legal-political space of liberal-democratic politics, it is important to keep in mind that our engagement with “law” is not only with the content of positive laws, but also with the idea and practice of legality.

In this regard, Honig's Rousseauian formulation of the paradox of politics might look quite different, once one moves away from its positivistic assumption that the exclusive issue is the determinate content of the law—men will only be good if they engage with good legal content, but legal content will be good only if good men make the law—to a position that sees that the politics of producing goodness has much or more to do with legality. Consider that the writ of habeas corpus requires an official justification of a detention, while the norms of due process in a criminal trial in a civil court require a rigorous demonstration of the basis for a finding of guilt. In just these two examples, we see that legality, on the liberal conception of that idea, creates a process of deliberation and argument in which only a limited class of reasons is deemed acceptable: those supported by the legal record of the political community. This is what a South African lawyer, on the cusp of South Africa's transition to democracy, called a “culture of justification.”Footnote ⁴⁴ It is even arguable that engagement in this culture does not require that all its participants be good in order to produce good laws, since its claim is that a liberal-legalist culture of justification exerts a discipline on the production of law and valid official decisions that will tend to produce good laws and good decisions.

It is possible that, in the US, appreciation of these points is limited by a rather parochial focus on the American political and legal order, a factor to which Lazar alone is alert.Footnote ⁴⁵ It is significant that no major European nation other than the United Kingdom found it necessary to enact major special legislation to deal with the new “war on terror.” And even in the United Kingdom, the legislative regimes that have been put in place have been subject to the legal checks of derogation and proportionality. The former is a much more sophisticated mechanism for temporary departures from rights in times of stress than that offered by the Congressional suspension of habeas corpus, since it requires that the derogation be precise and limited to its express terms, while the latter imposes a structure of rationality on constitutional argument that permits policy and rights to be brought into an explicit balance, reviewable by courts.

In other words, the allegedly absolutist nature of rights protection in the US Bill of Rights might lead to an all-or-nothing approach to emergencies, akin to Locke's naïve legal positivism: either rights with all their force, or no rights at all. In addition, it might stifle attempts at institutional innovation. For example, in the United Kingdom and Canada, there is an experiment taking place that attempts to subject to due process review claims that are made to courts and other tribunals by the security services and which the services base on confidential information. Such an experiment requires the cooperation of the legislature, the executive, and the judiciary, and it is not guaranteed to succeed.Footnote ⁴⁶ But it is an attempt in the spirit of liberal legalism to be responsive to political experience, including the understanding that liberal states will use law in response to emergencies, while at the same time living up to liberal ideals. In this respect, these efforts seem much more pragmatic than any Lockean approach.

There is thus an important difference between legal and extra-legal ways of responding to an emergency, at least from the liberal-legalist point of view. For that position will lead to interpretations of general delegations of authority that are consistent with its liberal principles, thus requiring a government that does not want to abide by such principles openly to declare that it wishes to govern by illiberal means. It is, I think, significant that none of the three authors discussed here undertakes an empirical demonstration of the necessity or the desirability for a liberal state to adopt either extra legal or illiberal means, or even relies to any great extent on such empirical studies. Rather, they adopt a set of assumptions about law that should, as I have argued, be rejected both by liberals and agonistic democrats. For law, and legality more generally, is best understood not as a determinate set of state-sanctioned prescriptions, but as a complex set of norms, principles, and practices centered on rendering public power visible, intelligible, and subject to review and revision.