Deference, Security and Human Rights
html xmlns=”http://www.w3.org/1999/xhtml” xml:lang=”en”>
6
Deference, Security and Human Rights
INTRODUCTION
THIS CHAPTER EXPLORES the implications of a public commitment to a ‘culture of justification’ for the roles of different constitutional actors in relation to laws and decisions that both concern national security and affect human rights. I take as my point of departure the central tension in the House of Lords judgment in A v Secretary of State for the Home Department1 between the majority’s approach to whether there existed a public emergency, which came close to treating the question as non-justiciable, and its approach to whether indefinite detention of foreign nationals without trial was a proportionate response, which involved substantive review for justification.
As I will show, these approaches rely on different conceptions of deference. I then ask whether the lack of a sufficiently theoretically grounded conception of deference has been responsible for traditional judicial timidity in the area of national security. I argue for a concept of due deference that is grounded in an understanding of the constitutional order in terms of a culture of justification distinct from both liberal constitutionalism and democratic positivism. Finally, I consider some of the practical and institutional implications of embracing such a culture of justification: the need for more detailed information from the executive about the nature and scale of the threat from terrorism, the need for greater parliamentary scrutiny of executive claims about national security, and the need for courts to treat parliamentary materials as relevant to their own determinations of compatibility.
I. TWO CONCEPTIONS OF DEFERENCE
After 11 September 2001, the UK Parliament introduced in section 23 of the Antiterrorism, Crime and Security Act 2001 (the ‘Anti-terrorism Act’) the power to detain indefinitely non-nationals who are determined by the Secretary of State for the Home Department to be a security risk but who cannot be deported because of some practical consideration or because deportation would subject them to a risk of torture. The power thus does not extend to British nationals, and the Government conceded throughout the litigation that ensued that this meant that nationals who were security risks escaped the indefinite detention visited on non-nationals.
Before the Bill was laid before Parliament, the Home Secretary notified the Secretary General of the Council of Europe that the Government intended to take measures derogating from Article 5 of the European Convention on Human Rights (ECHR),2 which precluded such indefinite detention. Here the Government relied on Article 15(1) of the Convention:
In time of war or public emergency threatening the life of the nation any High Contracting Party may take measures derogating from its obligations under this Convention, to the extent strictly required by the exigencies of the situation, provided that such measures are not inconsistent with its other obligations under international law.
And the Government made the Human Rights Act 1998 (Designated Derogation) Order 2001, which designated the detention powers under section 14(1) of the Human Rights Act 1998.
The Anti-terrorism Act did provide various safeguards: section 24 provided for the grant of bail by the Special Immigration Appeals Commission (SIAC); section 25 permitted a detainee to appeal to SIAC against his certification as a suspected international terrorist; section 26 provided for periodic reviews of certification to be conducted by SIAC; section 28 provided for periodic reviews of the operation of the detention scheme as a whole; section 29 provided for the expiry of the scheme subject to periodic renewal and the final expiry on 10 November 2006 unless renewed. Section 30 gave SIAC exclusive jurisdiction in derogation matters.
The detainees held under the Act claimed that there was no public emergency threatening the life of the nation and that indefinite detention was not strictly required by the exigencies of the situation. It followed, they argued, that there was no valid derogation under Article 15 of the ECHR. They also argued that the detention provisions were discriminatory in contravention of Article 14 of the ECHR, which had not been notified for derogation. That Article precludes discrimination on various grounds, including ‘national … origin’.
In A, the majority of the House of Lords found both that section 23 was not a proportional response to the emergency and that it was incompatible with the United Kingdom’s commitment to the principle of non-discrimination. A is thus rightly regarded as a victory for the rule of law. But the victory is not unqualified.
According to Lord Bingham’s summary of the Attorney General’s argument in A, the Government submitted
that as it was for Parliament and the executive to assess the threat facing the nation, so it was for those bodies and not the courts to judge the response necessary to protect the security of the public. These were matters of a political character calling for an exercise of political and not judicial judgment.3
In other words, the Government adopted the typical stance of governments that claim emergency powers by presenting an argument with two limbs: first, that the question whether there is an emergency is so quintessentially a matter for political judgement that courts must submit to the Government’s and Parliament’s assessment without examining the basis of that assessment; second, since the question of the most appropriate response to the emergency is no less quintessentially a matter for political judgement, courts must also submit to the Government and Parliament on that question, again without conducting any scrutiny of the justifications relied on. In effect, the Government was arguing that these are non-justiciable questions: questions not appropriate for or capable of judicial resolution. The House of Lords has a long tradition of accepting this argument, one that stretches from its decision in World War I in Halliday4 through its decision in World War II in Liversidge5 to its first post-11 September 2001 decision in Rehman.6
Lord Bingham’s response to the Attorney General was that while Parliament, the executive and the judges have ‘different functions’, ‘the function of independent judges charged to interpret and apply the law is universally recognised as a cardinal feature of the modern democratic state, a cornerstone of the rule of law itself’. It was thus wrong to ‘stigmatise judicial decision-making as in some way undemocratic’.7
It is significant that Lord Bingham did not find his ultimate ground in the Human Rights Act but in the constitutional nature of the democratic state with its inherent commitment to the rule of law. Put differently, his understanding of the judicial role does not look to any particular statute, not even the Human Rights Act itself, as the basis for the judicial authority to review legislation and executive decisions for their compliance with human rights and the rule of law, since the legal order is assumed to be a constitutional one, and thus premised on judges having such authority.
Lord Rodger elaborated the implications of this point:
If the provisions of section 30 of the 2001 Act are to have any real meaning, deference to the views of the Government and Parliament on the derogation cannot be taken too far. Due deference does not mean abasement before those views, even in matters relating to national security … The legitimacy of the courts’ scrutiny role cannot be in doubt.8
However, these remarks of Lords Bingham and Rodger are confined to the issue of judicial scrutiny of the appropriateness of the legislative response to the emergency. When it came to the issue of whether there was an emergency, the majority held that if there is some reason to suppose that there is an emergency, the test established in Rehman should apply.9 That is, as long as it not irrational to claim that there is an emergency, even if the judges doubt that there is, they still have to give the benefit of the doubt to the executive.
Thus the victory for the rule of law is qualified in that the majority of the House of Lords rejected outright only the second limb of the Government’s argument, that the appropriateness of the response to the emergency was a political question not for judicial determination. In relation to the first question, the existence of a public emergency, the House of Lords did not accept the Government’s non-justiciability argument but did accept that a more deferential stance was appropriate in light of the sorts of considerations relied on by the Government in support of its non-justiciability argument. The victory is, however, not merely qualified. It rests on an unarticulated tension and is thus unstable.
The remarks made by Lords Bingham and Rodger about the legitimacy of the judicial role are general in import. They apply equally to both limbs of the Government’s argument, as the Government suggested, and as Lord Walker accepted in the lone dissent. The instability arises because in deciding the second question, the majority had to decide whether the Government’s response was correctly calibrated to the emergency—whether it was a proportional response. Thus when they reviewed the appropriateness of the response, they had at least implicitly to review the judgement of the extent and kind of the emergency, indeed, whether there was an emergency at all in the sense that departures from the normal regime of law were justified.
The instability does not reside in the fact that the judges applied different standards of review to these two questions. Rather, it arises because the standard they applied to the question of whether there was an emergency did not seem to require the Government to make a proper justification for its claim. The risk here is that judges say that an issue is justiciable and so proceed to review, but their review is so unrigorous that it might have been better for them to assert non-justicability, that is, that the Government’s decision was not controlled by the rule of law.
I do not, however, underestimate the difficulty the majority faced in confronting the issue of reviewing the Government’s claim that there was an emergency. There was no doubt that the United Kingdom faced a serious threat of terrorist attack, and the events of July 2005 confirmed the Government’s claims in this regard. However, the issue of whether that threat, or indeed actual attacks, amounted to an emergency in accordance with the Article 15 ECHR definition was not so much debated but asserted, as one can gather from both the account in the judges’ speeches of the Government’s arguments and by Lord Hoffmann’s cursory dismissal of those same arguments in the only speech that decided for the applicants on the basis that there was no emergency. Indeed, Lord Hoffmann’s speech, which poured scorn on the claim that the situation was one in which there was a war or other public emergency threatening the life of the nation was a source of disquiet for the other judges. Three of the majority judges thus intimated or expressed their doubts about whether the Government had a serious case.10 However, they found justification for their more deferential approach on this question in two claims.
First, SIAC, in coming to the decision that it should defer to the Government’s claim that there was an emergency, had seen confidential material from the Government in closed session. The Attorney General, however, had declined to ask the House of Lords to read the same material. Still, the majority seemed to think that because SIAC had seen confidential material in closed session and come to a conclusion on its basis, the claim that there was an emergency must have been strengthened by that material.11 And they thought this despite the fact that SIAC had expressly not relied on the confidential material in coming to its conclusion in its open judgment.12
But even if it were thought appropriate that the higher courts should be ready to defer to SIAC should SIAC adequately justify its decision, and, correspondingly, that both SIAC and the courts should be ready to defer to the Government when it provides such a justification, such deference cannot be blind. As Lord Rodger has put it, ‘Due deference does not mean abasement’. To give the Government the ‘benefit of the doubt’, as one judge put it, at the same time as harbouring ‘very grave doubt’13 about the Government’s case seems peculiar, especially when the court itself could have asked to see evidence that might have removed some of that doubt and when what debate had taken place in Parliament tended to undermine rather than support the Government’s case.
Second, the majority relied on decisions of the European Court of Human Rights (ECtHR) that held that the Court should generally defer to a national Government’s determination that there is such an emergency.14 But such reliance fails to give proper effect to the gap some of the judges acknowledged15 between the situation on the one hand in which the European Court defers to a decision by a Government that has withstood challenges before that Government’s national courts and the situation on the other hand in which the highest national court has to evaluate the Government’s challenge. That is, a stricter standard is arguably appropriate in the latter situation, and the application of such a standard there would make more sense of the application of a more relaxed standard in the first situation.16
My argument in this chapter is not that the majority were necessarily wrong to defer on the first question, but that they failed to require that a proper case for deference be made by the Government and therefore failed to conduct anything approaching appropriate judicial scrutiny of the reasons underlying the Government’s assertions about the existence of an emergency. That argument has to address two different though closely connected sets of problems: an institutional set and a theoretical set.
The institutional problems pertain to the dearth of resources facing judges who are prepared to acknowledge the inextricability of the two issues—the existence of the emergency and the appropriateness of the response. Such judges would need a better justificatory basis to scrutinise than the House of Lords had available to it in A. For there to be such a basis, the Government would have to be prepared to treat Parliament parliament as more than a rubber stamp for legislation when the Government thinks it needs more powers to confront an alleged crisis. The Government would not only have to forego its standard (and nearly always unjustified) line that there is no time to debate properly both the extent of the emergency and the appropriate responses to it. It would also have to be prepared to allow into the public domain more detailed information about the precise nature and scale of the threat; or, to the extent that it genuinely cannot do so, the Government or Parliament itself would have to devise some system within Parliament whereby that part of the Government’s case that cannot be publicly debated can be heard. Thus more constitutional furniture would have to be put in place in order to ensure that the Government and Parliament could meet their justificatory responsibilities before the judges could carry out their duty properly to evaluate the Government’s and Parliament’s case. And for the judges to carry out that duty, they would of course have to be given by Parliament or devise themselves some means of testing the arguments made in any closed committee sessions, as well as those made in public.
While I will have more to say later about the institutional problems, my main task in this chapter is to confront the theoretical set of problems. Institutional reform requires a theoretical justification. If it is never appropriate for judges to defer to the executive or to parliamentary judgement about what the law requires, or if judges should always submit to such judgements, there is no need for these institutional resources.
It is important at this juncture to be clear about what I mean by deference to judgement. My conception of ‘due deference’ depends on a distinction between deference as respect and submissive deference, the equivalent of what Lord Rodger called deference as abasement.17 Deference as respect requires judges to pay respectful attention to the reasons offered by a primary decision maker as justification for a particular decision. It requires not only that there are reasons but that the reasons are capable of justifying the decision. Such deference has therefore both a procedural and a substantive or qualitative aspect to it—going through the hoop of providing reasons does not suffice to earn respect since respect requires demonstration of the adequacy of the reasons offered to justify the conclusion reached. Deference is due to a decision maker when he has offered reasons to justify the result, and the reasons are moreover adequate to justify the result. Scrutiny of reasons for their ‘adequacy’ therefore involves a qualitative appraisal of the substantive reasons for a particular decision, but is not the same as scrutiny for correctness: it imposes a high threshold that the reasons must satisfy basic standards of rationality and not be inconsistent with the principles that underpin democracy; but it leaves space for a conclusion that the reasons are ‘good enough’ in a democratic society.
In contrast, submissive deference requires judges to submit to the intention of the legislature, on a positivist understanding of what constitutes that intention. If the legislature’s intention is to delegate authority to an official to interpret and implement a statutory mandate, this conception of deference transfers the submissive stance it requires from the legislature to the official. Judges should defer to the primary decision maker’s decision, as long as the decision maker does not stray beyond the limits of his statutory authority, positivistically construed. If a reviewing judge embarks on a process of evaluating the decision maker’s reasons for decision, he or she is thus asking herself the wrong question. That question is likely to lead him or her to cross the line between the legitimate exercise of checking on legality or the limits of authority and the illegitimate exercise of substantive review, or review as to the merits. Since substantive review is to be avoided, there is no procedural requirement that the official give reasons. Indeed, the position that recommends submissive deference might go further than lack of support for a duty to give reasons. If the procedural duty invites substantive review, imposition of such a duty should be resisted.
I will come back to the distinction between deference as respect and submissive deference below. For the moment, I want to note two issues. First, one of the points of contention between these conceptions of deference is the idea of legal authority and legality. For deference as respect, legal authority is based on reasons, and legality is a matter of principles. A decision has authority when it is supported by reasons that show why all the relevant legal principles support the decision. For submissive deference, authority attaches automatically to any decision that is made within the space delimited by the delegating power. That is, legality is spatially conceived, so that its limits are only those limits that are explicitly stipulated by the delegating power.18
Second, it is important to keep in mind that debate so far about deference has been largely shaped by the assumption that in issue is the kind of deference, if any, due to primary decision makers, who have been delegated authority by Parliament to make decisions. But my concern in this chapter is deference to both primary decision makers and Parliament itself. If, as deference as respect presupposes, authority is reason-based and legality a matter of principles, then statutes—Parliament’s decisions—have authority only when they are supported by reasons that show why all the relevant legal principles support them.
In contrast, submissive deference presupposes that there are no limits on Parliament’s authority, other than the requirements of manner and form that attend the proper enactment of statutes. These requirements are the equivalents of the limits on delegated authority—they are formally conceived, determinate boundaries of a space in which there are no substantive principles to which Parliament is answerable. It is that presupposition that underpins the idea that Parliament’s intention is to be interpreted by positivistic tests as to what was in fact intended, so that the limits on the authority of primary decision makers are whatever and only the limits the omnipotent Parliament saw fit to stipulate. On this view, to conceive of Parliament as answerable to principles of legality is to conceive of its authority as substantively limited, which means that its delegates are similarly limited, thus providing the basis for judges to impose their understanding of those limits on the primary decision makers, an exercise in substantive review that amounts to a usurpation of the legislative role.
It is important to keep in mind that there is this connection between conceptions of deference to primary decision makers and some of the most fundamental issues in constitutional and legal theory about the nature of legal authority and legality. At stake in this debate is whether there is a conception of principle-based legality that does not collapse into either merits review (so that the conception of deference as respect turns out to be but a disguise for judicial supremacism, that is, for no deference at all) or judicial abdication (so that deference as respect turns out to be a non-justiciability doctrine, or submission by another name). As I will now try to show, developments in the last 10 years in the United Kingdom not only are more consistent with the idea of deference as respect than submissive deference, but also help to demonstrate that deference as respect is not a sham.
II. NATIONAL SECURITY: THE LAST PREROGATIVE?
At one time, decisions about national security, for example, who should be deported or detained as security risks, were regarded as matters so political and so based on sensitive information that they were not justiciable or amenable to judicial review. In other words, such matters fell within the prerogative of the executive, an extra-legal space of political decision. In the twentieth century, however, the idea that certain decisions of the political branch of Government have authority despite their extra-legal status became increasingly difficult to sustain. The claim, articulated by Dicey, that political authority is constituted by law and that the executive must therefore demonstrate a legal warrant for any of its decisions that impinge on fundamental rights and values gained an ever-stronger grip on the public mind.
The extension of law’s reach over the prerogative is good news for all those who regard it as a precondition of any Government’s claim to legitimacy that it exercises power in accordance with the rule of law. But there are dangers to this extension because of the possibility of confusing the rule of law with rule by law, or rule by statute. If one assumes that as long as there is a warrant in a statute for what an official does, he or she is acting in accordance with the rule of law and therefore legitimately, it seems to follow that an official who is authorised by a statute to act arbitrarily is by definition acting in accordance with the rule of law and therefore also legitimately. In other words, if the extension of law’s reach over the prerogative brings about the legal installation of the prerogative in all but name, it seems to legitimate arbitrary action by claiming that it is by definition legal. The rule of law is thus ultimately undermined.
I accept that this danger has to be taken seriously. In the context of an attempt to articulate a doctrine of due deference, it manifests itself in tests that amount to a stance of submissive deference. This stance allows the claim that certain acts of the executive are not justiciable to re-emerge as the claim that judges must defer to the executive unless the executive acts so irrationally that it is manifest that it has strayed beyond the limits of its legal authority, better known as ‘Wednesbury unreasonableness’. Unfortunately, Dicey’s constitutional theory, taken as a whole, is not only an unhelpful basis for responding to this problem; it in large part helps to create it.
Dicey’s advocacy of an ideology of legality, the content of which, he suggested, is to be found in common law presumptions about the liberty and equality of the individual, was undermined by his continuing support for the doctrine of legislative supremacy, which underpins submissive deference. While he thought that judges should maintain the rule of law by interpreting statutes in the light of common law presumptions, he also argued that judges must submit to clearly expressed legislative intentions, whatever their content. His theory of constitutionalism and the rule of law is thus radically unstable. It licenses judges to impose their views about what the rule of law requires on statutes, but only on condition that the statutes can bear these interpretations, a condition often stated as one to do with the ‘ambiguity’ of statutory language. However, if a statute unambiguously excludes such judicial presumptions, Dicey’s theory requires judges to submit to parliament’s will. Moreover, Dicey regarded the administrative state as in principle uncontrollable by the rule of law, since the statutes that set up that state granted to the frontline decision- makers an authority to act as they saw fit—to be a law unto themselves. Such statutes, in other words, create the equivalent of the prerogative within law.
It follows that, on the one hand, Dicey’s theoretical legacy is one in which there is no doctrine of deference—judges simply interpret the law in accordance with their sense of its fundamental values. On the other hand, his legacy also requires judges to adopt the stance of submissive deference to clear legislative intent, including the intent to delegate virtually unfettered discretion to administrative officials. Judges who work with a Diceyan understanding of legality tend to shuttle between these two inconsistent parts of his legacy. If they are tempted to try to restrict delegations of authority by interpreting them as confined by common law presumptions, legislatures can react through privative clauses that oust judicial review, by framing delegations so as to make it clear that officials’ decisions depends on their subjective judgment, or by a combination of both devices.
Despite these concerns, the extension of law’s reach over the prerogative must be seen as a potentially helpful first step in establishing the control of the rule of law over political decisions about national security. Moreover, during World War I and World War II, not only did the indefinite detention of individuals who were perceived to be risks to national security have to follow a procedure set out in statute and regulations, but each decision was in principle subject to appeal to an executive committee.
However, as Brian Simpson has shown, the committee lacked teeth.19 Not only did it fail to require the real reasons for detentions from the intelligence branch, but in any case the committee could advise the Home Secretary only if someone had been wrongly detained. One can view these facts from two very different perspectives: either the kind of legality contemplated by these detention regimes was a cynical façade designed to legitimate the illegitimate; or it represented genuine, though ineffective, attempts to ensure that the executive did not act arbitrarily.
There is an advantage to adopting the latter perspective—it permits judges to try to work up the regime of legality into something with more rule of law teeth, on the basis that the very attempt to design a regime of legality is evidence of Parliament’s intention to rule not only by law, but also in accordance with the rule of law. There are risks, namely that judges might end up giving the imprimatur of the rule of law to an administrative regime that does not provide those subject to it with any real protection. However, the more cynical perspective is not really available to judges. They have to try to make sense of the regime as a legal one, which means that they have only limited options. First, as suggested, they can try to make sense of the regime by attempting to give it rule of law teeth. Second, they can say that the regime is legal without making the attempt, in which case they give the regime the imprimatur of the rule of law by equating that rule with rule by law. Finally, they might find that the regime is illegal because it is incompatible with fundamental principles of legality.
The majority of the House of Lords in Halliday and in Liversidge adopted the second option. They said that the demands of legality were satisfied by the detention regimes under consideration and that such regimes were appropriate given the contexts, that is, periods of national emergency. In contrast, Lord Shaw in his dissent in Halliday started with the assumption that Parliament must be taken to intend that its delegates act in accordance with the rule of law, which means that it must explicitly authorise any departures from the rule of law. Since the Defence of the Realm Consolidation Act 1914 did not authorise a detention regulation, the regulation that brought the detention regime into play was invalid.
When civil servants subsequently put together the detention regime for World War II, they took note of Shaw’s dissent and so ensured that the authorising statute explicitly permitted the establishment of a detention regime by regulation. Again, one can view this step either as a progress towards the realisation of the rule of law or as yet another embellishment to the façade. But, in addition to this response to a dissenting judge, the Government responded to concerns raised in Parliament about the wording of the initial version of the detention regulation. It substituted ‘reasonable cause to believe’ when it came to the grounds for detention for the original ‘if satisfied that’. And it was on the basis of this substitution that Lord Atkin held in his famous dissent in Liversidge that a court was entitled to more than the Government’s say-so that an individual is a security risk, although the majority disagreed on the basis that it was inappropriate in wartime for judges to go beyond the mechanism explicitly put in place (that is, the toothless review committee).
Lord Atkin’s dissent was held up after the war as the correct stance for judges faced with executive decisions on national security grounds. However, an inspection of other judicial decisions in this area reveals a consistent pattern of executive-minded decisions by courts. In addition, when developments in ‘normal’ administrative law—where the issue was not national security—made it inevitable that judges asserted a comprehensive authority to control public power in general, including the prerogative, they also said eventually that the prerogative to deal with national security is subject to the rule of law and thus to judicial review. But in the same breath they said that the executive say-so as to what is required in the interests of national security must prevail. In other words, they reinvented the prerogative under the guise of a doctrine of judicial deference.20
The legal landscape changed in this regard when the ECtHR rejected the United Kingdom’s argument in Chahal v UK21 that national security grounds are inherently incapable of being tested in a court of law, holding that the advisory panel for those subject to deportation on national security grounds did not constitute the ‘effective remedy’ as required by Article 13 of the ECHR.
The UK Government responded through Parliament in 1997 with a statute that established SIAC. Of the three members on the SIAC panel, one member had to have held high judicial office, the second to have been the chief adjudicator or a legally qualified member of the Immigration Appeals Tribunal, while the third would ordinarily be someone with experience of national security matters. The 1997 statute gave individuals who would have had the right to appeal against a deportation order but for the fact that national security was involved a right to appeal to SIAC and granted SIAC the authority to review the Secretary of State’s decision on the law and the facts as well as the question whether the discretion should have been exercised differently. There was a further appeal to the Court of Appeal on ‘any question of law material to’ SIAC’s determination. In addition, the statute provided for the appointment of a special advocate who could represent appellants if parts of the proceedings before SIAC took place in closed session. SIAC decisions can be based on both closed and open sessions. It gives both open judgments, which do not disclose information from the closed sessions, and closed judgments, which are given to only the Government.
SIAC thus seemed to be, at least to some extent, an answer to concerns that the rule of law cannot control decisions on national security grounds. However, in Rehman the House of Lords held, on separation of powers grounds that, whilst the issue of what constitutes national security is a question of law, and hence for SIAC and the courts to interpret, it is for the executive to decide what is ‘in the interests of national security’.22 The result is little different from the majority’s claim in Liversidge that detention orders were not arbitrary since the courts could still check that they were made in good faith.
The question why this court did not give full effect to the legislative machinery contemplated by the establishment of SIAC becomes even more pressing when one notes that two of the judges on the bench—Lords Steyn and Hoffmann—are responsible for articulating a principle of legality in ordinary administrative law that requires all executive acts to be demonstrated as justifiable in law, where law is assumed to include fundamental values.23 It is puzzling that these two judges find that in some cases that they are driven to constitutional bedrock, which they find to be full of values and principles, whilst in other cases they treat the constitution as only a very rigid and formalistic doctrine of the separation of powers.