Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

The Rule of Law Project

Responding to Shirin Sinnar, Rule of Law Tropes in National Security

In my work on legal responses to national security, I have argued that it is important to distinguish between the “black holes” and the “grey holes” in the law that national security legislation tends to create. The former are a “lawless void, . . . in which the state acts unconstrained by law.”1×1. David Dyzenhaus, Schmitt v. Dicey: Are States of Emergency Inside or Outside the Legal Order?, 27 Cardozo L. Rev. 2005, 2006 (2006). The latter, by contrast, offer an individual “some procedural rights but not rights sufficient for him effectively to contest the executive’s case for his detention.”2×2. Id. at 2026. I suggested that grey holes are ultimately worse because they provide a “façade of legality”3×3. Id. at 2038. — the appearance without the substance of legality. Thus, I warned that a “little bit of legality can be more lethal to the rule of law than none.”4×4. Id. at 2026.

In Rule of Law Tropes in National Security, Professor Shirin Sinnar adopts this distinction and echoes my warning in her astute depiction of the legal problems raised by national security measures such as watchlisting people on the basis of their alleged threat to national security or their connection to such threats.5×5. Shirin Sinnar, Essay, Rule of Law Tropes in National Security, 129 Harv. L. Rev. 1566, 1570, 1614–15 (2016). Her powerful analysis of these problems deploys the tool of “rule of law tropes.” The executive claims to be relying on standards borrowed from constitutional or international law such as “reasonable suspicion,” “least intrusive method,” or “imminence,” but in fact acts in ways that are largely uncontrolled by such standards.6×6. Id. at 1573.

The power of Sinnar’s analysis comes from the fact that the tropes she identifies in the practices and rhetoric of watchlisting reshape not only those practices but also our understanding of them in ways that are troubling to anyone who values the rule of law. After all, tropes are standardly figures of speech that use a phrase in a sense other than its literal one in order to create an effect in an audience. Her account thus presupposes, as I do in my own work, that there is value and substance to the idea that the rule of law disciplines executive action. We also agree that this substance is often betrayed in the national security area, with the betrayal concealed by the tropes.7×7. See, for example, Sinnar’s discussion in section IV.C, id. at 1616–17, in which she argues that the original standards have “core, foundational features,” id. at 1617.

Sinnar vividly sketches several costs to this betrayal. First, people might mistakenly suppose that the executive’s invocation of rule of law–like standards makes the executive publicly accountable.8×8. Id. at 1609–10. Second, public awareness of these tropes can erode “trust in law itself” so that people doubt there is “any meaning left to the ‘rule of law.’”9×9. Id. at 1611. Third, the vacuous way these standards operate in the national security context might “seep into” other legal contexts, thus undermining the rule of law more generally.10×10. Id. at 1611–12.

On my understanding of Sinnar’s work, she and I share a commitment to what I call the “rule of law project”: the idea (1) that the rule of law has substantive and valuable content, (2) that there is no option but to extend rule of law’s control to the national security context, and (3) that extending such control can be effective. Thus, the attempt should be made to impose effective rule of law controls in the national security context. In our view, the rule of law project offers the best way to continue an ongoing, experimental process that allows courts and the political branches alike to address legal grey holes.

However, I must note at the outset that the optimism of the rule of law project is contested by two different accounts of the role of law in controlling responses to national security issues. First, realists argue that the national security context highlights dramatically the fact that the administrative state in other ordinary contexts is uncontrolled by the rule of law, so it is no surprise that attempts to impose controls in the national security context prove futile.11×11. See generally Eric A. Posner & Adrian Vermeule, Terror in the Balance (2007) (arguing that courts should defer to the executive in emergencies). Sinnar refers to this work as well. See Sinnar, supra note 5, at 5 n.28. Second, proponents of the extralegal measures model argue that it would be better not to make any attempt to impose rule of law controls but instead to admit frankly that, in the national security context, the executive acts extralegally because vain attempts bring the rule of law in the other ordinary contexts into disrepute.12×12. Oren Gross, Chaos and Rules: Should Responses to Violent Crises Always Be Constitutional?, 112 Yale L.J. 1011, 1097–99 (2003); see also Sinnar, supra note 5, at 1612 n.269 (citing Gross, supra, at 1097).

Notice that the rule of law project shares certain beliefs with both the realist and extralegal measures models. The extralegal measures model accepts the rule of law project’s first proposition, that the rule of law has substantive and valuable content, but rejects the project’s second and third propositions, that extending the rule of law’s control is necessary and effective. On the other hand, the realist view agrees with the second proposition, that there is no option but to extend the rule of law’s control to the national security context, though it rejects the first proposition, that the rule of law has value, and so also rejects the third proposition, that such an extension would be effective.

Put differently, the rule of law project and the realists recognize that governments of rule of law–abiding states answer to what I have termed “the compulsion of legality”: when public officials act in a way that affects individuals’ rights and interests, they must be able to show a legal warrant — an authorization in preexisting law — for their acts.13×13. David Dyzenhaus, The Compulsion of Legality, in Emergencies and the Limits of Legality 33 (Victor Ramraj ed., 2008). Compliance with legality is thus seen as a necessary condition for legitimate state action. Moreover, such compliance often seems to be regarded as a sufficient condition for state action, since (as Sinnar amply demonstrates) debate about the legitimacy of such action increasingly seems to focus on the answer to the question of whether the government can display a legal warrant for what it did.

That the compulsion of legality exists suffices to rule out as a political option the extralegal measures model. But that model does attend to a real problem. As I have explained, the compulsion of legality can set in motion two different cycles of legality: the virtuous cycle and the empty cycle:

In [the] “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 “empty” cycle, the content of legality is understood in an ever more formal or vacuous manner, resulting in the mere appearance or even the pretence of legality.14×14. David Dyzenhaus, Preventive Justice and the Rule-of-Law Project, in Prevention and the Limits of the Criminal Law 91, 98 (Andrew Ashworth, Lucia Zedner & Patrick Tomlin eds., 2013); see also id. at 103–14 (discussing examples of these cycles).

The crucial question for those who subscribe to the rule of law project is whether the empty cycle is inevitable either because the virtuous cycle is unavailable in any context (the realist view) or because there is something special about the national security context (the extralegal measures view).

The answer to this question starts with emphasizing the area of agreement and disagreement already noted between these two views. Both describe the national security context in the same way: as an area where rule of law controls are ineffective. Where they disagree is, first, on the normative implication of this description. The extralegal measures model argues that one should do without the rule of law in this context, while realists argue that we should see that the executive alone has the capacity and the legitimacy to act, even if it responds to the compulsion of legality by cloaking its actions with what Sinnar calls rule of law tropes. Second, they disagree on the effectiveness of rule of law controls outside of the national security context. The extralegal measures model regards such controls as both effective and appropriate, while realists regard the controls as effective only insofar as the executive finds it convenient to abide by them, in much the same way as realists regard international law as constraining state action only if a state finds it in its interest to abide by international law.

Once one sees this, one can also see that both of these views are complex combinations of descriptive and normative claims. In this respect, the rule of law project is no different. Where it differs is that it does not take its descriptive claims to be set in stone, as the foundation of its normative claims. Rather, in a deeply pragmatic spirit, the rule of law project takes its descriptive claims to be bets about how legal experience will turn out if institutional actors in the legal order make a good faith attempt to vindicate its normative claims in practice. Thus, the rule of law project is a proposal to institutional actors that aims to help them make sense of their roles as legal actors, as participants in a rule of law project, which — given the compulsion of legality — is surely the sense they must try to make.

I have detailed elsewhere how placing a bet on the virtuous cycle of legality can yield valuable dividends for the rule of law system as a whole.15×15. See id. at 103–14. If in the United Kingdom one looks back at the history of adjudicating the detention of individuals on the ground that they pose a threat to national security, one can see how Lord Shaw’s dissent in Rex v. Halliday,16×16. [1917] AC 260 (HL). during World War I, arguing that the legislation at issue did not explicitly authorize a detention regulation to be made led to a clear statement of such authority in the subsequent statute enacted on the eve of World War II. This new statute ignited parliamentary debate that prompted a change for the better in the wording of the detention regulation. This wording was in turn the basis for Lord Atkin’s dissent in Liversidge v. Anderson,17×17. [1942] AC 206 (HL). the major World War II decision on detention.

The Liversidge dissent paved the way fifty years later for a reform, prompted by the European Court of Human Rights in Chahal v. United Kingdom,18×18. 23 Eur. Ct. H.R. 413 (1996). that went some considerable distance to permitting effective review of detentions on the basis of national security. Chahal prompted the U.K. Parliament to create the Special Immigration Appeals Commission,19×19. Special Immigration Appeals Commission Act 1997, c. 68 (Eng.). a three-person tribunal, chaired by a High Court judge, with the other two members specializing in immigration and security. The tribunal, whose members are security-cleared, has access to all the information on which the executive bases its claims and to the services of a security-cleared special advocate to test the executive’s case on issues where the individual and his legal representative are excluded because of the sensitive nature of the information.

The reform does not go all the way because the special advocate is not permitted to communicate with the individual or his legal representative once he has seen the material considered by the Home Secretary to be “closed.” However, Canada has subsequently adopted a more refined version of the special advocates system with more scope for reviewing judges to order disclosure of information.20×20. Special Advocates Program, Canadian Dep’t of Just., http://justice.gc.ca/eng/fund-fina/jsp-sjp/sa-es.html (last updated May 15, 2015) [http://perma.cc/EKZ8-3JY6] (“Once the classified information is received, communication by the special advocate may only take place with a judge’s authorization and subject to any conditions the judge considers appropriate.”). The critical point here is that, by a judge’s insistence on adhering to clear legal norms and rules in the context of a particular case, progressive adjudications within a legal system can incrementally extend the writ of the rule of law.

At the crux of this virtuous cycle is the production of what I have described elsewhere as a “conversion” of public policy into legal form in a way that creates a kind of legal “surplus value”21×21. The theory that lies behind this idea is due to Professor Lon L. Fuller, both to his discussion of the eight desiderata of the rule of law that impart to law an “inner morality” in The Morality of Law 33–94 (rev. ed. 1969), and to his discussion of conversion in The Forms and Limits of Adjudication, in The Principles of Social Order 101, 111 (Kenneth I. Winston ed., 2001). For a more elaborate treatment of this theory, see David Dyzenhaus, Process and Substance as Aspects of the Public Law Form, 74 Cambridge L.J. 284, 296–97 (2015). : Public policy is turned into “public, legally applicable standards.” The conversion process thus adds value:

because it brings into being a particular type of public standard — one that permits the operation of principles [of legality] . . . and which enables claims of right based on legal principle to be adjudicated. And if, as is often (even usually) the case in the administrative state, the statute delegates in large part to public officials the task of developing the policy of the statute, the officials will be responsible for producing in appropriate form the public standards that will govern their administrative regime. Hence, there is also often an administrative conversion process that mediates between statute and judicial review of executive action.22×22. Dyzenhaus, supra note 21, at 297.

At its best, this process results in institutional reforms that — across judicial and executive branches alike — help to ensure that particular national security measures are both substantively and procedurally subject to the discipline of the rule of law. Perhaps this is an overly optimistic perspective. I nonetheless maintain that a virtuous cycle of legality is not only possible but also actual in at least some instances. And when it becomes actual through the cooperation of the institutions of legal order in the rule of law project, the vacuous and misleading tropes that Sinnar describes are made to live up to the aspirations on which they trade for their effect.

In my view, the lesson to be learned from virtuous cycles of legality is that if judges find that they are operating within what appears to be a grey hole, they should not insult their legislature by attributing to it the intention that judges should rubber-stamp executive decisions as legal. Rather, they must take the legal regime provided and read into it whatever legal protections they can. To the extent they cannot, it is incumbent on them to point out the problems that have been created for the rule of law. In this way they can prompt the legislature to design ever more effective means of disciplining the executive through institutional design that does not rely primarily on the judiciary, though the judiciary should retain a role as a quality check on how the institutions perform.23×23. For Sinnar’s own illuminating inquiries in this regard, see Shirin Sinnar, Institutionalizing Rights in the National Security Executive, 50 Harv. C.R.-C.L. L. Rev. 289 (2015); Shirin Sinnar, Protecting Rights from Within? Inspectors General and National Security Oversight, 65 Stan. L. Rev. 1027 (2013).

In sum, we can say that the idea of “project” in the rule of law project is meant to convey that the business of the rule of law is always unfinished and that as states respond in new ways to complex problems, new methods have to be found of ensuring that the public officials involved in the response act in accordance with the rule of law. Pessimism about the rule of law is not justified until such experiments have been undertaken.


* Professor of Law and Philosophy and Albert Abel Chair, University of Toronto.