AN v Secretary of State for the Home Department

Following the decision of the House of Lords in Secretary of State for the Home Department v AF(No.3) [2009] UKHL 28, the Secretary of State elected not to make further disclosure to a number of persons subject to control orders but to revoke the orders then in force. However, there remained pending proceedings in the Administrative Court pursuant to section 3(10) of the Prevention of Terrorism Act 2005 (the 2005 Act) in relation to the orders. The cases raised the questions: Were the control orders to be quashed with effect from the dates upon which they had been made? Or should their revocation only operate prospectively? In AN [2009] EWHC 1966 Admin, Mitting J concluded that only prospective revocation was required but in AE and AF [2010] EWHC 42 Admin Silber J ordered that the orders be quashed ab initio. AN now appeals in the first case and the Secretary of State appeals in the second case. The point is not merely academic. AN is the subject of criminal proceedings for breach of his control order. If it were to be quashed ab initio, the prosecution would fall at the first hurdle. AE and AF do not face criminal proceedings but they wish to claim damages in respect of the period of about 3½ years for which they were subject to their control orders prior to revocation.

By section 1(2) of the 2005 Act, the power to make a control order against an individual is exercisable:

“(a)

except in the case of an order imposing obligations that are incompatible with the individual’s right to liberty under Article 5 of the Human Rights Convention, by the Secretary of State; and

(b)

in the case of an order imposing obligations that are in or include derogating obligations, by the court on the application of the Secretary of State.”

To date, the Secretary of State has not made any application, in these or other cases, to the court pursuant to section 1(2)(b) but has sought to proceed by way of non-derogating control orders under section 1(2)(a).

Section 2 provides:

“(1)

The Secretary of State may make a control order against an individual if he –

(a)

has reasonable grounds for suspecting that the individual is or has been involved in terrorism-related activity; and

(b)

considers that it is necessary, for purposes connected with protecting members of the public from a risk of terrorism, to make a control order imposing obligations on that individual.

…

(3)

A control order made by the Secretary of State is called a non-derogating control order.”

Section 2(4) provides that a non-derogating control order has effect for a period of 12 months, but there is power given to the Secretary of State by section 2(6) to renew it.

Section 3 is headed “Supervision by court of making of non-derogating control orders”. It provides:

“(1)

The Secretary of State must not make a non-derogating control order against an individual except where –

(a)

having decided that there are grounds to make such an order against that individual, he has applied to the court for permission to make that order and has been granted that permission;

(b)

the order contains a statement by the Secretary of State that, in his opinion, the urgency of the case requires the order to be made without such permission; or

(c)

[not material]

(2)

Where the Secretary of State makes an application for permission to make a non-derogating control order against an individual, the application must set out the order for which he seeks permission and –

(a)

the function of the court is to consider whether the Secretary of State’s decision that there are grounds to make that order is obviously flawed;

(b)

the court may give that permission unless it determines that the decision is obviously flawed; and

(c)

if it gives permission, the court must give directions for a hearing in relation to the order as soon as reasonably practicable after it is made.”

If the Secretary of State makes an urgent order without permission pursuant to section 3(1)(b), he has an obligation under section 3(3) to refer the order immediately to the court, which has the function of considering whether the order was “obviously flawed”. If the court decides that it was obviously flawed, then by section 3(6)(a), “it must quash the order” or any condition in it that was obviously flawed. In the absence of an obvious flaw “it must confirm the order and give directions for a hearing in relation to the confirmed order” (section 3(6)(c)).

Other material provisions of section 3 are as follows:

“(10)

On a hearing in pursuance of directions under subsection (2)(c) or (6)(b) or (c), the function of the court is to determine whether any of the following decisions of the Secretary of State was flawed –

(a)

his decision that the requirements of section 2(1)(a) and (b) were satisfied for the making of the order; and

(b)

his decisions on the imposition of each of the obligations imposed by the order.

(11)

In determining –

(a)

what constitutes a flawed decision for the purposes of subsection (2), (6) or (8), or

(b)

the matters mentioned in subsection (10),

the court must apply the principles applicable on an application for judicial review.

(12)

If the court determines … that a decision of the Secretary of State was flawed, its only powers are –

(a)

power to quash the order;

(b)

power to quash one or more obligations imposed by the order; and

(c)

power to give directions to the Secretary of State for the revocation of the order or for the modification of the obligations it imposes.

(13)

In every other case the court must decide that the control order is to continue in force.”

Section 9(1) creates the criminal offence of contravention, without reasonable cause, of an obligation imposed by a control order. The Schedule contains provisions governing control order proceedings, as does Part 76 of the CPR. Although we received submissions about them, I do not find it necessary to set them out. They contain the procedural detail which gave rise to the difficulties resolved by AF(No.3).

Mitting J rejected the primary submissions on behalf of AN that the control order was a nullity. He said (at paragraph 5):

“If I had been persuaded that the order was a nullity, I agree that it would have to be quashed, like the order which the Secretary of State had no power to make in Secretary of State v JJ [2007] UKHL 45 … Article 6 applies to ‘control order proceedings’: see Lord Bingham’s summary of the Secretary of State’s concession in MB at paragraph 15. Whether or not the procedure used has involved significant injustice to the controlled person must be determined by looking at the process as a whole: paragraph 35. The making of the order by the Secretary of State is part of that process. But it is the Court which determines, when granting or withholding permission to make the order under section 3(2), whether the decision of the Secretary of State is obviously flawed. The obligation to disclose or gist to the controlled person the essence of the case only arises at the stage when the Secretary of State’s decisions are reviewed under section 3(10). Subject to the qualification made below, when the Secretary of State decides to apply for permission to make the order and makes it, he is not inhibited from relying on closed material which, in due course, he may elect to withdraw rather than to disclose or gist. Further, when the Secretary of State decided to make the order it was reasonable to suppose that she would be permitted to rely on the closed material without gisting or disclosing it [because of her reasonable understanding of what the law was, prior to later decisions of the House of Lords] … On the principle that a decision of a properly constituted Court on an issue within its jurisdiction is binding unless and until set aside, [the submission on behalf of AN] is untenable. I am satisfied that both elements of the proceedings at the inception of the control order (Collins J’s permission, and the Minister’s decision, to make the order) were lawful and that neither was a nullity. Taken together, that stage of the proceedings cannot be so described. It follows that I am not required by ordinary judicial review principles to quash the order.”

Mitting J accepted that he had a discretion whether to quash the order or to give directions for it to be revoked. However, on the basis that it had been “properly made and renewed on the basis of material which the Secretary of State and Collins J were entitled to take into account” it was not a nullity. It had become an order “which cannot now be sustained” as a result of a proper decision made on or shortly before 15 July 2009 “in the light of the law as it has now been declared to be by the House of Lords”. Accordingly, by reference to his discretion under section 3(12), he gave directions that an order “lawful at inception, but which can no longer be sustained” should be revoked.

Silber J, having repeatedly emphasised that he had come to the contrary conclusion on the basis of much fuller and more detailed submissions, explained that he had done so for a number of reasons. His primary view was that he had no discretion but was bound to quash ab initio. His reasoning can be summarised as follows:

(1)

he construed the speeches of Baroness Hale and Lord Bingham in Secretary of State for the Home Department v MB and AF [2008] 1 AC 440 as requiring that result;

(2)

he considered that the approach of Lord Hope (with whom Lord Scott agreed) in AF(No.3) was to like effect;

(3)

a “particularly important” reason was the “landmark decision on natural justice” of Ridge v Baldwin [1964] AC 40;

(4)

anything less than quashing ab initio would make the controlees’ Article 6 rights, as explained in AF(No.3), ineffectively secured; and

(5)

“if the Secretary of State had at any time during the life of any of the control orders been obliged to comply with the disclosure obligations in accordance with what was eventually decided in AF(No.3) by disclosing to the controlees the essence of the case against them, he would have refused as he actually did after the House of Lords had reached its decision”. (Paragraph 79).

Silber J went on to hold that, if he were wrong about the absence of discretion, he would have quashed the orders ab initio as an exercise of discretion for substantially the same reasons. He also rejected an alternative submission on behalf of the Secretary of State that any quashing order should not be ab initio but should only run from a later date in the judicial proceedings.

Until now, none of the control order cases in the Court of Appeal, the House of Lords, or the Supreme Court has called for a definitive ruling on the issue that lies at the heart of this appeal. However, there have been relevant obiter dicta.

In Secretary of State for the Home Department v JJ [2007] UKHL 45, [2008] 1 AC 385, the Secretary of State had purported to make a non-derogating control order but the obligations imposed by it were found to amount to an unlawful deprivation of liberty under Article 5. It was submitted on behalf of the Secretary of State that the appropriate course was to quash only the offending obligations and not the whole order. The argument was (at page 393c):

“The quashing of the whole order under section 3(12)(a) should be reserved for cases where it was clear that none of the obligations could be justified, or where the Secretary of State failed to establish the factual basis of his evaluation of terrorist-related activity or qualifying risk.”

In rejecting the submission, Lord Bingham said (at paragraph 27):

“An administrative order made without power to make it is, on well known principles, a nullity … It is true that, because public law remedies are generally discretionary, the court may in special circumstances decline to quash an order, despite finding it to be a nullity … But no such circumstances exist here, and it would be contrary to principle to decline to quash an order, made without power to make it, which had unlawfully deprived a person of his liberty.”

Baroness Hale, concurring, said that the judge “had no choice but to quash these orders” (paragraph 64). Lord Brown spoke to like effect (paragraph 109) as did Lord Carswell on this issue (at paragraph 85). In the present case, the submission on behalf of the Secretary of State is that JJ is distinguishable because there the Secretary of State had purported to make an order which could only have been made by the court as a derogating control order.

In MB and AF [2007] UKHL 46, [2008] 1 AC 440, Article 5 and Article 6 issues were raised. The Secretary of State advanced the same submission in the same language that was rejected in JJ (the decisions were handed down on the same day). In a passage heavily relied upon by the controlees on the present appeal, Baroness Hale said (at paragraph 72):

“Where the court does not give the Secretary of State permission to withhold closed material, she has a choice. She may decide that, after all, it can safely be disclosed … But she may decide that it must still be withheld. She cannot then be required to serve it. But if the court considers that the material might be of assistance to the controlled person in relation to a matter under consideration, it may direct that the matter be withdrawn from consideration by the court. In any other case, it may direct that the Secretary of State cannot rely upon the material. If the Secretary of State cannot rely upon it, and it is indeed crucial to the decision, then the decision will be flawed and the order will have to be quashed.”

This passage was adopted by Silber J in AN (at paragraph 29). The submission on behalf of the Secretary of State now is that the passage is not applicable outside the context of Article 5, that it stands alone and that Silber J gave it undue emphasis.

In AF(No.3), Lord Hope said (at paragraph 82):

“The difficulties that less than full disclosure gives rise to must be counterbalanced in such a way that the controlled person has the possibility effectively to challenge the allegations made against him. If that cannot be done, the judge must exercise the power that he is given by section 3(12) … and quash the control order.”

It does not seem that the quashing/revocation issue had been the subject of detailed submissions. The House of Lords remitted the cases to the High Court for further consideration. AN and AE were two of the cases.

The rival submissions on this appeal have ranged far and wide. I find it appropriate to consider the more significant of them in the following order.

Whereas a derogating control order under section 4 can only be made by the court on application by the Secretary of State, a non-derogating control order can only be made or renewed by the Secretary of State: section 2(1), (3) and (6). He must seek and obtain the permission of the court to make the order, either in advance of, or, in a case of urgency, immediately after, he makes the order: section 3(1) and (3). The function of the court is to consider whether, in a normal case, the decision of the Secretary of State that there are grounds to make the order is “obviously flawed”: section 3(2). In an urgent case, the function of the court is to consider whether the decision of the Secretary of State to make the order was “obviously flawed”: section 3(3)(b). In a normal case, when the court gives permission, it must give directions for a hearing in relation to the order as soon as is reasonably practicable: section 3(2)(a). In an urgent case, if the court does not quash the order as being obviously flawed, it must “confirm” the order and give directions for a hearing in relation to the “confirmed order”: section 3(b)(c). All this emphasises that at no stage does the court make a non-derogating control order. Moreover, when it proceeds to a substantive hearing pursuant to directions given under section 3(2)(c) or (b)(a), its only powers are those set out in section 3(10). At that stage, the test is whether the order is “flawed” rather than “obviously flawed”. Although judicial review principles apply at both stages, the intensity of the review increases by reason of the later omission of the adverb “obviously”. Ultimately, unless the court exercises one of the powers specified in section 3(12), it must decide that the order is to continue: section 3(13). Thus, the role of the court throughout in relation to non-derogating control orders is one of “supervision” – the word used in the heading to section 3. It does not and cannot make such an order, whether upon application or otherwise.

The corollary of this analysis is that the making of a non-derogating control order is an administrative act of the Secretary of State, albeit one for which, in a normal case, he requires the permission of the court, which will only be denied in the case of an obvious flaw. One of the submissions made on behalf of the Secretary of State, which found favour with Mitting J, is that a non-derogating control order so closely resembles a court order that, as with a court order, it retains its validity unless and until it is set aside by a court of competent jurisdiction. Reliance is placed on authorities concerning injunctions (Grafton Isaacs v Emery Robertson [1985] 1 AC 97) and anti-social behaviour orders (Director of Public Prosecutions v T [2006] EWHC 728 (Admin), [2007] 1 WLR 209). In my judgment, these are false analogies. The statute has vested the power to make a non-derogating control order exclusively in the hands of the Secretary of State and, in principle, if an order is in truth legally flawed, it attracts the usual consequence of a legally flawed administrative act, even if the legal flaw was not observed when the order was first considered by a court. The usual consequence is quashing ab initio: Ridge v Baldwin [1964] AC 40, per Lord Reid at page 80. That this reasoning should apply to control orders is supported by section 12 of the 2005 Act which entitles a person who has been convicted of a criminal offence of breach of a control order which is later quashed to appeal successfully against his conviction to the Court of Appeal Criminal Division which must allow his appeal: section 12(3). The fact that this applies also to court-ordered derogating control orders does not weaken the point.

At all material times before AF(No.3), the Secretary of State, in common with and in reliance upon decisions of the domestic courts, believed that the non-derogating control orders which he had been permitted to make by the court were valid and lawful. In AN, Mitting J considered this to be relevant to the issue of whether, after AF(No.3), the control order should be quashed ab initio. He said (at paragraph 3):

“… when the Secretary of State decided to make the order it was reasonable to suppose that she would be permitted to rely on the closed material without gisting or disclosing it.”

He observed that the hearing of MB and AF in the House of Lords did not begin until the following day and the decision was not handed down until some months later. Whilst one inevitably sympathises with the plight of any litigant caught up in a maelstrom of judicial development of the law, the Secretary of State can claim no special dispensation. The fact is that, although he was acting in good faith and in the interests of public protection, he was doing so on the basis of an interference with the human rights of the controlled persons. He cannot set up his, or the courts’, prior misapprehension of the extent of his legal obligations as a form of defence. Moreover, there is a fallacy at the heart of his claim to do so.

In order to advance a submission that a control order was valid when made but only succumbed to legal difficulty at a later date, the Secretary of State would have to establish that, in relation to the point for which he is asserting legality, he can satisfy the court as to the reasonable grounds for his suspicion of terrorism-related activity and the need for public protection. However, he could only do that by relying on the material that he is unwilling to disclose or gist. In other words, he would need to resort now to closed material in a manner not countenanced by AF(No.3). Whilst I accept Mitting J’s suggestion that, in court, the Secretary of State does not have to rely on all the material that led him to his view about terrorism-related activity and public protection, he does have to rely (with consequential disclosure obligations) on sufficient of it to satisfy the court that his decision to make a control order was not and is not flawed. In these cases, he has chosen not to do so. I shall assume he has reasonable grounds for exercising that choice. However, its consequence is that he has disabled himself from satisfying this appellate court that, throughout, he has been able to satisfy section 2(1). In essence, we are being invited to assume that, but without access to the relevant material. We are being asked to find that he acted reasonably when, in truth, that is something we cannot test against the material relied upon by the Secretary of State.

In the case of AN, Mitting J, having declined to quash the control order, directed that it be revoked pursuant to section 3(12)(c). He accepted the submission of the Secretary of State that a direction for revocation is one of a range of remedies made available in this situation by section 3(12) and that it is a matter for the judge whether to quash under section 3(12)(a) or to direct revocation under section 3(12)(c). The case for the Secretary of State is that, if this were not so, section 3(12)(c) would be emasculated. I do not agree. Section 3(12)(c) is of particular importance in relation to a control order that was valid and sustainable when made but which is no longer so because of a change of circumstances. AV v Secretary of State for the Home Department [2009] EWHC 902 (Admin) is an illustration. However, in essentially unchanged circumstances or where no change is asserted, the structure of the statute tends to assume that a flawed order is void ab initio. Thus, where the Secretary of State purports to make an urgent order without prior permission under section 3(1)(b) and the court proceeds to find it obviously flawed pursuant to section 3(6), it must quash it by reference to section 3(6)(a) – there is no option of revocation. In the case of a derogating control order, there is a power at the full hearing to revoke the order but that is in the context of the court bringing an end to its own order, not an order of the Secretary of State. A judge cannot normally quash an order of a judge of coordinate jurisdiction. Even so, the court is given a discretion, upon revocation, to direct that the Act have effect “as if the order had been quashed”. Also, section 12 (see paragraph 24, above) is again of relevance.

A point sought to be taken by the Secretary of State is that, accepting that the control orders cannot now be sanctioned, it does not follow that they were flawed from the outset. There was no obligation to disclose anything at all to the controlees until after they were served with the orders. Until then there was no contestation to which they were parties. The inter partes dispute only arose on service. Article 6 did not bite until then at the earliest. Consequently, the orders were not void ab initio. On the contrary, they were lawful at their inception.

I do not accept this submission. Although Article 6 may not bite at a stage of administrative determination, it is axiomatic that, when the Secretary of State decides that there are grounds for making a non-derogating control order (which, by definition, imposes restrictions on a person’s liberty), he knows that he will have to justify it so as to obtain the permission of the court under section 3(2) and, in due course, and subject to more intense scrutiny, at a hearing under section 3(10). Mitting J had this in mind in AN when he said (at paragraph 6):

“If, following the clarification of the law in AF(No.3), when applying for permission, the Secretary of State intended not to disclose or gist the material, it may be an abuse of the court’s process to make the application. What the Secretary of State would have done would have been to apply for permission to make a control order which he had no intention of seeking to sustain at the section 3(10) review. In those circumstances, it may well be right to quash the order.”

That he did not do so in AN was because of the view he took about the reasonableness of the Secretary of State’s erroneous understanding of the law at the time. I have explained in paragraph 26, above why I disagree with him about that. Moreover, I cannot escape the conclusion that it is unlawful for the Secretary of State to begin to move towards the making of a control order if, in order to justify it, he would need to rely on material which he is not willing to disclose to the extent required by AF(No.3), regardless of his understanding of the law at the time. If I were wrong about that it would mean that the Secretary of State could lawfully place significant restrictions on a person’s liberty without that person ever being able to discover the basis for the Secretary of State’s decision. It would be beyond scrutiny or challenge. This would run counter to the unappealed decision of the Court of Appeal in Secretary of State for the Home Department v MB [2007] QB 415 that the task of the court is to determine whether the decision to make the control order was flawed at the time it was made and thereafter (see paragraphs 40-46). For these reasons, I reject the submission that the orders only became flawed from the time of, or in the approach to, the section 3(10) hearings. That legal conclusion is accompanied, in the cases of AF and AE, by Silber J’s factual finding (at paragraph 79) that

“the Secretary of State would not have made any of the control orders against any of the controlees if he had appreciated the nature and extent of the disclosure of closed evidence which was required by AF(No.3).”

I consider such a finding to be inevitable in all of these cases. Indeed it was conceded on behalf of the Secretary of State at the hearing of these appeals that the Secretary of State would not have sought the permission of the court to make the orders if he or his predecessor had realised that the law was as AF(No.3) has now declared it to be.

It must be plain from what I have said that, in my judgment, the appropriate remedy in all these cases is one of quashing ab initio as held by Silber J and not simply revocation as determined Mitting J. I have not found it necessary to revisit all the issues canvassed by Silber J. Some of them did not feature in the submissions in this court. Moreover, the conclusion I have reached is based substantially on a construction of the statute and on general principles. I agree with the submission made on behalf of the controlees that, if the appropriate remedy were merely revocation, there is a risk that a breach of Convention rights would go substantially unremedied. This is best illustrated by the situation in AN, who still faces prosecution and potential imprisonment for breach of his control order prior to revocation. I say nothing about the prospective damages claims of controlees because they are not before us and it is said on behalf of the Secretary of State that they raise difficult issues at common law and under the Human Rights Act. I expressly reject the suggestion that this case is conceptually different from, and materially less serious than, an Article 5 case involving a deprivation of liberty. It seems to me that any difference is simply one of degree. In R v Governor of Brockhill Prison, ex parte Evans (No.2) [2001] 2 AC 19, in the context of an erroneous understanding of the release date of a prisoner which was based on earlier judicial decisions, Lord Steyn said (at page 28H-29A):

“In Eshugbayi Eleko v Officer Administering the Government of Nigeria [1931] AC 662, a habeas corpus case, Lord Atkin observed, at p.670, that ‘no member of the executive can interfere with the liberty or property of a British subject except on the condition that he can support the legality of his action before a court of justice’. Recently, with the approval of the other members of the House, I cited Lord Atkin’s observations in the Eleko case: Boddington v British Transport Police [1999] 2 AC 143, 173F. It represents the traditional common law view.”

I consider that the same principle applies here and is not to be confined to full deprivation of liberty cases.

For all these reasons, I would allow the appeal of AN and would quash his control order ab initio and I would dismiss the appeal of the Secretary of State in AF and AE.

I agree.

I also agree.