United Kingdom Supreme Court

Summary




Hilary Term
[2011] UKSC 12
On appeal from: [2010] EWCA Civ 111

JUDGMENT

Walumba Lumba (previously referred to as WL) (Congo) 1 and 2 (Appellant) v Secretary of State for the Home Department (Respondent)

Kadian Mighty (previously referred to as KM) (Jamaica) (Appellant) v Secretary of State for the Home Department (Respondent)

before

Lord Phillips, President
Lord Hope, Deputy President
Lord Rodger
Lord Walker
Lady Hale
Lord Brown
Lord Collins
Lord Kerr
Lord Dyson

JUDGMENT GIVEN ON

23 March 2011

Heard on 15, 16, 17 and 18 November 2010


 


Appellant (WL) Raza Husain QC Laura Dubinsky Tom Hickman Alex Goodman (Instructed by Public Law Project)		Respondent Michael Beloff QC Robin Tam QC Charles Bourne Jeremy Johnson (Instructed by Treasury Solicitors)
Appellant (KM) Raza Husain QC Martin Westgate QC Alex Goodman (Instructed by Lawrence Lupin Solicitors)		Respondent Michael Beloff QC Robin Tam QC Charles Bourne Jeremy Johnson (Instructed by Treasury Solicitors)
Intervener (JUSTICE) Rabinder Singh QC Elizabeth Prochaska (Instructed by Freshfields Bruckhaus Deringer LLP)		Intervener (Bail for Immigration Detainees) Michael Fordham QC Graham Denholm (Instructed by Allen & Overy LLP)

LORD DYSON

Introduction

1. These two cases raise a number of important issues in relation to the detention pending deportation of foreign national prisoners ("FNPs") following the completion of their sentences of imprisonment. Section 3(5)(a) of the Immigration Act 1971 ("the 1971 Act") provides that a person who is not a British citizen is liable to deportation from the United Kingdom if the Secretary of State "deems his deportation to be conducive to the public good". Schedule 3 to the 1971 Act provides in certain specified circumstances for the detention of such a person pending his deportation.
2. Walumba Lumba is a citizen of the Democratic Republic of Congo ("DRC") who entered the UK on 10 April 1994. He was convicted of a number of offences culminating in an offence of wounding with intent for which he was sentenced to 4 years' imprisonment on 12 January 2004. On 3 April 2006, the Secretary of State informed Mr Lumba of his intention to deport him under section 3(5)(a) of the 1971 Act. He was due to be released from prison on 23 June 2006, but by letter dated 22 June 2006 was notified that he was to be detained pending deportation. He left the United Kingdom voluntarily on 13 February 2011.
3. Kadian Mighty is a citizen of Jamaica. He arrived in the United Kingdom on 4 December 1992 and was given 6 months' leave to enter as a visitor. Thereafter, he made various unsuccessful applications for leave to remain. On 10 February 2003, however, he was granted indefinite leave to remain. He had been convicted of a number of offences, including possession of a Class A drug with intent to supply, for which on 27 June 2003, he was sentenced to 42 months' imprisonment. Following his release on licence, he committed a driving offence and was recalled to prison. He was finally released on 31 March 2006. On 10 May 2006, the Secretary of State informed Mr Mighty of his intention to deport him under section 3(5)(a) of the 1971 Act. On 19 May 2006, he was detained pending deportation because he was likely to abscond and his release was not conducive to the public good. He was released on bail on 28 July 2008.
4. Schedule 3 of the 1971 Act provides, so far as material:

"2. (1) Where a recommendation for deportation made by a court is in force in respect of any person, and that person is not detained in pursuance of the sentence or order of any court, he shall, unless the court by which the recommendation is made otherwise directs, or a direction is given under sub-paragraph (1A) below, be detained pending the making of a deportation order in pursuance of the recommendation, unless the Secretary of State directs him to be released pending further consideration of his case or he is released on bail.

(1A) Where--

(a) a recommendation for deportation made by a court on conviction of a person is in force in respect of him; and

(b) he appeals against his conviction or against that recommendation,

the powers that the court determining the appeal may exercise include power to direct him to be released without setting aside the recommendation.

(2) Where notice has been given to a person in accordance with regulations under section 105 of the Nationality, Immigration and Asylum Act 2002 (notice of decision) of a decision to make a deportation order against him, and he is not detained in pursuance of the sentence or order of a court, he may be detained under the authority of the Secretary of State pending the making of the deportation order.

(3) Where a deportation order is in force against any person, he may be detained under the authority of the Secretary of State pending his removal or departure from the United Kingdom (and if already detained by virtue of sub-paragraph (1) or (2) above when the order is made, shall continue to be detained unless he is released on bail or the Secretary of State directs otherwise)".

5. Between April 2006 and 9 September 2008, the Secretary of State's published policy on detention of FNPs under her immigration powers was that there was a "presumption" in favour of release, although detention could be justified in some circumstances. In fact, during this period the Secretary of State applied a quite different unpublished policy which was described as a "near blanket ban" by the Secretary of State, Ms Jacqui Smith, to the Prime Minister, Gordon Brown, on 19 September 2007 in a document entitled "Bail Proposal for Foreign National Prisoners" in which she said:

"Since April 2006, the BIA [("the Border and Immigration Agency")] has been applying a near blanket ban on release, regardless of whether removal can be achieved and the level of risk to the public linked to the nature of the FNP's original offence. By currently having no discretion to grant bail, the BIA has to regularly transfer FNPs around the Estate."

6. On 9 September 2008, the Secretary of State published a policy which included a "presumption" of detention and withdrew all references to a presumption of release. On 22 January 2009, following the decision of Davis J in the current proceedings, this policy was amended again to omit the reference to a presumption of detention and substitute a policy in favour of release from detention. It will be necessary to describe the policies and practices adopted from time to time in more detail later in this judgment.

The proceedings

7. Mr Lumba issued proceedings on 18 October 2007. He challenged the lawfulness of his detention on the grounds that he was no longer being detained pending deportation and that his continued detention was in breach of the principles stated by Woolf J in R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704 ("the Hardial Singh principles"). He also claimed a declaration that his detention was unlawful, a mandatory order that he be released and damages. On 4 July 2008, Collins J gave an interlocutory judgment on part of the claim: [2008] EWHC 2090 (Admin). He did not make a decision in relation to Mr Lumba's past detention and reserved for a further hearing inter alia the questions of whether the operation of an unpublished policy had been unlawful and the past detention had been unlawful as a consequence. On 28 July 2008, Mr Lumba's claim was joined to four other cases in which the same points arose. One of these was the claim of Mr Mighty which had been issued on 29 May 2008.
8. The five cases were heard by Davis J on 11-14 November 2008. In an impressive judgment given on 19 December 2008 [2008] EWHC 3166 (Admin), he granted the claimants declarations that (i) paragraph 2 of Schedule 3 to the 1971 Act prohibits the Secretary of State from operating any policy in relation to the detention of FNPs which contains a presumption in favour of detention and (ii) it was unlawful for the Secretary of State to operate the policy introduced in April 2006 in that it was not sufficiently published or accessible until its publication on 9 September 2008. He dismissed the other claims, in particular the claims for damages for unlawful detention.
9. The appellants appealed and the Secretary of State cross-appealed against the first declaration. In a judgment of the court delivered by Stanley Burnton LJ, the Court of Appeal (Lord Neuberger MR, Carnwath and Stanley Burnton LJJ) ([2010] 1 WLR 2168) allowed the cross-appeal and set aside the first declaration. They also varied the second declaration. Otherwise the appeals were dismissed.

The issues

10. The principal issues are as follows. (i) Were the detention policies that were applied to the appellants after April 2006 unlawful because (a) they were blanket policies (para 21 below) and/or (b) they were inconsistent with the published policies (para 26 below) and/or (c) they were not published policies (paras 27-38 below) and/or (d) they contained a "presumption" in favour of detention (paras 40-55 below)? (ii) If unlawful policies were applied to the appellants, was their detention unlawful in consequence (paras 56-89)? (iii) If their detention was unlawful, are the appellants entitled to more than nominal damages (paras 90-101 below)? (iv) Is Mr Lumba entitled to damages for unlawful detention on the grounds that, in his case, there has been a breach of the Hardial Singh principles? (paras 102-148 below) (v) Are the appellants entitled to an award of exemplary damages (paras 150-168 below)?

The policies in more detail

The published policies

11. The "presumption" of release had been entrenched in the Secretary of State's published policies since at least 1991. It appeared in the White Paper Fairer, Faster and Firmer: a Modern Approach to Immigration and Asylum (1998) (Cm 4018), which was published in 1998 and again in 2002 in the White Paper Secure Borders, Safe Haven: Integration Diversity in Modern Britain (2002) (Cm 5387) which stated at para 4.76:

"Our 1998 White paper set out the criteria by which Immigration Act powers of detention were exercised and confirmed that the starting point in all cases was a presumption in favour of granting temporary admission or release. The criteria were modified in March 2000 to include detention at Oakington Reception Centre if it appeared that a claimant's asylum application could be decided quickly. The modified criteria and the general presumption remain in place."

12. Chapter 38 of the Operational Enforcement Manual ("OEM"), which was a published document in force until April 2008, stated in its introductory section that the 1998 White Paper confirmed that "there was a presumption in favour of temporary admission or release and that, whenever possible, we would use alternatives to detention". Para 38.3 stated:

"1. There is a presumption in favour of temporary admission or temporary release.

2. There must be strong grounds for believing that a person will not comply with conditions of temporary admission or temporary release for detention to be justified.

3. All reasonable alternatives to detention must be considered before detention is authorised."

13. Identical wording was contained in Chapter 55 of the Enforcement Instructions and Guidance ("EIG") which replaced Chapter 38 of the OEM and came into force on 19 June 2008.
14. On 9 September 2008, Chapter 55 of the EIG was amended. With regard to FNPs, para 55.1.2 stated:

"Due to the clear imperative to protect the public from harm and the particular risk of absconding in these cases, the presumption in favour of temporary admission or temporary release does not apply where the deportation criteria are met. Instead the person will normally be detained, provided detention is, and continues to be lawful."

15. The EIG then gave guidance to caseworkers as to the factors which might make further detention unlawful. In particular, it stated that the presumption of detention "will be displaced where legally the person cannot or can no longer be detained because detention would exceed the period reasonably necessary for the purpose of removal." Following the decision of Davis J in the current proceedings, on 22 January 2009 this policy was changed again so as to replace a presumption in favour of detention with a presumption in favour of release from detention.

The unpublished policies

16. The true picture during the period from April 2006 until September 2008 was very different. Following the public disclosure on 25 April 2006 that 1,013 FNPs had been released from prison before consideration had been given to the question of whether they should have been deported, the Secretary of State adopted a new policy which he did not publish. I have already referred at para 5 above to the description of it contained in the 19 September 2007 Bail Proposal as a "near blanket ban".
17. The policy of blanket detention admitted of exceptions only on compassionate grounds. No formal guidance was given to caseworkers to give effect to this policy until on 8 November 2007 they were issued with a document (known as "Cullen 1") which set out criteria and guidance for the identification and release of FNPs who were considered to pose the lowest risks to the public and the lowest risks of absconding. Cullen 1 was not published to the outside world. It led at most to the release of a handful of FNPs. In March 2008, an amended guidance document (known as "Cullen 2") was issued to the caseworkers. It too was not published to the outside world. Attached to Cullen 1 and Cullen 2 was an extensive list of offences entitled "List of recorded crimes where release from immigration detention or at the end of custody will not be appropriate". In practice, almost all FNPs who had been sentenced to imprisonment were likely to have committed one or more of such offences. Both Mr Lumba and Mr Mighty had done so. The evidence of David Wood, Strategic Director of the Criminality and Detention Group, was that between December 2006 and July 2008, 15 FNPs were released from detention.
18. On 22 May 2008, the existence of an unpublished policy or practice was belatedly disclosed by the Secretary of State to Mitting J after he had given judgment in the case of R (Ashori) v Secretary of State for the Home Department [2008] EWHC 1460 (Admin). More detail as to the circumstances in which the policy came to be disclosed is given by Davis J at paras 21 to 26 of his judgment.

The unpublished policies were applied to Mr Lumba and Mr Mighty

19. It is now common ground that the unpublished policies were applied to the two appellants throughout their detention. It is, therefore, unnecessary to consider para 203 of the judgment of Davis J (where the judge stated that there was nothing to show that Mr Lumba was detained by application of the unpublished policy) or para 100 of the judgment of the Court of Appeal which upheld Davis J on this point.

Were these policies unlawful?

20. Here too, there is little dispute between the parties. Mr Beloff QC rightly accepts as correct three propositions in relation to a policy. First, it must not be a blanket policy admitting of no possibility of exceptions. Secondly, if unpublished, it must not be inconsistent with any published policy. Thirdly, it should be published if it will inform discretionary decisions in respect of which the potential object of those decisions has a right to make representations.
21. As regards the first of these propositions, it is a well established principle of public law that a policy should not be so rigid as to amount to a fetter on the discretion of decision-makers. Davis J held that the unpublished policy was not a "blanket" policy. The Court of Appeal disagreed. Basing themselves on the review by David Wood "of the failure to publish a revised FNP detention policy following the April 2006 crisis" approved on 3 August 2009, the Court of Appeal concluded that the policy, as applied at least from the time of Cullen 1 in November 2007, "effectively operated on a blanket basis" rather than (as held by the judge) "one of presumption". I agree with this assessment by the Court of Appeal but would go further. It seems clear to me that a blanket policy was also applied from April 2006 until the introduction of Cullen 1 in November 2007. During this earlier period, the only exceptions made to the policy of universal detention were on compassionate grounds and these were few and far between. Importantly, there were no releases on the basis of Hardial Singh principles. Indeed, Cullen 1 represented a modest relaxation of the previous position.
22. It is convenient to introduce the Hardial Singh principles at this stage, since they infuse much of the debate on the issues that arise on this appeal. It is common ground that my statement in R (I) v Secretary of State for the Home Department [2002] EWCA Civ 888, [2003] INLR 196 para 46 correctly encapsulates the principles as follows:

(i) The Secretary of State must intend to deport the person and can only use the power to detain for that purpose;

(ii) The deportee may only be detained for a period that is reasonable in all the circumstances;

(iii) If, before the expiry of the reasonable period, it becomes apparent that the Secretary of State will not be able to effect deportation within a reasonable period, he should not seek to exercise the power of detention;

(iv) The Secretary of State should act with reasonable diligence and expedition to effect removal.

23. Lord Phillips says that the first two of these principles cannot properly be derived from Hardial Singh. Since their correctness has not been put in issue by the parties to these appeals, I propose to deal with the points shortly. As regards the first principle, I consider that Woolf J was saying unambiguously that the detention must be for the purpose of facilitating the deportation. The passage quoted by Lord Phillips includes the following: "as the power is given in order to enable the machinery of deportation to be carried out I regard the power of detention as being impliedly limited to a period which is reasonably necessary for that purpose"(emphasis added). The first principle is plainly derived from what Woolf J said.
24. As for the second principle, in my view this too is properly derived from Hardial Singh. Woolf J said that (i) the power of detention is limited to a period reasonably necessary for the purpose (as I would say) of facilitating deportation; (ii) what is reasonable depends on the circumstances of the particular case; and (iii) the power to detain ceases where it is apparent that deportation will not be possible "within a reasonable period". It is clear at least from (iii) that Woolf J was not saying that a person can be detained indefinitely provided that the Secretary of State is doing all she reasonably can to effect the deportation.
25. It seems to me that, in relation to both the first and second principles, Lord Phillips is suggesting a different interpretation of paragraph 2(3) of Schedule 3 to the 1971 Act from that enunciated by Woolf J. I do not agree with this interpretation. But what is perhaps of more importance in the context of these appeals is that in my view it is not appropriate to depart from a decision which has been followed repeatedly for almost 30 years unless it is obviously wrong (which I do not believe to be the case), still less to do so without the benefit of adversarial argument.
26. As regards the second proposition accepted by Mr Beloff, a decision-maker must follow his published policy (and not some different unpublished policy) unless there are good reasons for not doing so. The principle that policy must be consistently applied is not in doubt: see Wade and Forsyth Administrative Law, 10^th ed (2009) p 316. As it is put in De Smith's Judicial Review, 6^th ed (2007) at para 12-039:

"there is an independent duty of consistent application of policies, which is based on the principle of equal implementation of laws, non-discrimination and the lack of arbitrariness."

The decision of the Court of Appeal in R (Nadarajah) v Secretary of State for the Home Department [2003] EWCA Civ 1768, [2004] INLR 139 is a good illustration of the principle. At para 68, Lord Phillips MR, giving the judgment of the court, said that the Secretary of State could not rely on an aspect of his unpublished policy to render lawful that which was at odds with his published policy.

27. As for the third proposition, the Court of Appeal dealt with the issue of whether there is a general rule of law that policies must be published at paras 70 to 79 of their judgment. Disagreeing with Davis J, they concluded that there is no such general rule and said that the fact that the appellants were detained pursuant to unpublished policies was not in itself a reason for holding that the decisions to detain them were unlawful. Mr Beloff did not feel able to support this conclusion. It is unfortunate that the Court of Appeal embarked on this topic at all, since it was not before them and was not, therefore, the subject of argument or citation of authority. As the point is of general importance, I need to say why in my view the judge was right and the Court of Appeal were wrong on this issue both as a matter of common law and ECHR law.
28. The Court of Appeal referred to a statement of Sedley LJ in R v Secretary of State for Education and Employment Ex P Begbie [2000] 1 WLR 1115, 1132C that there were "cogent objections to the operation of undisclosed policies affecting individuals' entitlements or expectations" and said at para 72 that they had no difficulty in accepting this as (no more than) a statement "of good administrative practice". They also said that the judge was wrong to rely on Sunday Times v United Kingdom (1979) 2 EHRR 245 and criticised the reasoning in Nadarajah at paras 64-67 which relied on the Sunday Times case in support of the proposition that a relevant policy is part of the law that must be accessible, so as to enable those affected by it reasonably to foresee the consequences of their actions. At para 73, they said that the relevant passage in the judgment of the ECtHR at para 49 of the judgment in the Sunday Times case is "not, as we read it, about policy as such, but is rather directed to the need for accessibility and precision, as requirements of law in the strict sense". They went on to say that, in the present context, the requirement for an accessible and precise statement of the relevant law is satisfied by paragraph 2 of Schedule 3 to the 1971 Act, taken with the Hardial Singh guidelines. In short, policy is not the same as law (para 57).
29. In support for their conclusion, they referred to what Laws LJ said in R (SK Zimbabwe) v Secretary of State for the Home Department [2008] EWCA Civ 1204; [2009] 1 WLR 1527, para 33. In that case, the Secretary of State had failed to carry out regular reviews following detention, as required by the Detention Centre Rules. As regards the requirement that any deprivation of liberty be "in accordance with a procedure prescribed by law" in article 5(1) of the ECHR, Laws LJ said that this was to ensure that any interference is not random and arbitrary, but governed by clear pre-existing rules. He continued:

"Here the 'rules' are the Hardial Singh principles. Their fulfilment in any given case saves a detention from the vice of arbitrariness. A system of regular monitoring is, no doubt, a highly desirable means of seeing that the principles are indeed fulfilled. But it is not itself one of those principles…."

30. But all that the Hardial Singh principles do is that which article 5(1)(f) does: they require that the power to detain be exercised reasonably and for the prescribed purpose of facilitating deportation. The requirements of the 1971 Act and the Hardial Singh principles are not the only applicable "law". Indeed, as Mr Fordham QC points out, the Hardial Singh principles reflect the basic public law duties to act consistently with the statutory purpose (Padfield v Minister of Agriculture, Fisheries and Food [1968] AC 997, 1030B-D) and reasonably in the Wednesbury sense (Associated Provincial Picture House's Ltd v Wednesbury Corporation [1948] 1 KB 223). But they are not exhaustive. If they were exhaustive, there could be no room for the public law duty of adherence to published policy, which was rightly acknowledged by the Court of Appeal at paras 51, 52 and 58 of their judgment. I therefore accept the submission of Mr Husain QC and Mr Fordham that the Court of Appeal's criticisms of Nadarajah were misplaced.
31. I should interpolate that there is in any event an obvious difference between rules which require the review of a detention to be undertaken at prescribed intervals and rules which prescribe the criteria by which a person is to be released or to be subjected to continuing detention. The fact that a policy states that only persons of a specified category will be considered for release is at least as substantively important as the Hardial Singh principles which determine, for example, that a person may not be detained for an unreasonable period.
32. There is further support in the ECtHR jurisprudence for the proposition that paragraph 2 of Schedule 3 to the 1971 Act and the Hardial Singh principles are not exhaustive of the "law". In Medvedyev v France (Appln no 3394/03, 29 March 2010), the Grand Chamber said at para 80: "where deprivation of liberty is concerned it is particularly important that the general principle of legal certainty be satisfied. It is therefore essential that the conditions for deprivation of liberty under domestic and/or international law be clearly defined." The case of Gillan v United Kingdom (2010) 50 EHRR 45 concerned the stop and search powers conferred on the police by the Terrorism Act 2000. For present purposes, the relevant issue was whether the powers were "in accordance with the law" within the meaning of article 8(2) of the ECHR. A Code of Practice was issued by the Secretary of State to guide police officers in the exercise of their powers of stop and search. The ECtHR said:

"77……Consequently, the law must indicate with sufficient clarity the scope of any such discretion conferred on the competent authorities and the manner of its exercise. The level of precision required of domestic legislation---which cannot in any case provide for every eventuality---depends to a considerable degree on the content of the instrument in question, the field it is designed to cover and the number and status of those to whom it is addressed."

33. The ECtHR noted at para 83 that the Code of Practice "governs essentially the mode in which the stop and search is carried out, rather than providing any restriction on the officer's decision to stop and search. That decision is, as the House of Lords made clear, one based exclusively on the 'hunch' or 'professional intuition' of the officer concerned". In the opinion of the court, there was a clear risk of arbitrariness in the grant of such a broad discretion to the police officer. At para 87, they concluded that, despite the existence of the Code of Practice, the statutory powers were not "in accordance with the law" because they were "neither sufficiently circumscribed nor subject to adequate legal safeguards against abuse".
34. The rule of law calls for a transparent statement by the executive of the circumstances in which the broad statutory criteria will be exercised. Just as arrest and surveillance powers need to be transparently identified through codes of practice and immigration powers need to be transparently identified through the immigration rules, so too the immigration detention powers need to be transparently identified through formulated policy statements.
35. The individual has a basic public law right to have his or her case considered under whatever policy the executive sees fit to adopt provided that the adopted policy is a lawful exercise of the discretion conferred by the statute: see In re Findlay [1985] AC 318, 338E. There is a correlative right to know what that currently existing policy is, so that the individual can make relevant representations in relation to it. In R (Anufrijeva) v Secretary of State for the Home Department [2003] UKHL 36, [2004] 1 AC 604, para 26 Lord Steyn said:

"Notice of a decision is required before it can have the character of a determination with legal effect because the individual concerned must be in a position to challenge the decision in the courts if he or she wishes to do so. This is not a technical rule. It is simply an application of the right of access to justice. "

36. Precisely the same is true of a detention policy. Notice is required so that the individual knows the criteria that are being applied and is able to challenge an adverse decision. I would endorse the statement made by Stanley Burnton J in R (Salih) v Secretary of State for the Home Department [2003] EWHC 2273 (Admin) at para 52 that "it is in general inconsistent with the constitutional imperative that statute law be made known for the government to withhold information about its policy relating to the exercise of a power conferred by statute." At para 72 of the judgment of the Court of Appeal in the present case, this statement was distinguished on the basis that it was made "in the quite different context of the Secretary of State's decision to withhold from the individuals concerned an internal policy relating to a statutory scheme designed for their benefit". This is not a satisfactory ground of distinction. The terms of a scheme which imposes penalties or other detriments are at least as important as one which confers benefits. As Mr Fordham puts it: why should it be impermissible to keep secret a policy of compensating those who have been unlawfully detained, but permissible to keep secret a policy which prescribes the criteria for their detention in the first place?
37. There was a real need to publish the detention policies in the present context. As Mr Husain points out, the Cullen policies provided that certain non-serious offenders could be considered for release. The failure to publish these policies meant that individuals who may have been wrongly assessed as having committed a crime that rendered them ineligible for release would remain detained, when in fact, had the policy been published, representations could have been made that they had a case for release.
38. The precise extent of how much detail of a policy is required to be disclosed was the subject of some debate before us. It is not practicable to attempt an exhaustive definition. It is common ground that there is no obligation to publish drafts when a policy is evolving and that there might be compelling reasons not to publish some policies, for example, where national security issues are in play. Nor is it necessary to publish details which are irrelevant to the substance of decisions made pursuant to the policy. What must, however, be published is that which a person who is affected by the operation of the policy needs to know in order to make informed and meaningful representations to the decision-maker before a decision is made.
39. For all these reasons, the policies which were applied to Mr Lumba and Mr Mighty were unlawful. But Mr Husain submits (with the support of Mr Rabinder Singh QC and Mr Fordham) that the policies were also unlawful because they included a "presumption" of detention.

Presumption of detention

40. Davis J held at paras 114 to 116 of his judgment that the provisions of paragraph 2 of Schedule 3 to the 1971 Act operate to prevent the Secretary of State from operating a policy of a presumption in favour of detention of FNPs pending deportation. He applied R (Sedrati) v Secretary of State for the Home Department [2001] EWHC Admin 418 in which, by consent, Moses J had granted a declaration that the terms of paragraph 2 of Schedule 3 do "not create a presumption in favour of detention upon completion of the sentence". On the Secretary of State's cross-appeal against the declaration, the Court of Appeal said at para 65:

"....... there is no reason in principle why paragraph 2.1 of Schedule 3 to the 1971 Act, which clearly does require continued detention unless the Secretary of State otherwise orders (i.e. a presumption of detention), should not be construed as a presumption of detention pending deportation. Equally, the Secretary of State may lawfully adopt a policy for the purposes of paragraph 2(2) or (3) that involves a presumption. A presumption that those who have committed serious crimes (e.g. most of those listed in Cullen 1 and 2) should be detained is unobjectionable. "

They went on at para 66 to say that for these reasons the declaration granted by Moses J was wrong and allowed the Secretary of State's cross-appeal.

41. Mr Husain, supported by Mr Singh and Mr Fordham, say that the judge was right and the Court of Appeal wrong on this issue. The primary case advanced by Mr Husain is that the policy that was applied between April 2006 and September 2008 was not properly described as a "presumptive policy" at all, but rather was a blanket policy. But whether that is right or not, Mr Lumba continued to be detained between September 2008 (when the Secretary of State published a policy in favour of detention) and 22 January 2009 (when the order of Davis J was implemented). It follows that even if, as I have accepted, the Court of Appeal was justified in holding that the policy was a blanket policy until September 2008, the "presumption of liberty" issue is of more than academic interest in this appeal.
42. It is important at the outset to define clearly what a "presumption" means in this context. It is the Secretary of State's case that paragraph 2 of Schedule 3 to the 1971 Act permits the operation of a policy in which she states that a FNP will normally be detained in certain prescribed circumstances. Such a policy serves as a guide to the caseworkers who make the decisions on the ground and as an indication to the FNPs of what they can normally expect in the circumstances specified in the policy. I shall refer to such a policy as "normal practice". It need not, and usually does not say anything about the burden of proof. Normal practice is to be distinguished from presumptions in the strict sense. Phipson on Evidence 17^th ed (2009) at 6-16 to 6-31 categorises "presumptions" in this sense into rebuttable presumptions of law, irrebuttable presumptions of law and rebuttable presumptions of fact. Such a presumption usually regulates the burden of proof in legal proceedings. Thus, a presumption that a deprivation of liberty is unlawful regulates the burden of proof in relation to that issue: the burden is on the detainer to show that there was a power to detain. I shall refer to a presumption in the strict sense as a "legal presumption".
43. The distinction between normal practice and a legal presumption is fundamental to the present issue. The fact that in legal proceedings the burden of proving a certain issue is allocated to one party rather than the other does not assist in deciding whether the Secretary of State may, in principle, lawfully give guidance that when certain factors are present, the decision should normally be to detain. This distinction was not articulated in the courts below.
44. A further preliminary point needs to be made. The legality of a decision may be considered at two stages: first at the administrative stage when the decision is taken and secondly, if the decision is challenged, at the stage of legal proceedings. At the administrative stage, the individual against whom the decision is taken often plays no part. It is not appropriate to talk of a burden of proof at this stage: see, for example, R v Lichniak [2003] 1 AC 903 at para 16 per Lord Bingham. At the stage of legal proceedings, the Secretary of State rightly accepts that the burden of proof is on her to justify the detention. This has long been established: Allen v Wright (1838) 8 C & P 522 and Lord Atkin's dissenting speech in Liversidge v Anderson [1942] AC 206, 245 "every detention is prima facie unlawful and that it is for a person directing imprisonment to justify his act."
45. Mr Husain submits that there is a "presumption of liberty" both under the jurisprudence of the ECtHR or at common law. I shall start with the jurisprudence on article 5 of the ECHR which Mr Husain submits establishes that there is such a presumption. He refers to Ilijkov v Bulgaria (Application No 33977/96) (unreported) 26 July 2001 where the ECtHR said:

"84. The court reiterates that continued detention can be justified in a given case only if there are specific indications of a genuine requirement of public interest which, notwithstanding the presumption of innocence, outweighs the rule of respect for individual liberty. Any system of mandatory detention on remand is per se incompatible with article 5(3) of the Convention…..Where the law provides for a presumption in respect of factors relevant to the grounds for continued detention…..the existence of the concrete facts outweighing the rule of respect for individual liberty must be nevertheless convincingly demonstrated.

85. Moreover, the court considers that it was incumbent on the authorities to establish those relevant facts. Shifting the burden of proof to the detained person in such matters is tantamount to overturning the rule of article 5 of the Convention, a provision which makes detention an exceptional departure from the right to liberty and one that is only permissible in exhaustively enumerated and strictly defined cases…. "

46. This was a decision in relation to an alleged violation of article 5(3) which provides:

"Everyone arrested or detained in accordance with the provisions of paragraph 1(c) of this article shall be brought promptly before a judge or other officer authorised by law to exercise judicial power and shall be entitled to trial within a reasonable time or to release pending trial…."

47. Mr Husain submits that, although these principles were articulated in the context of detention pending trial, they are more widely applicable as expressions of the right to liberty protected by article 5: see also Bykov v Russia (Application No 4378/02, 10 March 2009) at para 61 and Bordikov v Russia (Application no 921/03, 9 October 2009) at para 88. However, these cases only concern legal presumptions that regulate burdens of proof in legal proceedings. They are not concerned with normal practice contained in a policy of the kind with which these appeals are concerned.
48. In Ilijkov, the national courts rejected a series of applications for bail pending trial. They did so relying on the Bulgarian Code of Criminal Procedure, which provided that, for certain crimes, detention on remand was mandatory in the absence of exceptional circumstances. The ECtHR held that the initial decision to detain was lawful, but that the continuing application of the presumption of detention by the national judicial authorities was unlawful (paras 78-9 and 87). The case was not concerned with the lawfulness of any decision to detain taken at an administrative stage. It is clear from para 84 of the judgment that the court held that there was a breach of article 5(3). There is, however, no provision in article 5(1)(f) corresponding with article 5(3) and there is nothing to indicate that the court intended its ruling in relation to article 5(1)(c) and 5(3) to apply to article 5(1)(f).
49. The decisions in Bykov and Bordikov do not advance the argument. Para 61 of the judgment in Bykov merely reiterates what the court has repeatedly said in relation to article 5(3). The reference to the presumption of release under article 5 is a reference to the second limb of article 5(3). The case is concerned with the decisions of "judicial authorities" and not administrative decision-makers. Similarly, para 88 of the judgment in Bordikov makes it clear that the court in that case too was only concerned with the decisions of the courts.
50. The principal ECtHR authority on article 5(1)(f) is Saadi v UK (2008) 47 EHRR 17. The applicant sought judicial review of the decision to detain him for a short period while his asylum claim was being subject to fast-track processing. The decision was made pursuant to a policy under which all asylum claimants falling within defined criteria (usually by nationality) were normally detained at Oakington while their claims were determined in an accelerated process. This was a normal practice case and not a case about legal presumptions. The ECtHR held inter alia that deprivation of liberty must not be "arbitrary". It must comply with the substantive and procedural rules of national law and the detention must be in good faith. At para 69, the court said:

"The condition that there be no arbitrariness further demands that both the order to detain and the execution of the detention must genuinely conform with the purpose of the restrictions permitted by the relevant sub-paragraph of article 5(1). There must in addition be some relationship between the ground of permitted deprivation of liberty relied on and the place and conditions of detention."

51. And at para 72:

"Similarly, where a person has been detained under article 5(1)(f), the Grand Chamber, interpreting the second limb of this sub-paragraph, held that, as long as a person was being detained 'with a view to deportation', that is, as long as 'action [was] being taken with a view to deportation,' there was no requirement that the detention be reasonably considered necessary, for example, to prevent the person concerned from committing an offence or fleeing. The Grand Chamber further held in Chahal (1997) 23 EHRR 413 that the principle of proportionality applied to detention under article 5(1)(f) only to the extent that the detention should not continue for an unreasonable length of time…"

52. It was this statement which led the Court of Appeal to conclude at para 62 that "a national law that authorises detention with a view to deportation may be compatible with article 5 even if it imposes a presumption of detention pending deportation." It is not clear in what sense the Court of Appeal was using the phrase "presumption of detention" in this passage. But if it is construed as a reference to normal practice, I agree with it, provided that it requires the Government to justify the detention with reasons that are closely connected to the statutory purpose of effecting deportation.
53. I turn to the position at common law. It is not in dispute that the right to liberty is of fundamental importance and that the courts should strictly and narrowly construe general statutory powers whose exercise restricts fundamental common law rights and/or constitutes the commission of a tort. The Court of Appeal said at para 63 that there is no rule of our domestic law "that precludes the application of a presumption in favour of detention pending deportation, subject, of course, to the limitations in Ex P Hardial Singh [1984] 1 WLR 704, none of which involves consideration of risk of reoffending or absconding". Such risks, they said, are relevant to the reasonableness of the period during which it is lawful to detain a FNP, but the absence of such a risk "does not of itself render detention unlawful". If by "presumption in favour of detention" the Court of Appeal meant "the normal practice as to the circumstances in which a FNP will be detained", then I would agree with them, provided that it is understood that (i) the Hardial Singh principles are observed and (ii) each case is considered individually.
54. The Court of Appeal set aside the declaration granted by Davis J. They were right to do so. For the reasons I have given, it is lawful for the Secretary of State to operate a policy which sets out the practice that she will normally follow in deciding whether or not to detain FNPs pending their deportation provided that the criteria that I have set out at para 53 above are satisfied. Such a policy is not prohibited by paragraph 2 of Schedule 3 to the 1971 Act.
55. The Court of Appeal also held at para 66 that the declaration granted by Moses J in Sedrati [2001] EWHC Admin 418 was wrong in law. I find this somewhat puzzling. The declaration stated that the terms of paragraph 2 of Schedule 3 do not create a presumption in favour of detention. Whatever the position may be in relation to paragraph 2(1) and the parenthesis in paragraph 2(3), paragraph 2(2) and the remainder of paragraph 2(3) do not create any presumption at all. They simply give the Secretary of State a discretion to detain. In relation to paragraph 2(2) and (3), therefore, so far as it goes, the declaration granted by Moses J is correct.

Were the detentions unlawful?

56. In summary, the appellants' case is that their claims in false imprisonment should have succeeded: the Secretary of State's unlawful unpublished policy which operated between April 2006 and September 2008 influenced the initial decisions to detain them and the subsequent decisions to continue to detain them. Davis J accepted the argument advanced on behalf of the Secretary of State that, where the unlawful policy was of no causative effect because the claimants could and would have been lawfully detained if the published policy had been applied, their detention was not unlawful.
57. The Court of Appeal agreed. They distinguished Christie v Leachinsky [1947] AC 573 and Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662 on the footing that in those cases there was no lawful authority to detain the plaintiff. In the present case, however, they said at para 87:

"there is no doubt that the statutory powers relied on by the Secretary of State were apt for the purpose, and the case is not based on the breach of any specific regulation on which the legality of detention was dependent. Rather it is about the manner in which the power was exercised."

58. And at para 89:

"The mere existence of an internal, unpublished policy or practice at variance with, and more disadvantageous to the FNP than, the published policy will not render a decision to detain unlawful. It must be shown that the unpublished policy was applied to him. Even then, it must be shown that the application of the policy was material to the decision. If the decision to detain him was inevitable, the application of the policy is immaterial, and the decision is not liable to be set aside as unlawful."

59. In short, since Mr Lumba and Mr Mighty would inevitably have been detained even if the published policy had been applied to them, their detentions were lawful. The court therefore applied what it is convenient to call "the causation test".
60. Davis J and the Court of Appeal were right to hold that the detention of the appellants would have been inevitable in the light of the risk of absconding and re-offending that they both posed. This appeal therefore raises the important question of whether it was right to apply the causation test and for that reason to hold that the detentions were lawful.
61. A somewhat similar problem arose in R (SK Zimbabwe) v Secretary of State for the Home Department. In that case the unlawfulness lay in the failure of the Secretary of State to comply with her policy which prescribed the procedural requirements for reviews of FNPs who are already in detention. The present case concerns the substantive requirements for the initial detention of FNPs as well as their continued detention.
62. What follows is to a considerable extent based on the submissions of Mr Husain. The introduction of a causation test in the tort of false imprisonment is contrary to principle both as a matter of the law of trespass to the person and as a matter of administrative law. Neither body of law recognises any defence of causation so as to render lawful what is in fact an unlawful authority to detain, by reference to how the executive could and would have acted if it had acted lawfully, as opposed to how it did in fact act. The causation test entails the surprising proposition that the detention of a person pursuant to a decision which is vitiated by a public law error is nevertheless to be regarded as having been lawfully authorised because a decision to detain could have been made which was not so vitiated. In my view, the law of false imprisonment does not permit history to be rewritten in this way.
63. The Court of Appeal were right to say at para 89 that the mere existence of an unlawful policy is not sufficient to establish that any particular exercise of a statutory discretion is unlawful. The decision to detain and/or continue detention will not be vitiated on the grounds of an unlawful policy unless the policy has been applied or at least taken into account by the decision maker. But this does not shed any light on the correctness of the causation test.
64. Trespassory torts (such as false imprisonment) are actionable per se regardless of whether the victim suffers any harm. An action lies even if the victim does not know that he was imprisoned: see, for example, Murray v Ministry of Defence [1988] 1 WLR 692, 703A where Lord Griffiths refused to redefine the tort of false imprisonment so as to require knowledge of the confinement or harm because "The law attaches supreme importance to the liberty of the individual and if he suffers a wrongful interference with that liberty it should remain actionable even without proof of special damage." By contrast, an action on the case (of which a claim in negligence is the paradigm example) regards damage as the essence of the wrong.
65. All this is elementary, but it needs to be articulated since it demonstrates that there is no place for a causation test here. All that a claimant has to prove in order to establish false imprisonment is that he was directly and intentionally imprisoned by the defendant, whereupon the burden shifts to the defendant to show that there was lawful justification for doing so. As Lord Bridge said in R v Deputy Governor of Parkhurst Prison, Ex p Hague [1992] 1 AC 58, 162C-D: "The tort of false imprisonment has two ingredients: the fact of imprisonment and the absence of lawful authority to justify it."
66. The causation test shifts the focus of the tort on to the question of how the defendant would have acted on the hypothesis of a lawful self-direction, rather than on the claimant's right not in fact to be unlawfully detained. There is no warrant for this. A purported lawful authority to detain may be impugned either because the defendant acted in excess of jurisdiction (in the narrow sense of jurisdiction) or because such jurisdiction was wrongly exercised. Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147 established that both species of error render an executive act ultra vires, unlawful and a nullity. In the present context, there is in principle no difference between (i) a detention which is unlawful because there was no statutory power to detain and (ii) a detention which is unlawful because the decision to detain, although authorised by statute, was made in breach of a rule of public law. For example, if the decision to detain is unreasonable in the Wednesbury sense, it is unlawful and a nullity. The importance of Anisminic is that it established that there was a single category of errors of law, all of which rendered a decision ultra vires: see Boddington v British Transport Police [1999] 2 AC 143, 158D-E.
67. Mr Beloff submits that there are inherent problems in what I consider to be the correct and principled approach. First, it fails to take account of the nature or extent of the public law error. For example, he suggests that it treats for the purposes of liability as equally unlawful a decision to detain made by an official one grade lower than that specified in the detention policy (but which is otherwise unimpeachable) and a decision to detain for which there is no statutory authority at all. Secondly, it allows what is in essence a public law challenge to be made under the guise of a private law action without any of the procedural safeguards which apply in a judicial review application. In particular, the normal time limits for judicial review proceedings are circumvented. Thirdly, judicial review is a discretionary remedy. A minor public law error may result in no substantive relief being granted at all in judicial review proceedings, whereas a claimant can bring proceedings for false imprisonment as of right.
68. I do not consider that these arguments undermine what I have referred to as the correct and principled approach. As regards Mr Beloff's first point, the error must be one which is material in public law terms. It is not every breach of public law that is sufficient to give rise to a cause of action in false imprisonment. In the present context, the breach of public law must bear on and be relevant to the decision to detain. Thus, for example, a decision to detain made by an official of a different grade from that specified in a detention policy would not found a claim in false imprisonment. Nor too would a decision to detain a person under conditions different from those described in the policy. Errors of this kind do not bear on the decision to detain. They are not capable of affecting the decision to detain or not to detain.
69. Lord Walker and Lord Hope would prefer the more demanding test of the wrongful use of a statutory power amounting to an abuse of power. It is true that the phrase "abuse of power" is used in certain contexts in public law. For example, it has been held that the court will in a proper case decide whether to frustrate the legitimate expectation of a substantial benefit "is so unfair that to take a different course will amount to an abuse of power": see R v North and East Devon Health Authority, Ex p Coughlan [2001] QB 213 para 57. In that context, the phrase "abuse of power" denotes a degree of unfairness. It is not clear to me in what sense the phrase "abuse of power" is being suggested in the present context. Suppose that a detention policy states that no FNP who has been sentenced to less than 12 months' imprisonment is to remain in detention pending deportation for more than 6 months. Suppose further that, by an administrative oversight, a FNP who has been sentenced to 9 months' imprisonment is detained for 12 months. There can be little doubt that the FNP would have a good claim for substantial damages for false imprisonment in respect of the period of 6 months when he should not have been detained. It would be odd to say that his detention during that period was the result of an "abuse of power". I would say that the FNP would be the victim of a material public law error. The error was one which was relevant to the decision to detain him. It was capable of affecting the decision to continue to detain him and did in fact do so.
70. As for Mr. Beloff's other points, such force as they have derives from the fact that the detention in these cases is unlawful because it is vitiated by a public law error. The significance and effect of that error cannot be affected by the fortuity that it is also possible for a victim to challenge the decision by judicial review proceedings (which are subject to tighter time limits than private law causes of action) and that judicial review is a discretionary remedy. It is well established that a defendant can rely on a public law error as a defence to civil proceedings and that he does not need to obtain judicial review as a condition for defending the proceedings: see, for example, Wandsworth London Borough Council v Winder [1985] AC 461. The same applies in the context of criminal proceedings: see Boddington v British Transport Police [1999] 2 AC 143. Mr Beloff submits that the position of a claimant who relies on a public law error to found his cause of action and a defendant can sensibly be differentiated. But it is difficult to see how or why.
71. I can see that at first sight it might seem counter-intuitive to hold that the tort of false imprisonment is committed by the unlawful exercise of the power to detain in circumstances where it is certain that the claimant could and would have been detained if the power had been exercised lawfully. But the ingredients of the tort are clear. There must be a detention and the absence of lawful authority to justify it. Where the detainer is a public authority, it must have the power to detain and the power must be lawfully exercised. Where the power has not been lawfully exercised, it is nothing to the point that it could have been lawfully exercised. If the power could and would have been lawfully exercised, that is a powerful reason for concluding that the detainee has suffered no loss and is entitled to no more than nominal damages. But that is not a reason for holding that the tort has not been committed.
72. Both Mr Husain and Mr Beloff have referred to much authority. I shall refer to some of it. But there is nothing in the cases which shows that the conclusion which I consider is dictated by principle is questionable, still less that it is wrong.
73. Mr Husain relies on dicta of Lord Diplock in Holgate-Mohamed v Duke [1984] AC 437. Lord Diplock recognised that a claim for false imprisonment may be made out where police powers of arrest are unlawfully exercised by reference to common law principles. The statutory power for an arrest without warrant (section 2(4) of the Criminal Law Act 1967)) made it a "condition precedent" that the constable should have reasonable cause to suspect the person to be guilty of the arrestable offence in respect of which the arrest was being made. On the facts, the condition precedent was made out. Lord Diplock said at p 443B that this left the officer with an executive discretion whether to arrest or not. The lawfulness of the way in which the discretion had been exercised could not be challenged except on Wednesbury grounds. He then continued:

"The Wednesbury principles…are applicable to determining the lawfulness of the exercise of the statutory discretion of a constable under section 2(4) of the Criminal Law Act 1967, not only in proceedings for judicial review but also for the purpose of founding a cause of action at common law for damages for that species of trespass to the person known as false imprisonment, for which the action in the instant case is brought." (emphasis added)

74. Lord Diplock then applied the Wednesbury principles and concluded that the officer's action was not unlawful. It follows that his comments about false imprisonment were obiter dicta. Nevertheless, it is clear that, if he had concluded that the officer had exercised his statutory discretion unlawfully, Lord Diplock would have held that he was liable in tort for false imprisonment. I accept, however, that these are no more than dicta, albeit from a source of high authority, and that the issue does not seem to have been the subject of much if any argument in the House of Lords.
75. Other authorities relied on by Mr Husain as rejecting the causation test include Christie v Leachinsky [1947] AC 573, Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662 and Langley v Liverpool City Council [2006] 1 WLR 375. In addition, Mr Singh relies on Cooper v The Board of Works for the Wandsworth District (1863) 14 C.B. (N.S) 180M.
76. In Christie v Leachinsky, Viscount Simon explained that where an arrest was unlawful because it did not comply with the procedural requirements imposed by the common law—communication of the true and good ground of arrest to the detainee—there would be a false imprisonment notwithstanding that the arrest could have been effected in a proper manner. At p 588H he said:

"I entertain no doubt that in the present case the appellants are not exonerated from liability for false imprisonment by satisfying the judge that they had a reasonable suspicion that the respondent had been guilty of theft or of receiving stolen goods knowing they had been stolen, when they never told the respondent that this was the ground of his arrest. Instead of doing so, they gave a different ground which, as Christie admitted, was not a good excuse for arresting him at all."

77. Mr Beloff submits that this case should be distinguished on the basis that it concerned the giving of reasons for detention which was a condition precedent to a lawful arrest. He argues that it says nothing about the causation test in cases where the alleged error is not a failure to satisfy a condition precedent. I do not see why the failure to provide a detainee with the reasons for the arrest should be regarded as a failure to satisfy a condition precedent to lawful arrest rather than an unlawful exercise of the power to arrest. In any event, it would be remarkable if the question whether a cause of action in false imprisonment exists should depend on such fine distinctions of classification. More fundamentally, such distinctions have no justification in the light of Anisminic.
78. In Roberts v Chief Constable of the Cheshire Constabulary, the Court of Appeal held that a failure by the custody officer to conduct a review as required by section 40 of the Police and Criminal Evidence Act 1984 rendered the plaintiff's continued detention unlawful until the next review. The defence was raised that the plaintiff could only prove false imprisonment if he could show that, if the review had been carried out at the appropriate time, he would have been released. This "causation defence" was rejected by Clarke LJ (with whom Stuart-Smith and Schiemann LJJ agreed) at p 667B as being "nothing to the point".
79. Mr Beloff emphasises the fact that the plaintiff was not being detained in accordance with the relevant statutory provisions and that the statute stipulated an express condition precedent to the lawful continuation of the detention, namely a review of detention, and that condition was not satisfied. This argument has no more force than Mr Beloff's corresponding argument in relation to Christie.
80. Langley v Liverpool City Council [2006] 1 WLR 375 concerned child protection. The Court of Appeal held that a constable who had wrongfully removed a child under section 46 of the Children Act 1989 was liable in false imprisonment. He should instead have facilitated the exercise of a different power of removal through the execution of an emergency protection order ("EPO") obtained by the local authority under section 44 of the 1989 Act. I gave the main judgment (with which Thorpe and Lloyd LJJ agreed).
81. I held (para 32) that the power to remove a child under section 46 can be exercised even where an EPO is in force. I said (para 36) that where a police officer knows that an EPO is in force, he should not exercise the power of removing a child under section 46 unless there are compelling reasons to do so. On the facts of the case, there were no compelling reasons for the constable to exercise the section 46 power. The constable was in error in failing to ask himself whether there were compelling reasons why he should invoke section 46 rather than leave it to the council to execute the EPO. I held, therefore, that the removal of the child was unlawful. It was not in issue that, if the removal of the child was unlawful, the Chief Constable was liable to the child in false imprisonment.
82. Mr Beloff submits that the effect of my reasoning was that the constable had no jurisdiction (in the narrow pre-Anisminic sense) to do what he did. As an EPO was in force, it was in effect mandatory to invoke section 44 rather than section 46. I do not accept this analysis. I drew a clear distinction between the existence of the statutory authority to use the section 46 powers (which the constable had) and the exercise of those powers (which was wrongful on the facts of that case).
83. In Cooper v The Board of Works for the Wandsworth District (1863) 14 C.B. (N.S) 180M, the defendant board had the statutory power to demolish a house that was in the course of construction. It was held that this power was subject to the common law qualification that it should not be exercised without giving the builder notice and an opportunity to be heard. It was held that the board had exercised its statutory power unlawfully and that the builder was entitled to damages for trespass to property. But I agree with Mr Beloff that this decision does not shed any light on the question whether detention pursuant to an unlawful exercise of a power to detain is itself unlawful. As Byles J put it at p 195, the board "contravened the words of the statute". In effect, therefore, the court held that the decision to demolish the house was one which the board had no jurisdiction to make in the narrow pre-Anisminic sense.
84. I should deal with the authorities relied on by Mr Beloff. In R (Saadi) v Secretary of State for the Home Department [2002] 1 WLR 3131 as I have set out at para 50 above, the issue was whether the policy of detaining certain categories of asylum seekers whose claims could be processed quickly at the Oakington detention centre was lawful. The House of Lords held that it was. At first instance, Collins J had also considered the fact that when detaining the claimants the Home Office had used standard forms which did not reflect the new policy, and that therefore the true reasons for the detention had not been given. Collins J said that this did not affect the lawfulness of the detention. Lord Slynn at para 48 agreed that "the failure to give the right reason for detention and the giving of no or wrong reasons did not in the end affect the legality of the detention." But para 48 is not part of the ratio of the decision of the House. In any event, in so far as it was argued at all that the giving of untrue reasons for the detention rendered the detention unlawful, the Secretary of State did not advance a causation defence and contend that the giving of untrue reasons was immaterial because the true reasons were lawful.
85. In Nadarajah v Secretary of State for the Home Department [2004] INLR 139, the Secretary of State's published policy was that, if legal proceedings were initiated, removal would not be treated as imminent even if it otherwise was. The Secretary of State also had an unpublished policy, namely that information that proceedings were about to be initiated would be disregarded, however credible that information might be. At paras 68 and 69 of the judgment of Lord Phillips MR, the Court of Appeal said that the Secretary of State could not rely on the unpublished policy "as rendering lawful that which was, on the face of it, at odds with his policy, as made public" and for that reason the detention of N was unlawful. There is no hint of the causation test here. But the court went on to say that, if N's solicitors had been aware of the unpublished policy, they would have instituted proceedings sooner. N therefore suffered because his solicitor could not foresee the consequences of her conduct and his detention did not satisfy the requirement of lawfulness imposed by article 5(1)(f) of the ECHR. I accept that this was a causation point. But it was unnecessary for the court to adopt this additional reason for holding that the detention was unlawful. Further, it requires a huge leap to argue from this that the causation test must be satisfied as an element of the tort of false imprisonment. In short, neither Saadi nor Nadarajah bears the weight that Mr Beloff seeks to place on them.
86. Recognising that the court might reject the causation test, Mr Beloff suggested a number of alternative approaches. The first is that false imprisonment should be confined to "no authority" cases ie cases in which there was in fact no authority to detain, without recourse to the legal "fiction" that, because of a public law error, an authority to detain which was in fact given should be treated as if it had no legal effect because it was ultra vires. The second is that detention should be vitiated only by pre-Anisminic error of law. The third is that vitiating circumstances should be restricted to bad faith and improper purpose. The fourth is that authority to detain should be vitiated only by failure to have regard to a material consideration which had an effect on the detention. The fifth is that it should be a requirement that the claimant shall have successfully applied in judicial review proceedings for the decision to detain to be quashed.
87. The first two of these suggestions seek to put the clock back to the pre-Anisminic era. For reasons given earlier, this is unwarranted. As regards the third and fourth suggestions, I have accepted at paras 68 and 69 above that the detention must be vitiated by an error which is material in public law terms and have attempted to explain there what I mean by that. The fifth seeks unjustifiably to impose a procedural hurdle where none currently exists either at common law or in statute. To accede to this suggestion would be to engage in an unacceptable exercise of judicial legislation.
88. To summarise, therefore, in cases such as these, all that the claimant has to do is to prove that he was detained. The Secretary of State must prove that the detention was justified in law. She cannot do this by showing that, although the decision to detain was tainted by public law error in the sense that I have described, a decision to detain free from error could and would have been made.

Overall conclusion on liability on the basis that the policies applied were unlawful

89. I conclude, therefore, that since it is common ground that the unlawful policies in force between April 2006 and September 2008 were applied to Mr Lumba and Mr Mighty, they were unlawfully detained and their claims in false imprisonment must succeed. I turn to consider the assessment of damages.

Compensatory or nominal damages?

90. Having found that there was no liability in false imprisonment, the Court of Appeal did not need to decide whether the claimants were entitled to damages. They did, however, say at para 96:

"If, on the evidence, it was clear that, even assuming a lawful consideration, there was no realistic possibility of a different decision having been reached, and no realistic possibility of earlier release, then we do not see why that should not be reflected in an award of nominal damages only."

91. Mr Husain and Mr Westgate submit that, even if it was inevitable that the appellants would have been detained if the statutory power to detain had been lawfully exercised, they are nevertheless entitled to substantial and not merely nominal damages. They emphasise that false imprisonment is a tort of strict liability which is actionable without proof of special damage. The focus is on the claimant's right rather than the culpability of the defendant's conduct. They rely on two authorities in support of their argument. The first is Roberts v Chief Constable of the Cheshire Constabulary to which I have already referred at para 75 above. The plaintiff issued proceedings for false imprisonment arising from his detention by the police between 5.25 a.m. (when his detention should have been reviewed as required by statute) and 7.45 a.m. the same morning when it was reviewed and his continued detention authorised. It was found by the judge that the detention between 5.25 a.m. and 7.45 a.m. was unlawful, but that, if a review had taken place at 5.25 a.m., his continued detention would have been authorised. The judge awarded the plaintiff £500. The defendant's appeal on both liability and damages was dismissed by the Court of Appeal.
92. Clarke LJ gave the leading judgment. He said at p 668 D that there was an infringement of the plaintiff's legitimate right to have his case reviewed "and that, although the outcome may not have been affected by the failure to review in time, this infringement cannot be regarded as a purely nominal matter or a matter compensatable by entirely nominal damages. There are rules, the police must stick to them". He added at p 668G:

"As a matter of general principle such a plaintiff is entitled to be put into the position in which he would have been if the tort had not been committed. It is therefore important to analyse what the tort is. The plaintiff's claim was not for damages for breach of duty to carry out a review at 5.25 a.m. but for false imprisonment.

As I tried to explain earlier, the reason why the continued detention was unlawful was that no review was carried out. The wrong was not, however, the failure to carry out the review but the continued detention. If the wrong had not been committed the plaintiff would not have been detained between 5.25 a.m. and 7.45 a.m."

93. I do not consider that this case was correctly decided on the issue of damages. I agree that the plaintiff was entitled to be put into the position in which he would have been if the tort of false imprisonment had not been committed. But I do not agree that, if the tort had not been committed, the plaintiff would not have been detained between 5.25 a.m. and 7.45 a.m. On the judge's findings, if the tort had not been committed, he would have been detained during this period. It seems to me that the fallacy in the analysis in Roberts is that it draws no distinction between a detainee who would have remained in detention if the review had been carried out (and therefore no tort committed) and a detainee who would not have remained in detention if the review had been carried out. But the position of the two detainees is fundamentally different. The first has suffered no loss because he would have remained in detention whether the tort was committed or not. The second has suffered real loss because, if the tort had not been committed, he would not have remained in detention.
94. The second authority relied on in support of the appellants' case is Kuwait Airways Corpn v Iraqi Airways Co (Nos 4 and 5) [2002] UKHL 19, [2002] 2 AC 883. It was held there that it is no answer to a claim in conversion of goods for a defendant to say that the goods were or would have been subsequently converted by a third party: see the discussion by Lord Nicholls at paras 81 to 84 of his speech. But questions of causation in relation to cases of successive conversion by different tortfeasors have no application in the present context.
95. The question here is simply whether, on the hypothesis under consideration, the victims of the false imprisonment have suffered any loss which should be compensated in more than nominal damages. Exemplary damages apart, the purpose of damages is to compensate the victims of civil wrongs for the loss and damage that the wrongs have caused. If the power to detain had been exercised by the application of lawful policies, and on the assumption that the Hardial Singh principles had been properly applied (an issue which I discuss at paras 129-148 below), it is inevitable that the appellants would have been detained. In short, they suffered no loss or damage as a result of the unlawful exercise of the power to detain. They should receive no more than nominal damages.
96. I should add that this approach is consistent with the observation by Lord Griffiths in Murray v Ministry of Defence [1988] 1 WLR 692, 703A –B: "if a person is unaware that he has been falsely imprisoned and has suffered no harm, he can normally expect to recover no more than nominal damages."

Vindicatory damages

97. Mr Westgate submits that, if the appellants are entitled to no more than nominal damages, then they should also be awarded vindicatory damages. It has been said that the award of compensatory damages can serve a "vindicatory purpose": see, for example, Ashley v Chief Constable of Sussex Police [2008] UKHL 25, [2008] 1 AC 962 per Lord Scott at para 22 and Lord Rodger at para 60. But "vindicatory damages" serve a wider purpose than simply to compensate a successful claimant. The phrase "vindicatory damages" seems to have been coined by Sharma CJ in the Court of Appeal of Trinidad and Tobago in Attorney General of Trinidad and Tobago v Ramanoop as a head of loss in claims for breach of constitutionally protected rights and freedoms: see address given by the President of the Caribbean Court of Justice to a Symposium entitled "Current Developments in Caribbean Community Law" in Port of Spain on 9 November 2009. Lord Collins has traced the history of the use of the phrase in other contexts.
98. The concept of vindicatory damages was explained and endorsed by the Privy Council in the appeal in the Ramanoop case [2005] UKPC 15, [2006] 1 AC 328. Lord Nicholls said:

"18. When exercising this constitutional jurisdiction the court is concerned to uphold, or vindicate, the constitutional right which has been contravened. A declaration by the court will articulate the fact of the violation, but in most cases more will be required than words. If the person wronged has suffered damage, the court may award him compensation. The comparable common law measure of damages will often be a useful guide in assessing the amount of this compensation. But this measure is no more than a guide because the award of compensation under section 14 is discretionary and, moreover, the violation of constitutional right will not always be coterminous with the cause of action at law.

19. An award of compensation will go some distance towards vindicating the infringed constitutional right. How far it goes will depend on the circumstances, but in principle it may well not suffice. The fact that the right violated was a constitutional right adds an extra dimension to the wrong. An additional award, not necessarily of substantial size, may be needed to reflect the sense of public outrage, emphasise the importance of the constitutional right and the gravity of the breach, and deter further breaches. All these elements had a place in this additional award. 'Redress' in section 14 is apt to encompass such an award if the court considers it is required having regard to all the circumstances. Although such an award, where called for, is likely in most cases to cover much the same ground in financial terms as would an award by way of punishment in the strict sense of retribution, punishment in the latter sense is not its object. Accordingly, the expressions 'punitive damages' or 'exemplary damages' are better avoided as descriptions of this type of additional award".

99. It will be seen, therefore, that the Privy Council endorsed the principle of vindicatory damages for violation of constitutional rights. Should this principle be extended further? In Ashley at para 22 Lord Scott obiter said that vindicatory damages might be awarded for the tort of battery or trespass to the person by the police resulting in the death of the victim. But the issue in that case was whether a claimant should be allowed to continue with an action in order to establish whether an assault had been committed, where there could be no award of further compensatory damages because these had already been paid in full as a result of a concession by the police. Lord Scott's view that vindicatory damages have a role in the compensation for civil wrongs and the breach of ECHR rights was endorsed, at least to some extent, in Mosley v News Group Newspapers Ltd [2008] EWHC 1777 (QB). In awarding damages for breach of the claimant's right to privacy, after recognising the compensatory nature of damages for infringements of privacy, Eady J said at paras 216-7 that there was another factor which "probably" had to be taken into account, namely vindication to mark the infringement of the right.
100. It is one thing to say that the award of compensatory damages, whether substantial or nominal, serves a vindicatory purpose: in addition to compensating a claimant's loss, it vindicates the right that has been infringed. It is another to award a claimant an additional award, not in order to punish the wrongdoer, but to reflect the special nature of the wrong. As Lord Nicholls made clear in Ramanoop, discretionary vindicatory damages may be awarded for breach of the Constitution of Trinidad and Tobago in order to reflect the sense of public outrage, emphasise the importance of the constitutional right and the gravity of the breach and deter further breaches. It is a big leap to apply this reasoning to any private claim against the executive. McGregor on Damages 18^th ed (2009) states at para 42-009 that "It cannot be said to be established that the infringement of a right can in our law lead to an award of vindicatory damages". After referring in particular to the appeals to the Privy Council from Caribbean countries, the paragraph continues: "the cases are therefore far removed from tortious claims at home under the common law". I agree with these observations. I should add that the reference by Lord Nicholls to reflecting public outrage shows how closely linked vindicatory damages are to punitive and exemplary damages.
101. The implications of awarding vindicatory damages in the present case would be far reaching. Undesirable uncertainty would result. If they were awarded here, then they could in principle be awarded in any case involving a battery or false imprisonment by an arm of the state. Indeed, why limit it to such torts? And why limit it to torts committed by the state? I see no justification for letting such an unruly horse loose on our law. In my view, the purpose of vindicating a claimant's common law rights is sufficiently met by (i) an award of compensatory damages, including (in the case of strict liability torts) nominal damages where no substantial loss is proved, (ii) where appropriate, a declaration in suitable terms and (iii) again, where appropriate, an award of exemplary damages. There is no justification for awarding vindicatory damages for false imprisonment to any of the FNPs.

Application of Hardial Singh principles to the appeal of Mr Lumba

102. In addition to the issues that are raised which are common to both appeals, it is submitted on behalf of Mr Lumba (but not Mr Mighty) that his detention is unlawful on the grounds that there has been a breach of the Hardial Singh principles. This part of the appeal raises a number of points about the reach of those principles. I refer to the encapsulation of the principles set out in my judgment in R (I) [2003] INLR 196 at para 22 above.
103. A convenient starting point is to determine whether, and if so when, there is a realistic prospect that deportation will take place. As I said at para 47 of my judgment in R (I), there may be situations where, although a reasonable period has not yet expired, it becomes clear that the Secretary of State will not be able to deport the detained person within a period that is reasonable in all the circumstances, having regard in particular to time that the person has already spent in detention. I deal below with the factors which are relevant to a determination of a reasonable period. But if there is no realistic prospect that deportation will take place within a reasonable time, then continued detention is unlawful.
104. How long is a reasonable period? At para 48 of my judgment in R (I), I said:

"It is not possible or desirable to produce an exhaustive list of all the circumstances that are, or may be, relevant to the question of how long it is reasonable for the Secretary of State to detain a person pending deportation pursuant to paragraph 2(3) of Schedule 3 to the Immigration Act 1971. But in my view, they include at least: the length of the period of detention; the nature of the obstacles which stand in the path of the Secretary of State preventing a deportation; the diligence, speed and effectiveness of the steps taken by the Secretary of State to surmount such obstacles; the conditions in which the detained person is being kept; the effect of detention on him and his family; the risk that if he is released from detention he will abscond; and the danger that, if released, he will commit criminal offences."

105. So far as I am aware, subject to the following qualifications, the relevance of these factors has not been questioned. The qualifications are first that the relevance of the risk of offending on release is challenged on behalf of the appellants in the present case. Secondly, "the nature of the obstacles" begs two questions that have been raised on this appeal, namely what is the relevance, if any, of delays attributable to the fact that a detained person (i) is challenging the decision to deport him by appeal or judicial review and will generally not be deported until his challenges have been determined; and (ii) has refused to return voluntarily to his country of origin?

Risk of re-offending

106. Mr Husain accepts that, where there is a risk that the detained person will abscond, the risk of re-offending is relevant to the assessment of the duration of detention that is reasonably necessary to effect deportation. But he submits that, where there is no real risk of absconding, the risk of re-offending cannot of itself justify detention. Where there is no such risk, detention is not necessary to facilitate deportation, because it will be possible to effect the deportation without the need for detention. The underlying purpose of the power to detain is not to prevent the commission of criminal offences, but to facilitate the implementation of a deportation order.
107. I have some difficulty in understanding why the risk of re-offending is a relevant factor in a case where there is a risk of absconding, but not otherwise. It seems to me that it is possible to construe the power to detain either (more narrowly) as a power which may only be exercised to further the object of facilitating a deportation, or (more broadly) as a power which may also be exercised to further the object which it is sought to achieve by a deportation, namely, in the present case, that of removing an offender whose presence is not conducive to the public good. The distinction between these two objects was clearly drawn by the Court of Appeal in R (A) v Secretary of State for the Home Department [2007] EWCA Civ 804. Toulson LJ said at para 55:

"A risk of offending if the person is not detained is an additional relevant factor, the strength of which would depend on the magnitude of the risk, by which I include both the likelihood of it occurring and the potential gravity of the consequences. Mr Drabble submitted that the purpose of the power of detention was not for the protection of public safety. In my view, that is over-simplistic. The purpose of the power of deportation is to remove a person who is not entitled to be in the United Kingdom and whose continued presence would not be conducive to the public good. If the reason why his presence would not be conducive to the public good is because he has a propensity to commit serious offences, protection of the public from that risk is the purpose of the deportation order and must be a relevant consideration when determining the reasonableness of detaining him pending his removal or departure."

Para 78 of Keene LJ's judgment is to similar effect.

108. I acknowledge that the principle that statutory powers should be interpreted in a way which is least restrictive of liberty if that is possible would tend to support the narrower interpretation. But I think that the Court of Appeal was right in R (A) to adopt the interpretation which gives effect to the purpose underlying the power to deport and which the power to detain is intended to facilitate. Perhaps a simpler way of reaching the same conclusion is to say, as Simon Brown LJ said in R (I) at para 29, that the period which is reasonable will depend on the circumstances of the particular case and the likelihood or otherwise of the detainee re-offending is "an obviously relevant circumstance".
109. But the risk of re-offending is a relevant factor even if the appellants are right in saying that it is relevant only when there is also a risk of absconding. As Lord Rodger pointed out in argument, if a person re-offends there is a risk that he will abscond so as to evade arrest or if he is arrested that he will be prosecuted and receive a custodial sentence. Either way, his re-offending will impede his deportation.
110. The risk of re-offending is, therefore, a relevant factor.

Delay attributable to challenges to deportation

111. Mr Beloff submits that the time taken to resolve legal challenges brought by an individual against deportation should generally be left out of account in considering whether a reasonable period of detention has elapsed. He concedes that this general rule should be subject to two qualifications: (i) if the Secretary of State has caused delay in the resolution of the legal challenge, then that time may be taken into account; and (ii) the time during which a legal challenge is being resolved should be taken into account if removal is not possible for reasons unrelated to the legal challenge. I shall call this general rule "the exclusionary rule".
112. In support of this submission, Mr Beloff makes the following points. First, it is the individual's choice to challenge the removal and, if the time taken to resolve legal challenges were taken into account, the length of detention would be outside the control of the Secretary of State and would be entirely within the control of the detained person. Secondly, if the position were otherwise, those who (if at large) would be a danger to the public or who would be likely to thwart a deportation order by absconding, would be able to increase their prospects of release by pursuing every conceivable point by way of legal challenges and by doing everything possible to delay the legal process. Thirdly, if the legality of detention is capable of depending on the merits of a challenge to the decision to deport, it will be necessary for the High Court to decide for itself the merits of the underlying challenge, in advance of consideration of the case by the specialist tribunal appointed by Parliament to undertake that task. That is undesirable.
113. At para 102 of their judgment, the Court of Appeal said:

"In our judgment, the fact that a FNP is refusing to return voluntarily, or is refusing to cooperate in his return (for example, by refusing to apply for an emergency travel document, as initially did WL) is relevant to the assessment of the legality of his continued detention: see R (A) v Secretary of State for the Home Department [2007] EWCA Civ 804. So is the fact that the period of his detention had been increased, and his deportation postponed by his pursuit of appeals and judicial review proceedings seeking to challenge his deportation order or his application for asylum or leave to remain, particularly if his applications and appeals are obviously unmeritorious. In our judgment, as a matter of principle, a FNP cannot complain of the prolongation of his detention if it is caused by his own conduct."

114. It is not clear whether the Court of Appeal were accepting the exclusionary rule in its entirety. To say that the fact that the length of the detention is attributable to the pursuit of legal proceedings is relevant to the assessment of the legality of the detention suggests a rejection of the exclusionary rule. But to say that FNPs cannot complain of the prolongation of their detention caused by their own conduct suggests an acceptance of it.
115. The question of the relevance of the pursuit of legal proceedings has been considered in a number of authorities. I do not propose to analyse them. None is binding on this court. The discussion of the issue which I have found most helpful is that of Davis J in R (Abdi) v Secretary of State for the Home Department [2009] EWHC 1324 (Admin). In that case, the exclusionary rule (with the same two exceptions) was urged on the court on behalf of the Secretary of State. In rejecting it, Davis J pointed out at para 25 of his judgment, that it was undesirable, where the core question is an assessment of what is reasonable in all the circumstances, to be astute to look for mandatory restrictions or rules in what ought to be a fact-specific exercise. I agree. The Hardial Singh principles should not be applied rigidly or mechanically.
116. There are several problems with the exclusionary rule. First, it seems to require the exclusion of consideration of the individual circumstances of an applicant pending what may be a long appellate process. Suppose two FNPs who both embark on a meritorious appeal process which takes a number of years. The only difference between them is that A poses a very high risk of absconding and re-offending and B poses a very low risk. If the exclusionary rule is applied, no difference can be drawn between them from the time proceedings are commenced. In both cases, the several years during which they are detained while the appeal process is continuing are to be disregarded in assessing whether the period of detention is reasonable. Or suppose that the effect of detention on A is to cause serious damage to his health or that of members of his family, whereas there is no such effect in the case of B. I can see no warrant for such a mechanistic approach to the determination of what is reasonable in all the circumstances.
117. Secondly, the exclusionary rule seems to involve the exclusion from consideration of any delays occurring within the appeal process which are not the fault of the applicant or (as is conceded by Mr Beloff) the Secretary of State. I see no reason why such delays, for example, delays on the part of the tribunal or court, should be disregarded in a determination of whether the period of detention is reasonable.
118. Thirdly, the consequence of the exclusionary rule is that a person can be detained for many years while pursuing a prima facie meritorious appeal but he cannot by judicial review or habeas corpus challenge his detention on Hardial Singh or related article 5(1)(f) of the ECHR principles. It precludes such judicial scrutiny (i) however long the detention and appeals have lasted and (ii) regardless of the effects of prolonged detention on the detainee, provided that (iii) the appeals are being diligently pursued and there is no concurrent independent reason why deportation cannot be effected. I accept the submission of Mr Husain that bail is not a sufficient answer to the fundamental objection that the exclusionary rule constitutes an impermissible restriction on judicial oversight of the legality of administrative detention. Paragraph 29 of Schedule 2 to the 1971 Act gives the First Tier Tribunal power to grant bail pending an appeal, but this is subject to the restrictions stated in paragraph 30. Paragraph 30(1) provides that an appellant shall not be released under paragraph 29 without the consent of the Secretary of State if removal directions are currently in force. There is nothing in the schedule which requires the tribunal to apply the Hardial Singh principles in deciding whether or not to grant bail and, in particular, to have regard to the past and likely future length of a detention. Bail is not a determination of the legality of detention, whether at common law or for article 5(4) purposes.
119. Fourthly, the exclusionary rule is inconsistent with the decision of the ECtHR in Chahal v United Kingdom (1996) 23 EHRR 413. In deciding whether the detention complied with the requirements of article 5(1)(f), the court had regard to the length of the detention, including the time taken for the various domestic proceedings to be completed: see paras 114, 115 and 123 of the judgment of the court. There is a close analogy between the Hardial Singh principles and the article 5 requirement that detention for the purposes of deportation must not be of excessive duration.
120. I would reject Mr Beloff's in terrorem argument that, unless the exclusionary rule applies, detained FNPs will be able to procure their release from detention by the simple expedient of pursuing hopeless legal challenges. Time taken in the pursuit of hopeless challenges should be given minimal weight in the computation of a reasonable period of detention. Nor do I accept that it is undesirable (or indeed unduly difficult) to identify hopeless or abusive challenges. There exist statutory mechanisms to curb unmeritorious appeals. If a claim is "clearly unfounded", certification under section 94(2) of the Nationality, Immigration and Asylum Act 2002 precludes an in-country appeal. If a claim relies on a matter which could have been raised earlier in response to an earlier immigration decision or in response to a "one-stop notice", certification under section 96 of the 2002 Act precludes any appeal at all. In any event, a court considering the legality of a detention will often be able to assess the prima facie merits of an appeal. Where, as in the case of Mr Lumba, there have been orders for reconsideration, or where there has been a grant of permission to appeal to the Court of Appeal, the court will easily recognise that the challenge has some merit. Conversely, there may be one or more determinations from immigration judges dismissing claims as wholly lacking in credibility.
121. To summarise, I would reject the exclusionary rule. If a detained person is pursuing a hopeless legal challenge and that is the only reason why he is not being deported, his detention during the challenge should be given minimal weight in assessing what is a reasonable period of detention in all the circumstances. On the other hand, the fact that a meritorious appeal is being pursued does not mean that the period of detention during the appeal should necessarily be taken into account in its entirety for the benefit of the detained person. Indeed, Mr Husain does not go so far as to submit that there is any automatic rule, regardless of the risks of absconding and/or re-offending, which would compel an appellant's release if the appeals process lasted a very long time through no fault of the appellant. He submits that the weight to be given to time spent detained during appeals is fact-sensitive. This accords with the approach of Davis J in Abdi and I agree with it. The risks of absconding and re-offending are always of paramount importance, since if a person absconds, he will frustrate the deportation for which purpose he was detained in the first place. But it is clearly right that, in determining whether a period of detention has become unreasonable in all the circumstances, much more weight should be given to detention during a period when the detained person is pursuing a meritorious appeal than to detention during a period when he is pursuing a hopeless one.

Non-cooperation with return

122. The most common examples of non-cooperation are (i) a refusal by a person who does not have a valid passport to cooperate with the obtaining of travel documents to enable him to return and (ii) a person's refusal to avail himself of one of the Home Office schemes by which he may leave the United Kingdom voluntarily. Most of the discussion in the cases has centred on (ii).
123. It is common ground that a refusal to return voluntarily is relevant to an assessment of what is a reasonable period of detention if a risk of absconding can properly be inferred from the refusal. But I would warn against the danger of drawing an inference of risk of absconding in every case. It is always necessary to have regard to the history and particular circumstances of the detained person. What is, however, in issue is whether a failure to return voluntarily can of itself justify a period of detention which would otherwise be unreasonable and therefore unlawful. In R (I), I said in the Court of Appeal:

"50. As regards the significance of the appellant's refusal of voluntary repatriation, there appears to be agreement between Simon Brown LJ and Mummery LJ that this is a relevant circumstance, but Mummery LJ considers that it is decisively adverse to the appellant, whereas Simon Brown LJ considers that it is of relatively limited relevance on the facts of the present case. I too consider that it is a relevant circumstance, but in my judgment it is of little weight. [Counsel for the Secretary of State] submits that a refusal to leave voluntarily is relevant for two reasons. First, the detained person has control over the fact of his detention: if he decided to leave voluntarily, he would not be detained. Secondly, the refusal indicates that he would abscond if released from detention. It is this second feature which has weighed heavily with Mummery LJ.

51. I cannot accept that the first of [the Secretary of State's] reasons is relevant. Of course, if the appellant were to leave voluntarily, he would cease to be detained. But in my judgment, the mere fact (without more) that a detained person refuses the offer of voluntary repatriation cannot make reasonable a period of detention which would otherwise be unreasonable. If [the Secretary of State] were right, the refusal of an offer of voluntary repatriation would justify as reasonable any period of detention, no matter how long, provided that the Secretary of State was doing his best to effect the deportation.

52. I turn to [the Secretary of State's] second reason. I accept that if it is right to infer from the refusal of an offer of voluntary repatriation that a detained person is likely to abscond when released from detention, then the refusal of voluntary repatriation is relevant to the reasonableness of the duration of a detention. In that event, the refusal of voluntary repatriation is no more than evidence of a relevant circumstance namely the likelihood that the detained person will abscond if released.

53. But there are two important points to be made. First, the relevance of the likelihood of absconding, if proved, should not be overstated. Carried to its logical conclusion, it could become a trump card that carried the day for the Secretary of State in every case where such a risk was made out regardless of all other considerations, not least the length of the period of detention. That would be a wholly unacceptable outcome where human liberty is at stake.

54. Secondly, it is for the Secretary of State to satisfy the court that it is right to infer from the refusal by a detained person of an offer of voluntary repatriation that, if released, he will abscond. There will no doubt be many cases where the court will be persuaded to draw such an inference. I am not, however, satisfied that this is such a case. It is not at all surprising that this appellant has refused voluntary repatriation. He has not yet exhausted the asylum process, which, if successful, would permit him to remain in the United Kingdom. In these circumstances, why should one infer from the refusal of voluntary repatriation that, if released, he would abscond? In my judgment, the most that can be said is that there is a risk that if he is released the appellant will abscond. But that can be said of most cases. I do not consider that the fact that he has refused the offer of voluntary repatriation adds materially to the evidence that such risk is present in the instance case."

124. Simon Brown LJ adopted a somewhat different approach at para 31. He said that, since in Hardial Singh Woolf J had regarded it as a factor in the applicant's favour that he was "quite prepared to go to India", he could not see why the converse should not be relevant. The court should not "ignore entirely" the applicant's ability to end his detention by returning home voluntarily.
125. The point was considered again by the Court of Appeal in R (A) Somalia v Secretary of State for the Home Department [2007] EWCA Civ 804 Toulson LJ (with whom Longmore LJ agreed) said:

"54. I accept the submission on behalf of the Home Secretary that where there is a risk of absconding and a refusal to accept voluntary repatriation, those are bound to be very important factors, and likely often to be decisive factors, in determining the reasonableness of a person's detention, provided that deportation is the genuine purpose of the detention. The risk of absconding is important because it threatens to defeat the purpose for which the deportation order was made. The refusal of voluntary repatriation is important not only as evidence of the risk of absconding, but also because there is a big difference between administrative detention in circumstances where there is no immediate prospect of the detainee being able to return to his country of origin and detention in circumstances where he could return there at once. In the latter case the loss of liberty involved in the individual's continued detention is a product of his own making."

126. Keene LJ said at para 79:

"I am not persuaded by Mr Giffin that the refusal by this detainee to return to Somalia voluntarily when it was possible to do so is some sort of trump card. On this I see the force of what was said by Dyson LJ in R (I) at paragraph 52, namely that the main significance of such a refusal may often lie in the evidence it provides of a likelihood of the individual absconding if released. After all, if there is in a particular case no real risk of his absconding, how could detention be justified in order to achieve deportation, just because he has refused voluntary return? The Home Office in such a case, ex hypothesi, would be able to lay hands on him whenever it wished to put the deportation order into effect. Detention would not be necessary in order to fulfil the deportation order. Having said that, I do not regard such a refusal to return as wholly irrelevant in its own right or as having a relevance solely in terms of the risk of absconding. It is relevant that the individual could avoid detention by his voluntary act. But I do not accept that such a refusal is of the fundamental importance contended for by the Secretary of State"

127. It is necessary to distinguish between cases where return to the country of origin is possible and those where it is not. Where return is not possible for reasons which are extraneous to the person detained, the fact that he is not willing to return voluntarily cannot be held against him since his refusal has no causal effect. But what if return would be possible, but the detained person is not willing to go? Here it is necessary to consider whether the detained person has issued proceedings challenging his deportation. If he has done so, then it is entirely reasonable that he should remain in the United Kingdom pending the determination of those proceedings (unless the proceedings are an abuse). In those circumstances his refusal to accept an offer of voluntary return is irrelevant. The purpose of voluntary return is not to encourage foreign nationals to return to their countries of origin where, if their legal challenges succeed, it is likely to have been demonstrated that they would face a risk of persecution within the meaning of the Convention and Protocol relating to the Status of Refugees (1951) (Cmd 9171) and (1967) (Cmnd 3906) or treatment contrary to article 3 of the ECHR. Rather, it is to facilitate removal where that is justified because the FNPs have not proved that they would face the relevant risk on return. In accepting voluntary return, the individual forfeits all legal rights to remain in the United Kingdom. He should not be penalised for seeking to vindicate his ECHR or Refugee Convention rights and be faced with the choice of abandoning those rights or facing a longer detention than he would face if he had not been offered voluntary return.
128. What about those who have no outstanding legal challenges? Here, the fact that the detained person has refused voluntary return should not be regarded as a "trump card" which enables the Secretary of State to continue to detain until deportation can be effected, whenever that may be. That is because otherwise, as I said at para 51 of my judgment in R (I), "the refusal of an offer of voluntary repatriation would justify as reasonable any period of detention, no matter how long, provided that the Secretary of State was doing his best to effect the deportation." If the refusal of voluntary return has any relevance in such cases even if a risk of absconding cannot be inferred from the refusal, it must be limited. That was the view of Simon Brown LJ in R (I) and Keene LJ in R (A) and I agree with them.

The appeal of Mr Lumba

129. I can now turn to the particular case of Mr Lumba. He entered the United Kingdom unlawfully on 10 April 1994 and claimed asylum on the same day. His claim was refused on 20 February 2000. He was, however, granted exceptional leave to remain until 20 February 2004. He has a bad criminal record which includes convictions for assault occasioning actual bodily harm (six months' imprisonment); two offences of using threatening and abusive behaviour (twelve months' probation); two offences of shoplifting (non-custodial sentences); assault occasioning actual bodily harm (six months' imprisonment); assaulting a constable (four months' imprisonment); and wounding with intent contrary to section 18 of the Offences Against the Person Act 1861 (4 years' imprisonment).
130. He married MP, another national of the Democratic Republic of Congo ("DRC"). Their son, PAS, was born on 5 July 2003.
131. As I have already said, on 3 April 2006, the Secretary of State decided to deport Mr Lumba. Notwithstanding that his son is a British citizen, the Secretary of State also decided to deport his wife and son, although the orders against them were subsequently revoked. He was notified of the decision to detain him under the 1971 Act on 22 June 2006, the day before his release date. His first detention review was in February 2007. He appealed against the decision to deport him. On 7 September 2006, while the appeal was still pending, he refused to attend a travel document interview with the immigration authorities to facilitate his return to the DRC.
132. His appeal against deportation was dismissed by the Asylum and Immigration Tribunal ("AIT") on 15 December 2006. In about January 2007, he was transferred from prison to Dover Immigration Removal Centre, since when he remained in detention.
133. On 1 March 2007, he attended an interview with an immigration officer at which he explained that his passport was missing. He was asked to complete an application for an Emergency Travel Document ("ETD") but refused to do so, stating that he did not wish to be returned to the DRC because "the problem which formed the initial basis of his asylum claim is still true." By 20 March 2007, the Secretary of State's caseworkers had discovered that they had a copy of Mr Lumba's passport on file.
134. His indefinite leave to remain was terminated only after service on him of the signed deportation order on 21 March 2007. The discovery of the copy of his passport on file obviated the need for his cooperation in completing a travel document application. On 29 March, a completed application for an ETD, accompanied by copies of his passport and identity card, was sent to the Embassy of the DRC. The Embassy claimed to have lost the initial application and the application for an ETD was resubmitted on 26 April 2007. The application was granted by the DRC Embassy on 25 July 2007 and on the following day directions were set for Mr Lumba's removal from the United Kingdom on 20 August 2007.
135. On 15 August 2007, his new solicitors submitted fresh representations challenging his deportation. These noted that a country guidance case (BK (DRC)) concerning the safety of returns to the DRC of failed asylum seekers/ deportees was to be heard by the AIT on 17 September 2007. On 16 August 2007, the Secretary of State refused to treat those representations as a fresh asylum claim and on the following day, Mr Lumba issued judicial review proceedings challenging that refusal. On 14 September 2007, these proceedings were stayed by consent until the outcome of the pending country guidance case before the AIT was known.
136. On 23 October 2007, Mr Lumba issued the claim for judicial review challenging his detention which has given rise to the present appeal. On 18 December 2007, the AIT promulgated its determination in BK (Democratic Republic of Congo) v Secretary of State for the Home Department and decided that failed asylum seekers were not, as such, at risk in the DRC: [2007] UKAIT 98. Mr Lumba made applications for bail on 23 January 2008, 4 February 2008 and 4 March 2008. They were all rejected by the immigration judge principally on the ground that there was a real risk that if he was released he would re-offend.
137. On 23 April 2008, permission was granted to appeal the AIT's decision in BK (DRC) to the Court of Appeal.
138. Mr Lumba's claim for judicial review in respect of his detention was heard by Collins J on 4 July 2008. His judgment is reported at [2008] EWHC 2090 (Admin). He concluded that the detention was lawful on Hardial Singh principles and that it would continue to be so until the Court of Appeal gave judgment in BK (DRC), on the assumption that judgment was given by the end of December 2008. In fact judgment was given by the Court of Appeal on 3 December 2008 after the hearing before Davis J, but before he gave judgment: [2008] EWCA Civ 1322.
139. Davis J agreed with the reasoning and conclusion of Collins J as to the lawfulness of the detention. The findings made by Davis J are set out in full by the Court of Appeal at para 99 of their judgment. There is no need to repeat them in this judgment. In summary, Davis J said at para 203 that (i) Mr Lumba posed a high risk of absconding and a high risk of serious reoffending; (ii) at all stages there was a prospect of removing him within a reasonable period; and (iii) there had been no lack of due expedition on the part of the Secretary of State.
140. On 11 February 2009, Mr Lumba's solicitors made fresh representations on his behalf to the Secretary of State, and applied for the revocation of the deportation order. They said that his marriage had broken down and that he was seeking contact with his son. He relied on article 8 of the ECHR. This application was rejected by the Secretary of State on 1 July 2009. Mr Lumba appealed on 8 July 2009. The appeal was dismissed by the AIT on 28 September 2009. A reconsideration was ordered by a senior immigration judge on 26 October 2009. At the reconsideration, the Secretary of State conceded that the original tribunal had erred in law.
141. By a judgment given on 19 February 2010, the Court of Appeal dismissed Mr Lumba's appeals against the decisions of Collins J and Davis J. The Court of Appeal said at para 100 that they had seen nothing to justify interfering with Davis J's findings at para 203 of his judgment. They said in relation to the application of the Hardial Singh principles that the real attack was on the judgment of Collins J. There had been no material error in the approach or conclusions of Collins J who had taken into account the high risk of absconding and re-offending, the fact that Mr Lumba could have returned to the DRC voluntarily, and that his deportation had been delayed by his pursuit of several unsuccessful applications for asylum or leave to remain and appeals against their refusal.
142. The Court of Appeal had been asked to consider the period which had elapsed since the judgment of Davis J and subsequent evidence, including a psychiatric report, and to determine the legality of Mr Lumba's current detention in the light of it. They said at para 108 that it would be inappropriate for them to consider as a first instance decision whether Mr Lumba's mental condition rendered his continued detention unlawful. Apart from that, they said that:

"having reviewed the history of [Mr Lumba's] detention and the reasons given for continuing it, and for the refusals of bail, and his several hitherto unmeritorious appeals and applications, we are satisfied that his detention for the purposes of his deportation continues to be lawful".

143. On behalf of Mr Lumba it is submitted that the Court of Appeal should have concluded that his detention was unreasonably long by the time of the hearing before Collins J on 4 July 2008; alternatively by the time of the hearing before Davis J on 11-14 November 2008; alternatively by the time of the hearing before them between 30 November and 2 December 2009. By the time of the appeal, of particular importance were the facts that (i) Mr Lumba had been in detention for 41 months; (ii) he could not be deported while he was pursuing his appeal against the Secretary of State's refusal to revoke the deportation order; and (iii) there was evidence from the Croydon Mental Health Team and consultant psychiatrist Dr Dinakaran which showed that the risk of re-offending and psychotic relapse could be managed in the community.
144. Mr Lumba has now been in detention for 54 months. At first sight, his detention seems to have been of unreasonable duration. There must come a time when, however grave the risk of absconding and however the grave the risk of serious offending, it ceases to be lawful to detain a person pending deportation. Moreover, in certain respects the courts below have not applied the Hardial Singh principles correctly. In particular, they have wrongly regarded the fact that Mr Lumba has been able to delay his deportation by pursuing applications and appeals which, thus far, have been unsuccessful as being relevant to the assessment of a reasonable period of detention. It has not been suggested by the Secretary of State that any of these applications or appeals have been hopeless and abusive. For the reasons given above, the fact that the applications and appeals delayed the deportation should have been regarded as irrelevant. The courts below also appear to have taken into account Mr Lumba's refusal of voluntary return without making it clear how this is relevant to the assessment of a reasonable period. As I have said, it is of limited relevance as a free-standing reason (see paras 122-128 above). It would be legitimate to infer from the refusal of voluntary return that there is a risk of absconding. But it is not clear that Collins J or Davis J drew such an inference in this case.
145. Mr Husain submits that it was incumbent on the Court of Appeal to consider the legality of the continuing executive detention by reference to the situation current at the time of the appeal and that they erred in failing to do so. He says that they should have considered what the prospects were of removal within a reasonable period. I accept that, where the facts are the same as they were before the first instance judge and the only difference is the passage of further time, there is no reason why the Court of Appeal should not decide for themselves whether a continuing detention is unlawful. I also accept that, where there is fresh evidence, the Court of Appeal are entitled to apply the Hardial Singh principles and consider for themselves on the basis of all the up-to-date material whether a continuing detention is lawful. Indeed, that was the course that the Court of Appeal took in R (I). But there are some circumstances where that course is not appropriate.
146. In the present case, the Court of Appeal took the view that such a course was inappropriate. They said at para 108 that they "should not embark on a first instance decision on matters, such as whether the deportation order should be revoked on account of [Mr Lumba's] mental condition, that Parliament has entrusted to a specialist tribunal". That was an entirely proper reason for the Court of Appeal not making a first instance decision in this case.
147. Mr Husain submits that the Supreme Court should allow Mr Lumba's appeal. I would allow the appeal of both appellants for the generic reasons that I have stated earlier. But I would not decide whether the detention of Mr. Lumba for almost 56 months was in breach of the Hardial Singh principles. The reasons which led the Court of Appeal to refuse to make a first instance decision apply with even greater force in an appeal to the Supreme Court. It is only in the most exceptional case that this court should embark on a task that is normally to be performed by a court of first instance.
148. In view of the passage of further time since the decision of the Court of Appeal as well as the fact that the courts below failed to apply the Hardial Singh principles correctly, I would remit Mr Lumba's claim for damages for breach of those principles to the High Court for reconsideration in the light of all the evidence as to the current position.

The case of Mr Mighty

149. As I have already said, Mr Mighty was detained between 19 May 2006 and 28 July 2008. He issued proceedings on 29 May 2008 alleging that his detention was for longer than a reasonable period and inter alia that, on a proper application of the Hardial Singh principles, he should have been released. His Hardial Singh case was rejected by Davis J. There was no appeal on this aspect of the case to the Court of Appeal and the point has not been raised on behalf of Mr Mighty before this court.

Exemplary damages

150. The relevant principles are not in doubt. Exemplary damages may be awarded in three categories of case: see per Lord Devlin in Rookes v Barnard [1964] AC 1129. The category which is relevant for present purposes is that there has been "an arbitrary and outrageous use of executive power" (p 1223) and "oppressive, arbitrary or unconstitutional action by servants of the government" (p 1226). In this category of case, the purpose of exemplary damages is to restrain the gross misuse of power: see AB v South West Water Services Ltd [1993] QB 507, 529F per Sir Thomas Bingham MR. It must be shown that the "conscious wrongdoing by a defendant is so outrageous, his disregard of the plaintiff's rights so contumelious, that something more [than compensatory damages] is needed to show that the law will not tolerate such behaviour" as a "remedy of last resort": see per Lord Nicholls Kuddus v Chief Constable of Leicestershire Constabulary [2002] 2 AC 122 at para 63.
151. Both Davis J and the Court of Appeal addressed the question of exemplary damages, although in view of their findings on the issue of liability, it was not necessary for them to do so. Davis J said at para 205 of his judgment that:

"I add, briefly, that, even if I had concluded there was unlawful detention in any of these cases justifying an award of damages, I would not in any event have awarded exemplary damages on the footing of unconstitutional, oppressive or arbitrary conduct, in so far as sought. While the Home Office has, to put it mildly, not covered itself in glory in this whole matter of the new policy, I think the failings were in essence one of failing, promptly and directly, to confront and address a perceived legal difficulty: whether through concerns at being bearers of unwelcome news to the Ministers or through an instinct for ducking an apparently intractable problem or through institutional inertia or some other reason, I cannot really say. I am not prepared, however, to conclude on the material before me that there was a conscious decision within the Home Office to operate tacitly an unpublished policy, known to be highly suspect, in the hope it would not be uncovered or, if it was uncovered, against a plan, if the courts intervened, to present that reversal as being due solely to the courts or the Human Rights Act. In my view what happened here, in any of these five cases, cannot fairly, I think, be described as sufficiently outrageous to justify an award of exemplary damages. In any event, I emphasise that individual consideration was given to the cases of each of the claimants."

152. By the time of the appeal, the Secretary of State had disclosed more material than was before the judge. The Court of Appeal said that, even taking account of the additional material, they agreed with the assessment of the judge. They said:

"122. We give weight to that assessment by the judge at the end of his very careful and comprehensive judgment. It also accords with our own view, even taking account of the additional material which has been disclosed. We consider that there was a failure, which to put it very mildly indeed, was very regrettable, on the part of the department to face up to the basic problem that the published policy had not caught up with the much more restrictive approach implicit in ministerial statements on the subject. However, we find it difficult to describe such conduct as 'unconstitutional, oppressive or arbitrary', in circumstances where the Home Secretary had an undoubted power to detain for the purposes in question, and it has been held that on the facts of the case he could lawfully have exercised that power with the same effect; at any rate, if it can be so described, these circumstances mean that the conduct is at the less serious end of unconstitutional, oppressive or arbitrary. We also bear in mind also that the claimants had the right to apply for bail to an independent tribunal, at which it was possible for the continuing reasonableness of their detention to be challenged. An award of exemplary damages would be an unwarranted windfall for them, and it would have little punitive effect since it will not be borne by those most directly responsible. Rather it would be a drain on public resources which in itself is unlikely to add significantly to the remedial effect of a declaration of unlawfulness.

123. Moreover, it is difficult to see on what basis exemplary damages could be assessed in lead cases such as these. The conduct of the Home Secretary complained of in the present case was common to a large number of detainees who have brought proceedings against him. The selection of lead claimants such as [Mr Lumba] and [Mr Mighty] does not depend on the merits of their individual cases, which have not been assessed other than for the purposes of the grant of permission to apply for judicial review or permission to appeal. Other claimants may have equally or even more meritorious claims to damages, and if appropriate exemplary damages, than the present claimants. There would be no principled basis, therefore, to restrict an award of exemplary damages to the present lead claimants. If an award of exemplary damages is made to the present lead claimants, a similar award would have to be made in every case. Exemplary damages are assessed by reference to the conduct of the tortfeasor. The court would, we think, have to assess an appropriate sum as exemplary damages and divide it between all successful claimants. But we do not know how many successful claimants there will ultimately be. These considerations demonstrate that exemplary damages, in a case such as the present, may be ill suited to be a remedy in judicial review proceedings, and would be in the present cases."

153. Yet further material has been disclosed by the Secretary of State since the hearing before the Court of Appeal. Mr Westgate submits that it can now be seen that this is indeed one of those exceptional cases where awards of exemplary damages are merited.
154. His submissions are detailed and elaborate. I shall endeavour to concentrate on the essential points. He submits that the conditions for an award of exemplary damages have been established because (i) from April 2006 until September 2008 the Secretary of State operated a hidden blanket policy which did not give effect to the Hardial Singh principles; (ii) the Secretary of State actively discouraged disclosure of her true detention policy with the consequence that the integrity of written reasons for detention was compromised; (iii) there was a deliberate decision not to publish the hidden policy; and (iv) the Secretary of State and/or her officials knew that, or were reckless as to whether, their actions were unlawful, preferring for political reasons to leave it to the courts to remedy the illegality. In addition, Mr Westgate submits that the Secretary of State in this litigation has fallen short of the duties of candour owed to the courts in that (v) the courts have been intentionally or recklessly misled by the Secretary of State's officials; (vi) elementary safeguards necessary to promote compliance with a public authority's duty of candour in judicial review proceedings have not been observed: in particular, the Deputy Chief of Staff of the Chief Executive of the UK Border Agency was selected as the person responsible for overseeing disclosure, when the responsibility for disclosure was that of the Treasury Solicitor; and (vii) there remain significant lacunae in the disclosure.
155. I find it convenient to take (i) to (iv) together. I have already referred at paras 16-39 above to the hidden blanket policy which did not give effect to the Hardial Singh principles. There is no doubt that such a policy was operated between April 2006 and November 2007 when a slight relaxation was effected by the introduction of Cullen 1. To a large extent, the policy that was applied until September 2008 was a blanket policy. It certainly remained a hidden policy during the whole period. But that of itself comes nowhere near being sufficient to justify an award of exemplary damages. It is the reasons why the policy was not published that are the matters of real concern. There is undoubtedly evidence to support submissions (ii) to (iv). I shall refer to some of it. A more detailed description of the internal material relating to the period between April 2006 and September 2008 that was disclosed to Davis J appears at para 43 of his judgment.
156. At least from 17 May 2006, senior officers within the Home Office, including lawyers of the Home Office Legal Advisers' Branch ("HOLAB"), expressed concern to, among others, Lin Homer (Chief Executive of the Border and Immigration Agency) ("BIA") that the policy was unlawful on the grounds that it did not satisfy the Hardial Singh principles and that it differed from the published policy. Thus, for example, on 20 July 2006, Ms Rogerson (Head of Policy for the BIA) said in an email "we are increasingly vulnerable and we should probably publish revised criteria". She suggested that they should review the criteria and consider being prepared to release FNPs in some cases, with public protection as a priority. She added that Ministers' "preferred position may be to continue to detain all FNPs and let immigration judges take any hit which is to be had by releasing on bail."
157. On 19 January 2007, Joy Munro (Deputy Director, Border Control) wrote to Ms Rogerson asking for written evidence of the lawfulness of detaining FNPs whom they were unable to remove. She referred to there being "unrest in the CCD about the power in law to detain some of those held if they are not removable". Ms Rogerson replied: "We shouldn't be dealing with this—or any such policy discussions—on email in this way. I believe a meeting would be the most profitable way forward".
158. On 13 March 2007, Stephen Braviner-Roman of HOLAB wrote to Ken Sutton (Deputy Director of the Immigration and Nationality Directorate) saying: "if the courts were to find we had not been following our policy in these cases we would face criticism, but also claims for compensation." He also referred to the fact that Simon Harker (from the Treasury Solicitor's Department) had pointed out that they "have a duty of candour to the court and cannot mislead."
159. First-hand evidence of the attitude of the Secretary of State herself is to be found in an email dated 16 July 2007 sent to Lin Homer in response to a draft bail proposal: "Is this an issue primarily about legal vulnerability rather than capacity? If so, what is the reason for worrying about this now? Have we been threatened with legal action?"
160. This prompted a series of internal emails, some of which referred to the test case of R (A) (Somalia) v Secretary of State for the Home Department [2007] EWCA Civ 804. In an email dated 17 July 2007, Mr Braviner-Roman referred to the impending hearing and said that there was "an on-going legal vulnerability". If it materialised, "we would face a liability in damages as well as severe criticism if it was said that we have maintained a policy of detaining people which was unlawful (as opposed to having a lawful policy but just getting it wrong on the facts from time to time, which is inevitable)". But the unpublished policy was not disclosed in that litigation.
161. The response to the Secretary of State's question was given in an email dated 19 July 2007 which stated that the issue was one of legal vulnerability. To this the Private Secretary of the Secretary of State replied that there did not seem to be "a strong enough or immediate enough reason to be releasing or not detaining people at this point".
162. A further insight into the thinking of the senior officers can be derived from the draft policy submission that had been circulated in May 2007. It referred to the legal advice that the Secretary of State was open to legal challenge for the reasons to which I have earlier referred. It also said: "if we were to lose a test case, we could present any change in FNP detention practice as having been forced on us by the courts". I agree with what Davis J said about this document at para 43.12 of his judgment:

"That may or may not be good politics: but it is deplorable practice, especially when it is seen that almost from day one the new unpublished policy was perceived in virtually all quarters within the department to be at least legally 'vulnerable' and in some quarters positively to be untenable and legally invalid. The tone of this draft is further confirmed by the subsequent comments that the longer the delay the more likely it would be that a court judgment 'would force us to pay out significant sums in compensation to FNPs whose detention was held to be unlawful' as well as exposing the department to criticism in the media and to reputational damage."

163. Mr Westgate is able to point to clear evidence that caseworkers were directed to conceal the true reason for detention, namely the unpublished policy, and to give other reasons which appeared to conform with the published policy. This material was disclosed seven days before the start of the hearing before the Court of Appeal. It comprises an exchange of emails in relation to a Liverpool case. On 21 March 2007, Philip Muirhead of the Criminal Casework Directorate ("CCD") in Liverpool said of the FNP under consideration that there was no valid reason to detain him. Nichola Samuel, a lawyer of HOLAB, said that if detention was appropriate in that case, this suggested that they were applying a different policy, ie that all FNPs should be detained pending deportation. Gareth Lloyd (Assistant Director of the CCD) responded that he had an email from Mandie Campbell (Director of CCD) "telling me that we must detain all FNPs until removal". In an email dated 22 March 2007, Mr Lloyd said that this was not only a Liverpool issue and in an email sent on the following day, he said: "We just detain as instructed and choose the most defendable option in our opinion."
164. From the above, it seems to me to be clear that there was a deliberate decision not to publish the hidden policy. The material that has now been disclosed suggests that the assessment made by Davis J at para 205 may have been somewhat generous to the Secretary of State and her senior officials. It is true that they did not know and could not have known that the policy that was being applied was bound to be struck down as unlawful, but they certainly knew that it was "vulnerable to legal challenge" and that it did not accord with the published policy.
165. Nevertheless, the question remains whether, regrettable though the behaviour of the Secretary of State and her senior officials may have been, it was sufficiently outrageous and sufficiently "unconstitutional, oppressive or arbitrary" to merit awards of exemplary damages. I approach this question without regard to the allegations of lack of candour in the litigation. This is because the role of exemplary damages is to punish the commission of the underlying tort and not the subsequent conduct of the litigation. Any disapproval of the conduct of the litigation can be marked by an appropriate order for costs or by an increased award of (compensatory) aggravated damages: Thompson v Commissioner of Police of the Metropolis [1998] QB 498, 517D per Lord Woolf MR.
166. Whether the high threshold for the award of exemplary damages has been crossed in any particular case is ultimately a matter of judgment. Opinions can reasonably differ on whether a defendant's conduct has been so outrageous and so unconstitutional, oppressive or arbitrary as to justify the imposition of the penalty of exemplary damages. An appellate court should not interfere with the judgment of the court below unless that judgment is plainly wrong. On the material that was before him, Davis J was entitled to reach the conclusion that he reached. In my view, the Court of Appeal were also entitled to reach the conclusion that they reached on the more extensive material that was before them. Both the judge and the Court of Appeal applied the correct test. In particular, the Court of Appeal were right to place some weight on the fact that the Secretary of State had the statutory power to detain the appellants pending deportation and that, although she in fact exercised that power unlawfully, she could have done so lawfully. They were also right to say that, if her conduct is properly to be described as "unconstitutional, oppressive or arbitrary", it was at the less serious end of the scale. It is material that there is no suggestion that officials acted for ulterior motives or out of malice towards the appellants. Nevertheless, there was a deliberate decision taken at the highest level to conceal the policy that was being applied and to apply a policy which, to put it at its lowest, the Secretary of State and her senior officials knew was vulnerable to legal challenge. For political reasons, it was convenient to take a risk as to the lawfulness of the policy that was being applied and blame the courts if the policy was declared to be unlawful.
167. The Court of Appeal identified at para 123 of their judgment a further point which militated against awards of exemplary damages to the appellants. Where there is more than one victim of a tortfeasor's conduct, one award of damages should be made which should be shared between the victims, rather than separate awards of exemplary damages for each individual: see Riches v News Group Newspapers Ltd [1986] QB 256. This is because the purpose of the award is to punish conduct rather than compensate the claimants. In Riches, the victims of the tort were a small class who were all before the court. But where (as in the present case) there is potentially a large number of claimants and they are not all before the court, it is not appropriate to make an award of exemplary damages: see AB v South West Water Services Ltd [1993] QB 507, 527B-D per Stuart-Smith LJ and p 531D-E per Sir Thomas Bingham MR. Unless all the claims are quantified by the court at the same time, how is the court to fix and apportion that punitive element of the damages? If the assessments are made separately at different times for different claimants, how is the court to know that the overall punishment is appropriate? The Court of Appeal were right to regard this a further reason why it was not appropriate to award exemplary damages in the present case.
168. There is yet one further point. It is unsatisfactory and unfair to award exemplary damages where the basis for the claim is a number of serious allegations against named officials and Government Ministers of arbitrary and outrageous use of executive power and those persons have not been heard and their answers to the allegations have not been tested in evidence. In a private law action, they would almost certainly have been called to give evidence. But oral evidence is rarely adduced in judicial review proceedings and, understandably, it was not adduced in the present case.

Overall conclusion

169. I would, therefore, allow these appeals. For the reasons that I have given, the Secretary of State is liable to both appellants in the tort of false imprisonment on the narrow ground that she unlawfully exercised the statutory power to detain them pending deportation because she applied an unpublished policy which was inconsistent with her published policy. The appellants are, however, only entitled to nominal damages because, if the Secretary of State had acted lawfully and applied her published policy, it is inevitable that both appellants would have been detained. As regards the discrete question whether the detention of Mr Lumba was in any event unlawful on the grounds of a breach of the Hardial Singh principles, I would remit this to a High Court judge. Finally, neither appellant is entitled to exemplary damages.

LORD HOPE

170. In agreement with Lord Walker, Lady Hale, Lord Collins, Lord Kerr and Lord Dyson I would hold that the Secretary of State is liable to the appellants in the tort of false imprisonment because she applied to them an unpublished policy which was inconsistent with her published policy, and I too would remit to a High Court judge the question whether Mr Lumba's detention was unlawful as being in breach of the Hardial Singh principles. As I am anxious to avoid adding to the length of the court's judgment, I shall simply say that I am in full agreement with all the reasons that Lord Dyson has given on these issues except that I do not have the same difficulty as he has with the use of the phrase "abuse of power" by Lord Walker (see para 69, above). There are only three points on which I wish to comment.
171. First, as Lord Phillips has noted (see para 258, below), it was common ground in these appeals that Lord Dyson correctly summarised the effect of Woolf J's judgment in R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704. Nevertheless he has indicated that he does not consider that the first and second principles that Lord Dyson has extracted from it can properly be derived from what Woolf J said in that judgment. For my part, I think that Lord Dyson's summary, which he has taken from his judgment in R (I) v Secretary of State for the Home Department [2003] INLR 196, para 46, is accurate and I would endorse it.
172. We can, of course, read what Woolf J said for ourselves, and there are no doubt various ways of expressing what Lord Dyson has taken from it. The essential point, as Lady Hale has explained (see para 199, below), is that the detention must be for the statutory purposes of making or implementing a deportation order and for no other purpose. Lord Phillips says that he can see no justification for reading the terms of the statute in this way, that it places an unjustified restriction on the Secretary of State's power of detention and that he does not believe that Woolf J intended to say what Lord Dyson has attributed to him. He would read Hardial Singh as concerned only with the length of time that was reasonably necessary to effect deportation and not the purpose of the detention. I am unable, with respect, to agree with this interpretation of Woolf J's judgment.
173. There are two points that need to be considered. The first is, what do the words that Woolf J used actually mean? The second is whether, if they have the meaning that Lord Dyson has attributed to them, this limitation on the statutory power can properly be read into the statute. As to the first point, in the second sentence of the relevant paragraph (see para 262 of Lord Phillips's judgment) Woolf J makes it plain that in his view the statutory power can only authorise detention if the individual is being detained pending the making of a deportation order or pending his removal "and for no other purpose". I think that his judgment could not be clearer on this issue, and that Lord Dyson has captured the essential points that Woolf J made accurately in his summary.
174. As for the second, it must be recognised that until now Woolf J's statement of the limitations to the statutory power has never been questioned. I think that there are good reasons for this. To confine the limitation to the period of the detention only and release the Secretary of State from the limitation as to purpose which Woolf J identified would greatly reduce the protection which, as I read it, his judgment was designed to give to the detainee. For obvious reasons that, if it were to happen, would be regrettable. It would, as Lord Phillips indicates in para 269, enable the Secretary of State to exercise the power to detain simply because he believed that the detainee would, if released, pose a threat to national security. I do not think that article 5(1)(f) of the European Convention on Human Rights permits this interpretation of the statutory power. If Parliament had intended that the power to detain could be used for a purpose other than the making or effecting of a deportation order, it would have had to have said this expressly and it has not done so. It is hard also to see how the limitation as to time which Lord Phillips accepts can be read into the statute can be tested without having regard to the purpose for which the detainee is being held. The limitation as to time and the limitation as to purpose are really two sides of the same coin. They cannot be separated from each other. I think that Woolf J was right to recognise this and that there are sound reasons for all that he said about the limitations that must be understood to qualify the statutory power.
175. Second, I cannot accept Lord Phillips's conclusion that the application of the secret policy did not render the detention of the appellants unlawful. The basis for that view is that, if the published policy had been applied they would have been detained anyway and that, had they challenged their continued detention, they would have had no legitimate expectation of obtaining an order for their release. This is the causation argument which, for the reasons Lord Dyson gives in paras 62-68 with which I agree, he rejects. The key point, as I see it, are that we are dealing in this case with the tort of false imprisonment. Torts of this kind are actionable per se regardless of whether the individual suffers any harm. While not every breach of public law will give rise to a cause of action on this ground, the history of this case shows that there was here a serious abuse of power which was relevant to the circumstances of the appellants' detention. If the rule of law is to be sustained, the detention must be held to have been unlawful. The appellants were being detained without regard to the purpose for which the Secretary of State was authorised to exercise the power by the statute. The court must insist that powers of detention are exercised according to law. If they are not, those who have abused their powers must accept the consequences. It is no answer for them to say that they could, had they put their mind to it, have achieved the same result lawfully by other means.
176. Third, I agree that this is not a case for exemplary damages. But, for the reasons given by Lord Walker and Lady Hale, I would hold that the breach of the appellants' fundamental rights that has occurred in these cases should not be marked by an award only of nominal damages. An award on ordinary compensatory principles is, of course, out of the question. It is plain that the appellants would not have had any prospect of being released from detention if the Secretary of State had acted lawfully. So they cannot point to any quantifiable loss or damage which requires to be compensated. But the conduct of the officials in this case amounted, as Lord Walker says (see para 194, below), to a serious abuse of power and it was deplorable. It is not enough merely to declare that this was so. Something more is required, and I think that this is best done by making an award of damages that is not merely nominal.
177. The principles on which damages for breaches of fundamental rights are to be assessed in situations such as this are not greatly developed, as Elias CJ pointed out in the Supreme Court of New Zealand in Taunoa and others v Attorney General and another [2007] 5 LRC 680, [2007] NZSC 70 , para 108. But some guidance is available from judgments which the Judicial Committee of the Privy Council has given where a constitutional right has been infringed. In Attorney General of Trinidad and Tobago v Ramanoop [2006] 1 AC 328 Lord Nicholls of Birkenhead added his own words to those of the Board in Harrikissoon v Attorney General of Trinidad and Tobago [1980] AC 265 as to how an award of damages should be assessed in such a case. In para 18 he observed that when exercising its constitutional jurisdiction the court is concerned to uphold, or to vindicate, the constitutional right. In para 19 he said that an award, not necessarily of substantial size, might be needed to reflect the sense of public outrage, emphasise the gravity of the breach and deter further breaches. The law on this matter is still in the process of being worked out, so I should like to say just a little more about it.
178. Although such an award is likely in financial terms to cover much the same ground as an award by way of punishment in the sense of retribution, punishment in that sense is not its object. The expressions "punitive damages" or "exemplary damages" are therefore best avoided. Allowance must be made for the importance of the right and the gravity of the breach in the assessment of any award. Its purpose is to recognise the importance of the right to the individual, not to punish the executive. It involves an assertion that the right is a valuable one as to whose enforcement the complainant has an interest. Any award of damages is bound, to some extent at least, to act as a deterrent against further breaches. The fact that it may be expected to do so is something to which it is proper to have regard.
179. As for the amount to be awarded, an award is referred to as a conventional award when it is incapable of being calculated arithmetically as there is no pecuniary guideline which can point the way to a correct assessment: Lim Poh Choo v Camden and Islington Area Health Authority [1980] AC 174, 189G-H, per Lord Scarman. In most cases the sum to be awarded can be derived from experience and from awards in similar cases: Ward v James [1966] 1 QB 273, 303, per Lord Denning MR. But that cannot be said of this case. So I would turn for guidance to what Lord Bingham of Cornhill said in Rees v Darlington Memorial Hospital NHS Trust [2003] UKHL 52, [2004] 1 AC 309, para 8. The conventional award that he had in mind in that case to mark the injury and loss due to the unwanted child was not, and was not intended, to be compensatory. It was not the product of calculation, nor was it derived from awards in other similar cases. But it was not a nominal, let alone a derisory, award. Its purpose was to afford some measure of recognition of the wrong done.
180. In agreement with Lord Steyn, I regarded the idea of a conventional award under the tort system in that case as contrary to principle: Rees v Darlington Memorial Hospital NHS Trust paras 46, 70-77. But I do not think that it is open to the same objection in the present context. In this case the factors referred to by Lord Nicholls in Ramanoop must be the primary consideration. There must be some recognition of the gravity of the breach of the fundamental right which resulted in false imprisonment, and account should be taken of the deterrent effect of an award lest there be the possibility of further breaches. But account should also be taken of the underlying facts and circumstances which indicate that it should not be more than a modest one. It should do no more than afford some recognition of the wrong done, without being nominal or derisory. Lord Walker has suggested that an award of £1,000 to each appellant would be appropriate. We have no yardstick by which that sum can be measured to test its accuracy. Given the purpose of the award, I see no reason to disagree with his assessment although I, for my part, would have arrived at a substantially lower figure.

LORD WALKER

181. The issue on which Lord Brown differs from Lord Dyson is one of high importance and great difficulty. Its high importance is obvious. Lord Dyson cites Lord Bridge in R v Deputy Governor of Parkhurst Prison Ex p Hague [1992] 1 AC 58, 162, "The tort of false imprisonment has two ingredients: the fact of imprisonment and the absence of lawful authority to justify it". It is a species of trespass to the person and as such a tort actionable without the need for proof of special damage. The notion that no more than nominal damages should ever be awarded for false imprisonment by the executive arm of government sits uncomfortably with the pride that English law has taken for centuries in protecting the liberty of the subject against arbitrary executive action. It would in Lord Brown's view seriously devalue the whole concept of false imprisonment. The difference of opinion between two Justices with so much expert knowledge in the field of public law, on a point of such high importance, demonstrates its difficulty. I hardly need say that it is with diffidence that I make any contribution to the debate.
182. Lord Brown's approach to the solution to the problem is to distinguish between cases where there is no lawful authority to detain a person (including cases where a precondition to lawful detention has not been satisfied) and cases where there is a power to detain, but in the exercise of that power the decision-maker has been in breach of some public law duty. In cases (or at any rate some cases) of the flawed exercise of a power to detain there is (so the argument goes) no false imprisonment at all, and so the question of awarding damages, whether nominal or otherwise, simply does not arise.
183. This solution has considerable attractions. The proposed distinction is based on the difference between the existence (or rather non-existence) of a lawful authority to detain, and a defective exercise of an authority which does exist: see the observations of Lord Brown in R (Khadir) v Secretary of State for the Home Department [2006] 1 AC 207, para 33. The difficulty that I feel is whether the distinction, though clear enough in theory, can cope with the variety and complexity of the problems that arise in practice, as illustrated by the numerous decided cases cited to the court. I also have difficulty (or perhaps this is another way of putting the same point) in reconciling the basic existence/exercise distinction with the four categories which Lord Brown extracts from his analysis of successful claims for false imprisonment by executive action. His four categories are (1) no power to detain; (2) failure to satisfy a precondition to exercising the power to detain; (3) detention beyond the scope of the power to detain; and (4) power to detain limited by published official policy.
184. The distinction is clear enough in extreme cases. R v Governor of Brockhill Prison Ex p Evans (No 2) [2001] 2 AC 19 was a case (in Lord Brown's category (1)) in which there was simply no lawful authority for the claimant's detention after the date of expiration of her term of imprisonment, when properly adjusted for time spent on remand (the complicated statutory provisions as to allowances for multiple periods spent on remand had been misinterpreted). At or near the other extreme was the breach of a public law duty to see that an asylum-seeker detained at a detention centre received a medical examination within 24 hours. The omission to provide a medical examination was an administrative failing but it did not render the detention unlawful: R (HK (Turkey)) v Secretary of State for the Home Department [2007] EWCA Civ 1357.
185. Lord Brown supplements his category (1) by category (2) so as to let in a range of cases in which there is for the time being no lawful authority to detain because some precondition has not been satisfied. A well-known example of this is Christie v Leachinsky [1947] AC 573, where a man arrested without a warrant was not correctly informed of the offence which gave rise to the power of arrest without a warrant. Moreover categories (1) and (2) must be expanded to cover cases in which detention, although initially lawful, has become unlawful because of a failure to carry out some procedure or satisfy some condition of precedent fact required by statute. A procedural example is Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662, where the provisions of the Police and Criminal Evidence Act 1984 required detention in police custody to be reviewed every six hours. The claimant had been detained at 11.25pm and the police officers' failure to review his detention by 5.25 am on the next day made his detention unlawful until it was reviewed (and his detention continued) at 7.45 am, and the Court of Appeal refused to overturn or reduce the award of £500 damages, even though it accepted that his detention would have continued if a review had been carried out at the right time. By contrast in R (Saadi) v Secretary of State for the Home Department [2002] 1 WLR 3131 the statement of incorrect and inappropriate reasons on an official form handed to detained asylum seekers was not treated as a failure to satisfy a condition precedent affecting the legality of their continued detention.
186. The distinction between these two cases is that the relevant statutes were interpreted in one case as imposing a condition which had to be satisfied if continued detention was to be lawful, and as not imposing such a condition in the other case. It is, as Laws LJ emphasised in R (SK (Zimbabwe)) v Secretary of State for the Home Department [2009] 1 WLR 1527, paras 21 and 25, an issue of statutory construction. In that case the Court of Appeal held that compliance with rule 9 of the Detention Centre Rules 2001 (calling for a monthly review of detention with written reasons given to the detained person) was not a precondition to the continuation of lawful detention. (See Laws LJ at paras 31-35 and Keene LJ agreeing, with some hesitation, at para 47.)
187. Lord Brown's category (3) is detention beyond the "scope" of the relevant power. Laws LJ in SK (Zimbabwe), para 21, referred to the "reach" of the power. These expressions, as I understand them, approximate to the object or purpose for which Parliament has conferred the power. The importance of the statutory purpose has been recognised since Padfield v Minister of Agriculture, Fisheries and Food [1968] AC 997 and (as Lord Dyson notes in para 30 of his judgment) the Hardial Singh principles (see R v Governor of Durham Prison Ex p Hardial Singh [1984] 1 WLR 704) reflect the application of the wide general principle of not deviating from the statutory purpose to the particular field of the detention of foreign nationals pending deportation. Woolf J made this clear in Hardial Singh itself at p 706.
188. The wide general principle of not deviating from the statutory purpose is of such fundamental importance in public law that it can be seen as going to the existence of the power, rather than merely to its exercise. In law the power exists only for the purposes for which Parliament has conferred it on the executive. In Tan Te Lam v Superintendent of Tai A Chau Detention Centre [1997] AC 97, 113, the Privy Council put it very simply: "If removal is not pending . . . the director has no power at all." So I would see Lord Brown's category (3) as fitting reasonably comfortably into his basic existence/exercise classification.
189. Determining the purpose for which Parliament has conferred a power is also a process of statutory construction. That process should not, in my opinion, be minutely elaborated. I entirely accept the exposition of the Hardial Singh principles by Dyson LJ (as he then was) in R (I) v Secretary of State for the Home Department [2003] INLR 196, para 46. It goes no further than spelling out clearly what is necessarily implicit in the purpose of detention with a view to deportation, and it has been widely cited and followed. I have more of a problem in seeing Langley v Liverpool City Council [2006] 1 WLR 375 as a case where the police departed from the statutory purpose in exercising their power of removal under section 46 of the Children Act 1989 when an emergency protection order under section 44 of that Act had already been made. The purpose of both statutory powers (one conferred on the police, the other on the court) was child protection in an emergency situation. I regard Langley as a finely-balanced decision in which the Court of Appeal held that well-intentioned police action, directed to child protection and taken under pressure of circumstances, was nevertheless an unlawful manner of exercising the section 46 power. Thorpe LJ (at para 79) reached that conclusion with some reluctance.
190. It is with Lord Brown's category (4), however, that I have the greatest difficulty. Here the issues are concerned with official policies – how rigid or flexible they may be, whether and in how much detail they should be published, whether (in these appeals) a policy with a presumption towards detention is permissible. Official decision-makers need policies for obvious reasons. Although decisions in the field of immigration law are all taken in the name of the Secretary of State, only a tiny handful of cases are actually decided by the Secretary of State personally. Decisions are taken by a small army of officials at different levels, and they need guidance in order to achieve consistency in decision-making. Members of the public, or those of the public liable to be affected, should know where they stand, and so they are entitled to know, at least in general terms, the content of the official policies. This is not a matter of being faithful to the purposes of statutory powers, but of seeing that they are exercised consistently and fairly. There is a helpful discussion of these points in the judgment of the Court of Appeal in these appeals, prepared by Stanley Burnton LJ, at paras 53-58.
191. It is here that Lord Diplock's dictum in Holgate-Mohammed v Duke [1984] AC 437 calls for consideration. In a passage (at p.444) quoted by both Lord Brown and Lord Dyson, Lord Diplock expressed the view that Wednesbury principles are applicable to determining the lawfulness of the exercise of a statutory power of arrest not only in proceedings for judicial review but also in an action for damages for false imprisonment. As Lord Dyson says, there seems to have been little argument on this point in the House of Lords. Nor has there been much discussion of it in later authorities. It was cited and followed by the Court of Appeal in D v Home Office (Bail for Immigration Detainees intervening) [2006] 1 WLR 1003 (see especially Brooke LJ at para 111). In SK (Zimbabwe), Holgate-Mohammed was cited by counsel but not referred to in the judgments in the Court of Appeal. Holgate-Mohammed and D v Home Office (Bail for Immigration Detainees intervening) were both discussed at some length in the Court of Appeal in these appeals (paras 50-52, and, in relation to causation, paras 82-84). The Court of Appeal rightly regarded itself as bound by the latter decision.
192. This court is not bound to follow the Court of Appeal's acceptance of Lord Diplock's dictum, and for my part I would refrain from giving it unequivocal approval. Mr Beloff QC (appearing for the Secretary of State in this court) put forward some persuasive submissions in favour of an alternative approach. They are noted in paras 76 and 86 of Lord Dyson's judgment. The first two submissions would make a qualification or exception, for the purposes of a private law claim for damages for false imprisonment, to the Anisminic equation of any significant public law error with lack of jurisdiction (see Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147). Lord Dyson, in para 87, dismisses these in a single sentence as putting the clock back to pre-Anisminic days.
193. With great respect, I think there may be more to be said about it. Anisminic was one of the seminal cases in the development of modern public law. But its full implications are still open to debate. The context in which it equated wrongful exercise of jurisdiction with excess of jurisdiction (the court's response to an ouster clause in the Foreign Compensation Act 1950) was far removed from a private law claim for damages for false imprisonment. It is a big step to extend the principle to a claim for damages for false imprisonment, where a defendant may have his professional reputation at stake and may not enjoy the procedural protections which attend judicial review (strict time limits, and the discretionary nature of the remedy granted). I would prefer a more demanding test, that in a case where an extant statutory power to detain has been wrongly used there would be a private law claim only if the misuse amounted to an abuse of power (including but not limited to cases of misfeasance or other conscious misuse of power).
194. However, it is in my opinion unnecessary to decide the point in these appeals because the conduct of officials, including some senior officials, of the Home Office between April 2006 and September 2008 amounted to a serious abuse of power. Lord Dyson has in paras 154-165 of his judgment described in restrained language how senior officials were well aware of the risk (indeed the likelihood) of challenge and decided to run the risk, (including the proposal to "let immigration judges take any hit"), and how further damaging facts were disclosed by stages, some before Davis J, some before the Court of Appeal, and some only in this Court. Wherever the line is to be drawn (if, as I think, a line does need to be drawn between public law errors in detention policies which do or do not give rise to an action for false imprisonment) these appeals must in my view fall on the wrong side of the line from the Secretary of State's point of view.
195. I agree with Lord Dyson (paras 165-168) that despite the deplorable official conduct this is not a case for exemplary damages. But in my view it is not a case for nominal damages either. Apart from cases concerned with constitutional rights in the Caribbean, (the line of authority starts with Attorney-General of St Christophers, Nevis and Anguilla v Reynolds [1980] AC 607), the common law has always recognised that an award of more than nominal damages should be made to vindicate an assault on an individual's person or reputation, even if the claimant can prove no special damage. (See Mayne & McGregor on Damages, 18^th ed. (2009) paras 42-008 to -009). In these appeals, each claimant had a very bad criminal record and would undoubtedly have been kept in custody under the Secretary of State's published policies. They cannot therefore establish a claim to special damages. But the argument on causation does not completely defeat their claims. I would award each claimant the sum of £1,000 damages. I would remit the case of Mr Lumba as Lord Dyson proposes. On every point on which I have not expressed disagreement or doubt I am in respectful agreement with the judgment of Lord Dyson.

LADY HALE

196. I agree entirely with Lord Brown that far and away the most important issue in this case, as it is in the case of SK (Zimbabwe) [2009] 1 WLR 1527, is whether the breach of a public law duty on the part of the person authorising detention is capable of rendering that detention unlawful. If it is, the second question is which breaches of public law duties have that consequence; and the third question is whether it makes any difference that the person authorising the detention both could and would have done so lawfully, without breaching the public law duty in question, had the point been drawn to his attention. If that does not make a difference to liability, a fourth question is whether the fact that the person detained both could and would have been lawfully detained is of any relevance to the assessment of his damages for false imprisonment.
197. But I differ from Lord Brown in his view that the answer to the last of these questions should govern the answer to the first, second or third question: in other words, that if we take the view that no compensatory damages are payable in a case such as this it should follow that there is no liability in the first place. Forcefully and attractively though that argument is made, it does put the cart before the horse. It also fails to acknowledge that false imprisonment is a trespass to the person, actionable per se without proof that the claimant has suffered any harm for which the law would normally grant compensation.
198. As to the first question, this is a stronger case than is still before the Court in SK (Zimbabwe) because the illegality alleged (and now admitted) went to the criteria for detention rather than to the procedure for authorising it. The statutory power to detain under paragraph 2(2) and (3) of schedule 3 to the Immigration Act 1971 (quoted by Lord Dyson at paragraph 4 of his judgment) is, on its face, very broad. Provided that the detainee has been notified of a decision to make a deportation order against him, and he is not detained in pursuance of the sentence or order of a court, he may be detained pending the actual making of the order (para 2(2)). Once the deportation order is made, he may be detained pending his removal or departure from the United Kingdom (para 2(3)). However, since at least the case of R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704, it has been recognised that there are limitations implicit in these powers: the detention must be for the statutory purpose of making or implementing a deportation order and for no other purpose; hence it cannot be continued once it becomes clear that it will not be possible to effect deportation within a reasonable period; the Secretary of State must act with reasonable diligence and expedition to bring this about; and in any event the detention cannot continue for longer than a period which is reasonable in all the circumstances.
199. These limitations were devised long before the Human Rights Act and have been accepted without question ever since. They stem from the long-established principle of United Kingdom public law that statutory powers must be used for the purpose for which they were conferred and not for some other purpose: Padfield v Minister of Agriculture, Fisheries and Food [1968] AC 997. They were not inspired by article 5(1)(f) of the European Convention for the Protection of Human Rights and Fundamental Freedoms and it does not follow that, because detention would be permissible under article 5(1)(f), it is also permissible under United Kingdom law.
200. The last restriction referred to above has not hitherto been questioned but it is the most questionable, for the Secretary of State may genuinely be doing all that she can to effect deportation, and deportation may still be a realistic possibility, but even so the deportee may have been detained for so long that it is no longer lawful to keep him there. That this has never until now been questioned indicates how strong are the objections to indefinite detention by order of the executive. But it undoubtedly gives rise to some difficult questions, as is amply shown by Lord Dyson's discussion, in paragraphs 102 to 128 of his judgment, of the matters to be taken into account in deciding whether or not the period of detention is reasonable.
201. In addition to such substantive limitations, the law has also imposed procedural requirements upon apparently open-ended statutory powers. In common with Lord Dyson, I do not think that it matters whether these are characterised as implied conditions precedent or implied procedural requirements. The effect is the same. The best known example is Cooper v Wandsworth Board of Works (1863) 14 C.B. (N.S) 180M, 143 ER 414. The Board had power to alter or demolish a house if the builder had not given 7 days' notice of his intention to build. The Court held that the common law imposed upon the Board a requirement to notify the builder before they decided to demolish the house, so that he could present arguments why they should not. As the Board had not given the builder such an opportunity before demolishing the house, they were liable to him in damages for trespass. It is true that Byles J founded his opinion partly on the basis that the Board had also failed in their express statutory duty to notify the builder of their demolition order: but the majority based their opinions on the broader principle that he had a right to be heard before the order was made: in other words, there were public law duties inherent in the apparently open-ended statutory power. Another example of the same principle is Christie v Leachinsky [1947] AC 573 where the common law implied a duty, when exercising a power of arrest, to tell the arrested person the power under which he was being arrested, so that he might know whether or not he could resist arrest. Once again, the police were liable in false imprisonment.
202. The question is whether the same principles apply where the requirement in question is the duty, imposed by the common law, for the Secretary of State and his officials to comply with a published policy, unless there is good reason not to do so. As I understand it, Lord Brown accepts that they may indeed do so, for he agrees that if the published policy further narrows and defines the circumstances in which the power will be exercisable, the Secretary of State may not lawfully depart from that. It is on that basis that he considers R (Nadarajah) v Secretary of State for the Home Department [2004] INLR 139 to have been correctly decided.
203. Support for the proposition that the ordinary requirement to observe public law duties may restrict the lawful exercise of a statutory (or common law) power which would otherwise authorise a trespass can be found in the dictum of Lord Diplock in Holgate-Mohammed v Duke [1984] 1 AC 437, 443, quoted by Lord Dyson at para 73 of his judgment. On the face of it, this was a lawful arrest. The argument was that the police officer had arrested the claimant for the wrong reason: to get her to confess. There can be little doubt that, had the House of Lords concluded that this was indeed an impermissible consideration, they would have held that the arrest was unlawful and that the claimant was entitled to the damages she sought. This was, after all, an action for damages for false imprisonment in which the claimant had succeeded in the county court. If the House had thought that she would not have been entitled to damages in any event, it would surely have said so.
204. I agree with Lord Walker that it is not necessary to hold that every single departure from policy, or even failure in the decision-making process, attracts these consequences in order to hold that they apply in this case. The discrepancy between what the policy said should happen in these cases and what was actually happening is stark. The claimants were being dealt with, not under the published, lawful policy, but under an unpublished, unlawful policy or practice. Yet it is difficult not to have some sympathy for the officials involved. The Government had been hit by a perfect storm in April 2006 when the popular press discovered that foreign national prisoners were being released after serving their sentences without any consideration being given to whether or not they should be deported. It had cost the then Home Secretary his job. The immediate answer was not to let any of them go. This was at odds with the published policy, which presumed against the use of detention powers. Officials knew this and they also knew that the policy needed amendment. But they found it very difficult to devise a policy for publication which would be both lawful and acceptable to ministers. Ministers wanted a near blanket ban on release, whereas the law requires some flexibility to respond to the circumstances of the particular case. So the situation dragged on for many, many months.
205. These are just the sort of circumstances, where both Ministers and their civil servants are under pressure to do what they may know to be wrong, in which the courts must be vigilant to ensure that their decisions are taken in accordance with the law. To borrow from the civil servants' correspondence, the courts must be prepared to take the hit even if they are not. The law requires that decisions to detain should be made on rational grounds and in an open and transparent way and not in accordance with arbitrary rules laid down by Government and operated in secret. One of the most disturbing features of this sorry tale is that the case-handling officials had to give reasons for their decisions which were not what their real reasons were.
206. The European Convention on Human Rights and the Strasbourg Court have not imposed the same requirements of proportionality upon detention with a view to deportation under article 5(1)(f) as they have upon detention under other provisions in article 5(1). But any deprivation of liberty has to be 'in accordance with a procedure prescribed by law'. Unless the 'law' has certain essential characteristics, there is a risk that detentions may be arbitrary. That is why the open-ended common law power to detain people who lack the capacity to make decisions for themselves on grounds of necessity was found incompatible with article 5(1)(e): see L v United Kingdom (2005) 40 EHRR 32. There is every reason to think that Strasbourg would find a secret policy which presumed in favour of the detention of every foreign national prisoner open to the same objections. The common law is just as respectful of the liberty of the person, and just as distrustful of arbitrary and secret decision-making by officials acting on behalf of Government, as is the Convention.
207. I would therefore answer 'yes' to the first question. I would also answer the second question in the way proposed by Lord Dyson. In other words, the breach of public law duty must be material to the decision to detain and not to some other aspect of the detention and it must be capable of affecting the result – which is not the same as saying that the result would have been different had there been no breach.
208. The third question is whether it makes any difference that, in these particular cases, if the officials had been operating the published policy they both could and would have authorised the detention of these appellants. This would not normally make any difference. In Christie v Leachinsky, the officers could have made a lawful arrest and only chose to make an unlawful arrest 'for convenience', but they were still liable for false imprisonment. In Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662, the custody officer could have reviewed the case when he should have done and would no doubt have authorised the continued detention then, but there was still liability in false imprisonment. For all the reasons given by Lord Dyson, there is no basis for drawing a distinction according to the reason why the detention is unlawful, permitting what has been referred to as a 'causation defence' in some cases but not in others.
209. The most difficult question is whether this should make any difference to the measure of damages awarded. I quite agree with Lord Brown that the importance of strict adherence to the law when depriving people of their liberty should not be devalued. Awarding the same measure of damages, irrespective of whether or not the person could and would have been lawfully detained, serves to reinforce the importance of this principle. Also, if no distinction, according to the reason why the detention is unlawful, is to be drawn in relation to the second question, there should be no such distinction in relation to damages. If we are to hold that a person who could and inevitably would have been detained lawfully had the correct criteria or procedures been applied is not to be compensated for the loss of liberty, then this must apply irrespective of the reason why depriving him of his liberty was unlawful. We cannot single out these public law cases for special treatment.
210. In most cases of false imprisonment, the problem will not arise, because the detainer does not have a choice between acting lawfully and acting unlawfully. The prison governor in R v Governor of Brockhill Prison, Ex p Evans (No 2) [2001] 2 AC 19 had no power to detain the prisoner beyond the properly calculated term of her imprisonment: the fact that he was acting in compliance with the law as it had previously been thought to be was neither here nor there. The police officer in Langley v Liverpool City Council [2006] 1 WLR 375 had, as the Court of Appeal found in what I agree with Lord Walker was a finely-balanced decision, no power to use his power to take a child into police protection under section 46 of the Children Act 1989 when the child could and should have been protected by social workers' implementing an emergency protection order under section 44. The immigration officers in Kuchenmeister v Home Office [1958] 1 QB 496 had no power to detain the claimant in such a way as to prevent his transiting from one aircraft to another at London airport.
211. However, where the defendant has failed to comply with a procedural requirement, there is always the possibility that the deed might have been done lawfully. But the whole point of procedural requirements, such as those in Cooper v Wandsworth Board of Works or Christie v Leachinsky, is that the person whose rights are being infringed should have an opportunity of challenging this. So it will rarely be possible to be confident that, had the correct procedure been followed, the outcome would have been the same. Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662 is an example where it was possible. This case is another, because it would appear that, had the decision-makers applied the published, lawful policy rather than the unpublished, unlawful policy, they would inevitably have reached the same conclusion.
212. Insofar as damages for false imprisonment are meant to compensate for the loss of liberty, it is difficult to see why a claimant should be compensated for the loss of something which he would never have enjoyed. But, left to myself, I would not regard this as the end of the story. Trespass, whether to person or property, has always been actionable per se, without proof of anything which the law regards as damage. The tort is complete when a direct interference with person or property without lawful justification is established. Usually, there will also be some harm done which the wrongdoer must remedy, either by damages or in some other way. Take, for example, the case of the neighbour who put a row of ridge tiles on his neighbour's roof. They did no harm to the roof; they did not diminish the value of the house in any way; indeed many might think them an enhancement; but the claimant did not want them there and successfully sued for trespass. The obvious remedy was to remove the tiles or pay the cost of the claimant's doing so.
213. But suppose there is no such harm. The claimant has nevertheless been done wrong. Let us also assume, as is the case here, that the circumstances are not such as to attract punitive or exemplary damages. Is our law not capable of finding some way of vindicating the claimant's rights and the importance of the principles involved? A way which does not purport to compensate him for harm or to punish the defendant for wrongdoing but simply to mark the law's recognition that a wrong has been done?
214. As Lord Collins explains, the concept of vindicatory damages has been developed in some Commonwealth countries with written constitutions enshrining certain fundamental rights and principles and containing broadly worded powers to afford constitutional redress (and also in New Zealand, which has no written Constitution but does have a Bill of Rights: Taunoa v Attorney General [2008] 1 NZLR 429). In an early article on the Canadian Charter, 'Damages as a remedy for infringement of the Canadian Charter of Rights and Freedoms' (1984) 62(4) Canadian Bar Review 517, Marilyn Pilkington argued that an award of damages under section 24(1) of the Charter should not be limited by the common law principles of compensation. In a proper case it might be designed to deter repetition of the breach, or to punish those responsible or to reward those who expose it. In Attorney General of St Christopher, Nevis and Anguilla v Reynolds [1980] AC 607, the Privy Council upheld a modest award of exemplary damages for breach of a constitutional right. But there can be a middle course between compensatory and exemplary damages. In Jorsingh v Attorney General (1997) 52 WIR 501, de la Bastide CJ and Sharma JA in the Court of Appeal of Trinidad and Tobago both said, albeit obiter, that the remedies available under section 14(2) of the Constitution were not limited by common law principles. Sharma JA said, at p 512, that

"The court is mandated to do whatever it thinks appropriate for the purpose of enforcing or securing the enforcement of any of the provisions dealing with the fundamental rights. . . . Not only can the court enlarge old remedies; it can invent new ones as well, if that is what it takes or is necessary in an appropriate case to secure and vindicate the rights breached."

215. Since then, the concept of vindicatory damages for breach of constitutional rights has been recognised by the Judicial Committee of the Privy Council, in Attorney General of Trinidad and Tobago v Ramanoop [2005] UKPC 15, [2006] 1 AC 328 and Merson v Cartright and Attorney General [2005] UKPC 38 (Bahamas); applied to breach of constitutional provisions other than the fundamental rights and freedoms, in Fraser v Judicial and Legal Services Commission [2008] UKPC 25 (St Lucia) and Inniss v Attorney General [2008] UKPC 42 (St Kitts), which involved the dismissal of respectively a magistrate and a High Court registrar in breach of the procedures laid down in the Constitution; and applied to the breach of fundamental rights in Takitota v Attorney General [2009] UKPC 11 (Bahamas), where the Board quoted from Lord Nicholls in Ramanoop, at para 19:

"An award of compensation will go some distance towards vindicating the infringed constitutional right. How far it goes will depend on the circumstances, but in principle it may well not suffice. The fact that the right violated was a constitutional right adds an extra dimension to the wrong. An additional award, not necessarily of substantial size, may be needed to reflect the sense of public outrage, emphasise the importance of the constitutional right and the gravity of the breach, and deter future breaches. . . . Although such an award, where called for, is likely in most cases to cover much the same ground in financial terms as would an award by way of punishment in the strict sense of retribution, punishment in the latter sense is not its object. Accordingly, the expressions 'punitive damages' or 'exemplary damages' are better avoided . . ."

216. We are not here concerned with a written constitution with a broadly drawn power to grant constitutional redress. But neither are we concerned with a statutory provision, such as section 8(3) and (4) of the Human Rights Act 1998, with a narrowly drawn power to award damages. We are concerned with a decision taken at the highest level of Government to detain certain people irrespective of the statutory purpose of the power to detain. The common law has shown itself capable of growing and adapting to meet new situations. It has recently invented the concept of a conventional sum to mark the invasion of important rights even though no compensatory damages are payable. In the view of the majority of the House of Lords in Rees v Darlington Memorial Hospital NHS Trust [2003] UKHL 52, [2004] 1 AC 309, there were sound reasons of public policy why damages should not be recoverable for the cost of bringing up a healthy child born as a result of a negligent sterilisation. Nevertheless, an award limited to the stress and trauma associated with the pregnancy and birth was insufficient to mark the legal wrong which had been done to the mother. This should be marked by a fixed, non-negligible, conventional sum (in that case £15,000).
217. Rees was concerned with the rights to bodily integrity and personal autonomy, the right to limit one's family and to live one's life in the way planned: see Lord Bingham of Cornhill at paragraph 8. These are important rights indeed. But no-one can deny that the right to be free from arbitrary imprisonment by the state is of fundamental constitutional importance in this country. It is not the less important because we do not have a written constitution. It is a right which the law should be able to vindicate in some way, irrespective of whether compensatable harm has been suffered or the conduct of the authorities has been so egregious as to merit exemplary damages. Left to myself, therefore, I would mark the false imprisonment in these cases with a modest conventional sum, perhaps £500 rather than the £1000 suggested by Lord Walker, designed to recognise that the claimant's fundamental constitutional rights have been breached by the state and to encourage all concerned to avoid anything like it happening again. In reality, this may well be what was happening in the older cases of false imprisonment, before the assessment of damages became such a refined science.
218. I therefore agree, and (subject to the additional points made above) for the reasons given by Lord Dyson, that both these appeals should be allowed. When considering what was a reasonable period for which to detain Mr Lumba in accordance with the Hardial Singh principles, however, I would stress that his psychiatric condition must be among the factors to be taken into account.

LORD COLLINS

219. I agree with Lord Dyson that the appeals should be allowed, substantially for the reasons which he gives. This is a case in which on any view there has been a breach of duty by the executive in the exercise of its power of detention. Fundamental rights are in play. Chapter 39 of Magna Carta (1215) said that "no free man shall be seized or imprisoned … except … by the law of the land" and the Statute of Westminster (1354) provided that "no man of what state or condition he be, shall be … imprisoned … without being brought in answer by due process of the law." That the liberty of the subject is a fundamental constitutional principle hardly needs the great authority of Sir Thomas Bingham MR (see In re S-C (Mental Patient: Habeas Corpus) [1996] QB 599, 603) to support it, but it is worth recalling what he said in his book The Rule of Law (2010), at p 10, about the fundamental provisions of Magna Carta: "These are words which should be inscribed on the stationery of the … Home Office."
220. The evidence shows that concern was expressed in the Home Office from an early stage about the lawfulness of the policy, and that a deliberate decision was taken to continue an unlawful policy. As Lord Dyson says, caseworkers were directed to conceal the true reason for detention, namely the unpublished policy, and to give other reasons which appeared to conform with the published policy. Home Office officials recognised that "Ministers' preferred position may be to continue to detain all FNPs and let the immigration judges take any hit which is to be had by releasing on bail." The draft policy submission circulated in May 2007 recommended a change in policy, but also set out continued detention as one of the options, recognising that legal advisers considered that the department would lose on any legal challenge. The draft added: "… we could present any change in our approach as having been forced on us by the courts".
221. I am satisfied that the serious breach of public law in this case has the result that the detention of the appellants was unlawful. Any other result would negate the rule of law. Christie v Leachinsky [1947] AC 573 shows that where an arrest was unlawful because it did not comply with the requirements imposed by the common law there would be a false imprisonment even if the arrest could have been effected in a proper manner. Holgate-Mohammed v Duke [1984] AC 437, 443, is high authority for the proposition that breach of principles of public law can found an action at common law for damages for false imprisonment.
222. Are they entitled to more than nominal damages? In particular are they entitled to "vindicatory damages"? The expression "vindicatory damages" has been in common use in the context of proceedings for violation of constitutional rights since Attorney General of Trinidad and Tobago v Ramanoop [2005] UKPC 15, [2006] 1 AC 328 and Merson v Cartright and Attorney General [2005] UKPC 38.
223. It would seem that the expression had its origin in the United States, where it was sometimes used as a synonym for exemplary or punitive damages (e.g. Cole v Tucker, 6 Tex 266 (1851); Blair Iron & Coal Co v Lloyd, 3 WNC 103 (Pa (1874)), but at other times used to mean damages designed to vindicate a right but which were compensatory in nature (e.g. McBride v McLaughlin, 5 Watts 375 (Pa 1836); Hallmark v Stillings, 648 SW 2d 230 (Mo 1983)). In England the expression first emerged in a sense somewhat different from, but in a sense related to, that in which it is now used. In Broome v Cassell & Co Ltd [1972] AC 1027, 1071, Lord Hailsham of Marylebone LC said:

"In actions of defamation and in any other actions where damages for loss of reputation are involved, the principle of restitutio in integrum has necessarily an even more highly subjective element. Such actions involve a money award which may put the plaintiff in a purely financial sense in a much stronger position than he was before the wrong. Not merely can he recover the estimated sum of his past and future losses, but, in case the libel, driven underground, emerges from its lurking place at some future date, he must be able to point to a sum awarded by a jury sufficient to convince a bystander of the baselessness of the charge. As Windeyer J well said in Uren v John Fairfax & Sons Pty Ltd (1966) 117 CLR 115, 150: 'It seems to me that, properly speaking, a man defamed does not get compensation for his damaged reputation. He gets damages because he was injured in his reputation, that is simply because he was publicly defamed. For this reason, compensation by damages operates in two ways—as a vindication of the plaintiff to the public and as consolation to him for a wrong done. Compensation is here a solatium rather than a monetary recompense for harm measurable in money.'"

See also, e.g. Sutcliffe v Pressdram Ltd [1991] 1 QB 153; Rantzen v Mirror Group Newspapers (1986) Ltd [1994] QB 670; Purnell v Business F1 Magazine Ltd [2007] EWCA Civ 744, [2008] 1 WLR 1.

224. Any consideration of the Privy Council decisions on vindicatory damages must be prefaced by three points. First, they were concerned with alleged violations of constitutional rights. Second, the constitutions contained provision in relation to such violations for "redress" (Trinidad and Tobago, section 14(1) (without prejudice to any other action which is lawfully available); Bahamas, article 28(1) (but not if adequate means of redress are available under any other law), or "relief" together with "such remedy" as the court considers appropriate (Saint Christopher & Nevis, section 96(1), (3); Saint Lucia, section 105(1), (3)). Third, although the distinction has sometimes been blurred (as perhaps in Takitota v Attorney General [2009] UKPC 11, 13), the decisions are concerned with two heads of damage, compensatory damages and vindicatory damages.
225. In Attorney General of Trinidad and Tobago v Ramanoop [2005] UKPC 15, [2006] 1 AC 328, at 18-19 Lord Nicholls, speaking for the Board, dealt with both heads of damages in this way:

"[18] When exercising this constitutional jurisdiction the court is concerned to uphold, or vindicate, the constitutional right which has been contravened. A declaration by the court will articulate the fact of the violation, but in most cases more will be required than words. If the person wronged has suffered damage, the court may award him compensation. The comparable common law measure of damages will often be a useful guide in assessing the amount of this compensation. But this measure is no more than a guide because the award of compensation under section 14 is discretionary and, moreover, the violation of the constitutional right will not always be coterminous with the cause of action at law.

[19] An award of compensation will go some distance towards vindicating the infringed constitutional right. How far it goes will depend on the circumstances, but in principle it may well not suffice. The fact that the right violated was a constitutional right adds an extra dimension to the wrong. An additional award, not necessarily of substantial size, may be needed to reflect the sense of public outrage, emphasise the importance of the constitutional right and the gravity of the breach, and deter further breaches. All these elements have a place in this additional award. 'Redress' in section 14 is apt to encompass such an award if the court considers it is required having regard to all the circumstances. Although such an award, where called for, is likely in most cases to cover much the same ground in financial terms as would an award by way of punishment in the strict sense of retribution, punishment in the latter sense is not its object. Accordingly, the expressions 'punitive damages' or 'exemplary damages' are better avoided as descriptions of this type of additional award."

226. In Merson v Cartright and Attorney General [2005] UKPC 38 the gravity of police misconduct was such as to attract an award of $100,000 for violations of the Constitution in addition to $90,000 in damages for assault, battery and false imprisonment, and $90,000 for malicious prosecution. It was held by the Privy Council that the awards were not duplicative. Lord Scott said (at 18):

"… The purpose of a vindicatory award is not a punitive purpose. It is not to teach the executive not to misbehave. The purpose is to vindicate the right of the complainant … to carry on his or her life in the Bahamas free from unjustified executive interference, mistreatment or oppression. The sum appropriate to be awarded to achieve this purpose will depend upon the nature of the particular infringement and the circumstances relating to that infringement. It will be a sum at the discretion of the trial judge. In some cases a suitable declaration may suffice to vindicate the right; in other cases an award of damages, including substantial damages, may seem to be necessary."

227. A vindicatory award of $50,000 was made in Inniss v Attorney General [2008] UKPC 42. In Subiah v Attorney General of Trinidad and Tobago [2008] UKPC 47, at 11 Lord Bingham noted that when deciding whether to award vindicatory damages, the answer "is likely to be influenced by the quantum of the compensatory award, as also by the gravity of the constitutional violation in question to the extent that this is not already reflected in the compensatory award". See also Durity v Attorney General of Trinidad and Tobago [2008] UKPC 59, 35. But in Suratt v Attorney General of Trinidad and Tobago (No 2) [2008] UKPC 38 and James v Attorney General of Trinidad and Tobago [2010] UKPC 23 declaratory relief was sufficient.
228. The availability of damages for constitutional wrongs, and in particular, exemplary or vindicatory damages, is, or has been, controversial in many countries. In the United States, nominal damages can be awarded for the deprivation of a constitutional right without proof of actual injury, but substantial damages can be awarded only to compensate for actual injury: e.g. Elkins v District of Columbia, 710 F Supp 2d 53, 63-64 (DDC 2010), citing Carey v Piphus, 435 US 247 (1978); Memphis Community School District v Stachura, 477 US 299 (1986).
229. In Ntandazeli Fose v Minister of Safety and Security in the Constitutional Court of South Africa [1998] 1 LRC 198, [1997] ZACC 6 Ackermann J considered whether "appropriate relief" for infringements of the Interim Constitution of South Africa justified, in addition to compensatory damages for assault, an award for vindication of the rights and for punitive damages. After a full account of the law in other countries he said that he had considerable doubts whether, even where the infringement of the right caused no damage, an award of constitutional damages in order to vindicate the right would be appropriate, and suggested that the court might conclude that a declaratory order combined with a suitable order for costs would be a sufficiently appropriate remedy to vindicate the right even in the absence of an award of damages. But in any event there was no place for constitutional punitive damages: 68, 69.
230. In Taunoa v Attorney General [2008] 1 NZLR 429 the Supreme Court of New Zealand was more sympathetic to vindicatory damages. Elias CJ said (at para 109) that damages in such cases should be limited to what is adequate to mark any additional wrong in the breach and, where appropriate, to deter future breaches. See also Tipping J at 317 (the interests of the victim require the court to consider what compensation is due, but society is a victim also, and the court must consider also what is necessary by way of vindication to protect fundamental rights and freedoms); also Blanchard J at 258; McGrath J at 370.
231. The most recent consideration of the question was by the Supreme Court of Canada in Vancouver (City of) v Ward [2010] 2 SCR 28, in relation to damages for breach of the Canadian Charter of Rights and Freedoms. McLachlin CJ said that Charter damages had the functions of compensation, vindication and deterrence. By vindication she meant the affirmation of constitutional values, focusing on the harm which breach of the Charter did to society. The fact that the claimant had not suffered personal loss did not preclude an award of damages where the objectives of vindication or deterrence clearly called for an award, and the view that constitutional damages were only available for pecuniary or physical loss had been widely rejected in other constitutional democracies: 28, 30.
232. The present claims are not, of course, for constitutional damages. Exemplary damages are available where the executive has acted in a way which is oppressive, arbitrary or unconstitutional. In Kuddus v Chief Constable of Leicestershire Constabulary [2001] UKHL 29, [2002] 2 AC 122, 63 Lord Nicholls said:

"The availability of exemplary damages has played a significant role in buttressing civil liberties, in claims for false imprisonment and wrongful arrest. From time to time cases do arise where awards of compensatory damages are perceived as inadequate…. The nature of the defendant's conduct calls for a further response from the courts. On occasion conscious wrongdoing by a defendant is so outrageous, his disregard for the plaintiff's rights so contumelious, that something more is needed to show that the law will not tolerate such behaviour. Without an award of exemplary damages, justice will not have been done. Exemplary damages, as a remedy of last resort, fill what would otherwise be a regrettable lacuna."

233. But this is not a case for exemplary damages falling within the first head of Rookes v Barnard [1964] AC 1129. Nor do I consider that the concept of vindicatory damages should be introduced into the law of tort. In truth, despite the suggestions to the contrary in the Privy Council in Ramanoop and Merson, vindicatory damages are akin to punitive or exemplary damages (as in Attorney General of St Christopher, Nevis and Anguilla v Reynolds [1980] AC 607).
234. In Ashley v Chief Constable of Sussex Police [2008] UKHL 25, [2008] AC 962 the question was whether there was an abuse of process to allow a claim against the police for the wrongful death of James Ashley to proceed where the police admitted liability for all damages. The House of Lords held by a majority that for the claim to proceed was not an abuse. Lord Scott (obiter) suggested that the claim should proceed in order that vindicatory damages could be available. He referred to Lord Hope's observation in Chester v Afshar [2004] UKHL 41, [2005] 1 AC 134, 87 that "the function of the law is to enable rights to be vindicated and to provide remedies when duties have been breached". Lord Scott said at para 22:

"Although the principal aim of an award of compensatory damages is to compensate the claimant for loss suffered, there is no reason in principle why an award of compensatory damages should not also fulfil a vindicatory purpose. But it is difficult to see how compensatory damages can could ever fulfil a vindicatory purpose in a case of alleged assault where liability for the assault were denied and a trial of that issue never took place.

…

Damages awarded for the purpose of vindication are essentially rights-centred, awarded in order to demonstrate that the right in question should not have been infringed at all. In Attorney General of Trinidad and Tobago v Ramanoop [2006] 1 AC 328 the Privy Council upheld an award of vindicatory damages in respect of serious misbehaviour by a police officer towards the claimant. These were not exemplary damages; they were not awarded for any punitive purpose. They were awarded, as it was put in Merson v Cartwright [2005] UKPC 38, another case in which the Privy Council upheld an award of vindicatory damages, in order 'to vindicate the right of the complainant … to carry on his or her life in the Bahamas free from unjustified executive interference, mistreatment or oppression': para 18. The rights that had been infringed in the Ramanoop case and in Merson v Cartwright were constitutional rights guaranteed by the respective constitutions of the countries in question. But the right to life, now guaranteed by article 2 of the European Convention [on Human Rights] and incorporated into our domestic law by the Human Rights Act 1998, is at least equivalent to the constitutional rights for infringement of which vindicatory damages were awarded in Ramanoop and Merson v Cartwright. …"

235. But what Lord Hope said in Chester v Afshar was not said in the context of damages, and it seems clear that neither Lord Bingham nor Lord Rodger agreed. In particular Lord Rodger said that the right to bodily integrity was protected by the tort actionable per se of trespass to the person, where "the law vindicates that right by awarding nominal damages" (para 60).
236. To make a separate award for vindicatory damages is to confuse the purpose of damages awards with the nature of the award. A declaration, or an award of nominal damages, may itself have a vindicatory purpose and effect. So too a conventional award of damages may serve a vindicatory purpose. That is the basis of Rees v Darlington Memorial Hospital NHS Trust [2003] UKHL 52, [2004] 1 AC 309. As a result of a failed sterilisation negligently performed, the claimant gave birth to a child. The House of Lords held by a majority that although the health authority was not liable to compensate for the child's upbringing, compensation in respect of stress, trauma and the costs associated with pregnancy and birth were recoverable. In addition the claimant was awarded an additional sum of £15,000 of which Lord Bingham said: "[the] award would not be, and would not be intended to be, compensatory. It would not be the product of calculation. But it would not be a nominal, let alone a derisory, award. It would afford some measure of recognition of the wrong done" (para 8). See also Lords Nicholls, Millett, and Scott: at 17, 123, 148, and the critical views expressed in McGregor, Damages 18^th ed (2009), paras 35-297-35-299.
237. Neither the minority dicta in Ashley v Chief Constable of Sussex Police [2008] UKHL 25, [2008] 1 AC 962 nor the award in Rees v Darlington Memorial Hospital NHS Trust [2003] UKHL 52, [2004] 1 AC 309 justify a conclusion that there is a separate head of vindicatory damages in English law. Consequently I do not consider that there is any basis in the present law for such an award. Nor do I consider that there is a basis in policy for the creation of a head of vindicatory damages at common law, distinct from the existing law of compensatory or exemplary damages. I would therefore restrict the remedy in this case to nominal damages for the reasons given by Lord Dyson.

LORD KERR

238. For the reasons given by Lord Dyson, with which I agree, I too would allow this appeal.

A causation test

239. False imprisonment is established if there has been a detention and an absence of lawful authority justifying it. The question whether lawful authority exists is to be determined according to an objective standard. It either exists or it does not. It is for this elementary – but also fundamental – reason that a causation test can have no place in the decision whether imprisonment is false or lawful. By a "causation test" in this context I mean a test which involves an examination of whether the persons held in custody could have been lawfully detained. The fact that a person could have been lawfully detained says nothing on the question whether he was lawfully detained.
240. The Court of Appeal in the present case decided that, since the claimants could have been detained lawfully had the published policy been applied to them, the fact that an unpublished and unlawful policy was in fact applied was immaterial. With great respect, this cannot be right. The unpublished policy was employed in the decision to detain the appellants. It was clearly material to the decision to detain. Indeed, it was the foundation for that decision. An ex post facto conclusion that, had the proper policy been applied, the appellants would have been lawfully detained cannot alter that essential fact.
241. The inevitability of the finding that the detention was unlawful can be illustrated in this way. If, some hours after making the decision to detain the appellants (based on the application of the unpublished policy), it was recognised that this did not constitute a legal basis on which they could be held, could their detention be said to be lawful before any consideration was given to whether the application of the published policy would have led to the same result? Surely, at the moment that it became clear that there was no lawful authority for the detention and before any alternative basis on which they might be detained was considered, their detention was unlawful.
242. It is, I believe, important to recognise that lawful detention has two aspects. First the decision to detain must be lawful in the sense that it has a sound legal basis and, secondly, it must justify the detention. This second aspect has found expression in a large number of judgments, perhaps most succinctly in the speech of Lord Hope in R v Governor of Brockhill Prison Ex p Evans (No 2) [2001] 2 AC 19, 32 D where he said "it is of the essence of the tort of false imprisonment that the imprisonment is without lawful justification". It seems to me to be self evident that the justification must relate to the basis on which the detainer has purported to act, and not depend on some abstract grounds wholly different from the actual reasons for detaining. As Mr Husain QC put it, the emphasis here must be on the right of the detained person not to be detained other than on a lawful basis which justifies the detention. Detention cannot be justified on some putative basis, unrelated to the actual reasons for it, on which the detention might retrospectively be said to be warranted. Simply because some ground for lawfully detaining may exist but has not been resorted to by the detaining authority, the detention cannot be said, on that account, to be lawful.
243. This point was clearly made in Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662. In that case the plaintiff had been lawfully arrested and detained in a police cell. A review of his detention as required by statute was not carried out within the prescribed time. At p 667 B the submission of counsel for the Chief Constable was recorded as being that if circumstances existed which were or would be sufficient to justify continued detention the plaintiff could not fairly be said to be detained without lawful excuse. That submission was roundly – and, in my view, rightly – rejected, Clarke LJ saying, "As I see it, it is nothing to the point to say that the detention would have been lawful if a review had been carried out or that there were grounds which would have justified continued detention". Likewise it is nothing to the point in this case that if the decision had been taken on the basis of the published policy, it would have been immune from challenge. As Professor Cane put it in The Temporal Element in Law (2001) 117 LQR 5, 7 "imprisonment can never be justified unless actually [as opposed to hypothetically] authorised by law". (The emphasis and the words enclosed in square brackets are mine).
244. The matter might be considered on the following hypothetical basis. Suppose that there were two policies, one lawful published policy for the detention of foreign national prisoners sentenced to more than 5 years' imprisonment, the other an unlawful secret policy for detention of those sentenced to more than 2 years' imprisonment. On the respondent's case an individual detained under the second policy, who would have been detained under the first policy if it had been applied, has not been detained unlawfully. I do not consider that such an argument is viable.
245. A policy may lawfully be devised for the purpose of dealing generally with a regularly occurring species of case but it must always be possible to depart from the policy if the circumstances of an individual case warrant it. As the author of Wade & Forsyth on Administrative Law 10^th ed (2009) at page 270 states:

"It is a fundamental rule for the exercise of discretionary power that discretion must be brought to bear on every case: each one must be considered on its own merits and decided as the public interest requires at the time.

246. In the mooted example, consideration would be given to departing from a policy which is different from that which the individual is entitled to have applied to him. The possible justification for departing from the policy would be considered on a different basis from that which ought properly to determine the question. This is, in my view, impermissible in public law terms.
247. A minister exercising his discretion by applying a published policy is acting lawfully. But if the policy which is applied is unlawful, the exercise of discretion is unlawful. The individual has not had applied to his case the proper exercise of discretion to which he is entitled. The application of an unlawful policy will therefore ipso facto render the decision to detain unlawful. In this context, I consider that it matters not whether the decision is said to be in violation of a public law principle or ultra vires the power to make the decision. To draw such a distinction would mark a radical departure from how error of law has long been understood. Again, a short extract from Wade & Forsyth at p 255 makes the point decisively:-

"'Void or voidable' was a distinction which could formerly be applied without difficulty to the basic distinction between action which was ultra vires and action which was liable to be quashed for error of law on the face of the record. That distinction no longer survives since the House of Lords [in Anisminic and subsequent cases] declared all error of law to be ultra vires."

The nature of the public law breach required to invalidate the detention

248. In R (SK Zimbabwe) v Secretary of State for the Home Department [2009] 1 WLR 1527 it was accepted by the appellant that not every type of public law breach, committed after an initially valid detention, would render continued detention unlawful. On the present appeal the argument on behalf of the detained persons is put thus: a public law error that bears directly on the decision to detain will mean that the authority for detention is ultra vires and unlawful, and will sound in false imprisonment. But breaches which have no direct bearing on the decision to detain do not have that effect. Since, therefore, for instance, statutory obligations to permit a detainee to consult with his legal advisers (Cullen v Chief Constable of the Royal Ulster Constabulary [2003] 1 WLR 1763) or to be provided with food or clothing, or to be held in certain conditions (R v Deputy Governor of Parkhurst Prison Ex p Hague [1992] 1 AC 58) did not bear on the legality of detention, breach of those obligations did not render detention unlawful nor did it give rise to a claim for false imprisonment.
249. Lord Walker has analysed the existence/exercise of power to detain dichotomy in a way that I find compelling. This has led him to the conclusion that the essential test as to the validity of continued detention which is said to be beyond the scope of the power to detain is whether there has been a departure from the statutory purpose. Again, I find his reasoning on this wholly convincing. I do not agree, however, that it is necessary to establish abuse of power in order to show that the decision is beyond the scope of the power to detain, if by 'abuse of power' it is meant that some deliberate misuse of power is required. If a review of a person's detention was inadvertently overlooked and it subsequently became clear that, had the review taken place, he would certainly have been released, it surely could not be suggested that the detention that had in the meantime occurred did not constitute false imprisonment.
250. The statutory purpose of the power to detain foreign nationals after the expiry of their sentence is to facilitate their deportation. (In this connection I agree fully with Lord Dyson in his analysis of the Hardial Singh (R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704) principles and with what he had to say about those principles in R (I) v Secretary of State for the Home Department [2002] EWCA Civ 888, [2003] INLR 196 para 46.) Where the statutory purpose no longer exists, the power to detain falls away. The means of ascertaining whether the statutory purpose remains achievable is the system of review. Where that system is operated on the basis of a policy, it is of obvious – and critical – importance that the policy be transparent and that those who may be detained on foot of it have the opportunity to make informed representations on its application.
251. Breach of a public law duty which has the effect of undermining the achievement of the statutory purpose will therefore, in my opinion, render the continued detention invalid. A claim of false imprisonment is the natural, indeed inevitable, entitlement of a person whose detention is no longer justified. Since the appellants in the present case were detained by the operation of a secret, unpublished policy, an effective system of review of the justification for their detention was not possible. As a consequence their detention could no longer be said to be justified. As it seems to me, this approach approximates to the way in which the case was put for the appellants but links it more closely to the vital consideration of the statutory purpose of the power to detain.

Does the award of nominal damages devalue the tort of false imprisonment?

252. As various members of the court have pointed out, the fact that false imprisonment is a species of trespass to the person and is actionable without proof of special damage must be carefully taken into account in deciding whether nominal damages can ever be considered appropriate. The impact of a finding that the State has been guilty of false imprisonment (whether or not it is also ordered to pay compensation) should not be underestimated, however. Such a finding has the effect, in the words of Lady Hale, of "mark[ing] the law's recognition that a wrong has been done". And it is in the unambiguous recognition and declaration by the law that an individual has been falsely imprisoned that the essential value of the entitlement to assert that claim lies. I do not believe, therefore, that the award of nominal damages will, of itself and as a matter of automatic consequence, bring about a devaluation of the tort.
253. On the question whether the award of nominal damages or some other measure of compensation is required in false imprisonment claims, I believe that a distinction is clearly merited between those cases where it is plain that the detainees would have been released and those where it can be shown that they would have been lawfully detained, had the correct procedures been followed. Because false imprisonment is a trespassory tort, it is said that the 'vindicatory' dimension to the assessment of compensation is important. I shall examine that claim presently but, whatever may be said about its correctness, it is surely right that the actual impact on the individual who has been falsely imprisoned (or perhaps more importantly, the impact that could have been avoided) should feature prominently in the assessment of the appropriate amount of compensation.
254. Traditionally, the primary function of damages has been to compensate the individual for the loss that he or she has suffered ('compensatory damages'). More recently the concept of restitutionary damages has been recognised where damages for the tort are measured according to the gain that the defendant has obtained or the value that the right infringed might have had to the claimant where, for instance, unknown to the claimant, the defendant has used the claimant's property. This category of damages is not relevant here. A third type of damages (vindicatory damages) may be. In a number of recent decisions the Judicial Committee of the Privy Council has awarded what might be classified as vindicatory damages where there has been a breach of constitutional rights. Attorney General for Trinidad and Tobago v Ramanoop [2006] 1 AC 328 is perhaps the leading of these cases. At para 19 Lord Nicholls, delivering the judgment of the Committee, said :-

"An award of compensation will go some distance towards vindicating the infringed constitutional right. How far it goes will depend on the circumstances, but in principle it may well not suffice. The fact that the right violated was a constitutional right adds an extra dimension to the wrong. An additional award, not necessarily of substantial size, may be needed to reflect the sense of public outrage, emphasise the importance of the constitutional right and the gravity of the breach and deter further breaches. All these elements have a place in this additional award. … Although such an award, where called for, is likely in most cases to cover much the same ground in financial terms as would an award by way of punishment in the strict sense of retribution, punishment in the latter sense is not its object. Accordingly, the expressions 'punitive damages' or 'exemplary damages' are better avoided as descriptions of this type of additional award."

255. Lord Nicholls' recognition that this type of award covered much the same ground as that involved in exemplary or punitive damages is reflected in the more recent decision of the Privy Council in Takitota v Attorney General [2009] UKPC 11 where, at para 15, Lord Carswell said :-

"… it would not be appropriate to make an award both by way of exemplary damages and for breach of constitutional rights. When the vindicatory function of the latter head of damages has been discharged, with the element of deterrence that a substantial award carries with it, the purpose of exemplary damages has largely been achieved."

256. For the reasons given by Lord Dyson an award of exemplary damages is not warranted in these cases. If there is any scope for the award of vindicatory damages where exemplary damages are not appropriate, it must be, in my opinion, very limited indeed. Such an award could only be justified where the declaration that a claimant's right has been infringed provides insufficiently emphatic recognition of the seriousness of the defendant's default. That situation does not arise here. The defendant's failures have been thoroughly examined and exposed. A finding that those failures have led to the false imprisonment of the appellants constitutes a fully adequate acknowledgement of the defendant's default. Since the appellants would have been lawfully detained if the published policy had been applied to them, I agree that no more than a nominal award of damages is appropriate in their cases.

DISSENTING JUDGMENTS

LORD PHILLIPS

Introduction

257. The appellants are foreign nationals who have served sentences of imprisonment in this country ("FNPs") They were detained pursuant to Schedule 3 of the Immigration Act 1971 ("Schedule 3"). They challenge the legality of their detention. At the times of the decisions to detain them there existed a policy published by the Secretary of State setting out the circumstances in which her power to detain immigrants would be exercised. Had the decision-maker applied this policy he would have detained each of the appellants. In the event, however, he decided to detain each by the application of a policy which Mr Beloff QC for the Secretary of State has conceded was "unlawful". Whether the "lawful" or "unlawful" policy had been applied the decision would have been the same. The principal common issues raised by these appeals are first whether, in these circumstances, the detention of each of the appellants was "unlawful"; secondly whether, if it was "unlawful," the result was that the detention of each of the appellants constituted the tort of false imprisonment and; if so, thirdly, whether and on what basis the appellants are entitled to damages. I have placed the words "lawful" and "unlawful" in parentheses because these appeals raise the question of whether there is a material difference between a policy, or a decision, or an act which is "unlawful" because it violates principles of public law and a policy, or a decision, or an act which is "unlawful" because it is ultra vires.
258. In the case of Mr Lumba there is a second issue. This is whether his detention was or became unlawful because it infringed what have become known as the Hardial Singh principles which date back to the decision of Woolf J in the case of that name over a quarter of a century ago. Lord Dyson at para 22 of his judgment rightly states that it has been common ground in these appeals that he correctly summarised the effect of Hardial Singh in the four principles which he set out in R (I) v the Secretary of State for the Home Department [2003] INLR 196. As I shall explain I have reservations about the first two principles which, so far as I am aware, have never been the subject of debate.
259. Lord Dyson has set out the relevant facts and statutory provisions and I can turn immediately to the common issues raised by these appeals. Lord Dyson has held that the Secretary of State committed the tort of false imprisonment in relation to each appellant because the decision to detain him was reached in violation of public law. The violation was the failure to apply the Secretary of State's published policy and the application of a policy to which there were various objections of public law. He has concluded that, because the reasoning offended the requirements of public law, the acts that the decision-maker decided upon were beyond his powers, or ultra vires. I have come to a different conclusion.
260. I propose in this judgment to address the following questions. First, what restrictions are implicit, as a matter of statutory interpretation, in the power to detain conferred on the Secretary of State by Schedule 3? Second, what were the policies published by the Secretary of State in relation to the detention of immigrants? Third, what were the practical implications of those policies? Fourth, what were the legal implications of those policies? Fifth, was the detention of each of the appellants contrary to those policies? Sixth, what were the defects in the policy applied when deciding whether the appellants should be detained? Seventh, what were the circumstances in which this policy was applied? Eighth, did the application of that policy render the detention of the appellants unlawful? If so, ninth, are the appellants entitled to damages for false imprisonment?

Implied limitations on the power to detain conferred by Schedule 3

261. I refer to the four principles that Lord Dyson states at para 22 of his judgment are derived from Hardial Singh. The third and fourth principles were an essential part of the reasoning that led Woolf J to the decision that he reached in that case. They are not open to question. This is not true of the first two. The first is that "the Secretary of State must intend to deport the person and can only use the power to detain for that purpose". Lord Dyson explains that by this he means that the power to detain must be exercised "for the prescribed purpose of facilitating deportation". The second principle is that "the deportee may only be detained for a period that is reasonable in all the circumstances". Neither of these principles was stated in these terms in Hardial Singh, although I accept that they are possible interpretations of the words used by Woolf J. Neither of these principles was essential to the conclusion that he reached. I do not myself consider that either principle can properly be derived from his judgment.
262. The applicant in Hardial Singh sought a writ of habeas corpus. He was an Indian who had entered the United Kingdom lawfully and been granted indefinite leave to remain. He had been convicted of offences of burglary and been sentenced to a total of two years' imprisonment. Before he was due to be released he was served with a deportation order on behalf of the Secretary of State. He was due for release on 20 July 1983 but was then detained by the Secretary of State pursuant to paragraph 2(3) of Schedule 3. The reason given for his detention was the risk that, if released, he would abscond. Because of delay on the part of the Secretary of State in making arrangements for his return to India he was still detained in December 1983. In these circumstances Woolf J, at p 706, said this about the power of detention under Schedule 3:

"Although the power which is given to the Secretary of State in paragraph 2 to detain individuals is not subject to any express limitation of time, I am quite satisfied that it is subject to limitations. First of all, it can only authorise detention if the individual is being detained in one case pending the making of a deportation order and, in the other case, pending his removal. It cannot be used for any other purpose. Second, as the power is given in order to enable the machinery of deportation to be carried out, I regard the power of detention as being impliedly limited to a period which is reasonably necessary for that purpose. The period which is reasonable will depend on the circumstances of the particular case. What is more, if there is a situation where it is apparent to the Secretary of State that he is not going to be able to operate the machinery provided in the Act for removing persons who are intended to be deported within a reasonable period, it seems to me that it would be wrong for the Secretary of State to seek to exercise his power of detention.

In addition, I would regard it as implicit that the Secretary of State should exercise all reasonable expedition to ensure that the steps are taken which will be necessary to ensure the removal of the individual within a reasonable time."

263. There is a degree of ambiguity in the earlier part of this passage. "Pending the making of a deportation order" is not a purpose. Nor is "pending his removal". What then did Woolf J mean when he said that the power to remove "cannot be used for any other purpose"? He goes on to say that the power is given "to enable the machinery of deportation to be carried out" and that the power of detention is limited to such period as is "reasonably necessary for that purpose". If one takes these two passages together it is possible to interpret Woolf J as saying that you can only detain a person for the purpose of facilitating deportation, as Lord Dyson has done. It is, however, also possible to read him as saying that you can only detain a person while you are pursuing the objective of deporting him and that is how I interpret what he said. I believe that the interpretation given by Lord Dyson places an unjustified restriction on the Secretary of State's power of detention. It is obvious that detention will almost always make the practical task of deporting the detainee easier to arrange. Most deportees will be in this country through choice and cannot reasonably be expected to do anything to facilitate their deportation even if they do not try actively to prevent this. It is open to the Secretary of State to detain a person in order to facilitate his deportation and this is often the, or one of the, reasons for doing so. But, as I shall explain, I do not consider that detention of a deportee will only be lawful if used for this purpose.
264. The second principle identified by Lord Dyson is that "the deportee may only be detained for a period that is reasonable in all the circumstances". This I understand to be derived from Woolf J's statement "The period which is reasonable will depend upon all the circumstances". But that sentence was immediately preceded by the statement that the power to detain was impliedly limited to a period that was reasonably necessary for the purpose of enabling the machinery of deportation to be carried out. Thus I believe that the "circumstances" that Woolf J had in mind were restricted to those that related to the task of effecting deportation. I am fortified in this belief by the fact that Woolf J went on to cite R v Governor of Richmond Remand Centre, Ex p Asghar [1971] 1 WLR 129. In that case the Secretary of State had detained two persons who were awaiting removal with the object that they should testify in a pending criminal trial. Lord Parker CJ rejected the suggestion that the detention could be justified as reasonable in these circumstances, stating at p 132

"it does seem to me that while a reasonable time is contemplated between the giving of the directions and the final removal, that is a reasonable time necessary to effect the physical removal".

265. In Hardial Singh Woolf J was concerned only with the length of time that was reasonably necessary to effect deportation and the relationship that this bore to the power to detain. He was not concerned with the question of whether there were further implied restrictions on the power to detain during that period.
266. The extent of the power to detain pending deportation was an important, albeit not the most important, issue in Chahal v United Kingdom where the nature of the domestic proceedings is apparent from the judgment of the European Court of Human Rights when the case reached Strasbourg (1996) 23 EHRR 413. Mr Chahal was a Sikh separatist leader who had been granted indefinite leave to remain in the United Kingdom. On 14 August 1990 the Secretary of State decided that he ought to be deported because his continued presence in the United Kingdom was unconducive to the public good for reasons of national security and other reasons of a political nature, namely the international fight against terrorism. Two days later he was served with a notice of intention to deport. He was then detained pursuant to Schedule 3 and remained in detention up to the time of the judgment of the Strasbourg Court. During this time he pursued an unsuccessful attempt to be granted asylum. He also, by an application for judicial review, challenged his proposed deportation on the ground that this would violate article 3 of the Convention because it would expose him to the risk of torture and persecution if returned to India. He was unsuccessful in the Divisional Court and the Court of Appeal and was refused leave to appeal to the House of Lords. He then applied to Strasbourg, alleging breaches of articles 3 and 5 of the Convention. In November 1995, while he was awaiting a hearing at Strasbourg, he challenged his continued detention by seeking from the Divisional Court a writ of habeas corpus and judicial review. The Secretary of State opposed his application on the grounds that he could not safely be released because of the substantial threat that he posed to national security. It does not appear to have been suggested that his lengthy detention was necessary to facilitate his deportation. His application was refused on the ground that there was no reason to believe that the Secretary of State did not have good reason for his apprehension. MacPherson J ruled that

"the detention per se was plainly lawful because the Secretary of State [had] the power to detain an individual who [was] the subject of a decision to make a deportation order" ( para 43).

267. It is relevant to see how the Strasbourg Court addressed this matter, if only because any interpretation of Schedule 3 must, if possible, be compatible with the requirements of the Convention. Article 5(1) of the Convention provides, in so far as material:

"Everyone has the right to liberty and security of person. No one shall be deprived of his liberty save in the following cases and in accordance with a procedure prescribed by law:

…

(f) the lawful arrest or detention…of a person against whom action is being taken with a view to deportation…"

The court said this as to the effect of that provision:

"112. The court recalls that it is not in dispute that Mr Chahal has been detained 'with a view to deportation' within the meaning of article 5(1)(f). Article 5(1)(f) does not demand that the detention of a person against whom action is being taken with a view to deportation be reasonably considered necessary, for example to prevent his committing an offence or fleeing; in this respect article 5(1)(f) provides a different level of protection from article 5(1)(c).

Indeed all that is required under this provision is that 'action is being taken with a view to deportation'. It is therefore immaterial, for the purposes of article 5(1)(f), whether the underlying decision to expel can be justified under national or Convention law.

113. The court recalls, however, that any deprivation of liberty under article 5(1)(f) will be justified only for as long as deportation proceedings are in progress. If such proceedings are not prosecuted with due diligence, the detention will cease to be permissible under article 5(1)(f)."

These observations echo the Hardial Singh principles, as I would interpret them. The court went on to consider whether the asylum proceedings, which had delayed the stage at which Mr Chahal would have been deported, had taken an excessive length of time, and concluded that they had not. This indicates that the Strasbourg Court, when considering the time reasonably needed for deportation, accepted that this would be prolonged by delay reasonably attributable to attempts to obtain asylum.

268. The court then considered the requirement that the detention should be "lawful". It observed at para 119 that there was no doubt that it was lawful under national law but that, because of the extremely long period during which Mr Chahal had been detained it was also necessary to consider whether "there existed sufficient guarantees against arbitrariness". At para 122 the court concluded that the domestic procedure under which Mr Chahal's appeal against deportation had been considered by an advisory panel

"provided an adequate guarantee that there were at least prima facie grounds for believing that if Mr Chahal were at liberty, national security would be put at risk and thus, that the executive had not acted arbitrarily when it ordered him to be kept in detention. "

269. I am not able to accept that under domestic law it had been an implicit requirement of Schedule 3 that Mr Chahal's detention was necessary to facilitate his deportation. Provided that he was being detained with a view to his removal as soon as reasonably possible I consider that the Secretary of State was entitled to detain him pending that removal on the ground that he would pose a terrorist threat if released.
270. The Hardial Singh principles were applied by analogy by the Judicial Committee of the Privy Council when considering the legitimacy under Hong Kong legislation of the detention of four "boat people" from Vietnam in Tan Te Lam v Superintendent of Tai A Chau Detention Centre [1997] AC 97. Tens of thousands of these people had arrived unlawfully in Hong Kong. They were steadily being repatriated, but this was taking a long time. Section 2 of the Immigration (Amendment) Ordinance 1991 added to the relevant legislation a provision designed expressly to deal with this situation:

"The detention of a person under this section shall not be unlawful by reason of the period of the detention if that period is reasonable having regard to all the circumstances affecting that person's detention, including – (a) in the case of a person being detained pending a decision under section 13A(1) to grant or refuse him permission to remain in Hong Kong as a refugee – (i) the number of persons being detained pending decisions under section 13A(1) whether to grant or refuse them such permission; and (ii) the manpower and financial resources allocated to carry out the work involved in making all such decisions; (b) in the case of a person being detained pending his removal from Hong Kong – (i) the extent to which it is possible to make arrangements to effect his removal; and (ii) whether or not the person has declined arrangements made or proposed for his removal….."(p 106).

Lord Browne-Wilkinson, giving the advice of the Board, said this, at p 111, under the heading "The Hardial Singh principles":

"Section 13D(1) confers a power to detain a Vietnamese migrant 'pending his removal from Hong Kong'. Their Lordships have no doubt that in conferring such a power to interfere with individual liberty, the legislature intended that such power could only be exercised reasonably and that accordingly it was implicitly so limited. The principles enunciated by Woolf J in the Hardial Singh case [1984] 1 WLR 704 are statements of the limitations on a statutory power of detention pending removal. In the absence of contrary indications in the statute which confers the power to detain 'pending removal' their Lordships agree with the principles stated by Woolf J. First, the power can only be exercised during the period necessary, in all the circumstances of the particular case, to effect removal. Secondly, if it becomes clear that removal is not going to be possible within a reasonable time, further detention is not authorised. Thirdly, the person seeking to exercise the power of detention must take all reasonable steps within his power to ensure the removal within a reasonable time."

This accords with my reading of Hardial Singh. His Lordship went on to say, however:

"Their Lordships are unable to agree with the Court of Appeal of Hong Kong that there is any conflict between the Hardial Singh principles and the provisions of section 13D. Section 13D(1A), which was inserted in 1991, expressly envisages that the exercise of the power of detention conferred by section 13D(1) will be unlawful if the period of detention is unreasonable. It expressly provides that 'The detention…shall not be unlawful by reason of the period of the detention if that period is reasonable having regard to …' (Emphasis added.) What section 13D(1A) does is to provide expressly that, in deciding whether or not the period is reasonable, regard shall be had to all the circumstances including (in the case of a person detained pending his removal from Hong Kong) 'the extent to which it is possible to make arrangements to effect his removal' and 'whether or not the person has declined arrangements made or proposed for his removal.' Therefore the subsection is expressly based on the requirement that detention must be reasonable in all the circumstances (the Hardial Singh principles) but imposes specific requirements that in judging such reasonableness those two factors are to be taken into account."

The shorthand summary of the Hardial Singh principles as "detention must be reasonable in all the circumstances" was made in the context of those circumstances that affected the time reasonably necessary to effect removal and, just as in the case of Hardial Singh itself, I would restrict its ambit to those circumstances. This I believe was, and remained, the understanding of some, at least, of the judges dealing with claims in respect of the detention of immigrants in the Administrative Court. Thus in R (Konan) v Secretary of State for the Home Department [2004] EWHC 22 (Admin) Collins J, who had appeared as counsel in Hardial Singh, held at para 21:

"The power to detain pending removal is not dependent on a fear of absconding or of any other misconduct by the person in question. Provided it is exercised for the purpose of removal, it is lawful. It must be exercised reasonably, but reasonableness in this context relates to whether removal can be achieved within a reasonable time: see R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704 and Tan Te Lam v Superintendent of Tai A Chau Detention Centre [1997] AC 97 as applied in R (Saadi) v Secretary of State for the Home Department [[2002] 1 WLR 356]."

271. R (Saadi) v Secretary of State for the Home Department [2001] EWCA Civ 1512; [2002] 1 WLR 356 raised the question of the legality of the detention at Oakington Reception Centre for up to 10 days of aliens seeking leave to enter whose cases appeared susceptible of speedy processing. The power to detain that was relied on was that afforded by paragraph 16(1) of Schedule 2 to the 1971 Act. Paragraph 2 provides that an immigration officer may examine any person arriving in the United Kingdom to determine whether he should be given leave to enter. Paragraph 16(1) provides:

"A person who may be required to submit to examination under paragraph 2 above may be detained under the authority of an immigration officer pending his examination and pending a decision to give or refuse him leave to enter."

The Court of Appeal, in a judgment which I delivered, considered both the scope of the power to detain afforded by this paragraph and the effect of article 5(1)(f) of the Human Rights Convention. As to the former the court made the following observations:

"14. Collins J concluded that the only limitation on the power to detain pending examination and the decision whether to grant or refuse leave to enter is that the detention must be for a reasonable time. For the Secretary of State, the Attorney General supported this conclusion. He argued that the power to detain persisted for so long as was reasonably necessary to conduct the examination and to reach a decision whether or not to grant leave to enter. As a matter of statutory interpretation we accept this submission. Were it not correct, the power to grant temporary admission would also be liable to come to an end before an examination could reasonably be completed and a decision whether to grant or refuse leave to enter reasonably be taken.

15. We are not aware that it has ever been the policy of the Secretary of State that applicants for leave to enter should be detained pending the decision of their applications, however long that might take. A more liberal policy has been adopted whereby he has approved the exercise of the power to grant temporary admission in place of detention. If the basis upon which immigration officers are detaining asylum seekers at Oakington is in conflict with this policy, then, under established principles of public law, they are acting unlawfully."

As to the Convention the court held:

"66. We consider that the test of proportionality required by article 5(1)(f) requires the Court simply to consider whether the process of considering an asylum application, or arranging a deportation, has gone on too long to justify the detention of the person concerned having regard to the conditions in which the person is detained and any special circumstances affecting him or her. Applying that test no disproportionality is demonstrated in this case."

This was not a test of proportionality that the Strasbourg Court had laid down in Chahal and it received no support from that Court when Saadi reached it, as I shall show.

272. Giving the only reasoned speech in a unanimous decision of the House of Lords [2002] UKHL 41; [2002] 1 WLR 3131 Lord Slynn of Hadley referred at para 18 to a statement by the Oakington Project Manager that he accepted that an important consideration in relation to detention powers was that no detention should be longer than reasonably necessary. Lord Slynn went on to express the view at para 22:

"As the judge and the Court of Appeal stressed, paragraph 16 of Schedule 2 gives power to detain 'pending' examination and a decision; that in my view means for the period up to the time when the examination is concluded and a decision taken. There is no qualification that the Secretary of State must show that it is necessary to detain for the purposes of examination in that the examination could not otherwise be carried out since applicants would run away. Nor is it limited to those who cannot for whatever reason appropriately be granted temporary admission. The period of detention in order to arrive at a decision must however be reasonable in all the circumstances."

The last sentence reflected Government policy, as accepted by the Project Manager.

273. One of the applicants in Saadi took his case to Strasbourg (2008) 47 EHRR 17. He claimed that his detention at Oakington had infringed article 5(1)(f). Liberty, and other interveners, contended that a test of necessity and proportionality should be applied to article 5(1)(f), so that an asylum seeker could only be detained if, but for such detention, he would attempt to effect an unauthorised entry into the country. The Grand Chamber rejected this submission. Dealing first with the interpretation of the express provisions of article 5(1)(f) it said:

"64. Whilst the general rule set out in article 5(1) is that everyone has the right to liberty, article 5(1)(f) provides an exception to that general rule, permitting states to control the liberty of aliens in an immigration context. As the court has remarked before, subject to their obligations under the Convention, states enjoy an 'undeniable sovereign right to control aliens' entry into and residence in their territory'. It is a necessary adjunct to this right that states are permitted to detain would-be immigrants who have applied for permission to enter, whether by way of asylum or not. It is evident from the tenor of the judgment in Amuur that the detention of potential immigrants, including asylum seekers, is capable of being compatible with article 5(1)(f).

65. On this point, the Grand Chamber agrees with the Court of Appeal, the House of Lords and the Chamber, that until a state has 'authorised' entry to the country, any entry is 'unauthorised' and the detention of a person who wishes to effect entry and who needs but does not yet have authorisation to do so, can be, without any distortion of language, to 'prevent his effecting an unauthorised entry'. It does not accept that, as soon as an asylum seeker has surrendered himself to the immigration authorities, he is seeking to effect an 'authorised' entry, with the result that detention cannot be justified under the first limb of article 5(1)(f). To interpret the first limb of article 5(1)(f) as permitting detention only of a person who is shown to be trying to evade entry restrictions would be to place too narrow a construction on the terms of the provision and on the power of the state to exercise its undeniable right of control referred to above."

As to the argument that a test of proportionality applied to the detention, the Court, referring to Chahal, held:

"72. Similarly, where a person has been detained under article 5(1)(f), the Grand Chamber, interpreting the second limb of this sub-paragraph, held that, as long as a person was being detained 'with a view to deportation', that is, as long as 'action [was] being taken with a view to deportation', there was no requirement that the detention be reasonably considered necessary, for example to prevent the person concerned from committing an offence or fleeing. The Grand Chamber further held in Chahal that the principle of proportionality applied to detention under article 5(1)(f) only to the extent that the detention should not continue for an unreasonable length of time; thus, it held that

"[A]ny deprivation of liberty under article 5(1)(f) will be justified only for as long as deportation proceedings are in progress. If such proceedings are not prosecuted with due diligence, the detention will cease to be permissible."

73. With regard to the foregoing, the court considers that the principle that detention should not be arbitrary must apply to detention under the first limb of article 5(1)(f) in the same manner as it applies to detention under the second limb. Since states enjoy the right to control equally an alien's entry into and residence in their country (see the cases cited in para 63 above), it would be artificial to apply a different proportionality test to cases of detention at the point of entry than that which applies to deportation, extradition or expulsion of a person already in the country.

74. To avoid being branded as arbitrary, therefore, such detention must be carried out in good faith; it must be closely connected to the purpose of preventing unauthorised entry of the person to the country; the place and conditions of detention should be appropriate, bearing in mind that:

"[T]he measure is applicable not to those who have committed criminal offences but to aliens who, often fearing for their lives, have fled from their own country;"

and the length of the detention should not exceed that reasonably required for the purpose pursued."

274. This passage reinforces the conclusions of the court in Chahal. Where a person is detained pending deportation, the only proportionality requirement that Strasbourg imposes, if indeed it is right so to describe it, is that the detention should not be for longer than is reasonably necessary to effect the deportation. Added to this, however, is the important requirement that the exercise of the power to detain must not be arbitrary.
275. Applying this principle the European Commission of Human Rights held manifestly inadmissible an application of infringement of article 5 by a man who had been detained for five years while he used every means to avoid extradition to Hong Kong. See Osman v United Kingdom (Application No 15933/89) (unreported) 14 January 1991.
276. The most recent pronouncement of the Grand Chamber on article 5(1)(f) is to be found in A v United Kingdom ((2009) 49 EHRR 29, where it was held that article 5(1)(f) did not justify detention of the famous Belmarsh detainees. At para 164 the Grand Chamber stated:

"To avoid being branded as arbitrary, detention under article 5(1)(f) must be carried out in good faith; it must be closely connected to the ground of detention relied on by the Government; the place and conditions of detention should be appropriate; and the length of the detention should not exceed that reasonably required for the purpose pursued."

277. Against this background of Strasbourg jurisprudence, I return to domestic consideration of the Hardial Singh principles. In R(I) v Secretary of State for the Home Department [2003] INLR 196 the appellant was an Afghani asylum seeker who had been given exceptional leave to remain. He was then convicted of indecent assaults and sentenced to 3 years' imprisonment with a recommendation for deportation. The Secretary of State made a deportation order and detained him pursuant to paragraph 2(3) of Schedule 3 from February 2001 to May 2002. The delay occurred because practical difficulties had caused the Secretary of State to cease removing nationals to that country. He claimed that his further detention was unlawful as there was no reasonable possibility of his being deported within a reasonable period. This raised the question of how such a reasonable period fell to be calculated. Simon Brown LJ addressed three issues of principle that had arisen. The first was the relevance of the fact that he was likely to "go to ground" and re-offend if released. His counsel contended that this was irrelevant to the question of whether removal would be possible within a reasonable time. Simon Brown LJ disagreed. He held at para 29:

"The likelihood or otherwise of the detainee absconding and/or re-offending seems to me to be an obviously relevant circumstance. If, say, one could predict with a high degree of certainty that, upon release, the detainee would commit murder or mayhem, that to my mind would justify allowing the Secretary of State a substantially longer period of time within which to arrange the detainee's removal abroad."

278. The second issue was whether it was relevant that the appellant refused to accept voluntary repatriation. Simon Brown LJ held that it was. The third issue was whether the calculation of the reasonable period should take account of the fact that the appellant had been making asylum applications. Simon Brown LJ held that it should not, because it would not have been possible to deport him in any event. The conclusion that he formed at para 37 was that because the Secretary of State could not establish more than a hope of being able to remove him by the summer

"substantially more in the way of a risk of re-offending (and not merely a risk of absconding) than exists here would in my judgment be necessary to have justified continuing his detention for an indeterminate further period."

279. Mummery LJ gave a short dissent on the facts rather than the applicable principles.
280. Dyson LJ agreed with Simon Brown LJ. He set out the four principles that he derived from Hardial Singh in the same form as he has in his judgment in the present case. He then made the following observations about the application of those principles.

"47. Principles (ii) and (iii) are conceptually distinct. Principle (ii) is that the Secretary of State may not lawfully detain a person 'pending removal' for longer than a reasonable period. Once a reasonable period has expired, the detained person must be released. But there may be circumstances where, although a reasonable period has not yet expired, it becomes clear that the Secretary of State will not be able to deport the detained person within a reasonable period. In that event, principle (iii) applies. Thus, once it becomes apparent that the Secretary of State will not be able to effect the deportation within a reasonable period, the detention becomes unlawful even if the reasonable period has not yet expired.

48. It is not possible or desirable to produce an exhaustive list of all the circumstances that are, or may be, relevant to the question of how long it is reasonable for the Secretary of State to detain a person pending deportation pursuant to paragraph 2(3) of Schedule 3 to the Immigration Act 1971. But in my view they include at least: the length of the period of detention; the nature of the obstacles which stand in the path of the Secretary of State preventing a deportation; the diligence, speed and effectiveness of the steps taken by the Secretary of State to surmount such obstacles; the conditions in which the detained person is being kept; the effect of detention on him and his family; the risk that if he is released from detention he will abscond; and the danger that, if released, he will commit criminal offences.

49. Simon Brown LJ has identified the three main points of principle which were in issue in the present appeal, namely, the risk of absconding and reoffending, the appellant's refusal to accept voluntary repatriation, and the asylum claim and appeal. As I have already stated, the risk of absconding and offending or reoffending is relevant to the reasonableness of the length of a detention pending deportation. It is, as Simon Brown LJ says 'an obviously relevant circumstance' (at para 29): see also per Lord Phillips of Worth Matravers MR in R v (Saadi) v Secretary of State for the Home Department [2001] EWCA Civ 1512, [2002] 1 WLR 356, at paras 65-67."

281. I have accepted that it is possible to derive Lord Dyson's first two principles from the language used by Woolf J in Hardial Singh, and explained why I would not do so. The passage from the judgment of the Court of Appeal in Saadi that he cites was dealing with the effect of article 5(1)(f) of the Convention and, as I have shown, advanced a test of proportionality which Strasbourg did not endorse. The problems that have been raised by these appeals suggest that Strasbourg may have had very good reason not to do so.
282. The conceptual difficulty inherent in the first two principles identified by Lord Dyson is apparent in the passage that he has quoted at para 107 from the judgment of Toulson LJ in A. If the risk of re-offending can be the determinant factor in deciding how long it is reasonable to detain a FNP why should it not be the determinant, or even the sole reason for detaining him in the first place? Why should it be a pre-condition to the power to detain that its use is to facilitate deportation, even if this is not the real reason for detention, as in the case of Chahal? It is that logical inconsistency that underlies the challenge that is made on behalf of Mr Lumba in the present case to the legitimacy of having regard to the risk of his re-offending.
283. The interpretation that I would give to the power to detain conferred by the 1971 Act is as follows. The scheme of the Act is simple, and reflects article 5(1)(f) of the Convention. The Secretary of State is not required to permit an immigrant who has unlawfully entered this country to roam free. Schedule 3 permits her to detain the immigrant for as long as she reasonably requires in order to decide whether he should have leave to enter. If he is not given leave to enter she may detain him for as long as she reasonably requires to effect his deportation, provided always that deportation is a practical possibility.
284. If the 1971 Act confers powers as wide as this on the Secretary of State, she has not availed herself of them. She adopted a policy under which, on her calculation, only 1.5% of those who were liable to detention under her immigration powers were actually detained, see para 285 below. Having chosen to discriminate between those whom she detains and those whom she does not, she is subject to the established principles of public law in choosing between the two. It is these principles which constrain the exercise of her power to detain rather than restrictions to be implied into the 1971 Act as a matter of statutory interpretation. They include the obligation to act rationally, an obligation also imposed by article 5 of the Convention. It is rational and lawful to detain a FNP pending deportation to prevent his re-offending or because he would pose a security risk if at large, just as it is rational and lawful to detain him because of the risk of his absconding. Public law principles include the restraint that a published policy imposes on executive action, a topic that I am about to consider. As I shall show, the guidance published by the Secretary of State includes a requirement to comply with Lord Dyson's first two principles, so that to that extent their enunciation has been self-fulfilling.

The policies published by the Secretary of State in relation to the detention of immigrants.

285. Lord Dyson has referred to the two White Papers in which in 1998 and 2002 the Secretary of State published her policies in relation to detention. The first of these, Fairer, Faster, Firmer informed the reader at the beginning of Chapter 12 dealing with "Detention" that at any one time, only about 1.5% of those liable to detention under immigration powers were actually detained. The White Paper dealt with the criteria to be adopted in identifying this small minority of immigrants who were to be detained. FNPs awaiting deportation will have formed only a tiny proportion of those liable to detention under those powers. It is, perhaps, not surprising that the White Paper predominantly addressed the position of the vast majority of immigrants who were not criminals. Thus, in the passage quoted by Lord Dyson at para 11 of his judgment the White Paper spoke of a presumption in favour of granting temporary admission or release, terms that were not appropriate to those recommended for deportation. In dealing with "Detention Criteria" at 12.3 the White Paper identified three circumstances where detention would normally be justified. The first was where there was a reasonable belief that the individual would fail to keep to the terms of temporary admission or temporary release. The second was to clarify a person's identity and the basis of their claim on entry. The third was where removal was imminent.
286. 12.11 dealt with detention in relation to removals. It focussed entirely on detention to facilitate removals. Nothing in that White Paper gave any suggestion that those awaiting deportation might be detained because of concern as to the way they might behave if permitted to be at large. There was no focus on the provisions of Schedule 3.
287. These comments are equally true of the second White Paper, Secure Borders, Safe Haven, save that this had the following statement in para 4.80 under the heading "Serious Criminals":

"We will explore what more we can do, as other countries have done, to stop serious criminals abusing our asylum system by seeking to remain in the UK having completed a custodial sentence."

There is there no indication that such criminals would be liable to detention pending deportation.

288. The two White Papers dealt in broad terms with detention. They were supplemented by Chapter 38, headed "Detention and Temporary Release", of the Operational Enforcement Manual, which was a published document, available to the public on the internet. The court was provided with the version that was current in April 2006. This included guidance on the law as it was understood to be. Para 38.1.1.1 gave the following summary of the effect of article 5 and the domestic case law:

"(a) The relevant power to detain must only be used for the specific purpose for which it is authorised. This means that a person may only be detained under immigration powers for the purpose of preventing his unauthorised entry or with a view to his removal (not necessarily deportation). Detention for other purposes, where detention is not for the purposes of preventing unauthorised entry or effecting removal of the individual concerned, is not compatible with article 5 and would be unlawful in domestic law;

(b) The detention may only continue for a period that is reasonable in all the circumstances;

(c) If before the expiry of the reasonable period it becomes apparent that the purpose of the power, for example, removal, cannot be effected within that reasonable period, the power to detain should not be exercised ; and

(d) The detaining authority (be it the immigration officer or the Secretary of State), should act with reasonable diligence and expedition to effect removal (or whatever the purpose of the power in question is)."

This summary of the law reflected aspects of the decision of the Court of Appeal in I with which I have differed. None the less the principles that it expounded were consonant with the general policy of the Secretary of State that there was a presumption against detention.

289. Para 38.3 set out the factors that influenced a decision to detain. Those in favour of detention were all matters that bore on the likelihood that the individual would abscond or go to ground if not detained. The manual set out the contents of a standard form IS9IR. This set out 6 possible reasons for detention, with instructions that the Immigration Officer should tick the relevant reasons. In contrast to the general focus on the likelihood of absconding, one of these stated "Your release is not considered conducive to the public good". Factors forming the basis of the reasons also had to be ticked. These included

" • You are excluded from the United Kingdom at the personal direction of the Secretary of State.

• You are detained for reasons of national security, the reasons are/will be set out in another letter.

• Your previously unacceptable character, conduct or associates."

Para 38.5.2 was headed "Authority to detain persons subject to deportation action". It summarised the effect of Schedule 3 and required that decisions whether to detain pursuant to the provisions of the Schedule should be made "at senior caseworker level in CCT". No specific guidance was given, however, as to the criteria that should be applied when making those decisions.

290. In summary, the general message of these published policies was that detention should be used sparingly and, in the case of detention pending removal or deportation, only where necessary to facilitate this in order to prevent individuals from absconding or otherwise evading the immigration system. Officials were instructed that the law was as held by the Court of Appeal in I and, in particular, that detention could only continue for a period that was reasonable in all the circumstances. Form IS9IR raised, however, the possibility that detention could be used for reasons of national security or where the individual's previous character, conduct or associates were unacceptable. Despite this, there was no specific guidance as to the approach to be adopted to criminals whom the Secretary of State had decided to deport.
291. Not only was it open to the Secretary of State to decide to deport, and to detain pending deportation, criminals in respect of whom the court had made no recommendation, she also had to decide whether to accept recommendations for deportation made by the courts. Under paragraph 2(1) of Schedule 3 those subject to a recommendation remained detained pending the Secretary of State's decision whether to deport them unless released by the court or granted bail pursuant to section 54 of the Immigration and Asylum Act 1999, which came into force in February 2003. It may have been thought that paragraph 2(1) created a presumption in favour of detention of FNPs pending deportation but in R (Sedrati) v Secretary of State for the Home Department [2001] EWHC Admin 418, with the agreement of counsel for the Secretary of State, Moses J made a declaration that there was no such presumption.
292. What then, under the Secretary of State's published policies, was the position of FNPs whom the Secretary of State decided to deport? It seems to me that many of these were likely to tick the boxes of those who, exceptionally, could properly be detained in accordance with those policies. They were in this country because this was where they had chosen to live. They had committed offences that had led the Secretary of State to conclude that their continued presence was no longer conducive to the public good. Most would be unlikely willingly to submit to deportation. There would be a risk both of re-offending and of absconding.
293. As Lord Dyson has explained at paras 14 and 15 of his judgment, on 9 September 2008 the Enforcement Instructions and Guidance, which had replaced the Operational Enforcement Manual, was amended so as to lay down a presumption in favour of detaining immigrants where the deportation criteria were met "in order to protect the public from harm and the particular risk of absconding in these cases". I agree with the Court of Appeal and with Lord Dyson that this amendment to her policy was one that it was open to the Secretary of State to make. However, Davis J, at first instance, ruled to the contrary and this led the Secretary of State to withdraw this amendment.
294. The UK Borders Act 2007 has since made provision by section 32 for automatic deportation of "foreign criminals" in specified circumstances. Section 36 requires the Secretary of State to exercise a power of detention of those being deported under section 32 unless "in the circumstances the Secretary of State considers it inappropriate". These provisions were not in force at the time of the events that have given rise to these appeals.
295. There is thus a picture of a series of changes to policy, and of legislation, that has been influenced by decisions of the courts, not all of which have been sound.

The practical implications of the policies

296. Under this heading I propose to consider the practical implications of complying with the Secretary of State's published policies, including her directions as to the effect of article 5 and our domestic case law. Compliance with the Hardial Singh guidelines, even as I have interpreted them, gives rise to some practical difficulties. Detention pending deportation is permissible for a lengthy period provided that the Secretary of State is taking reasonable steps to effect deportation and provided that there is a reasonable prospect that deportation will be possible. It is the latter proviso that raises particular difficulties for the possibility of deportation may vary from time to time. R (Khadir) v Secretary of State for the Home Department [2005] UKHL 39; [2006] 1 AC 207 illustrates this problem. In the case of each detainee it is necessary to keep the practicability of deportation under review. This problem is made more difficult if there is a requirement to detain for no longer than is reasonable "in all the circumstances", where those circumstances include the nature of the crimes committed by the FNP and the degree of risk of re-offending. The assessment of what period of detention is reasonable in all the circumstances is not an easy one and there will inevitably be cases where, if subjected to judicial review, it will be held not to have been correctly answered. Furthermore the material circumstances are likely to be subject to frequent change, so that frequent reviews will be necessary.
297. This last fact was reflected by the requirements in relation to "Detention reviews" imposed under para 38.8 of the Operational Enforcement Manual which lie at the heart of the appeal in SK (Zimbabwe).
298. Where there are concerns about the risk of absconding, it may be possible to meet these by measures which fall short of detention. The Secretary of State may release a FNP who is subject to deportation under a restriction order setting out terms as to residence, employment or occupation and reporting to the police pursuant to paragraph 2(5) of Schedule 3. The court has the same power in respect of those recommended for deportation under paragraphs 4 to 6 of Schedule 3.
299. Detainees also have the right to apply for bail. In para 12.8 of Fairer, Faster, Firmer the Government explained that it believed that there should be a more extensive judicial element in the detention process and proposed that this should be by way of bail hearings, commenting on the resource implications that this would have. Para 4.83 of Secure Borders, Safe Haven stated that Part III of the Immigration and Asylum Act 1999 had created a complex system of automatic bail hearings at specific points in a person's detention, that this had never been brought into force and that most of it was to be repealed. There is now a comprehensive statutory scheme for release on bail produced through a series of amendments to Schedule 2 to the 1971 Act.
300. All of this illustrates the practical problems implicit in the implementation of a regime that attempts to give effect to the policy of using the power to detain only as a last resort. Despite efforts to implement this policy there will inevitably be cases where individuals are detained when, under the policy, they should not be. The question arises of whether those who find themselves in this position are entitled to claim damages for false imprisonment.

What are the legal implications of the Secretary of State's published policies?

301. The appellants in this case should have been detained had the Secretary of State's published policy been applied. They claim to be entitled to damages for false imprisonment because those considering their cases reached the right conclusions by applying the wrong policy. Their complaint is as to the manner in which the decisions to detain them were taken, not as to the substance of those decisions. Thus, the question of the legal effect of the Secretary of State's published policies is not directly in issue. Nonetheless, underlying the appellants' case is the premise that it would not have been lawful for the Secretary of State to reach a decision that was in conflict with her published policy. For this reason she was required to reach her decision by applying her published policy, not some other policy. It follows that it is material to consider the effect of the Secretary of State's published policies.
302. I agree with Lord Dyson that, under principles of public law, it was necessary for the Secretary of State to have policies in relation to the exercise of her powers of detention of immigrants and that those policies had to be published. This necessity springs from the standards of administration that public law requires and by the requirement of article 5 that detention should be lawful and not arbitrary. Decisions as to the detention of immigrants had to be taken by a very large number of officials in relation to tens of thousands of immigrants. Unless there were uniformly applied practices, decisions would be inconsistent and arbitrary. Established principles of public law also required that the Secretary of State's policies should be published. Immigrants needed to be able to ascertain her policies in order to know whether or not the decisions that affected them were open to challenge.
303. What is the effect of a decision to take action that falls within a power conferred by statute but which conflicts with a published policy as to the manner in which that power will be exercised? This is no easy question. It overlaps with the question of the nature and effect of a legitimate expectation. Is a decision that is contrary to policy unlawful, so that action taken pursuant to it is ultra vires? If so a published policy has the same effect as delegated legislation. Is this result dictated by the jurisprudence that has its origin in Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147? Mr Husain QC for the appellants submitted that it is. He relied on the oft-cited catalogue of matters rendering the decision of a tribunal void propounded by Lord Reid in Anisminic at p 171. This included:

"It may have refused to take into account something which it was required to take into account. Or it may have based its decision on some matter which, under the provisions setting it up, it had no right to take into account."

The words that I have emphasised are significant. Lord Reid's proposition cannot necessarily be extended to the situation where the decision-maker fails to have regard to his own policy. Anisminic is, however, only the start of the story.

304. The effect of Anisminic was the subject of observations by Lord Diplock, which have been treated as authoritative. In In re Racal Communications Ltd [1981] AC 374, at pp 382-383 he described Anisminic as a legal landmark. It established that when Parliament conferred on an administrative authority the power to decide particular questions defined by the Act conferring the power, and the authority asked itself and answered the wrong question, it did something that the Act did not empower and its decision was a nullity. In O'Reilly v Mackman [1983] 2 AC 237, at p 278 he observed that if a tribunal mistook the law it must have asked itself the wrong question and one that it had no jurisdiction to determine, so that its decision was a nullity. In R v Hull University Visitor, Ex p Page [1993] AC 682, at pp 701- 702 Lord Browne-Wilkinson endorsed his comment, adding that any error of law made by an administrative tribunal or inferior court in reaching its decision could be quashed for error of law. Earlier at p 701 he had observed that it was to be taken that Parliament had only conferred the decision-making power on the basis that it was to be exercised on the correct legal basis with the effect that an error of law rendered the decision ultra vires. This reasoning cannot readily be extended to a decision which departs from executive policy. It would be totally unrealistic to postulate that when Parliament passes an Act conferring a discretionary power it does so with the intention that if the decision-maker publishes a policy in relation to the exercise of that power he will abide by that policy unless he has good reason not to do so.
305. In Boddington v British Transport Police [1999] 2 AC 143 the House of Lords took Anisminic a significant step further. The issue was whether the appellant could raise by way of defence to a criminal charge a contention that the bye-law, or an administrative decision taken under it, pursuant to which he was prosecuted, was ultra vires. Lord Irvine of Lairg LC, giving the leading speech, said at p 155 that an order made by the Secretary of State in the purported exercise of a statutory power would be regarded as void ab initio if it had been made in bad faith, or as a result of taking into account an irrelevant, or ignoring a relevant, consideration. At p 158 he said:

"The Anisminic decision established, contrary to previous thinking that there might be error of law within jurisdiction, that there was a single category of errors of law, all of which rendered a decision ultra vires. No distinction is to be drawn between a patent (or substantive) error of law or a latent (or procedural) error of law."

Lord Irvine added, at p 159:

"Also, in my judgment the distinction between orders which are 'substantively' invalid and orders which are 'procedurally' invalid is not a practical distinction which is capable of being maintained in a principled way across the broad range of administrative action."

306. Other members of the House were not prepared to reject the possibility that an ultra vires act might have legal consequences before its invalidity was recognised by the court: see Lord Browne-Wilkinson, at p 164 and Lord Slynn, at p 165.
307. Boddington no longer judged the vires of the exercise of a discretionary power by the assumed intention of Parliament. It held that if a decision was vitiated by procedural impropriety it was ultra vires and a nullity.
308. In the light of Boddington these appeals raise two issues: (i) is a decision of the Secretary of State that, without good reason, conflicts with her published policy outside her powers, so that it is a nullity? (ii) is a decision reached by the Secretary of State by the application of a policy that conflicts with her published policy a nullity, even if the decision itself accords with her published policy? I am currently concerned with the first question. The proposition underlying the appellants' case is that if a minister, without good reason, acts in a way that is contrary to her published policy she acts outside her powers. Her action is unlawful and can found a claim for damages if it infringes a private law right. It is time to look at the law relating to policy and legitimate expectation.
309. Where a public authority gives an undertaking to an individual that a discretionary power will be used in a particular way, this creates a legitimate expectation in the individual that the authority will comply with that undertaking. The courts will require the authority to give due consideration to that legitimate expectation when exercising its power: see R (Bibi) v Newham London Borough Council [2001] EWCA Civ 607; [2002] 1 WLR 237. In an extreme case the courts can require the authority to comply with its undertaking: see R v North and East Devon Health Authority, Ex p Coughlan [2001] QB 213. The same principle applies where a minister publishes a policy that sets out the way in which he intends to exercise a discretionary power. This also creates a legitimate expectation in those affected by the power as to the manner in which it will be exercised. In either case the court can intervene in performance of its duty to ensure that the executive acts fairly and does not abuse the powers conferred on it by Parliament.
310. These principles have quite often been applied in relation to immigration decisions. I have already quoted my invocation of them in Saadi: see para 271 above. Lord Dyson at para 85 has referred to my judgment in Nadarajah v Secretary of State for the Home Department [2003] EWCA Civ 1768. In that case the Court of Appeal held that N's detention was unlawful because it was in conflict with the Secretary of State's policy. The court further held that as the detention was contrary to law, it infringed article 5(1)(f) of the Convention. The Court had earlier stated at para 15 that the consequence of these findings would be that N would be entitled to damages for unlawful detention. The Secretary of State had sought to show that N's detention was not contrary to his policy. He had not challenged the proposition that if the detention was contrary to his policy it would be unlawful, nor that this would lead to liability in damages. There was no discussion of the nature and effect of the doctrine of legitimate expectation in the context of detention under the 1971 Act. There was, however, a sequel to the case in which the nature of the doctrine of legitimate expectation received detailed consideration.
311. The development of the law of legitimate expectation was pellucidly set out at some length by Laws LJ, giving the only reasoned judgment in the Court of Appeal in R (Nadarajah) v Secretary of State for the Home Department, [2005] EWCA Civ 1363, a decision which, amazingly, does not seem to have found its way into the law reports. At the end of his judgment, in paras 68 and 69, Laws LJ set out his conclusions on the principles to be derived from these authorities. This merits citation at length:

"The search for principle surely starts with the theme that is current through the legitimate expectation cases. It may be expressed thus. Where a public authority has issued a promise or adopted a practice which represents how it proposes to act in a given area, the law will require the promise or practice to be honoured unless there is good reason not to do so. What is the principle behind this proposition? It is not far to seek. It is said to be grounded in fairness, and no doubt in general terms that is so. I would prefer to express it rather more broadly as a requirement of good administration, by which public bodies ought to deal straightforwardly and consistently with the public. In my judgment this is a legal standard which, although not found in terms in the European Convention on Human Rights, takes its place alongside such rights as fair trial, and no punishment without law. That being so there is every reason to articulate the limits of this requirement – to describe what may count as good reason to depart from it – as we have come to articulate the limits of other constitutional principles overtly found in the European Convention. Accordingly a public body's promise or practice as to future conduct may only be denied, and thus the standard I have expressed may only be departed from, in circumstances where to do so is the public body's legal duty, or is otherwise, to use a now familiar vocabulary, a proportionate response (of which the court is the judge, or the last judge) having regard to a legitimate aim pursued by the public body in the public interest. The principle that good administration requires public authorities to be held to their promises would be undermined if the law did not insist that any failure or refusal to comply is objectively justified as a proportionate measure in the circumstances.

This approach makes no distinction between procedural and substantive expectations. Nor should it. The dichotomy between procedure and substance has nothing to say about the reach of the duty of good administration. Of course there will be cases where the public body in question justifiably concludes that its statutory duty (it will be statutory in nearly every case) requires it to override an expectation of substantive benefit which it has itself generated. So also there will be cases where a procedural benefit may justifiably be overridden. The difference between the two is not a difference of principle. Statutory duty may perhaps more often dictate the frustration of a substantive expectation. Otherwise the question in either case will be whether denial of the expectation is in the circumstances proportionate to a legitimate aim pursued. Proportionality will be judged, as it is generally to be judged, by the respective force of the competing interests arising in the case. Thus where the representation relied on amounts to an unambiguous promise; where there is detrimental reliance; where the promise is made to an individual or specific group; these are instances where denial of the expectation is likely to be harder to justify as a proportionate measure. …On the other hand where the government decision-maker is concerned to raise wide-ranging or 'macro-political' issues of policy, the expectation's enforcement in the courts will encounter a steeper climb. All these considerations, whatever their direction, are pointers not rules. The balance between an individual's fair treatment in particular circumstances, and the vindication of other ends having a proper claim on the public interest (which is the essential dilemma posed by the law of legitimate expectation) is not precisely calculable, its measurement not exact."

312. This passage in Laws LJ's judgment was obiter, as is my approval of it as setting out a compelling analysis of the law. I have, however, some concerns as to the consequences of this extension of the principles of judicial review if a minister's unjustified failure to comply with his policy is to be treated as an unlawful act that subjects him to potential liabilities in private law independently of the discretionary remedies of judicial review. The facts of this case illustrate my concern. Assume that I am correct in concluding that Schedule 3 gives the Secretary of State wide ranging powers to detain those who are illegally in this country, whether pending the processing of their applications for permission to enter or pending their removal when such applications have been refused. The Secretary of State did not choose to exercise those powers to their full extent. The policy of only detaining immigrants as a last resort is a benevolent policy. That policy carries with it, however, as I have explained, the risk that, as a result of erroneous decisions in the application of a difficult test, some immigrants may be detained in breach of that policy. Not all would agree that it is fair that they should be entitled to compensation, in the form of damages for false imprisonment, on the same scale as those whose detention falls outside any statutory power.
313. Notwithstanding these concerns, I have concluded that the detention of a person in circumstances where, under the Secretary of State's published policies he should not have been detained, was both a violation of principles of public law and unlawful so as to exclude any justification for the detention under article 5(1)(f) of the Convention and to give rise to liability for false imprisonment.

Was the detention of each of the appellants contrary to the Secretary of State's published policies?

314. On this question all members of the court are agreed. Each of the appellants, by reason of the risk of re-offending and of absconding, fell into the exceptional category of those who were liable to detention under the Secretary of State's published policies. Had the decision-maker applied those policies each of the appellants would have been detained. On this topic I have nothing to add to the judgment of Lord Dyson.

What were the defects in the policy applied when deciding whether the appellants should be detained?

315. Mr Beloff conceded that the policy, which I shall call "the secret policy", applied by the decision-maker when deciding to detain the appellants was "unlawful" on three counts. The meaning of the word "unlawful" in this context needs clarification. It is not having a policy but implementing a policy that may infringe the law. A policy can be described as "unlawful" if action taken pursuant to it will, or may, be unlawful under private or public law. The first reason why the policy that was applied was unlawful was that it was a "blanket policy". It brooked of no exceptions save rare ones that might be made for compassionate reasons. It was a blanket policy in that it was applied to any FNP recommended for deportation. Thus it paid no regard to the severity of the offence for which the FNP had been convicted nor to the likelihood and consequence of his re-offending. In this I do not consider that it was open to objection on public law grounds. On my interpretation of Schedule 3 it was open to the Secretary of State to decide her own criteria for determining those she wished to detain pending deportation. Secondly the policy was a blanket policy in that it gave no scope for the application of discretion, save on compassionate grounds. I do not consider that this was objectionable on public law grounds. There is a principle of public law that a minister who is granted a discretionary power must not improperly fetter the exercise of his discretion. In a case such as this, however, a minister has to lay down a firm policy in order to ensure consistency of decision-making. The Secretary of State was laying down an exception to the general rule that there was a presumption against detention. I can see no reason why she should not lay down a policy under which all those whom she chose to deport should be detained pending their deportation save in circumstances where there were compassionate reasons for departing from the policy. Where the secret policy was objectionable was that it was to be applied without consideration of whether detention would violate the Hardial Singh principles. Under the secret policy FNPs were liable to be detained even if they were nationals of a country to which deportation was not a practical possibility. Thus application of the secret policy would, in some cases, result in detention that was outside the power conferred by the 1971 Act.
316. The second reason why Mr Beloff conceded that the secret policy was unlawful was that it was inconsistent with the Secretary of State's published policy. Inasmuch as the application of the secret policy would result in the detention of some who would not have been detained on the application of the published policy it was plainly open to objection on this ground.
317. The third reason why Mr Beloff conceded that the secret policy was unlawful was that it was secret. This concession was also properly made, for the reasons given by Lord Dyson.

What were the circumstances in which the secret policy was applied?

318. Lord Dyson has summarised the circumstances in which the secret policy was applied. Lord Walker and Lord Collins have described these as a serious abuse of power. Lady Hale has expressed some sympathy for the predicament in which the civil servants found themselves. Having considered the voluminous discovery, some of it very belated, I share both the concern expressed by Lord Walker and Lord Collins and the sympathy expressed by Lady Hale. It was clear to officials that the Secretary of State wished all FNPs to be deported, and to be detained pending deportation. They were concerned at the legality of such a policy, particularly because it conflicted with the published policy. There was a protracted period preparing and circulating draft advices to ministers in which each of a lengthy circulation list was given the chance to comment. Many did, whereupon the draft would be re-circulated. There seems to have been a reluctance to grasp the nettle of presenting advice that would be unpalatable. There were considerable delays caused by the perceived need to obtain counsel's opinion. The picture is of bureaucracy at its worst with the best proving constantly the enemy of the good. The lamentable fact is that approximately two years elapsed between the identification of the need to publish a revised policy on detention and the publication of such a policy in the new Enforcement Instructions and Guidance in September 2008. Although it was suggested that ministers might favour a policy that would appear to make the courts responsible for the release from detention of FNPs, this course was neither advocated nor adopted. There was muddle galore, but I am not persuaded having considered the considerable discovery that there was a deliberate attempt to deceive the courts as to the policy that was being applied.

Did the application of the secret policy render the detention of the appellants unlawful?

319. It is now accepted by the Secretary of State that the decision-maker in the case of each of the appellants decided that he should be detained by applying to him the secret policy. The power to detain that he purported to apply was that conferred by Schedule 3. Had the appellants been persons whose deportation would not be possible within a reasonable period, so that they fell outside the lawful application of Schedule 3, the application of the secret policy would have resulted in their detention none the less. Had the appellants been persons whose detention would have conflicted with the Secretary of State's published policy, so that their detention would be unlawful under the principles of public law discussed above, the application of the secret policy would have resulted in their detention none the less. In the event each appellant fell within the group of FNPs for whom detention was appropriate, indeed inevitable, if the Secretary of State's published policy was applied to them. Was their detention none the less unlawful because of the process of reasoning that had brought it about? Mr Husain submits that it was. The Court of Appeal held that it was not, because the application of the secret policy was not material; it produced the result that would have been produced had the right policy been applied.
320. This is I believe a novel question, not to be answered by the simple answer that the detention was "unlawful" because the decision that produced it was "unlawful". It is also a complex question because of the novel feature of the existence of a published policy that would have predetermined the decisions in relation to the two appellants had it been applied.
321. It is helpful to unpick the secret policy and consider its effect if each objection to it had been the only objection. I take first the objection that it was a blanket policy. Imagine the Secretary of State had publicly announced that all FNPs who were given deportation orders would henceforth be detained pending deportation, subject to exceptions on compassionate grounds. The application of this policy would have been objectionable in that it would have resulted in the detention of those FNPs whose deportation would not be possible within a reasonable period. Would this fact have rendered unlawful the detention of the majority of FNPs whose deportation was possible? I see no reason of principle why it should.
322. Next I take the objection that the policy was in conflict with the published policy. Had the secret policy not been secret, this objection would have melted away. The public policy would simply have been publicly altered. This would not have been objectionable.
323. What of the objection that the policy was secret? Had this been the only objection to it I do not see how this could have availed the appellants. They were already subject to a policy that would result in their detention. They had no legitimate expectation of being permitted to remain at large. If the secret policy had extended the category of those FNPs who would be detained, those who, without knowing it, were brought within the category of detainees might have had cause to object to their detention, but I do not see how those who were going to be detained under the previous published policy could legitimately complain.
324. If none of the individual objections to the secret policy would have afforded the appellants legitimate grounds for challenging their detention, does the position change when the objections are considered cumulatively? I can see no reason why it should. Both logically and intuitively my conclusion is that the introduction of the secret policy gave those whose detention resulted from it cause to challenge the legality of their detention, but not those whose legitimate expectation was that they would be detained under the application of the published policy.
325. This was also the reaction of those officials who questioned the application of the secret policy. Their concern was that those whose detention infringed the Hardial Singh guidelines or the published policy would have claims for illegal detention, not that every detainee would have such a claim.
326. I now turn to some of the authorities relied upon by the appellants to see how they bear on the unusual problem raised by these appeals.
327. In Christie v Leachinsky [1947] AC 573 the plaintiff, who claimed damages for false imprisonment, had been arrested by police officers on a charge of "unlawful possession" under the Liverpool Corporation Act 1921. That Act did not give a power to arrest for this offence. The defendants raised by way of defence a plea that, at the time of the arrest they reasonably suspected him of receiving stolen goods, which provided a valid ground for his arrest. The House of Lords held that this was no defence as at the time of his arrest he had been given a different ground of arrest. This decision is normally cited for the proposition that an arrest will be unlawful if the person arrested is not told the reason for his arrest at the time that he is arrested. It is also authority for the proposition that if a person is arrested for a reason which is not a valid statutory ground of arrest it is no defence to an action for false imprisonment that he could have been validly arrested on alternative grounds. I do not see that this decision bears on the very different facts of the present case.
328. In Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662 the Court of Appeal held that detention by the police without the review required by section 40(1)(b) of the Police and Criminal Evidence Act 1984 was unlawful and the fact that, if the review had taken place authorised detention would have continued was no answer to a claim for substantial damages for false imprisonment. That decision has no bearing on the issue that I am considering, which is whether the detention of the appellants was unlawful.
329. In Langley v Liverpool District Council [2005] EWCA Civ 1173; [2006] 1 WLR 375 a constable had purported to exercise a discretionary power under the Children Act 1989 in removing a child from its family. The Court of Appeal held that his exercise of discretion had been wrongful in that, in the circumstances prevailing, it ran counter to the statutory scheme. It followed that the removal of the child was unlawful and the constable had committed the tort of false imprisonment. I have found this decision of no assistance in deciding whether, on the unusual facts of this case, the detention of the appellants was unlawful.
330. One of the cornerstones of the appellants' case was the speech of Lord Diplock, with which all other members of the House agreed, in Holgate-Mohammed v Duke [1984] AC 437. A police officer had arrested the plaintiff on suspicion of the theft of jewellery. He did so pursuant to section 2(4) of the Criminal Law Act 1967, which gave him a discretionary power to arrest her. She alleged, however, that the exercise of this power had been unlawful because, when deciding whether to exercise his discretion the officer had been influenced by a consideration which was irrelevant, namely the likelihood that the fact that she had been arrested would be more likely to induce her to confess to her crime when interviewed. Lord Diplock held that this consideration was not irrelevant to the proper exercise of the officer's discretion. It was a matter to which he could legitimately have regard having regard to the objectives of the statutory power to arrest. Thus the comments upon which the appellants have relied were obiter. Those comments, at p 443 of Lord Diplock's speech, were as follows:

"The Wednesbury principles, as they are usually referred to, are applicable to determining the lawfulness of the exercise of the statutory discretion of a constable under section 2(4) of the Criminal Law Act 1967, not only in proceedings for judicial review but also for the purpose of founding a cause of action at common law for damages for that species of trespass to the person known as false imprisonment, for which the action in the instant case is brought.

The first of the Wednesbury principles is that the discretion must be exercised in good faith. The judge in the county court expressly found that Detective Constable Offin in effecting the initial arrest acted in good faith. He thought that he was making a proper use of his power of arrest. So his exercise of that power by arresting Mrs Holgate-Mohammed was lawful, unless it can be shown to have been 'unreasonable' under Wednesbury principles, of which the principle that is germane to the instant case is: 'He [sc the exerciser of the discretion] must exclude from his consideration matters which are irrelevant to what he has to consider'."

331. The way in which the appellants argue that this passage impacts on the facts of the present case is, as I understand it, as follows. The decision to detain the appellants was taken by application of the secret policy. This infringed the Wednesbury principles because it failed to have regard to relevant considerations, namely whether the Hardial Singh principles precluded detention and whether the appellants' detention complied with the published policy. The answer given by the Court of Appeal to this point is that, so far as the appellants were concerned, the failure to consider these matters was not material because, had they been considered, the conclusion would have been that neither matter posed an impediment to the appellants' detention and had they been considered the decision would inevitably have been the same.
332. The approach of the Court of Appeal involves a refinement of the Wednesbury principles. It is an application of the following reasoning of May LJ in R v Broadcasting Complaints Commissioner, Ex p Owen [1985] QB 1153, 1177:

"Where the reasons given by a statutory body for taking or not taking a particular course of action are not mixed and can clearly be disentangled, but where the court is quite satisfied that even though one reason may be bad in law, nevertheless the statutory body would have reached precisely the same decision on the other valid reasons, then this court will not interfere by way of judicial review."

333. I believe that at least in this new area of the effect of public policy, the approach of the Court of Appeal is a sound one. Where a minister publishes a policy as to the circumstances in which he will exercise a statutory power and then he or his officials apply a different policy which results in the exercise of that power in circumstances which range wider than those published, I do not consider that those whose cases fell within the published policy can automatically contend that the application of the power to them has been unlawful. If the facts are that no reasonable decision-maker applying the published policy could have done other than reach the decision which the decision-maker arrived at, the fact that he applied a more expansive, but unpublished, policy when reaching his decision will not invalidate that decision.
334. For these reasons, which accord I believe with the reasoning of both the Court of Appeal and Lord Brown, I have concluded that the application of the secret policy did not render the detention of the appellants unlawful.

Damages

335. In view of my conclusions on liability, the issue of damages does not arise. Had I agreed with Lord Dyson on liability, I would have shared his approach to damages. I also endorse Lord Collins' conclusions in relation to vindicatory damages.

Reviews

336. A word about reviews. If the majority are correct in concluding that the application of the secret policy rendered all decisions taken pursuant to it unlawful, then it seems to me that the moment that the secret policy was applied to reviews of the lawfulness of those detained, their continued detention would have been rendered unlawful, even if they were lawfully detained under the published policy before the secret policy was introduced and even if they remained subject to detention within the terms of the published policy. This would be an extraordinary result.

MR LUMBA'S HARDIAL SINGH APPEAL

337. Mr Lumba has now voluntarily left the country, but the question remains of whether his detention became unlawful in the period before he left. Because of the view that I take of the scope of the Hardial Singh principles, I find the issues in relation to this part of Mr Lumba's appeal easier to resolve than has Lord Dyson. The lengthy period during which Mr Lumba was detained largely resulted from his own efforts to avoid deportation. For most of the period his deportation had been a practical possibility. The Secretary of State had not been dragging her feet in her effort to deport Mr Lumba. I agree with the Court of Appeal that Mr Lumba could not be heard to say that it was impossible to deport him within a reasonable time when the difficulty in doing so resulted from his own attempts to avoid deportation and not from extrinsic problems in effecting his deportation.
338. On my view of the interpretation of Schedule 3, whether Mr Lumba posed a risk of re-offending was not relevant to the period for which he could lawfully be detained. It seems to me that para 107 of Lord Dyson's judgment lends support to my belief that the power to detain is not dependent upon an object of the detention being to facilitate deportation. The question remains, however, in Mr Lumba's case of whether his continued detention, having particular regard to his mental condition, remained consistent with the Secretary of State's published policy. He has now voluntarily left the country, but if the issue of the lawfulness of his detention is to be pursued I agree with Lord Dyson that this is a matter to be considered by the Administrative Court on remission.
339. Subject to this, for the reasons that I have given I would dismiss these appeals.

LORD BROWN (with whom Lord Rodger agrees)

340. Amongst the many issues to be decided on these appeals far and away the most important concerns the true nature of the tort of false imprisonment. Lord Dyson having set out all the relevant facts and the detailed legal context in which the many issues here arise for determination, I shall proceed without more to what I recognise is to be a dissenting judgment on this crucial question.
341. "Freedom from executive detention is arguably the most fundamental right of all." Thus Lord Bingham of Cornhill in his 2002 Romanes lecture. The tort of false imprisonment is, of course, the remedy provided by law for the violation of this freedom, for the unlawful deprivation of a person's liberty. The outcome of the appeals proposed by the majority of the court is to hold the appellants – and, indeed, a large number of others similarly placed – to have been unlawfully detained, in many instances for a period of years, and yet to compensate them by no more than a nominal award of damages. They are to be held unlawfully detained because, in his (or her) exercise of the undoubted power to detain them, the Secretary of State breached certain public law duties. But they are to be awarded only nominal damages because, whatever approach had been taken to the exercise of the detaining power, the appellants must inevitably have been detained in any event.
342. Whilst I share to the full the majority's conclusion that it would be quite wrong in the circumstances of these cases to award the appellants any substantial compensation in respect of their detention, for my part I would reach that conclusion by a very different route. I would hold that a public law breach of duty in the course of exercising an executive power of detention does not invariably, and did not here, result in the subsequent detention itself being unlawful - in short, that these appellants were not the victims of false imprisonment.
343. Naturally I recognise the beguiling simplicity and apparent purity of the majority's approach. Ever since the House of Lords decision in Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147, decisions made within the decision-maker's jurisdiction but containing a public law error have generally been regarded as precisely equivalent to decisions made entirely without jurisdiction. Thus, it is said, a decision to detain which is flawed by a public law error is ultra vires and the detention which follows is necessarily unauthorised and therefore unlawful. Logically, indeed, this must be so however minor the public law error involved in the making of the decision and however inevitable it is that the decision to detain would have been made in any event. (At one stage in the argument Mr Husain QC for the appellants accepted that some public law failures might be too inconsequential to require this conclusion but unsurprisingly he was unable to formulate any test by which to determine such cases and, indeed, he had argued before this court on behalf of the appellant in R (SK (Zimbabwe)) v Secretary of State for the Home Department [2009] 1WLR 1527 that false imprisonment must follow inexorably even from periodic failures to comply with the Secretary of State's self-imposed monthly review process.) Given, moreover, that the tort of false imprisonment is actionable per se – that "it is of the essence of the tort of false imprisonment that the imprisonment is without lawful justification" (Lord Hope in R v Governor of Brockhill Prison ex parte Evans [2001] 2 AC 19, 32) – logic also suggests that the notion of nominal damages should have no part to play in determining the compensation payable. Why should someone imprisoned without lawful justification be paid nominal damages only? If the answer is that they would have been imprisoned anyway, under the same power and in just the same way, then in reality the Court is saying that the tort may be committed merely in a technical way. I have to say that such an approach would to my mind seriously devalue the whole concept of false imprisonment.
344. It is true that in Murray v Ministry of Defence [1988] 1 WLR 692, whilst confirming that an action for false imprisonment lies even if the detainee does not know that he has been imprisoned, Lord Griffiths added (p703 A-B):

"If a person is unaware that he has been falsely imprisoned and has suffered no harm, he can normally expect to recover no more than nominal damages".

In my opinion, however, there is a very real difference between a detainee who is in fact unaware of being under physical restraint (perhaps because he is asleep or because he simply does not know that the door has been locked) and a detainee who is fully aware of his loss of freedom. To award the latter nominal damages only, on the basis that, even had he been dealt with lawfully he would still have been deprived of his freedom anyway, is really to say that he was in truth rightly in detention. That seems to me very different from saying that he was wrongly imprisoned but happily unaware of it. I have difficulty, therefore, with Lord Dyson's criticism (at paras 92 and 93) of the passage on damages in Clarke LJ's judgment in Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662, 668. To compensate (or rather to deny compensation) on the basis that the detainee "has suffered no loss because he would have remained in detention whether the tort was committed or not" is in my opinion the very negation of the tort: it is to hold that the detainee was at one and the same time both rightly and wrongly imprisoned.

345. Search as one may in the judgments both of the Court of Appeal and of the House of Lords in Christie v Leachinsky (respectively [1946] KB 124 and [1947] AC 573), there is no hint of a suggestion that the false imprisonment which followed upon the unlawful arrest in that case might properly attract only a nominal award of damages. I do not say that those falsely imprisoned should be compensated identically irrespective of how deserving they were of liberty rather than restraint during the relevant period of detention; I do say, however, that rather than hold a detainee simultaneously both rightly detained and falsely imprisoned, the law should instead recognise that, notwithstanding a flaw in the decision-making process such as to involve the breach of a public law duty, the decision-maker has not in those circumstances committed the tort of false imprisonment. A court which speaks with two voices risks bringing the law into disrepute.
346. Is this court then bound by established law to reach so unsatisfactory a conclusion as the majority of the court suggest: tortious liability but for nominal damages only? To my mind it is not. Assuming that a power to detain exists, that any preconditions to its exercise have been satisfied and any limitations upon its scope observed, I know of no case which holds the detainee to have been falsely imprisoned merely because, in the course of exercising the power, the decision-maker committed some public law breach of duty. On the contrary, and to my mind hardly surprisingly, the courts have consistently shied away from such a conclusion as is amply and consistently demonstrated by the series of Court of Appeal judgments in this very line of cases. Such was the decision of the Court of Appeal (Laws, Keene and Longmore LJJ) in SK (Zimbabwe), the Court of Appeal (Lord Neuberger of Abbotsbury MR, Carnwath and Stanley Burnton LJJ) in the cases from which the present appeals are brought, and the Court of Appeal (Maurice Kay, Longmore and Black LJJ) more recently still in Anam v Secretary of State for the Home Department [2010] EWCA Civ 1140. In Anam, indeed, Black LJ in giving the leading judgment characterised the appellant's argument thus at para 17:

"At its most radical, the submission advanced by [counsel] is that a failure by the Secretary of State to apply his published policy in making a decision to detain necessarily renders that decision a nullity and the resulting detention unlawful."

Have all these Lords Justices, many of them distinguished public law jurists, lost sight of the clear and basic principles of public law which, it is said, necessarily compel such a 'radical' conclusion?

347. So far from the authorities supporting such a surprising conclusion they can and should, I suggest, be analysed rather to the following effect.

(1) False imprisonment is the inescapable result of detention absent any power to detain R v Governor of Brockhill Prison (No 2) Ex p Evans [2001] 2 AC 19 is itself a good illustration of this basic proposition: the appellant having been kept in prison beyond the date when her custodial sentence, properly calculated, expired, there could hardly have been a clearer case of false imprisonment. Such would also be the position were someone arrested for a non-arrestable offence. Analysed in the way I analysed the power of detention under Schedule 2 to the Immigration Act 1971 (directly analogous to the Schedule 3 power under consideration in these appeals) in R (Khadir) v Secretary of State for the Home Department [2006] 1 AC 207, these are cases not about the exercise of the power, but rather about its existence.

(2) The tort of false imprisonment is committed too if someone is detained in breach of a condition precedent to the existence of the detention power. Christie v Leachinsky [1947] AC 573 illustrates the common law's imposition of such a condition precedent: a right of arrest only arises when the citizen is told why he is being arrested. Roberts v Chief Constable of the Cheshire Constabulary [1999] 1 WLR 662 illustrates the imposition of such a precondition by statute (in that case the requirement for review as a precondition of continued detention beyond six hours). R v Secretary of State for the Home Department Ex p Khawaja [1984] AC 74 provides another such illustration, the House of Lords there deciding that the power to detain and remove an immigrant as an illegal entrant under Schedule 2 to the 1971 Act was dependent upon establishing such illegality as a precedent fact.

(3) Detention beyond the scope of a detaining power similarly constitutes false imprisonment. For example, the limitations imposed by the courts following Woolf J's decision in R v Governor of Durham Prison Ex p Hardial Singh [1984] 1 WLR 704 (the Hardial Singh principles) undoubtedly operate to constrain the power of detention under paragraph 2 of Schedule 3 and it has long been recognised that detention in breach of the Hardial Singh principles gives rise to tortious liability. (Categories 2 and 3, I recognise, may not always be easily distinguishable. It could, for example, be said that the first Hardial Singh principle constitutes a precondition for the exercise of the detention power. It has seemed to me nonetheless worth attempting the distinction.)

(4) I would accept too that in certain circumstances a power of detention may be narrowed by a published policy as to how it will be exercised. The Court of Appeal's decision in R (Nadarajah) v Secretary of State for the Home Department [2004] INLR 139 is, I think, an illustration of that in the present context. The Secretary of State had in that case adopted a published policy which in substance narrowed the grounds on which the detaining power (in that case under schedule 2 to the 1971 Act) would be exercisable (the stated policy there being to release anyone whose removal was not imminent). Certainly it is on that basis alone that I would regard Nadarajah as correctly decided. And it must, of course, be recognised that, as with any other statement of policy (a policy being, by definition, no more than an advance indication of how it is proposed to exercise the particular discretionary power in question) it is always open to the holder of the power to change that policy – see, for example, in relation to the Immigration Rules themselves, MO (Nigeria) v Secretary of State for the Home Department [2009] 1 WLR 1230.

348. Every false imprisonment case on which the appellant relies can, in my opinion, be seen to fall within one or other of the above four categories. Besides those already mentioned, two cases in particular call for brief special mention.
349. Kuchenmeister v Home Office [1958] 1 QB 496 concerned a German national who landed at Heathrow en route to Dublin. The immigration officers, instead of refusing him leave to land (as they had been instructed to do), detained him at the airport until it was too late for him to catch the Dublin flight. Holding him to have been wrongfully imprisoned, Barry J said this (p 512):

"His liberty was restricted to a greater degree than the immigration authorities were entitled to restrict it under [the particular power they sought to rely upon]. The fact that they might have restricted his mobility by employing the powers conferred upon them by other articles of the Order seems to me to be immaterial. It is no answer, when a man says 'I have been unlawfully arrested without a warrant,' to say 'Well, had I (the person making the arrest) taken the trouble to go and ask for a warrant, I would undoubtedly have got it.' That would be no answer to a claim for unlawful arrest. Similarly here, although the [immigration officers] could have detained the plaintiff by refusing him leave to land, that does not entitle them to detain him on the grounds on which they did."

The case was to my mind correctly decided and can be seen to fall squarely under the first of the above four categories: the immigration officers simply had no power to restrict the plaintiff's movements in the way they did.

350. Langley v Liverpool City Council [2006] 1 WLR 375 concerned a child in obvious need of protection. Two relevant powers existed, respectively under sections 44 and 46 of the Children Act 1989. Section 44 provides for the grant of an emergency protection order (EPO) authorising the council to remove a child into the care of foster carers; section 46 gives the police power to take a child into police protection. The police officer there having removed the child in purported exercise of his power under section 46, the Court of Appeal held that he had been wrong to do so and accordingly that the child had been falsely imprisoned. Dyson LJ in giving the leading judgment said that "the statutory scheme clearly accords primacy to section 44 [which] is sanctioned by the court and . . . involves a more elaborate, sophisticated and complete process than removal under section 46" (para 38); that "section 46 should be invoked only where it is not practicable to execute an EPO" (para 40); and that in the result, albeit the officer "had jurisdiction to remove [the child] pursuant to section 46", he could not lawfully invoke that power unless there were "compelling reasons for exercising this power when, to his knowledge, an EPO was in force which authorised the council to remove [the child] into the care of foster carers" (paras 44 and 46). Again, I have no difficulty in accepting the correctness of this decision: just as the Hardial Singh line of cases imposed restrictions upon the power of detention under the 1971 Act, so the court in Langley thought it right to place a limitation upon the scope of the section 46 power. The case falls neatly into category 3 (or perhaps into category 2, on the basis that the impracticality of executing a concurrent EPO was held to be a precondition to the lawful exercise of the section 46 power).
351. In what circumstances, then, does the breach of a public law duty in the exercise of a power to detain result in the detainee being falsely imprisoned? I have already indicated, at paragraph 347(4) above, one such circumstance, namely when the Secretary of State by his published policy indicates that he will not exercise his power to its fullest extent but rather will confine its exercise within certain limits (for example, as in Nadarajah, by releasing anyone whose removal is not imminent). Such a published policy, unless and until it is changed, as a matter of public law requires the decision-maker to decide cases (subject always to reasoned exceptions) in accordance with it.
352. It is my clear present view (subject to any further argument on the point) that SK (Zimbabwe) provides a good illustration of circumstances where, the breach of a public law duty notwithstanding, the detainee should not be regarded as falsely imprisoned. The Secretary of State there breached what was his undoubted public law duty to review all detention cases monthly in accordance with his published policy on procedure. As it seems to me, however, it is one thing to breach a policy under which a detainee is entitled to be released; quite another to breach a policy under which he is entitled merely to be reviewed for release. Whereas the former will result in false imprisonment, the latter will not. Obviously, if the detainee on review would have been entitled to be released under the Hardial Singh principles (or, if more favourable, the published policy statements) then he has a claim for false imprisonment. But the claim in those circumstances arises from his continued detention beyond the date of such entitlement, not from the failure to review his case.
353. What, however, is the position in a case like the present when the Secretary of State's breach of public law duty consists of applying, in place of his published policy, an unpublished policy less favourable to those subject to the detaining power? On the appellant's case, of course, that automatically results in the false imprisonment of every single detainee whose continued detention has been considered under the wrong policy, irrespective of whether or not they would have had the least prospect of release whatever policy had been applied. Applying the wrong policy, the argument runs, means that the Secretary of State failed to have regard to a material consideration, instead had regard to an immaterial consideration, and therefore reached his decision in an unlawful manner. This renders it a nullity with the result that there was no lawful authority for the ensuing detention.
354. The closest this thesis comes to finding high-level support in the authorities is in a much quoted dictum of Lord Diplock in Holgate-Mohammed v Duke [1984] 1 AC 437. Following Mrs Holgate-Mohammed's arrest on suspicion of theft and her detention for six hours at a police station for questioning, she was found at first instance to have been wrongfully arrested and was awarded £1,000. This was on the basis that the arresting officer, in deciding that there would be a better prospect of her confessing if she were arrested and detained rather than merely interviewed under caution, wrongfully exercised his power of arrest. Upholding the Court of Appeal's decision to allow the Chief Constable's appeal, the House of Lords held that the better prospect of the plaintiff confessing her guilt if arrested and questioned at a police station was in fact a relevant matter so that the arrest was not after all unlawful. In the course of the only reasoned speech, Lord Diplock (at p443) observed that, the condition precedent to the officer's powers of arrest and detention having been fulfilled by his having reasonable cause to suspect the plaintiff to be guilty of an arrestable offence, "this left him with an executive discretion whether to arrest her or not." It is the next passage in the speech that is so strongly relied upon by the appellants:

"Since this is an executive discretion expressly conferred by statute upon a public officer, the constable making the arrest, the lawfulness of the way in which he has exercised it in a particular case cannot be questioned in any court of law except upon those principles laid down by Lord Greene MR in Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223, that have become too familiar to call for repetitious citation. The Wednesbury principles, as they are usually referred to, are applicable to determining the lawfulness of the exercise of the statutory discretion of a constable under section 2(4) of the Criminal Law Act 1967, not only in proceedings for judicial review but also for the purpose of founding a cause of action at common law for damages for that species of trespass to the person known as false imprisonment, for which the action in the instant case is brought."

Lord Diplock then noted that amongst the Wednesbury principles was that whoever exercises the discretion must (in Lord Greene's words) "exclude from his consideration matters which are irrelevant to what he has to consider".

355. It is, to my mind, critical to a proper understanding of Lord Diplock's speech to recognise that the relevant matter in question there, the matter which the judge at first instance held the officer to have wrongly taken into consideration, was a consideration clearly decisive of the way he had exercised his discretion. The whole reason for the officer exercising his power to arrest and detain the plaintiff was so that she would then be more likely to confess. If that had been an unlawful consideration, nothing could be clearer than that without it the officer would not have exercised the power at all. In short, on the facts of the case, the power would have been exercised for an improper reason and the resultant detention necessarily unlawful.
356. Certainly it would in my opinion be quite wrong to regard the case as authority for any wider proposition, least of all a thesis so absolute and unsatisfactory as that contended for here. There had been no argument on the point: the Chief Constable was not even called upon. Small wonder that in the quarter century following, despite the hugely increased scope of public law challenges, not a single case appears to have held an arrested detainee falsely imprisoned on Wednesbury grounds save where the arresting officer acted either in bad faith or for an improper purpose.
357. Perhaps the way to put the point is this. Classically most public law challenges go to the decision-making process rather than to the substance of the decision taken. The substantive decision is for the public body and not for the court to take and generally a successful challenge requires that the impugned decision be taken afresh. Obviously, if the substantive decision reached is outside the powers conferred upon the public body purporting to make it (ultra vires in the traditional, literal sense), then it must be nullified. So too if it is irrational in the public law (Wednesbury) sense – i.e. outside the range of permissible decisions which the public body is empowered to make. But if, as here, the challenge is to the process whereby decisions to continue a detainee's detention were taken – or, indeed, as in SK (Zimbabwe), if what is challenged is a failure to take decisions (as often as promised by the policy) as to whether a detainee's detention should continue – then the appropriate remedy is to require the decision-maker to act lawfully: to re-decide decisions that resulted from a flawed decision-making process or, as in SK (Zimbabwe), to take decisions that earlier the decision-maker had omitted to take. It simply does not follow from a flawed decision-making process that the detainee's continued detention immediately thereupon becomes unlawful so as to make him a victim of false imprisonment until a fresh decision comes later to be lawfully taken – so that, for example, a detainee whose detention is reviewed every second month instead of monthly as the published policy dictates, alternates yo-yo like between lawful detention and false imprisonment.
358. Nor does it seem to me that the absolutist approach to the consequences of public law error espoused here by the majority of the Court has received universal endorsement over recent years. The decision of the House of Lords in Boddington v British Transport Police [1999] 2 AC 143 is often cited as a modern example of the Anisminic principle being applied in the ordinary course of proceedings (there in the context of a criminal trial). But there are important dicta in the judgments which expressly leave open the question whether legal consequences may nonetheless follow from unlawful action. Consider this passage from the speech of Lord Slynn of Hadley, an acknowledged master in the field of public law:

"In our jurisdiction the effect of invalidity may not be relied on if limitation periods have expired or if the court in its discretion refuses relief, albeit considering that the Act is invalid. These situations are of course different from those where a court has pronounced subordinate legislation or an administrative act to be unlawful or where the presumption in favour of their legality has been overruled by a court of competent jurisdiction. But even in these cases I consider that the question whether the acts or byelaws are to be treated as having at no time had any effect in law is not one which has been fully explored and is not one on which it is necessary to rule in this appeal and I prefer to express no view upon it. The cases referred to in Wade & Forsyth, Administrative Law, 7^th ed. (1997), pp.323-324, 342-344 lead the authors to the view that nullity is relative rather than an absolute concept (p.343) and that 'void' is 'meaningless in any absolute sense. Its meaning is relative.' This may all be rather imprecise but the law in this area has developed in a pragmatic way on a case by case basis." (p.165 C-F)

(Wade & Forsyth, now in its tenth edition (2010), continues to recognise the relativity of the concept of nullity in this context.)

All I am saying is that if the law is to continue to develop in this area in a pragmatic way on a case by case basis, then in this particular case it should expressly recognise that not every decision to detain affected by a public law breach necessarily carries in its wake an unanswerable claim for false imprisonment.

359. That said, I readily acknowledge that the power of executive detention is one to be exercised with the greatest care and sensitivity and that it is of the first importance that those who exercise it should be ever-mindful not merely of the legal limits of the power but also of such public law duties as surround its proper exercise. Where, as here, a wrong and less favourable policy has been applied in deciding whether a person should be, or continue to be, detained, I accept that it must be for the Secretary of State as decision-maker to establish that this breach of his or her public law duty did not in fact prejudice the detainee. In short, it is for the Secretary of State in these circumstances, in order to avoid liability for the tort of false imprisonment, to establish that the detainee would have been detained in any event – even, that is, had the lawful published policy been followed. Assuming, however, that the inevitability of detention in any event can be established, it seems to me nothing short of absurd to hold the tort of false imprisonment nevertheless made out. After all, had a detainee in such circumstances sought to challenge his continued detention by judicial review (or habeas corpus), the court would have been likely to grant declaratory relief only, declining in its discretion to order the detainee's release. Assuming, indeed, that the Secretary of State's lawful policy in effect dictates the detainee's detention, why, one wonders, would a decision to release him not itself be in breach of the Secretary of State's public law duty?
360. Given, as the Court of Appeal held, and as I understand each member of this court to accept, that there was no realistic prospect whatever of these appellants having been released even had the Secretary of State applied his more favourable published policy, it follows that for my part, whilst, of course, deeply regretting the public law breaches of which the Secretary of State was here guilty, I would dismiss the appeals.
361. The bulk of this judgment was written upon my understanding that the essential choice facing the court was between (a) no false imprisonment and (b) false imprisonment but nominal damages only. It now appears that some members of the court favour a third outcome: (c) false imprisonment with damages of perhaps £500 - £1,000 by way of a "vindicatory" or "conventional" award. Describe such an award how one will, to my mind it cannot sensibly be justified here. Is the court really to award substantial damages to those conceded to have been rightly detained? I have made clear my difficulties with a nominal award of damages. A substantial award would appear to me more objectionable still. Lord Hope (at para 177 of his judgment) refers to Attorney General of Trinidad and Tobago v Ramanoop [2006] 1 AC 328 – a constitutional challenge based upon "some quite appalling misbehaviour by a police officer" (Lord Nicholls at para 2) – and calls here for "some recognition of the gravity of the breach of the fundamental right which resulted in false imprisonment." Properly critical though our judgments may be of the conduct of Home Office officials in these and similar cases, I find it quite impossible to recognise in them any breach (grave or otherwise) of the detainees' fundamental rights. The detainees, I can only repeat, were rightly detained and it would have been wrong to release them.
362. Save insofar as this judgment indicates the area of my disagreement with Lord Dyson's judgment, I wish to say that I am in respectful agreement with it and feel unable to add anything of value on the various other issues arising for decision.